Electrical Circuit Analysis and Design by Noel M. Morris

Electrical Circuit Analysis and Design
www.Technicalbookspdf.com
Other Macmillan titles of re la ted interest
W. A. Atherton, From Compass to Computer
B. R. Bannister and D. G. Whitehead, Fundamentals of Modern Digital
Systems
J. C. Cluley, Transducers for Microprocessor Systems
R. F. W. Coates, Modern Communication Systems, second edition
Donard de Cogan, Solid State Devices - A Quantum Physics Approach
C. W. Davidson, Transmission Lines for Communication, second edition
M. E. Goodge, Analog Electronics
M. E. Goodge, Semiconductor Device Technology
B. A. Gregory, An Introduction to Electrical Instrumentation and
Measurement Systems, second edition
Paul A. Lynn, An Introduction to the Analysis and Processing of Signals,
third edition
Paul A. Lynn, Electronic Signals and Systems
J. E. Parton, S. J. T. Owen and M. S. Raven, Applied Electromagnet-
ics, second edition
P. Silvester, Electric Circuits
Trevor J. Terrell, Introduction to Digital Filters, second edition
M. J. Usher, Sensors and Transducers
L. A. Warnes, El~ctronic Materials
B. W. Williams, Power Electronics - Devices, Drivers, Applications and
Passive Components
G. Williams, An Introduction to Electrical Circuit Theory
Macmillan New Electronics Series
Series Editor: Paul A. Lynn
Rodney F. W. Coates, Underwater Acoustic Systems
W. Forsythe and R. M. Goodall, Digital Control
Paul A. Lynn, Digital Signals, Processors and Noise
Paul A. Lynn, Radar Systems
A. F. Murray and H. M. Reekie, Integrated Circuit Design
F. J. Owens, Signal Processing of Speech
Dennis N. Pim, Television and Teletext
M. J. N. Sibley, Optical Communications
Martin S. Smith, Introduction to Antennas
P. M. Taylor, Robotic Control
G. S. Virk, Digital Computer Control Systems
Alan Waters, Active Filter Design
Electrical Circuit
Analysis and Design
N oel M. Morris
formerly Principal Lecturer
Staffordshire University
M
MACMILLAN
© Noel M. Morris 1993
All rights reserved. No reproduction, copy or transmission of

this publication may be made without written permission.
No paragraph of this publication may be reproduced, copied or
transmitted save with written permission or in accordance with
the provisions of the Copyright, Designs and Patents Act 1988,
or under the terms of any licence permitting limited copying
issued by the Copyright Licensing Agency, 90 Tottenham Court
Road, London W1P 9HE.
Any person who does any unauthorised act in relation to this
publication may be liable to criminal prosecution and civil
claims for damages.
First published 1993 by

THE MACMILLAN PRESS LTD
Houndmills, Basingstoke, Hampshire RG21 2XS
and London
Companies and representatives
throughout the world
ISBN 978-0-333-55483-8 ISBN 978-1-349-22560-6 (eBook)

DOI 10.1007/978-1-349-22560-6
A catalogue record for this book is available

from the British Library
Contents
Preface xii
Acknowledgements xv
1 Elements and Laws 1

1.1 Introduction 1
1.2 Electric current 1
1.3 Voltage, e.m.f. and p.d. 3
1.4 Power in d.c. circuits 4
1.5 Linear passive circuit elements 4
1.6 Resistance, conductance and Ohm's law 5
1.7 Power in a resistive circuit 6
1.8 Energy consumed in a resistive circuit 7
1.9 Independent and dependent supply sources 8
1.9.1 The operational amplifier 9
1.10 Kirchhoff's laws 11
1.11 The double-suffix voltage notation 14
1.12 Practical (non-ideal) sources 16
1.13 Transformation of practical sources 17
1.14 Resistance of aseries circuit 19
1.15 Voltage division in series-connected resistors 20
1.16 Resistance and conductance of a parallel circuit 21
1.17 Current division in a parallel circuit 22
Unworked problems 24
2 Circuit Analysis 28
2.1 Introduction 28
2.2 Definitions and terminology 28
2.3 Mesh analysis 31
VI Contents
2.4 General mIes for writing mesh equations 34

2.5 Solution of three simultaneous equations 34
2.6 Worked examples using mesh analysis 37
2.7 Nodal analysis 41
2.8 General mIes for writing nodal equations 44
2.9 Worked examples using nodal analysis 44
2.10 Network topology 49
2.11 Loop analysis 52
2.12 Duality 54
3 Circuit Theorems 61
3.1 Introduction 61
3.2 Linearity 61
3.3 Principle of superposition 61
3.4 Thevenin's theorem 63
3.5 Norton's theorem 65
3.6 Relationship between Thevenin's and Norton's
circuits 67
3.7 Reciprocity theorem 67
3.8 The maximum power transfer theorem 69
3.9 The parallel-generator (Millman's) theorem 70
3.10 Rosen's theorem or the general star-mesh
transformation 72
3.11 The star-delta, tee-wye or tee-pi transformation 73
3.12 The delta-star, wye-tee or pi-tee transformation 74
3.13 Summary of star-delta and delta-star transformations 74
4 Energy Storage Elements 79

4.1 Introduction 79
4.2 Capacitors 79
4.3 Energy stored in a capacitor 81
4.4 Capacitors in parallel 82
4.5 Capacitors in series 82
4.6 Potential division in series-connected capacitors 83
4.7 Inductance 84
4.8 Energy stored in an inductor 87
4.9 Inductors in se ries 87
4.10 Inductors in parallel 88
4.11 Duality between inductors and capacitors 90
4.12 Relationship between inductance, reluctance and the
number of turns on a coil 90
Contents vii
4.13 Mutual inductance 91

4.14 Direction of the mutually induced e.m.f. - the dot
notation 92
4.15 Coefficient of coupling 93
4.16 Mutually coupled coils in se ries and in parallel 94
5 The Sinewave, Phasors and Power 100

5.1 Introduction 100
5.2 Mean or average value of an alternating quantity 100
5.3 The effective value or r.m.s. value of a periodic wave 102
5.4 Phase angle 103
5.5 Phasors and phasor diagrams 105
5.6 Representation of a sinusoidal quantity as a complex
value 106
5.7 Impedance of elements 107
5.7.1 Resistance 108
5.7.2 Pure inductance 108
5.7.3 Pure capacitance 109
5.7.4 CIVIL - an a.c. mnemonic 110
5.8 The susceptance of elements 111
5.9 The admittance of elements 111
5.10 The impedance of elements in se ries 112
5.11 The admittance and impedance of elements in
parallel 113
5.12 Impedance and admittance of series-parallel circuits 115
5.13 Power and power factor 115
5.14 Power, VA and VAr absorbed by ideal elements 117
5.15 v-i waveforms 118
5.16 Power consumed in an a.c. circuit 118
5.17 Complex power 121
6 Sinusoidal Steady-state Analysis 126

6.1Introduction 126
6.2Nodal, mesh and loop analysis 130
6.3Principle of superposition 132
6.4Thevenin's theorem and Norton's theorem 133
6.5Millman's theorem 135
6.6Rosen's, star-delta and delta-star theorems 136
6.7Maximum power transfer theorem 137
6.8a.c. circuits with dependent sources 138
viii Contents
7 Polyphase Circuits 144

7.2 Three-phase generation 144
7.3 Star-connection or Y -connection 145
7.4 Phase sequence 146
7.5 Balanced and unbalanced systems 147
7.6 Phase and line voltages in a star-connected system 147
7.7 Delta-connected or mesh-connected three-phase
source 150
7.8 Three-phase, four~wire, star-star system 150
7.8.1 Balanced supply, balanced load 151
7.8.2 Three-phase, four-wire, star-connected
system with an unbalanced load 152
7.9 Three-phase, three-wire, star-star system 153
7.10 Delta-connected systems 155
7.10.1 Balanced delta-connected load with a
balanced supply 156
7.10.2 Unbalanced delta-connected load with a
balanced supply 158
7.10.3 Unbalanced delta-connected load with an
unbalanced supply 159
7.11 Delta-connected supply and a star-connected load 159
7.12 Star-connected supply and delta-connected load 159
7.13 Summary of balanced star- and delta-connected loads 160
7.14 Power consumed in a three-phase system 160
7.15 Power measurement in a three-phase system 162
7.16 The two-wattmeter method of power measurement 163
7.17 Introduction to symmetrical components 164
7.18 Analysis of unbalanced conditions 165
7.18.1 Power consumed by symmetrical
components 166
8 Two-port Networks 171

8.2 Input impedance, output impedance, voltage gain,
current gain and power gain 172
8.3 Admittance parameters or y-parameters 176
8.4 Impedance parameters or z-parameters 180
8.5 Hybrid parameters or h-parameters 183
8.6 Transmission parameters 186
8.7 Relationship between the y-, Z-, h- and t-parameters 189
Contents ix
8.8 Interconnection between two-port networks 191

9 The Transformer 198

9.2 The ideal transformer 198
9.3 Phasor diagram for an ideal transformer on no-load 201
9.4 Leakage flux in a transformer 203
9.5 Impedance matching with an ideal transformer 204
9.6 The ideal transformer as a two-port network 207
9.7 Thevenin's equivalent circuit of an ideal transformer
and voltage regulation 208
9.8 The linear transformer 211
9.9 T- and n-equivalent circuit of a linear transformer 214
10 Transient Solution of Electrical Circuits 218

10.2 Classical solution of first-order systems 219
10.2.1 The time constant i of a circuit 223
10.2.2 Circuits reducible to first-order format 223
10.3 Classical solution of second-order systems 224
10.4 The Laplace transform 226
10.5 Step, impulse and ramp functions 227
10.6 Laplace transforms of some useful functions 233
10.7 Properties of the Laplace transform 237
10.8 Representation of circuit elements in the s domain 242
10.9 Introduction to analysis of first-order systems using
the Laplace transform 244
10.10 Sinusoidal excitation of first-order systems 249
10.11 Solution of second-order circuits using Laplace
transforms 251
10.12 s-Domain transfer functions 258
10.13 Transients in magnetically coupled circuits 259
11 Complex Frequency, the s-Plane and Bode Diagrams 266

11.2 The exponential form of a complex number 267
11.3 Complex frequency 267
11.4 The s-plane 268
11.5 Transformation of impedance into the s-domain 269
x Contents
11.6 Frequency response as a function of w 271

11.7 Transfer functions 276
11.8 Bode diagrams 280
11.9 Bode diagram of H(jw) = K 282
11.10 Bode diagram of H(jw) = (jwrt 282
11.11 Bode diagram of H(jw) = (1 + jwr) 283
11.12 Bode diagram of H(jw) = (1 + jwr)n
10.13 Plot of H(jw) = 1 + 2' [ jm 1+ [jm
Wo Wo r 287
290
12 Resonance 299
12.2 The resonant condition 299
12.3 Series resonance 300
12.3.1 Resonant frequency and frequency response 300
12.3.2 The quality factor, Q s, of the series circuit 302
12.3.3 The peak voltage across R, Land C 305
12.3.4 Bandwidth and selectivity of aseries RLC
circuit 305
12.4 Parallel resonance 308
12.4.1 Resonant frequency and frequency response
of an ideal parallel circuit 308
12.4.2 The quality factor, Q p' of an ideal parallel
circuit 310
12.4.3 Bandwidth and selectivity of a parallel
circuit 311
12.4.4 Resonance in a practical parallel circuit 311
12.5 Scaling electrical circuits 315
12.5.1 Impedance or magnitude scaling 315
12.5.2 Frequency scaling 316
12.5.3 Combined scaling 317
12.5.4 Universal resonant circuits 317
12.6 Passive and active filters 318
12.6.1 Passive band-pass filter 318
12.6.2 Passive band-stop filter 319
12.6.3 Active RC filters 320
12.7 Selective resonance 320
12.8 Tuned coupled circuits 321
Contents xi
13 Harmonics and Fourier Analysis 325

13.2 Harmonics 325
13.3 Trigonometric Fourier series 326
13.4 Waveform symmetry 332
13.5 Line spectra 335
13.6 Circuit response to a non-sinusoidal forcing function 338
13.7 r.m.s. value of a complex wave and the power
supplied 340
13.8 Effect of harmonics in a.c. systems 343
13.9 Harmonic analysis 343
14 Computer Solution of Electric Circuits 349
14.2 Circuit description 351
14.3 Element description 351
14.4 Programming examples 357
15 Complex Numbers, Matrices, Determinants and
Partial Fractions 386
15.1 Imaginary numbers 386
15.2 Complex numbers 387
15.3 Representation of complex numbers 387
15.4 Conjugate of a complex number 388
15.5 Sum and difference of complex numbers 388
15.6 Multiplication of complex numbers 389
15.7 Division of complex numbers 389
15.8 Powers and roots of complex numbers 390
15.9 Matrix representation 391
15.10 Matrix addition and subtraction 392
15.11 Matrix multiplication 392
15.12 The determinant of a square matrix 394
15.13 Minors and cofactors 395
15.14 Evaluating a determinant 395
15.15 The rule of Sarrus 396
15.16 Cramer's rule 396
15.17 Matrices and determinants containing complex
numbers 397
15.18 Partial fractions 397
Bibliography 401
Index 402
Preface
Electrical Circuit Analysis and Design is intended for use with the early
years of a first degree course in Electrical, Electronic and Control En-
gineering, and for Higher National Diploma and Certificate courses in
Electrical and Electronic Engineering.
The main prerequisite to its use is a knowledge of the basic concepts of
electricity, magnetism and mathematics; an introduction to calculus is
more in the nature of a corequisite than aprerequisite.
The book has primarily been written for the student, and it is intended
that readers should be able to teach themselves the analytical techniques
involved. To this end, many fully worked examples are included in the
body of the text, and a large number of unworked problems (with solu-
tions) are included at the end of chapters. Throughout the book, both
'power' and 'electronic' circuit examples and problems have been included.
A 'plus' feature of the book is a chapter on the use of SPICE software
(Simulated Program with Integrated Circuit Emphasis) for circuit analysis.
Examples in this chapter range from resistive d.c. networks to a.c. solu-
tions and transient analysis, and illustrate the practical advantages of this
software, which is pre-eminent in the field of circuit analysis.
When writing the book, I decided that it should be written from a logical
teaching viewpoint. That is, as with a conventional course, the more
understandable parts of circuit theory are treated first, after which the less
easy but, technically, more interesting topics are covered.
Chapter 1 covers d.c. circuits and intro duces the concept of basic
elements and laws, including Kirchhoff's laws together with simple circuit
analysis, and described dependent and independent sources.
In chapter 2, we take a first look at network analysis using mesh, nodal
and loop analysis. In undergraduate and so me HND courses, the latter
usually involves a knowledge of network topology, which is also in-
troduced. Finally, an introduction to the duality between circuits having
similar mesh and nodal equations is given.
xii
Preface xiii
In order to understand circuit analysis fully, the reader should have a

grasp of a number of circuit theorems and this, for d.c. circuits, is provided
in chapter 3.
To move on to alternating current theory, the reader needs to under-
stand the basis of circuits containing energy storage elements, this informa-
tion being provided in chapter 4. Here we deal with capacitors, inductors
and mutual inductance. Engineers have devised the 'dot' notation to deal
with the latter, and this is fully explained in this chapter.
In chapter 5, we look at some of the many interesting aspects of
alternating current theory, including phasors and phasor diagrams, com-
plex impedance and admittance, together with series and parallel combina-
tions of elements and circuits. Also covered are power and power factor,
together with complex power. Next, in chapter 6, we apply a range of
circuit theorems to a.c. networks.
Power-based electrical engineers have a particular interest in polyphase
circuits, and this topic is comprehensively covered in chapter 7. This
chapter describes and analyses many aspects of three-phase systems,
including power measurement and symmetrical components.
Two-port networks are of great significance to electronics and tele-
communications engineers and, in chapter 8, the reader is introduced to y,
Z, hand transmission parameters, together with the relationship between
them.
In chapter 9 we meet the transformer , both 'ideal' and 'linear'. A
knowledge of these is vital to both electrical and electronic engineers
alike.
In chapter 10, we deal comprehensively with the transient analysis of
circuits. A practice in many courses is to deal with this topic using two or
sometimes three different techniques, each time covering very similar
ground! In this chapter we look, initially, at the process of solving first- and
second-order circuits by classical methods. These methods gene rally have a
number of disadvantages, which are overcome by the use of the Laplace
transform method; the latter is used throughout the remainder of the
chapter.
While the Laplace transform method has the minor drawback that we
need to spend a little time looking at the development of Laplace trans-
forms before moving on to circuit analysis, it has the great advantage that
the solution of circuits (both without and with initial conditions) becomes
relatively straightforward. This chapter covers step function (d.c.) and a.c.
analysis of first- and se co nd-order circuits, together with transients in
magnetically coupled circuits.
A feature of many electrical and electronic courses is the treatment of
the frequency response of circuits, and this is described in chapter. 11.
Additionally, an introduction to complex frequency and the s-plane is
provided and, equally importantly, the transformation of time-domain
xiv Preface
impedance into its equivalent s-domain impedance is covered. Frequency

response is described in terms of Bode diagrams, and the method of
drawing the diagrams is outlined in a straigthforward manner for both first-
and second-order circuits.
Resonance occurs both in electronic and power circuits, and compre-
hensive coverage of series and parallel resonance is provided in chapter 12.
Additional features in this chapter include frequency scaling, selective
resonance and tuned coupled circuits.
In chapter 13 the attention of the reader is directed to harmonics and
Fourier analysis. A knowledge of Fourier analysis is vital for all engineers,
and the chapter includes such topics as waveform symmetry, line spectra,
circuit response and the effect of harmonics in a.c. systems. Also included
is a section on harmonic analysis.
In chapter 14 we meet one of the most powerful software packages
available for analysis of electrical and electronic circuits, namely SPICE
(Simulated Program with Integrated Circuit Emphasis). This software,
which is both fast and versatile, is widely available both in full and in
educational versions, and can be used to solve almost any electrical
problem. The solution of a wide range of problems is included in this
chapter.
Chapter 15 is devoted to a number of mathematical 'tools' needed by
engineers and technicians, namely complex numbers, matrices, determi-
nants and partial fractions.
Noel M. Morris
Acknowledgements
I acknowledge with thanks the assistance I have received from Derek
Hopewell of Nottingham Polytechnic, Mr F. W. Senior, M.Sc., former
Senior Lecturer at the Staffordshire University and, in particular for his not
inconsiderable assistance, Lionel Warnes of Loughborough University. I
would also like to thank Mr A. Lewis, Director of the Open Terminal
Computer Centre, Stevenage.
Finally, grateful thanks are due to my wife, without whose assistance
and support the book would not have been possible.
xv
1
Elements and Laws
1.1 Introduction
In this chapter the basic relations in electric circuits are reviewed, including
current, voltage, resistance, Ohm's law, electric power, etc. Additionally,
other topics including the application of Kirchhoff's laws to circuits are
described.
The concepts of independent and dependent voltage and current
sources, are introduced, some of which may be new to some readers. One
application of dependent sources is in the operation al amplifier, which is
used in every sphere of electrical and electronic engineering. So vital is this
to all engineers it is introduced in this, the first chapter of the book.
1.2 Electric current
We are all familiar with electrostatic charge - and its most dramatic effect,
namely lightning discharge. It is the latter effect, namely the electrical
charge in motion or electric current which attracts our attention here.
The current in a circuit is a measure of the rate at which electric charge
passes through the circuit, and we define the instantaneous value of the
current, i, as
. dq
z=-
dt
The charge in motion is, gene rally , carried by electrons which move from a
low (that is, negative) potential to a higher potential (that is, a positive
potential). In this book, however, we adopt the more usual convention that
current flows trom a point ot positive potential to a point ot negative
potential.
2 Electrical Circuit Analysis and Design
4 2
'"
.0
E '"Q;
Q)
0
;; c.
"-
0 E t (s)
t (s) 2 6
~
<.>
'"
% "
-4 -2-
(a) (b)
Figure 1.1 (a) The charge passing a point in a circuit, and (b) the corresponding current in
the circuit.
The diagrams in figure 1.1 illustrate the concept of movement of charge

and current. The waveform of movement of charge is shown in figure
1.1(a), and the corresponding current waveform is in diagram (b).
Lower-case symbols (such as i) are used in electrical engineering to
represent an instantaneous value which varies with time. When steady-state
operating conditions are reached, we use a capital symbol (such as J) to
represent the quantity. We also use capital symbols to deal with the
steady-state (root mean square) value of alternating quantities (see also
chapter 5).
In the UK, current is represented by an arrow drawn on the wire in
which the current flows, as shown in figure 1.2. If it is an instantaneous
value of current, it is shown either as i(t) or simply as i. Depending on the
direction in which it is flowing, the current may be assigned either a
positive or a negative value, as indicated by the 5 A current in figure 1.2. In
the figure, a current of 5 A flows from left to right along the top wire;
alternatively, -5 A flows from right to left. Engineers refer to a current as
flowing either 'in' or 'through' a circuit.
In many US texts, the current is often shown as an arrow by the side of
the wire or element in which it flows.
5A -5A
Supply Circuit
source
i(t)
Figure 1.2 Representation 01 current on a circuit diagram.
Elements and Laws 3
1.3 Voltage, e.m.f. and p.d.
The voltage 'between' or 'across' a pair of terminals is a measure of the

work required to move charge through the element or circuit connected
between the terminals. The voltage across an element is, in fact, the work
required to move acharge of 1 C from one terminal to the other. The SI
unit is the volt (V) - named after the Italian physicist Giuseppe Antonio
Anastasio Volta.
The instantaneous voltage is represented by the symbol v, and steady-
state voltage (or r.m.s. voltage if alternating) by V.
Since voltage represents the potential needed to move acharge between
a pair of terminals, a voltage can exist even if no current ftows.
The energy converted per unit charge in an electrical source is known as
the electromotive force (e.m.f.) of the source, and the electrical potential
difference (p.d.) between two points in a circuit is a measure of the work
required to move charge through the element. The general name given to
both e.m.f. and p.d. is voltage.
Voltage may be represented on a circuit diagram either by a '+' and '-'
pair of symbols (see figure 1.3(a» , or by an arrow pointing from one
terminal to another (see diagram (b». Where + and - symbols are used,
the + terminal assumes the voltage wntten between the terminals with
respect to the - terminal. For example, in figure 1.3(a), terminal A is
+8 V with respect to terminal B, and in diagram (c) terminal B is -8 V
with respect to terminal A.
(a) (b) (c) (d)
Figure 1.3 Representation 0/ voltage. In each case, terminal A is 8 V positive with respect
to terminal B.
When an arrow is used, the terminal which the arrowhead points to

assumes the voltage written by the arrow with respect to the tail of the
arrow. For example, in figure 1.3(b), terminal A is +8 V with respect to
terminal B; in diagram (d) terminal B is -8 V with respect to terminal A.
In some cases both sets of symbols are used.
1.4 Power in d.c. circuits
Power is the rate of transfer of energy; the unit of power is the watt (W) or
joule per second (J s -1). Since voltage has the dimensions of joules per
coulomb, and current has the dimensions of coulombs per second, then
power, p = vi
Once again, Iower-case pis reserved for the instantaneous value of power
in the circuit, and upper-case Pis reserved for average power. That is
P = VI
Figure 1.4 illustrates various combinations of electrical source and power
absorbing element (or load). In general, current ftows out 01 the positive
terminal 01 apower source, and into the positive terminal 01 a load (see
figure 1.4(a)).
In diagram (b), current flows out of the positive terminal of the left-
hand block and into the positive terminal of the right-hand block. Conse-
quently, the left-hand block is the power source and the right-hand block is
the load. The power transferred from the source to the load is 10 x 5 =
50W.
In figure 1. 4( c) the lower line is - 7 V with respect to the upper line,
and the power transferred from the source (the right-hand block) to
the load is 7 x 3 = 21 W. In diagram (d) the power absorbed by the load is
8 x 4 = 32 W.
+A A
.,0
Vst
'"-Gl
'tJ
tvs
GlU
(a) :i:'"-
0'"
a..5l ....J
- B B
+5A - 3A -4A
Gl
.,0 .,0 Gl
.,0
ts v
Gl
~- 7 V
~ 'tJ 'tJ ~ 'tJ
~
0'"
(f)
t10 V ....J ....J '0"
(f)
0'"
(f)
....J
+
(b) (c) (d)
Figure 1.4 Block diagrams showing apower source and its load.
1.5 Linear passive circuit elements
The major part of electric circuits comprises passive elements, which can
only dissipate or store electrical energy. A linear circuit element is one in
Elements and Laws 5
which the voltage across the element varies linearly with the current
through it. A number of elements are non-linear, in which case the voltage
across the element does not vary linearly with the current through it.
There are two broad categories of passive element, namely those which
dissipate energy and those which store energy. The former include resist-
ance, which is the subject matter of this chapter, and the latter include
inductance and capacitance which are described in chapter 4.
An ideal circuit element is one which is both linear and either dissipates
energy or stores it. U nfortunately, most practical circuit elements have
some degree non-linearity.
1.6 Resistance, conductance and Ohm's law
The outer electrons in the atoms of metallic conductors are loosely bound
to the parent atom, and are relatively 'free' to move around the lattice
structure at normal room temperature. When a p.d. is applied to the
conductor, free electrons drift towards the positive pole of the supply,
resulting in current flow in the conductor.
As the electrons move through the crystallattice, they collide with other
electrons and lose so me of their energy. That is, there is some 'resistance'
to current flow, and the resulting loss of energy is usually converted to
heat. The resistance of the element enables us to broadly classify materials
as folIows:
Conductors: these have a low resistance, and include met als such as
copper , brass, manganin, etc.
Insulators: these have a high resistance to current flow, and include
wood, plastic and glass.
Semiconductors: these have resistance between that of conductors and
insulators but, as students of electronics will know, it is not quite that
simple! These include germanium, silicon, cadmium sulphide, etc.
Superconductors: when the temperature is within a few degrees of
absolute zero, the resistance of these materials falls to zero (or nearly
so). These m&terials include tin, lead, thallium, etc.
In 1827 George Simon Ohm published a pamphlet describing the result of
experiments in electrical circuits, one relationship in the pamphlet being
what we know as Ohm's law, and is
v = iR or V = IR
where v is the p.d. in volts across the linear element, i is the current in
amperes through the element and R is a constant of proportionality called
the resistance. The unit of resistance is the ohm (symbol Q). An alternative
form of Ohm's law is
Figure 1.5 Ohm's law.
i = vG or 1= VG
where G is the conductance of the element or circuit, and has the unit of
the siemens (symbol S).
A resistance (or a conductance) is represented in a circuit by means of a
rectangular box as shown in figure 1.5.
1. 7 Power in a resistive circuit
In a linear resistive circuit the voltage is given by v = iR, and the power
consumed is
or, alternatively
p = vi = vZG = iZ/G W
In many installations the power is measured in kilowatts (1 kW = 1000 W),
and in large systems it is measured in megawatts (1 MW = 10 6 W = 1000
kW). In electronic circuits the power may be measured in milliwatts (1 mW
= 10- 3 W) or in microwatts (1 I-lW = 10- 6 W).
Worked example 1.7.1
If v = 10 V and i = -15 mA, calculate the power consumed.
Solution
p = 10 X (-15 X 10- 3 ) = -0.15 W

that is, the element can either be regarded as consuming -0.15 W, or as
supplying or generating 0.15 W.
If P = 200 mWand I = 10 mA, calculate the resistance of the circuit.
Elements and Laws 7
Solution
R = P/J2 = 200 x 10- 3/(10 X 10-3 )2 = 2000 Q

A current of 0.1 A produces 5 W of power in a circuit. Calculate the

conductance of the circuit.
Solution
G = P/P = 0.1 2/5 = 0.002 S
1.8 Energy consumed in a resistive circuit
Power is the rate of energy transfer, that is

dw
p=-
dt
f' pdt
and energy is given by
2
W=
'1
where W is the energy transferred in the interval t 1 to t 2 • If the power

supply is constant for time t, then
W= Pt
In a d.c. circuit where steady-state conditions exist
P = VI
hence
W= VIt
and since V = IR in a d.c. circuit, then
W = J2Rt = V 2t/R
In most systems, energy is measured in kilowatt-hours (1 kWh =
10 3 X 60 x 60 = 3.6 X 10 6 J), and in large systems it is measured in
megawatt-hours (1 MWh = 10 3 kWh).
1.9 Independent and dependent supply sources
A source is an active element, that is, it can deliver energy to an external

device; sources include batteries, alternators, oscillators, etc. Sources can
be classified into 'ideal' or 'practical' as folIows.
An ideal source is one wh ich does not represent a 'real' device (other
than special electronic circuits which have been engineered to provide an
'ideal' characteristic over a limited operating range); such sources can,
theoretically, deliver an infinite amount of power. A practical source is one
having limitations on its output voltage, current and power.
There are two types of source (ideal or otherwise), namely voltage
sources and current sources . An ideal voltage source could, theoretically,
maintain a constant voltage across a load of any resistance. An ideal
current source could, also theoretically, maintain a constant current in a
load of any resistance. An independent source is one whose source quantity
(voltage or current) is independent of the remainder of the circuit.
The symbols in figure 1.6 depict typical independent sources. In diagram
(a), the 6 V source supplies a current of 1.5 A to its load while, in diagram
(b), the independent source (a battery) receives a current of 3 A, that is, it
is being charged. Diagram (c) shows an independent current source, the
voltage across its terminals depending on the resistance of the connected
load.
3A
12vtC
-L
(a) (b) (c)
Figure 1.6 An ideal or independent voltage souree (a) delivering apower of 9 W,

(b) absorbing 36 W. (e) An ideal or independent 4 A eurrent souree.
There is another group of voltage and current sources, namely depen-

dent sources or controlled sources , which are characterised by the
diamond-shaped symbols in figure 1.7. The output voltage or current
produced by the controlled source is dependent on a voltage or current at
some other point in the circuit.
The dependent source in figure 1.7(a) produces a voltage which is
dependent on the voltage VI existing at some other point in the circuit. We
would describe this source as a voltage-controlled voltage source. The
dependent voltage source in diagram (b) is controlled by a current 12 which
Elements and Laws 9
(a) (b) (c) (d)
Figure 1.7 Diagrams (a) and (b) represent dependent or eontrolled voltage sourees, (e)
and (d) represent dependent eurrent sourees.
exists at another point in the eireuit. Diagrams (e) and (d) respeetively
show a voltage-controlled current source and a current-controlled current
source.
The reader should note, onee again, that dependent sourees are ideal
sourees.
1.9.1 The operational amplijier
The operational amplifier (often abbreviated to op-amp) is of fundamental

importanee in praetieally all forms of eleetrieal eircüit. The elements whieh
eomprise an ideal operational amplifier are shown in figure 1.8(a).
The amplifier has two input terminals marked VI and v 2' and one output
terminal, Vo' We refer to the terminal marked '+' (to whieh V 2 is eon-
neeted) as the non-inverting input, and to the terminal marked '-' as the
inverting input; this is beeause the output voltage, vo, is in phase wirh V 2
I'
I "
v, I "
I ", v,
v, IR , Vo
Vo = Kv"
v,
- v""""
(a) (b)
Figure 1.8 (a) Idealised operational amplijier, (b) cireuit representation.
and antiphase with VI' We can, in fact, connect a signal of any potential
(either positive or negative) to either input terminal.
The voltage gain of the amplifier is K, and the output voltage is V o =
K(v 2 - VI) = Kv x . The value of K approaches infinity in an ideal op-amp
and, if V 0 is to be finite then the value of V x = V 2 - V I must approach zero.
That is, in an ideal op-amp, v 2 = V I which implies that the input current
must be practically zero! Also, in an ideal op-amp, the input resistance R is
infinite, so that the input current can be regarded as being zero (this also
confirms that v x must be zero!). The reader will also observe that the
output from the op-amp is provided by a voltage-controlled voltage source,
whose output resistance is zero (or nearly so!).
Deduce an expression for the voltage gain of the operational amplifier

circuit in figure 1.9. The op-amp is ideal.
v,
v,
i,
Figure 1.9 Circuit Jor worked example 1.9.1.
Solution
Since the op-amp is ideal, we can make the simplifying assumptions that
v x = 0 and i = 0, hence
and
that is
Elements and Laws 11
that is
Vo =
R R+ R2v
1
1
2 =
[1 +/i;
R2] V2
and the overall voltage gain is
~=1+~
V2 R1
1.10 Kirchhotf's laws
At about the same time as Ohm was carrying out his experimental work,
Gustav Robert Kirchhoffwas born in Germany; he was to revolutionise the
work on electric circuit theory.
It is a simple fact that the rate of ftow of charge into any point or node
(a formal definition of anode is given in section 2.2) in a circuit is equal to
the rate of ftow of charge out of it. That is, charge cannot accumulate at a
given point in the circuit. This is not a mathematical proof, but is the basis
of Kirchhoff's current law, or KCL. That is
the algebraic sum of currents entering any node is zero
At the node in figure 1.10, KCL states that
A simple way of expressing this is ~i = 0 or, more precisely

N
n
L = 1
in = 0
where N is the number of wires meeting at the node. Alternatively, we can

re-write KCL as follows
i2
i3
Figure 1.10 Kirchhoff's current law (KeL).
the algebraic sum 01 the current entering any node is equal to the
algebraic sum 01 the current leaving the node
which, in figure 1.10 gives
Also, if we proceed around the complete circuit, and return to the starting
point, there is no change in electrical potential. This is the basis of Kirch-
hoff's voltage law or KVL, and can be stated in the form
the algebraic sum 01 the e.m.fs. and p.d.s around any closed
circuit is zero
If, for example, we proceed around the closed path ABCA in figure 1.11
we get
or, if we take the path ACBA we get
which, effectively, gives the same equation. A simple way of writing this is
~v = 0 or, expressed in mathematical form, KVL says
N
~ Vn = 0
n = 1
r---.C
Figure 1.11 Kirchhaff's valtage law (KVL).
where N is the number of elements in the closed loop. Alternatively, KVL

may be written in the following form
in any closed path the algebraic sum 01 the e.m.f.s is equal to the
algebraic sum 01 the p.d.s
--
12 A
--
12 A 12 A
R, @ 10 Q I. R, @ 100
@ G) @l
I.
120 V t VA BA
R,
BA
R,
0.5A
(a) (b)
10 Q @ le --®
1Oie
BA
Figure 1.12 Figure for worked example 1.10.1.
Calculate VA' 1B' R 1 and R 2 in the circuit in figure 1.12(a).
Solution
The relevant section of the drawing is shown in diagram (b). Applying

KCL to node 3 gives
12 = JA + 8
or lA = 4 A
hence, from Ohm's law
VA = 10J A = 40 V
Applying KCL to node 2 yields
JA + 0.5 = JB
or I B = 4.5 A
At node 0, KCL shows that
8 = 0.5 + Je
or Je = 7.5 A
Applying KVL to the closed-Ioop on the left of figure 1. 12(b) shows that
the voltage at node 2 with respect to node 1 is 10J e + 120, which is equal to
VB' That is
VB = (10 x 7.5) + 120 = 195 V
Hence, from Ohm's law
R1 = -VB = -195 = 43.33 Q

JB 4.5
Also, the voltage at node 3 with respect to node 0 is
120 + VA = 160 V
Hence R 2 = 160/8 = 20 Q
1.11 The double-suffix voltage notation
Either when specifying or calculating the voltage of one node in a circuit

with respect to another node, a double-suffix voltage notation is very
useful.
Referring to the circuit diagram in figure 1.13, the voltage V AN is read as
'the voltage of node A with respect to node N', and may be written in the
form
c
- - - V NC
Figure 1.13 The double-suffix voltage notation.
VAN=VA-V N
If node N is the reference node (whose potential may be regarded as zero),
we can define V AN simply as VA. Thus
V AS = V AN - V SN = (VA - V N) - (V S - V N) = VA - VS
also
V SA = V SN - V AN = (V S - V N) - (VA - V N) = - V AS
If in figure 1.13 V AN = 10 V, V BN = 20 V, V CN = 30 V and V AS = -10 V,
then
V SN = V AN - V AS = 10 - (-10) = 20 V
Alternatively, we may say V SA = - V AS = 10 V, and
V SN = V AN + V SA = 10 + 10 = 20V
If we need to calculate the voltage between two nodes, say node A and
node B, then
V AB = voltage of node A with respect to node B
= V AN - VBN = 10-20= -lOV
or
V BA = V BN - V AN = 20 - 10 = 10 V
that is
V AB = -V BA
Also V AB =V AS - V BS =V AS - (V BN - V SN)
=-10 - (20 - 20) = -10 V
Calculate the volta ge V AB in figure 1.14.
Solution
In this case there are two closed meshes (a formal definition of a mesh
is given in section 2.2) linked by a 4 Q resistor. Since there is no return
path for the current through the 4 Q resistor, no current can ftow in it, and
V 2 = O. The current 1 1 circulating around the left-hand mesh is
A -- v, --v,
5Q
_ _10_ _ = 1 A
2+3+5
hence VI = 2/ = 2 V
1
The current /2 in the right-hand mesh is

_ _40_ _ =2A
5+ 6+ 9
and V 3 = 6/ 2 = 12 V
To evaluate V AB, we start at node Band proceed to node A (using any
path), and we simply add (or subtract) voltages as we meet them. Selecting
one of the available paths we get
VAB =-V 3 + 40 - V2 - VI = -12 + 40 - 0 - 2
=26 V
1.12 Practical (non-ideal) sources
An ideal voltage source can, theoretically, supply any current to a load,

giving the ftat v-i characteristic (shown dotted) in figure 1.15(a). However,
a practical voltage source cannot do this, and the current it can deliver is
limited. We account for this fact by introducing an internal resistance, or
source resistance, or output resistance, R, in series with the ideal voltage
source, es, as shown in figure 1.15(b).
The terminal voltage of the source is equal to es only when i = 0, that is,
when the terminals of the practical source are open-circuited.
An ideal current source can, theoretically, supply any load with a
constant current; the corresponding v-i characteristic is shown by the
dotted line in figure 1.16(a). A practical current source cannot do this; the
u r-------.
Idea I sou rce
~r ____
-=-:- - - - -- I
I
I
I
I
I
o I _____ ...l
L
(al (bl
Figure 1.15 (a) v-i characteristic of ideal and practical voltage sources, (b) an equivalent
circuit of a practical voltage source.
u r-------..,
I
I;'
I
G, r
I _ _ _ _ _ _ -l
L
(a) (b)
Figure 1.16 (a) v-i characteristic of ideal and practical current sourees, (b) an equivalent
circuit of a practical current source.
v-i characteristic of a practical current source is shown in full line in

diagram (a), and an equivalent circuit for this type of source is shown in
diagram (b). The change in characteristic is accounted for by the internal
conductance, G s , which is also known as the source conductance, or the
output conductance.
1.13 Transformation of practical sources
When analysing circuits, it is often convenient to convert all the sources in

the network either to voltage sources or to current sources. In the follow-
ing we will consider the process of source trans!fJrmation.
To simplify the process of analysis, let the internal conductance G s of
the practical current source be replaced by its equivalent resistance r s,
where r s = 1/G s. Consider the practical voltage and current sources in
figure 1.17(a) and (b), respectively, which are connected to an identical
load resistor.
The current i u in the load connected to the voltage source is (figure
1. 17(a))
(a) (b)
Figure 1.17 Source transformation.
Rs = l/G s
es = islGst
(a)
Figure 1.18 Equivalent sourees.
or
The current flowing in the load connected to the current source is (figure
1. 17(b))
. . . . VL isr s - V L
lL = ls - l = ls - - =
rs rs
or
Since the two sources are identical, we can equate the voltage across the
load in the two cases as follows
Equating terms in i L shows that

R s = r s = lIG s
Also, for equality, it follows that
That is, the voltage source and the current source in diagrams (a) and (b)
are equivalentas far as measurements of voltage and current at the
terminals are concerned.
When converting sources, the reader should be careful to ensure that
the polarity of the voltage source acts in the correct direction to produce
the correct current in the current source, and vice versa.
Calculate the value of 1 in figure 1.19(a).
Solution
Initially, we will convert the 3 A practical current source into its equivalent
voltage source. The result is shown in diagram (b). The equivalent voltage
source has an intern al resistance of 1/0.2 = 5 Q, and a source voltage of
[slG s = 310.2 = 15 V.
The net voltage in figure 1.19(b) acting in the direction of [is 10 - 15 =
-5 V, so that
[ = -5/(10 + 5) = -0.333 A
that is, its value is 0.333 A flowing in the opposite direction to that shown in
figure 1. 19(a).
f ·3
10 Q 10 Q , 5 Q
3A1ov t
(a) (b)
1.14 Resistance of aseries circuit
Elements which carry the same current are connected in series; the three
resistors in figure 1.20 are connected in series.
Since the current, [, is common to each resistor, then
Applying KVL to the circuit gives
,----
v, V, V3
R, R, R3
Figure 1.20 Series-connected resistors.
Es = V l + V2 + V3 = I(R l + R2 + R 3 ) = IRE
where RE = R l + R2 + R3
and is the equivalent resistance of the series circuit. If there are n resistors
in series, the equivalent resistance of the circuit is
RE = R l + R 2 + ... + Rn
To summarise, the equivalent resistance 01 a series-connected circuit is
always greater than the largest value 01 individual resistance in the circuit.
Resistors in series are often described as a string of resistors.
1.15 Voltage division in series-connected resistors
If there are n resistors in series, the voltage across the nth resistor is
V n = IRn> where I is the current flowing through the resistors. If the voltage
across the series circuit is V s, then V s = IRE, where RE is the equivalent
resistance of the circuit. That is
Vn IR n Rn
-=--=-
Vs IRE RE
or
Calculate the equivalent resistance of the series circuit in figure 1.21, and
determine I, V l1 V 2 and V 3 •
Solution
The equivalent resistance of the circuit is
100 200 700
2OV' T'--_ _-----IT

Figure 1.21 Figure Jor worked example 1.15.1.
RE = R I + R 2 + R 3 = 10 + 20 + 70 = 100 Q
Applying KVL to the circuit shows that
V s = 20 - 10 = 10 V
where V s produces I in the direction shown in the figure, hence
1= VS/R E = 10/100 = 0.1 A
hence
= IR I = 0.1 X 10 = 1 V
VI
V 2 = IR 2 = 0.1 x 20 = 2 V
V 3 = IR 3 = 0.1 x 70 = 7 V
and from KVL, V s = VI + V 2 + V 3 = 1 + 2 + 7 = 10 Vor, alternatively,
from the theory developed above we can say
VI = VsR/R E = 10 x 10/100 = 1V
V2 = VsRiR E = 10 x 20/100 = 2 V
V3 = VsR/R E = 10 x 70/100 = 7 V
It can be seen from the above results that the value of the voltage across any
resistance in aseries circuit is proportional to the value of the resistance.
1.16 Resistance and conductance of a paraDei circuit
Elements are said to be connected in parallel with one another when they
are connected between the same pair of terminals in a circuit. The resistors
in figure 1.22 are connected in parallel.
The solution of parallel circuits can be approached either by considering
resistances in parallel or conductances in parallel. Since the former is the
most popular approach, we will adopt it here.
By definition, each branch in a parallel circuit supports the same
voltage, and the current in the nth branch is
" "
R, R, R3
Figure 1.22 Parallel-connected resistors.
In = VrJR n
that is 1 1 = V rJR}> V2 = V rJR 2, etc. Applying KCL to the circuit gives
Vp Vp Vp
I = I 1 + I 2 + I 3 =-
R 1 + -R 2 + -R 3
=V
P
[1...
R 1
+ 1...
R
+ 1...]
R 2 3
but I = V rJR E, where RE is the equivalent resistance of the circuit.

That is
1 1 1 1
-=-+-+-
RE R1 R2 R3
If there are n resistors in parallel then
1 1 1 1
- = - + - + ... + -
RE R1 R2 Rn
In the special case of two resistors in parallel
R E_- R 1R 2
R1 + R2
Rad we considered the case of parallel-connected conductances, and
allowed G n = lIR n and GE = liRE, then for the general case of n
parallel-connected conductances
GE = GI + G 2 + ... + G n
1.17 Current division in a parallel circuit
Possibly the simplest method of calculating the way in which current

divides in a parallel circuit is to consider the case of a parallel circuit
containing conductances. The current in branch n is
In = VpG n
where V p is the voltage applied to the parallel circuit. The total current
drawn by the circuit is
1= VpG E
where GE is the equivalent conductance of the parallel circuit, hence
In VpG n Gn
-=---=-
I VpG E GE
or
1 = 1 Gn
n GE
Alternatively, considering parallel-connected resistors, it can be shown

that
where REis the equivalent resistance of the parallel circuit. In the special
case of a two-branch parallel circuit
GI R2
II = 1 = I---=--
GI + G2 RI + R2
Calculate the value of 11> 12 , 13 and V p in figure 1.23.
Solution
The total current entering the top node of the parallel circuit is (14 - 4) =
10 A, and this divides between the three branches. The total conductance
of the parallel circuit is
1 1 1 1 1 1
G=G+G+G=-+-+-=-+-+-
E I 2 3 R R R 2 4 5 I 2 3
= 0.5 + 0.25 + 0.2 = 0.95 S

hence
'3
5Q
/1 = / G/G E = 10 x 0.5/0.95 = 5.263 A

/2 = / GiG E = 10 x 0.25/0.95 = 2.632 A
/3 = / GiG E = 10 x 0.2/0.95 = 2.105 A
It can be seen from the result that the current in each branch is proportional
to the conductance 01 the branch (or is inversely proportional to the
resistance of the branch).
Also
v = (14 -
p 4)/G E = 10/0.95 = 10.562 V
Unworked problems
1.1. The repetitive waveform of current entering a circuit is shown in

figure 1.24. (a) Wh at is the value of the current at t = 0.01 s? (b)
Wh at charge enters the circuit between t = 80 ms and t = 150 ms? (c)
What total charge has entered the circuit at t = 210 ms?
[Ca) 0.05 A; (b) 15.25 mC; (c) 0.05025 C]
05AZVl
t (ms)
o 100 200
Figure 1.24
1.2. The electrical charge ente ring a terminal in a circuit is 10 sin 100:n:t
mC. (a) What charge has ente red between -3 ms and 3 ms? (b)
Calculate the current at t = 2 ms.
[Ca) 16.18 mC; (b) 2.54 A]
1.3. In figure 1.25 calculate V BC ' V CA ' V BD , V EB and V DC •
[30.93 V; -20.93 V; -20 V; -110 V; -9.07 V]
1.4. Calculate the power consumed by each resistor in figure 1.25.
[10 0 resistor 95.7 W; 200 resistor 4.11 W; 25 0 resistor 250 W; 40
o resistor 10.95 W]
1.5. Calculate the power absorbed by each of the circuit elements in
figure 1.26.
[Ca) 2.4 W; (b) -180e- St ; (c) -20 W; (d) 30 W]
---
10 V
B 10 Q
A
---
-30V
D 20 Q
C
---
-100 V
E 25 Q
40 Q
Figure 1.25
-
12 V
o----c:::J-----
0.2 A
--- tQ
20 V
+
-10V
_
A
~
6V
---
5A
Figure 1.26
1.6. The resistanee, R, of a eonduetor in ohms is given by the equation

R =pLlA, where pis the resistivity (or volume resistivity) of the
conduetor in n m, L is the length of the eonduetor in m, and A is the
eross-seetional area of the eonduetor in m2 • If a eonduetor of resistiv-
ity 0.027 fln m, wh ich is 100 km long and of diameter 1 em earries a
eurrent of 20 A, ealculate the p.d. aeross the length of the wire, the
power eonsumed in the wire and the energy eonsumed in 20 mins.
[687.5 V; 13.75 kW; 4.58 kWh]
1.7. For the eireuit in figure 1.27. ealculate (a) I, (b) Vand (e) the power
absorbed by eaeh element. The ealculation should verify the prinei-
pIe of eonservation of energy.
[Ca) 0.75 A; (b) 3.75 V; (e) 5 n resistor 90.31 W, 10 n resistor 5.625
W, resistor R 2.8125 W, 20 V souree 85 W, 30 V souree 22.5 W, 5 A
souree -206.25 W]
1.8. Calculate I in figure 1.28, and determine the power supplied by the
10 V independent souree.
[-0.4 A; -4 W]
Figure 1.27
t 10V
Figure 1.28
- v
2 kQ
Figure 1.29
2Q
Ao-"'-C=:J'--4~C::J-""-QB
7Q
Figure 1.30
1.9. In figure 1.29, calculate V and the power supplied by the 50 mA

source.
[20 V; 1 W]
1.10. Calculate the resistance between A and B in figure 1.30.
[5.95 n]
1.11. IF 10 V is applied between A and B in figure 1.30, calculate the
voltage across and the current in each element.
[2 n resistor 2.617 V, 1.308 A; 3 n resistor 1.121 V, 0.374 A; 4 n
resistor 1.495 V, 0.374 A; 5 n resistor 2.07 V, 0.414 A; 6 n resistor
2.07 V, 0.345 A; 7 n resistor 5.313 V, 0.76 A; 8 n resistor 7.384 V,
0.923 A]
2
Circuit Analysis
2.1 Introduction
By now the reader is familiar with the use of Ohm's and Kirchhoff's laws in
the analysis of simple d.c. circuits. One of the primary goals of circuit
theory is the attainment of knowledge and experience of analysing more
practical systems. Among the methods adopted are mesh analysis, nodal
analysis and loop analysis.
Any one of these methods can be applied to almost any circuit and, in
many cases there is no simple way of saying which is the 'best' method of
solution; we cannot lay down simple roles to determine the 'best'
approach. A knowledge of each method can only by gained by acquiring a
sound understanding of the features of each type of solution.
2.2 Definitions and terminology
An electrical network is a system of interconnected circuit elements contain-

ing, for example, resistors, inductors, capacitors, voltage and current
sources, transformers, amplifiers, etc. If the network contains at least one
closed path or mesh it is an electrical circuit; every circuit is a network, but
not a11 networks are circuits (see figure 2.1). Despite this academic differ-
ence, engineers use the term network and circuit without differentiating
between them.
Terminals A and Bin figure 2.1(a) are not electrica11y connected, and an
open-circuit is said to exist between them. Terminals C and D in figure
2.1(b) are connected (idea11y) by a resistanceless piece of wire, so that
current can pass from C to D without power loss; a short-circuit is said to
exist between C and D. Strictly speaking, figure 2.1(a) is a network, and
figure 2.1(b) is a circuit.
An ideal circuit element is one which does not, strictly speaking, rep-
28
Circuit Analysis 29
B D
(a) (b)
Figure 2.1 (a) A network and (b) a circuit.
resent a practical element. For example, a practical resistor has an ideal

resistance element as part of its make-up, but it also has so me inductance
because the current passing through it pro duces a magnetic field, and it has
some capacitance because it is insulated from, say, earth (the insulation
acting as a dielectric). Fortunately it is usually possible either to neglect
subsidiary effects or to compensate for them, but not in every case.
It is fortunate that the major feature of an element (such as the resist-
ance of a resistor, or the capacitance of a capacitor) can be thought of as
being at one point within the element; if this is the case, then we say we are
dealing with a lumped-constant element. In other cases, such as a transmis-
sion line, we need to think of the element as an infinite number of infinitely
sm all interconnected elements; in this case we say that we are dealing with
a distributed-constant element. This latter type of network is dealt with in
more advanced texts.
Anode is a point in a circuit which is common to two or more circuit
elements; the circuit in figure 2.2 has four nodes (numbered 0, 1,2 and 3).
A junction or principal node is a point in the circuit where three or more
elements are connected together; nodes 0 and 2 in figure 2.2 are principal
nodes. In many cases we drop the term 'principal', and refer to them as
'nodes' . In the majority of cases of circuit analysis, we choose one node
(usually a principal node) to be a reference node or datum node, so that the
voltage of other nodes can be defined with respect to it.
For example if, in figure 2.2, we define node 0 as the reference node,
then the voltage at node 3 with respect to node 0 is V 30 ; however, since
node 0 is the reference node, we simply say that the voltage of node 3 with
respect to node 0 is V 3 •
A branch in a circuit is a path containing one circuit element, and which
connects one node to another. Thus R 1 in figure 2.2 is the element in the
branch connecting node 1 to node 2. A path in a network is a set of
connected elements that may be traversed without passing through the
same node twice. For example, elements R 2 and R 4 in figure 2.2 are in the
path which commences at node 0 and, after passing through node 2, arrives
at node 3. A loop in a circuit is a closed path within the network; the closed
path containing the current source Is , and the resistors RH R4 and R3 in
(j) ....--{==t---.......
Figure 2.2 A circuit with tour nodes and jive branches.
figure 2.2 is a loop (that is, the loop passing through nodes 0, 1, 2, 3 and
back to node 0)
An element which can be connected in a circuit in either direction (this
assumes that it is a two-terminal element) without changing the electrical
performance of the circuit is known as a bilateral element. Examples,
include resistors, inductors and capacitors. The majority of networks are
bilateral networks, that is, they contain only bilateral elements. Certain
network theorems, such as the reciprocity theorem (see chapter 3 for
details), are only applicable to bilateral networks.
A planar network is one which may be drawn on a ftat surface, so that
none of the branches passes over or under any other branch. When this
cannot be done, the network is non-planar. The circuit drawn in fullline in
figure 2.3 is a planar network but, if the branch containing R 6 is introduced
(shown broken), it becomes a non-planar circuit.
A mesh is a loop wh ich does not contain any other loops within it. For
example, the circuit in fulliine in figure 2.3 contains the meshes ABCDA
(R 1 , R 2 , R 3 , V s ), BCDB (R 2 , R 3 , R 4 ) and ABDA (RH R 4 , Vs ). The loop
containing, for example, R 1 R 2 R 3 V s in figure 2.3 is not a mesh because it
contains two loops. However, the reader should note that, in some cases
R.
/,-----0--------,
....-~j_~._-~J_-.© I
I
I
R3 I
I
L -_ _......_ _ _ _ _. . . ._ _ _ _- - J __ JI
Figure 2.3 The circuit drawn in tulliine is a planar network.
Circuit Analysis 31
(referring to figure 2.3), it is possible to re-draw the eireuit so that a loop

whieh is not a mesh in one version of the cireuit ean beeome a mesh in
another version.
2.3 Mesh analysis
Mesh analysis involves the eoneept of mesh current, whieh is introdueed

via the two-mesh, four-node, five-braneh planar eireuit in figure 2.4; the
reader should note that the four nodes eomprise two 'nodes' and two
'principal nodes'. The eireuit is shown with a mesh eurrent cireulating
around the periphery of eaeh mesh in a clockwise direction (11 and 12 in the
figure). The relationship between the mesh eurrent and the branch currents
(1A' 1B and 1d are
Let us apply KVL to eaeh mesh in turn. For mesh 1 (the left-hand mesh),
the equation is (proeeeding in the direetion of 11 )
20 - 101\ - 3011 + 3012 = 0
or
20 = 1011 + 3011 - 3012 = 4011 - 3017
and for mesh 2 the equation is (proeeeding in the direetion of 12 )
-10 + 3011 - 3012 - 2012 = 0
or
-10 = - 3011 + 3012 + 2012 = -3011 + 5012
The equations to be solved are, therefore
20 = 4011 3012
-10 = -3011 + 5012
Node 2
Node 1 10 Q IA 18 20 Q Node 3
:0 ::0
le
20 V t 30Q
Node 4
Figure 2.4 Mesh analysis.
R2
R.
Figure 2.5 A three-mesh, jive-node, seven-branch circuit.
Solving the two simultaneous equations is fairly straightforward, and the

answers are 11 = 0.636 A and 12 = 0.182 A. While the above problem is
useful in illustrating a simple application, it does not help us to 'look' and
simply write down the mesh equations.
Before we can do this, we need to determine the equations of a more
complex circuit, such as that shown in figure 2.5. The equations of this
circuit will enable us to draw some general conclusions, which allow us to
write down the mesh current equations by observation. The reader should
carefully study the following passages.
Applying KVL to mesh 1 (in which 11 circulates), the mesh equation is
seen to be
or
V 1 + V 2 = (R 1 + R 4 )11 - R 112 - R 413 (2.1)
Similarly for mesh 2 (in wh ich 12 circulates), the equation is
- V2 + l 1R 1 -:- 12R 1 - 12R 2 - 12R 3 + 13R 3 = 0
or
- V2 = -RJ1 + (R 1 + R 2 + R 3)12 - Ri3 (2.2)
and for mesh 3 the equation is
l 1R 4 - 13R 4 + 12 R 3 - 13R 3 - 13R s = 0
that is
(2.3)
The equations (2.1) to (2.3), inclusive, allow us to solve the circuit, and
are grouped below
Circuit Analysis 33
V 1 + V 2 = (R 1 + R 4 ) /1 - Ri2 R 4/ 3
- V2 - Ri1 + (R 1 + R 2 + R 3)/2 - R 3/ 3
o -R 4/ 1 R 3/ 2 + (R 3 + R 4 + R 5 ) /3
The equations can be written in the following generalised form
E1 R n / 1 + R 12/ 2 + R 13/ 3
E2 R 21 / 1 + R 22/ 2 + R 2i3
E3 R 3i1 + R 32/ 2 + R 33/ 3
where E 1 is the sum of the source voltages driving 11 in a clockwise
direction, that is, E 1 = V 1 + V 2 ; E 2 is the sum of the source voltages driving
12in a clockwise direction, that is, E 2 = - V 2 (that is, V 2tries to drive 12in a
counterclockwise direction); E 3 is the sum of the source voltages driving 13
in a clockwise direction, that is, E 3 = O.
R w R 22 and R 33 are, respectively, the self-resistances of the mesh in
which 11 , 12 and 13 ftow. That is, Rn is the sum of all the resistances around
the perimeter of mesh 1, that is Rn = R 1 + R 4 ; R 22 is the sum of the
resistances around the perimeter of mesh 2, that is R 22 = R 1 + R 2 + R 3, etc.
If we write down R ij as the resistance in row i and column j of the matrix
containing the resistance elements, we see that R ij is (for =F j)
( -1) x the resistance in the branch
which is mutual to the meshes in
which li and I j ftow
that is, R 13 = - R 4 , R 23 = - R 3, etc.
We can write down the resistance values in the above equation in wh at is
known as matrix form as follows; the square brackets around the symbols
tell us that we are dealing with a matrix (see chapter 15 for details).
R13]
R 23
R 33
The resistance matrix is a square matrix, that is it has as many rows as it
has columns. For a bilateral network it is symmetrical about the major
diagonal, that is R ij = Rji , for example R 23 = R 32 , R 13 = R 31> etc. In the case
of abilateral network, that is, a resistive network, all elements on the
major diagonal of the resistance matrix are positive; elements not on the
major diagonal are either zero or negative.
These comments do not always apply to a non-bilateral network or to
networks containing sources other than independent voltage sources, for
example, current sources (see worked example 2.6.2).
2.4 General rules for writing mesh equations
Depending on the circuit, there are either three or four general rules which
need to be followed:
1. Draw a carefully labe lIed circuit diagram.
2. Assign mesh currents II> 12 , • • • , Im to each mesh flowing in a clockwise
direction in the circuit.
3. If the circuit contains only voltage sources, apply KVL to each mesh
and solve the resulting simultaneous equations for the unknown mesh
currents (if there are m meshes, there are m equations). If the circuit
contains dependent voltage sources, relate the dependent source volt-
ages to the unknown mesh currents.
4. If the circuit contains one or more current sources, we cannot deal with
it in the normal way because the internal resistance of these sources is
infinity. The following rule explains how to deal with it. However, since
it is fairly technical, the reader should study worked example 2.6.2 in
association with the following.
It is first necessary to replace each such source by an open-circuit (note: the
mesh currents assigned in step 2 must not be changed). Each source
current should then be related to the mesh currents assigned in step 2. The
resulting simultaneous equations should then be solved
Where a circuit contains practical current sources (see the work on
Thevenin's and Norton's theorems in chapter 3), each can be converted
into its equivalent practical voltage source, and the problem solved as
outlined in step 3 above.
2.5 Solution of three simultaneous equations
Solution of circuits containing two unknowns is relatively simple because

we are dealing with only two simultaneous equations. Unfortunately, many
practical circuits contain three or more unknowns. We look here at the
principles involved in solving for three unknowns. Let us suppose that the
three simultaneous equations representing the circuit are of the form
VI = AX + BY + CZ
V2 DX + EY + FZ
V3 GX + HY + JZ
where VI> V 2 and V3 are three numerical values, A to Hand J ('I' is omitted
for the reason that it may be confused with current) are coefficients, and X,
Y and Z are the unknown variables.
The solution of three simultaneous equations is no more difficult than
Circuit Analysis 35
solving two simultaneous equations, but it simply involves a few more steps
and takes a little longer. We will list the general procedure here, and then
solve a set of three simultaneous equations.
1. Eliminate one of the variables (say X) from, say, the 1st and 2nd
equations.
2. Eliminate the same variable (X) from two other equations (say the 1st
and the 3rd). This leaves two simultaneous equations with two un-
knowns (Y and Z).
3. Solve for Yand Z from these two equations.
4. Insert the values of Y and Z into one of the original equations to
determine the value of X.
Consider the equations
-20 = + 2Y - 4Z
3X (2.4)
32 = 1.5X - 3Y + 4Z (2.5)
-11 = X + Y - 2Z (2.6)
Step 1: Eliminate X from equations (2.4) and (2.5) by multiplying equation
(2.5) by 2 and subtracting it from equation (2.4).
-20 = 3X + 2Y - 4Z «2.4) re-written)
64 = 3X - 6Y + 8Z «2.5) x 2)
SUBTRACT -84 = 8Y -12Z (2.7)
Step 2: Eliminate X from equations (2.4) and (2.6) by multiplying equation
(2.6) by 3 and subtracting it from equation (2.4).
-20 = 3X + 2Y - 4Z «2.4) re-written)
-33 = 3X + 3Y - 6Z «2.6) x 3)
SUBTRACT 13= Y +2Z (2.8)
Step 3: Solving between equations (2.7) and (2.8) for Yand Z gives Y = -3
and Z = 5.
Step 4: Re-writing equation (2.4) in terms of X gives
X = (-20 - 2Y + 4Z)/3 = (-20 -2(-3) + 4(5))/3
=2
Alternatively, we can solve the simultaneous equations by determinants
(which are fully described in chapter 15).
Yet another method is to solve three simultaneous equations by means
ofthe BASIC language program given in listing 2.1. This also uses determi-
nants to solve the equations. Some versions of BASIC do not use line
numbers but, generally speaking, they are advanced forms of the language
and will accept this program directly (including the line numbers) ..
Listing 2.1
BASIC program for the solution of three simultaneous equations.
10 CLS
20 PRINT TAB(3); "Solution of three simultaneous equations"
30 PRINT TAB (15) ; "of the form": PRINT
40 PRINT TAB (11); "V1 A*X + B*Y + C*Z"
45 PRINT TAB (11) ; "V2 D*X + E*Y + F*Z"
50 PRINT TAB(11); "V3 G*X + H*Y + J*Z": PRINT
60 PRINT TAB(3); "Where V1, V2 and V3 are numerical values, "
70 PRINT TAB (3) ; "A to H and J are numerical coefficients,"
80 PRINT TAB(3); "and X, Y and Z are the variables."
90 PRINT
100 INPUT "Vl = ", V1
110 INPUT "A = " , A
120 INPUT "B ", B
130 INPUT "C = " , C
140 PRINT
150 INPUT "V2 = ", V2
160 INPUT "D ", D
170 INPUT "E ", E
180 INPUT "F ", F
190 PRINT
200 INPUT "V3 = ", V3
210 INPUT "G ", G
220 INPUT "H = ", H
230 INPUT "J = ", J: PRINT
240 D1 = (A * E * J) + (B * F * G) + (C * D * H)
250 D2 = (G * E * C) + (H * F * A) + (J * D * B)
260 Det = D1 - D2
270 REM ** There is no solution if Det = 0 **
280 IF Det = 0 'IHEN PRINr TAB(3); '''lbe equatioos cannJt be solved.": E2ID
290 REM ** Calculate Det X, Det Y and Det Z **
300 D1 = (V1 * E * J) + (B * F * V3) + (C * V2 * H)
310 D2 = (V3 * E * C) + (H * F * V1) + (J * V2 * B)
320 DetX = D1 - D2
330 D1 = (A * V2 * J) + (V1 * F * G) + (C * D * V3)
340 D2 = (G * V2 * C) + (V3 * F * A) + (J * D * V1)
350 DetY = D1 - D2
360 D1 = (A * E * V3) + (B * V2 * G) + (V1 * D * H)
370 D2 = (G * E * V1) + (H * V2 * A) + (V3 * D * B)
380 DetZ = D1 - D2
390 REM ** Calculate the value of the variables **
400 PRINT TAB(3); "X "., DetX / Det
410 PRINT TAB (3) ; "Y "; DetY / Det
420 PRINT TAB(3); "z "; DetZ / Det
430 END
Circuit Analysis 37
2.6 Worked examples using mesh analysis
In this section we look at four examples - three of them using the same
basic network - respectively involving independent voltage sources only,
mixed voltage and current sources, and independent and dependent volt-
age sources. The fourth involves an operational amplifier circuit.
Using mesh analysis, analyse the circuit in figure 2.6.
Solution
Since the circuit contains independent voltage sources only, a solution can
be obtained using the first three steps outlined in section 2.4. The first two
steps are already performed in figure 2.6. Next, we need to write down the
three mesh equations, which we may do by observation in the manner
outlined in section 2.3. The three equations are
10 13/1 - 6/2
-8 -6/1 + 16/2
o -5/1 - 7/2 + 16/3
We can solve for the three unknown mesh currents by any of the methods
outlined in section 2.5, and the answers are
/1 0.789 A
/2 -0.119A
/3 0.195 A
3Q
Figure 2.6 Circuit tor worked example 2.6.1.

t
sv
5Q
(a) (b)
Figure 2.7 Mesh analysis 0/ a circuit containing an independent current source.
The reader will note that 12 is negative. This merely means that the current
in mesh 2 circulates in a counterclockwise direction.
In this case we use the basic circuit in figure 2.6, but have inserted a 0.5 A
independent source in the branch containing the 6 ohm resistor, as shown
in figure 2.7(a). The mesh currents can be evaluated as shown in the
solution.
Solution
Since we employ KVL in mesh analysis, it is necessary to remove the

current source from the network and use the method outlined in step 4 in
section 2.4. That is, the independent current is replaced by its internal
resistance, namely an open-circuit, as shown in figure 2.7(b). This has the
effect of reducing the circuit to two meshes, one carrying 13 , and the other
(which is called a supermesh) carrying 11 and 12 , However, the reader
should note that since 11 and 12 both exist, we must maintain their indepen-
dent identities in the analysis.
Applying KVL to the supermesh in figure 2.7(b), we get
10 - 8 = 211 + 312 + 7(12 - 13 ) + 5(11 - 13 )
or
2 = 711 + 1012 - 1213
A second equation is obtained in the usual way by applying KVL to the
mesh in which 13 ftows as follows
Circuit Analysis 39
o= -511 - 712 + 1613

Finally, the third equation is obtained by relating the current in the 0.5 A
independent current source to the unknown mesh currents as follows
0.5 = 11 - 12
The three mesh current equations for the circuit in figure 2.7(a) are,
therefore
2 711 + 1012 1213
o = -511 - 712 + 1613
0.5 11 - 12
Solving these equations by any of the methods described earlier gives
11 = 0.547 A
12 = 0.047 A
13 = 0.191 A
This example illustrates a method of solving a circuit which includes both

independent and dependent voltage sources. In this case, the current
source in figure 2.7 is replaced by a voltage-controlled voltage source. The
circuit is shown in figure 2.8.
Solution
The dependent voltage source, 3Vx is regarded as anormal voltage source,

with the exception that the voltage V x across the 5 ohm resistor must be
Figure 2.8 Mesh analysis 01 a circuit containing independent and dependent voltage
sources.
related to the unknown mesh currents. By inspection, the voltage Vx is

Vx = 5(11 - 13)
Applying KVL to the mesh in which 11 ftows gives
10 + 3Vx = 1311 - 612 - 513
or
10 = -21 1 - 612 + 1013
Applying KVL to mesh 2 gives
-3Vx - 8 = -611 + 1612 - 713
that is
-8 = 91 + 161
1 2 - 2213
and the equation for mesh 3 is
o= -511 - 712 + 1613
That is, the three simultaneous equations are
10 = --':211 - 612 + 1013
-8 = 9/1 + 16/2 - 22/3
Solving gives the results

11 = 4.776 A
12 = - 2.847 A
13 0.247 A
The operation al amplifier in the circuit in figure 2.9(a) has infinite input
resistance, a voltage gain of -100, and an output resistance of 20 kilohms.
Calculate the voltage gain (ViVI) of the circuit.
Solution
Since the voltage gain and the output resistance of the operational ampli-
fier differ from that of an 'ideal' op-amp, we need to solve the circuit
completely. The equivalent circuit is shown in diagram 2.9(b); since there
is only one mesh, its solution is
1 + 100Vx = (1000 + 10 000 + 20000)1
Cireuit Analysis 41
10 kQ 10 kQ
~----------e-----------~
I v,
(a) (b)
Figure2.9 Diagram tor worked example 2.6.4.
but V x = 1 - 1000/, that is

1 + 100(1 - 10001) = 31 0001
or I = 7.71 X 1O- 4A
The output voltage is
V2 = -l00Vx + 20 0001 = -100(1 - 10001) + 20 0001
= -100 + 1200001 = -7.48 V
and the overall voltage gain is
ViV l = -7.48/1 = -7.48
This value contrasts with the gain of -10, which would prevail if an ideal
operational amplifier was used. Note: if a load resistance of 10 kilohms is
connected to the output of the circuit in figure 2.9(a), the overall gain falls
to -6.4! The reader would find it an interesting exercise to verify this using
mesh analysis.
2.7 Nodal analysis
Nodal analysis uses KCL to evaluate the voltage at each principal node in
the circuit, and is valid for all circuits, both planar and non-planar. In this
case we write down and solve a set of simultaneous equations in terms of
the unknown voltage at each node. We will illustrate this initially by means
of the simple three-node example in figure 2.10.
Of the three principal nodes in the circuit, we must choose one to be a
reference or datum node; we select node O. Generally speaking, if the
circuit has n principal nodes, we need (n - 1) simultaneous equations to
solve the circuit. Applying KCL to each non-reference node in turn,
starting with node 1, we get
2 = 2Vl + 3V 12 = 2Vl + 3(Vl - V 2 ) = 5Vl - 3V2
v12
--
- ---
-----
1._
-. V21
NOdeW NOdeW
Reference
node ~
'ri'
Figure 2.10 Simple example o[ nodal analysis.
Reference
node
Figure 2.11 Nodal analysis.
In the above equation, the current entering the node appears on the
left-hand side of the equals sign, and the current leaving the node appears
on the right-hand side. For node 2 we have
-3 = 3V2I + 4V2 = 3(V2 - VI) + 4V2 = -3VI + 7V2
Solving the resulting pair of simultaneous equations gives VI and V2,
enabling other data to be calculated. We merely quote the solutions VI =
0.192 V and V 2 = -0.346 V here, and will proceed to a more useful circuit
which enables us to write down the circuit equations by observation.
Consider the circuit in figure 2.11. When applying KCL to node 1, we
see that JA ftows towards it and a current of (GE (VI - Vo) + GA (VI - V2))
ftows away from it. Since Vo = 0, the node equation is
Circuit Analysis 43
I A = (GA + GE)VI - G A V 2
At node 2, a current of (ID - I A - le) enters the node and (G A (V2 - VI)
+ G D(V2 - V o) + G e(V2 - V 3» leaves it. As before, V o = 0, hence the
equation at this node is
ID- IA - le = -GAVI + (GA + GD + Ge)V2 - G eV 3
The current entering node 3 is le, and the current leaving it is (Ge(V3 - V2 )
+ G D (V3 - Vo». The equation for node 3 is, therefore
le = -GeV 2 + (Ge + G D )V3
The three node voltage equations describing the circuit are therefore
IA = (GA + GE)VI - G A V 2
ID - IA - le = -GAVI + (GA + GD + Ge)V2 - G eV 3
le - G eV 2 + (Ge + G D )V3
These can, conveniently, be written in the following generalised form
I1 G ll VI + G 12 V2 + G 13 V 3
12 G21 VI + G22 V 2 + G23 V 3
13 = G31 VI + G32 V 2 + G33 V 3
where I 1 (= I A ) is the current entering node 1, 12 (= I D - I A - Ie) is the
current entering node 2, and 13 (= Je) is the current entering node 3.
We may write the conductance elements in matrix form (see chapter 15
for details) as folIows.
Gu G 12 G13]
[ G 21 G 22 G 23
G 31 G 32 G 33
Each of the terms on the major diagonal of the conductance matrix
(which is a square matrix) is the sum of the conductances terminating on
node 1, 2 and 3, respeetively. That is G l l (= GA + GE) is the total
eonduetanee terminating on node 1, G 22 (= GA + GD + Ge) is the total
eonduetanee terminating on node 2, etc.
The voltages VI> V2 and V3 are, respeetively, the unknown voltage at
node 1, 2 and 3.
If G ij is the eonductance in row i and column j of the matrix eontaining
the conductanee elements, we see that Gij (for i -:1= j) is
(- 1) x the eonduetanee linking node i
to node j
That is, G 12 = -GA' G 32 = - Ge, ete.
The conductance matrix is a square matrix, and is symmetrical about the
major diagonal, that is for all ij (i =1= j), G;j = G j;, that is, G 12 = G w G 23 =
G32 , etc.
In the case of abilateral network, all the elements on the major diagonal
of the conductance matrix are positive: elements not on the major diagonal
are negative or zero.
2.8 General roles for writing nodal equations
Depending on the circuit, there are either three or four steps to be carried
out:
1. Draw a carefully labelIed circuit diagram.
2. Mark the principal nodes on the circuit, and select a reference node. If
there are n principal nodes, (n - 1) simultaneous equations are needed
to solve the circuit.
3. If the circuit contains only independent current sources, apply KCL to
each non-reference node. If the circuit contains dependent current
sources, relate the source current to the unknown node voltages.
4. If the circuit contains voltage sources, we cannot deal with it in the
normal way because its internal resistance is zero. The following rule
explains how to handle it, and is fairly technical; the reader should study
worked example 2.9.2 in association with the following.
Replace each voltage source by a short-circuit; the voltages assigned in step
2 should not be changed. Each source voltage should then be related to the
unknown node voltages. If the circuit contains a practical voltage source, it
can be converted to its equivalent practical current source (see chapter 3)
and dealt with as anormal current source
2.9 Worked examples using nodal analysis
In the following we analyse a circuit and illustrate its solution firstly when it
contains only independent current sources, secondly when it contains
independent current and voltage sources and, thirdly, when it contains
independent and dependent current sources. Finally we will analyse an
operation al amplifier circuit.
Determine the node voltages in figure 2.12.

Circuit Analysis 45
2A
Figure 2.12 Circuit for worked example 2.9.1.
Solution
This circuit is relatively straightforward, as it contains only independent

current sourees. We nominate node 0 as the reference node (as we do for
the other circuits of this kind). The first two steps outlined in section 2.8
have already been carried out, and are shown in figure 2.12. Next we will
write down by inspection (see section 2.7) the node voltage equations. For
node 1 the equation is
o= 12VI - 2V2 - 4V3
and for node 2 the equation is
-1 = -2VI + 5V2 - 3V3
For node 3 the equation is
3 = -4VI - 3V2 + 12V3
The equations are
o 12VI - 2V2 4V3
-1 = -2VI + 5V2 3V3
3 -4VI - 3V2 + 12V3
Solving by any of the methods described earlier yields
VI = 0.0963 V
V 2 = 0.0092 V
V 3 = 0.2844 V
Note: all voltages are relative to node O.
In the second worked example in this section, an independent 4 V voltage

source is connected in parallel with the 4 S conductance in figure 2.12 (see
also figure 2.13). One method of solving this type of circuit is described
below.
Solution
Since the nodal analysis deals with current sources, it is not possible to
handle voltage sources directly. In order to deal with this type of circuit
element, the notion of a supernode is introduced. What we do in this case is
to regard nodes which are connected by the voltage source as though they
were connected by the internal resistance of the voltage source, namely
zero ohms. That is, nodes 1 and 3 in figure 2.13 become a supernode
(which is enclosed by a broken line). The total current flowing towards the
supernode is 3 A, and the total current leaving it is
2(V! - V z) + 6(V! - Vo) + 3(V3 - V z) + 5(V3 - Vo)
The reader should note that although nodes 1 and 3 are combined in the
supernode, the voltage of both nodes is maintained in the above equation.
The nodal equation (remember, node 0 is the reference node, so that
Vo = 0) for the supernode is therefore
3 = 2(V! - V z) + 6V! + 3(V3 - V z) + 5V3
= 8V! - 5Vz + 8V3
,
2A t 1~
I
I Supernode
_J
I
Figure 2.13 The concept 01 a supernode.

Circuit Analysis 47
Applying KCL to node 2 gives (see also worked example 2.9.1)

-1 = -2VI + 5V2 - 3V3
The process of combining two no des into one supernode means that we
have lost one of the nodes to which we could apply KCL! However all is
not lost because, at the supernode, nodes 1 and 3 are actually separated by
a potential of 4 V, that is
4 = VI - V3
which leaves us with three equations which are written below
3 = 8Vl - 5V2 8V3
-1 = -2VI + 5V2 - 3V3
Solving by one of the methods outlined earlier gives

VI = 2V V2 = -0.6 V V3 = -2 V
The 'supernode' is a convenient fiction which allows us to get out of some
difficult situations. However, study of the circuit shows that if node 1 is
chosen as the reference node, the analysis follows the normal pattern. Not
all lecturers are convinced of the usefulness of the supernode concept.
In this case, the 4 S conductance in figure 2.12 is shunted by a voltage-

dependent current source, as shown in figure 2.14. The analysis is as
folIows.
Solution
Since all sources in the circuit are current sources, we can apply KLC
directly to each node. At node 1 we have
-1.5V32 = 12Vl - 2V2 - 4V3
However V 32 = V 3 - V 2 • Inserting this into the above equation gives, for
node 1
o= 12Vl - 3.5V2 - 2.5V3
Applying KCL to node 2 gives
-1 = -2Vl + 5V2 - 3V3
Finally, at node 3
2A t
Figure 2.14 Nodal analysis of a circuit containing adependent current source.
or
3 = -4V j - 1.5V2 + 1O.5V3
and the circuit equations are
o= 12VI - 3.5V2 - 2.5V3
-1 = -2VI + 5V2 - 3V3
3 = -4VI - 1.5V2 + 1O.5V3
and the solution is
VI = 0.0707 V
V 2 = 0.0174 V
V 3 = 0.3151 V
The operation al amplifier in the circuit in figure 2.15(a) has the following
parameters
input resistance = 50 kilohms
voltage gain = -1000
output resistance = 1.5 kilohms
Calculate the overall gain (V/VI) of the circuit.
Circuit Analysis 49
v, = 0.3 V,
(a) (b)
Figure 2.15 Figure tor worked example 2.9.4.
Solution
The equivalent circuit of the amplifier is shown in figure 2.15(b); since the
values differ from those of an ideal operation al amplifier, we will analyse
the circuit using nodal analysis.
In this case, we place the 0.3 V independent source and the 1000Vx
dependent source (corresponding to the amplifier gain) in the supernode in
diagram (b), thereby reducing the number of nodes by two. Applying KCL
to node 2 yields
Now at the supernode VI = 0.3 V, and

V 3 = 1000Vx = 1000(VI - V 2 ) = 1000(0.3 - V 2 )
That is
V - 0.3 V2 - 1000(0.3 - V 2 ) V2 0
---"---+
2
+--=
50 000 1500 1000
giving V2 = 0.299 V. The voltage gain therefore is
ViV I = 0.299/0.3 = 0.997
2.10 Network topology
It was stated earlier that mesh current analysis is applicable only to planar
networks. However there is a similar approach - known as loop analysis
W 30
a
20
..._--.b
(a) (b)
Figure 2.16 The circuit in (a) has the graph in (b).
(see also seetion 2.11) - whieh allows us to solve non-planar networks. In

order to understand some of the teehniques involved we need to have a
brief introduetion to the subjeet of network topology.
Network topology is eoneerned with a mathematieal diseipline known as
graph theory with special referenee to eleetrieal cireuits. Tbe 'graph' refer-
red to here is not a eonventional graph, but is a eolleetion of points (nodes)
and eonneeting lines (branehes). When drawing the 'graph' of a network,
the nature of the element in the braneh between a pair of nodes is
suppressed, and is replaced by a line or 'edge'.
The eireuit in figure 2.16(a) has four nodes and six branehes; the
eorresponding graph (or connected graph) is shown in diagram (b).
Given a graph, we define a tree (or spanning tree) as any set of branehes
in the graph whieh eonneet every node to all other nodes in the graph, but
not neeessarily direetly. Moreover, the tree does not contain a loop 01 any
kind.
If the graph has N nodes, eaeh tree has (N-1) branehes in it. Tbe
four-node graph in figure 2.16(b) eontains sixteen trees, four of which are
shown in fulliine figure 2.17. In loop eurrent analysis, we seleet anormal
tree, that is one containing all the voltage sources in the network, together
with the maximum number of voltage-controlled dependent sources.
A cotree is a set of branehes whieh do not belong to a tree; the eotrees
corresponding to eaeh of the four trees in figure 2.17 are shown in broken
lines. A braneh in a eotree is known as a link. A eotree is the eomplement
of a tree, and a tree and its cotree form the complete graph of a network.
An N-node network eontains a number of trees, eaeh with (N-1)
branehes; if B is the number of branehes in the network, and L is the
number of links in the eotree, the relationship between them is
B = L + (N - 1)
Circuit Analysis 51
a a a a
,, /
,,
/
I
I
b
, b
I
I d b
\
\
\
\ ,
c c c c
Figure 2.17 Examples oftrees and cotrees within the graph offigure 2.16(b).
For example, in a four-node (N = 4), six-branch (B = 6) network, there

are a number of trees each containing (N - 1) = 3 branches, and the
number of links in each cotree is
L = B - (N - 1) = 6 - (4 - 1) = 3
If we re-position any one of the links from the cotree in the tree, we will
form a loop in the tree. Consider for the moment the left-hand tree in
figure 2.17. If we add the link connecting d to ewe get the independent loop
bcdb. This process is repeated throughout the tree as folIows. If we add
link ad we get the independent loop adbca; adding link ab gives the
independent loop abca.
Figure 2.18 The graph of a six-node non-planor network.
Since adding a link to the tree forms a loop, the number of links is equal
to the number of independent voltage equations we need to form the loop
equations of the network. For example, the six-node nonplanar network in
figure 2.18 contains a number of trees, each having five branches, and we
need L = 9 - (6 - 1) = 4 independent equations to solve the network. In
some cases, however, the situation is not always that difficult because (as
we shall see from an example in section 2.11) the number of equations
needed to solve the circuit is reduced when there are current sources in the
links.
2.11 Loop analysis
The following steps allow us to write a set of loop current equations for a
circuit:
1. Draw a graph of the network and identify anormal tree.
2. Ensure that all voltage sources and, if possible, all control-voltage
branches for voltage-controUed dependent sources are in the tree.
3. Ensure that all current sources and, if possible, aU control-current
branches for current-controUed dependent sources are in the cotree.
4. Reposition in the tree, one at a time, each link in the cotree. Using
KVL, write down for each loop the associated loop current equation;
solve the equations.
To illustrate the general principles involved we will solve, using loop
analysis, the non-plan ar circuit in figure 2.19. The circuit has six nodes and
nine branches, and requires a set of four loop current equations for its
solution. Since the circuit is non-planar, it cannot be solved by mesh
analysis. The selected normal tree is shown in fuUline in diagram (b).
Since the voltage source must be included in the tree, the branch be
must be included in the tree. Thereafter we can select any four other
connected nodes (provided that they do not include the current source) to
complete the tree.
Next we insert the links (shown broken in figure 2.19(b», one at a time,
from the cotree into the tree in order to produce four fundamentalloops in
the graph; these will provide us with the required equations. These loops
are shown in figure 2.20 for loop currents I A , I B , Ie. and 10 • We have
decided that each current shaU ftow in a clockwise direction; we need not
choose this direction for loop analysis and can, alternatively, be counter-
clockwise, or in either direction. The currents are also shown on the
branches of the circuit in figure 2.19(a). We now write down the loop
current equations.
For loop I B (the loop abcda)
It will be seen that, in figure 2.19(a), each branch ofthe normal tree carries
more than one loop current, and that each link of the cotree carries only
one loop current. Since the 3 A current source is in link da, we can simply
say that
IA = 3 A
This has the effect of reducing by unity the number of equations required
to solve the circuit.
For loop IB (the loop abcfa)
The equation is
0= 7IB + 5(IA + I B - 10 ) + 4(IA + I B + I e ) + 3(IB + I e )
Circuit Analysis 53
W 40 W
IA + 10 + le
W W
le + 10 /
10 + le
10 I
~ I I
\7 70
IA - 10
\.ß>I
......
60 (b)
[Y ~
-
(a)
Figure 2.19 The solution of a non-planar circuit using loop analysis.
(a) (b) (e) (d)
Figure 2.20 The fundamentalloops for figure 2.19.
or
91A = -27 = 191B + 71e - 51D
For loop 1e (the loop abefa)
0=5 + 4(IA + 1B + 1e) + 3(IB + 1e) + 2Ie
or
-5 - 4/A = -17 = 7/B + 9/e
For loop 10 (the loop bedca)
o= 5 - 5(/A + I B - 10 ) - 6(/A - 10 ) + 8/0
or
-5 + 11IA = 28 = -5/B + 19/0
The equations are regrouped below
-27 = 19/B + 7/e - 5/0
-17 = 7/ + 9/e B
28 = -5/B + 19/0
Solving gives
I B = -0.5236 A
le = -1.4816 A
10 = 1.336 A
2.12 Duality
Two eireuits are duals if the mesh equations of one eireuit have the same
mathematieal form as the nodal equations of the other. They are exact
duals if the mesh equations of one cireuit are numerieally identical to the
nodal equations of the other. A limitation of duality is that it is only
possible to produee the dual of a network if it is a planar network.
Consider the cireuit in figure 2.21; the equations for the two cireuits are
Circuit 2.21(a) Circuit 2.21(b)
d· dv
= Ri + L....!:.. = Gv + C-
dt dt
where R is a resistanee, G is a eonduetance, L is an induetanee* , and Cis a
eapaeitanee.
The cireuits are duals of one another beeause the mesh eurrent equation
of cireuit (a) has the same form as the node voltage equation of cireuit (b).
• A tull deseription of induetance and capacitance is given in chapter 4.

Circuit Analysis 55
i,
L G
(al (bl
Figure 2.21 Introduction to dual circuits.
The two would be exact duals if the voltage of v in volts in circuit (a) was
equal to the current i in circuit (b), and if the value of R in ohms in (a) was
equal to the value of G in siemens in (b), and if the value of L in henrys in
(a) was the same as the capacitance of C in farads.
Clearly, once the (say) mesh equations of one circuit have been solved,
then the nodal equations of the exact dual have also been solved.
The relationship between a circuit and its dual are summarised in table
2.1. Multiples and submultiples are also exchanged between the cireuit and
its dual; for example, mV in a eireuit beeome mA in the dual, !J,F in the
eireuit beeome !J,H in the dual, ete.
Table 2.1
Circuit element Dual circuit element
Series connection Parallel connection

Parallel connection Se ries connection
Voltage source Current source
Current source Voltage source
Voltage of n volts Current of n amperes
Current of n amperes Voltage of n volts
Resistance of n ohms Conductance of n siemens
Conductance of n siemens Resistance of n ohms
Inductance of n henrys Capacitance of n farads
Capacitance of n farads Inductance of n henrys
The reader should note that the dual is not the equivalent 0/ the original
circuit. That is, if a eurrent of 5 A in an element in the original eireuit produees
a voltage of 2 V aeross that element then, in the dual eireuit, a voltage of 5 V
aeross the dual of the element produces a eurrent of 2 A in it.
We will now study how a planar eircuit in figure 2.22(a) is eonverted into
its dual. The steps followed are:
1. Plaee anode (we use node 0) in the spaee outside the eireuit.
2. Place anode (no des 1-4) inside eaeh mesh of the cireuit.
r----l
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I Q volts I I
--
4f:~-~4-~13~Q--~13F I I
L ____ ~'::'__ e&---":::-.J___ ...JI
\} )
(a)
2S
M
amperes
(b)
Figure 2.22 (a) An electrical network and (b) its dual.

Circuit Analysis 57
3. Draw a broken line joining adjacent pairs of nodes, each li ne passing

through one circuit element on the perimeter of each mesh.
4. Draw the dual circuit by inserting the dual of each element in the
original circuit on the broken line linking a pair of nodes.
5. Assign directions to current and voltage sources in the dual using the
method outlined below. This method can also be used to assign initial
values of current and voltage in the dual.
Taking a look at the broken links between nodes 2 and 3 in the original
circuit we see that, in the dual, the circuit comprises a conductance of 4 S in
parallel with a capacitor of capacitance 6 F. This process is repeated until
all the elements have been replaced by their dual. Finally, it only remains
to assign voltage and current direction in the dual, as follows.
Rotate the voltage (or current) arrow in the original circuit in a clock-
wise direction until it lies in the broken line linking the nodes in the meshes
on the original circuit. The new direction of the arrow indicates the
direction of the corresponding current (or voltage) arrow on the dual.
Unworked problems
2.1. Calculate the currents 11 and 12 in figure 2.4, and determine the total
power consumed.
[11 = 0.636 A; 12 = 0.182 A; 10.9 W]
2.2. If, in figure 2.5, VI = 10 V, V 2 = 20 V, R 1 = 1 ohm, R 2 = 2 ohm, R 3
= 3 ohm, R 4 = 4 ohm and R 5 = 5 ohm, calculate the mesh currents.
[11 = 7.7 A; 12 = -0.875 A; 13 = 2.35 A]
2.3. Calculate VI and V 2 in figure 2.10, and compute the total power
consumed.
[VI = 0.192 V; V 2 = - 0.346 V; 1.42 W]
2.4. Using mesh analysis, calculate 11 and 12 in figure 2.23.
[11 = 0.555 A; 12 = - 0.803 A]
v, 6Q v, 4Q
5Q :0
Figure 2.23
60
Figure 2.24
--
60
v
2A
30 50
Figure 2.25
2.5. Using nodal analysis, calculate V 1 and V 2 in figure 2.23.

[V1 = 10.12 V; V2 = 6.79 V]
2.6. Construct the dual of the circuit in figure 2.23.
2.7. The mesh equations of a network are as folIows:
-10.7 = 1111 - 412 - 213
14.9 = -411 + 1412
-5.3 = -211
-1.1 =
Draw the corresponding circuit diagram and construct its dual.
2.8. Using mesh analysis, calculate v in figure 2.24.
[0.27 V]
2.9. Use nodal analysis to calculate v in figure 2.25.
[0.36 V]
2.10. In figure 2.24, the 3 ohm, 5 ohm and 6 ohm resistors forin a tree.
Use loop analysis with respect to this tree to calculate v.
[0.27 V]
Circuit Analysis 59
30
I, 40 50
v,t
Figure 2.26
100kO
v, = 0.1 vf
1
V2
50 !!S 10 kQ
c c
Figure 2.27
2.11. Use mesh analysis to calculate v in figure 2.25.

[0.36 V]
2.12. For the circuit in figure 2.26, use mesh analysis to calculate (a) the
voltage gain ViV I and (b) the input resistance (= V/I I ) of the
circuit.
[Gain = -0.32; input resistance = 4.61 ohms]
2.13. For the simplified emitter folIower equivalent circuit in figure 2.27,
use mesh analysis to calculate (a) the voltage gain of the circuit
(= ViV I ) and (b) the input resistance (= VI/lI).
[(a) 0.823; (b) 10.9 kilohm]
2.14. Use nodal analysis to solve problem 2.12.
[gain = -0.32; input resistance = 4.61 ohm]
2.15. Solve problem 2.13 using nodal analysis.
[(a) 0.823; (b) 10.9 kilohm]
I,
3fl
I, 4fl 13
2fl
5fl ~v,
Figure 2.28
Figure 2.29
2.16. Use nodal analysis to calculate 1}> 12 and 13 in the circuit in figure
2.28.
[11 = -0.87 A; 12 = -1.128 A; 13 = 0.071 A]
2.17. Use nodal analysis to determine the node voltages VI and V 2 in

figure 2.29.
[VI = 6.154 V; V2 = -21.54 V]
3
Circuit Theorems
3.1 Introduction
All electrical and electronic circuits can be solved by the application of

basic circuit laws such as Ohm's law, Kirchhoff's laws, etc. However, it is
useful to have a collection of theorems which, for a particular application,
encapsulate appropriate laws; this allows us to obtain a speedy solution to
these problems. We look at the more important theorems in this chapter.
3.2 Linearity
Many network theorems are based on the concept of linearity; an element,

or system is said to be linear if the effect (such as the output voltage from a
circuit) is directly proportional to the stimulus (such as the input voltage).
3.3 Principle of superposition
The principle 0/ superposition states that in a linear system, having more

than one independent source, the response (either a voltage or a current)
can be obtained from the sum of the responses produced by each source
acting alone.
While it is not possible here to provide a formal proof of the principle of
superposition, a simple demonstration of the principle is given.
Consider the linear resistor R connected to the independent voltage
source VI in figure 3.1(a). The resulting current is 11 = V/R. When VI is
removed and replaced by its internal resistance (a short-circuit), and a
second independent source V 2 is connected in the circuit (see diagram (b),
the resulting current in the circuit is 12 = ViR. Finally, when VI is
61
v, , /, /, v,t /, + /,
R R R
v,t v, t
(a) (b) (c)
Figure 3.1 The principle 0/ superposition.
re-connected in the circuit (diagram (c», the current in the circuit is (V} +
V 2 )!R = I} + 12 , and
Vs = R(I} + 12 ) = V} + V2
That is, the response is the sum of the stimulating signals.
Moreover, if the stimulus is scaled by a factor K, then the response is
also scaled by the same factor. That is, if the applied voltage is KV, then
the response is KI, where KV = R(Kl). The principle of superposition can
be stated as folIows:
In any linear bilateral network containing several independent sources, the
voltage across (or the current in) any element or source is the sum 0/ the
individual voltages (ar currents) produced by each individual source
acting alone (other sources in the network meanwhile being replaced by
their internal resistance)
While the principle of superposition can be applied to many d.c. circuits, it
is particularly useful in its application to a.c. circuits where, for example,
sources of differing frequencies are involved.
Calculate I in figure 3.2 using the superposition theorem.
Solution
We need to calculate the current in the 10 Q resistor produced by the

individual resources.
Current I} produced by the IO V source: Initially the 5 A current source is
removed and replaced by its internal resistance, namely an open-circuit;
this is illustrated in figure 3.3. The current I} (ftowing downwards) is
I} = 10/(5 + 10) = 0.6667 A
Circuit Theorems 63
Figure 3.3 Curren! 11 produced in the 10 Q resistor by the 10 V source acting alone.
Current /2 produced by the 5 A source: In this case the 10 V independent

voltage source is replaced by its internal resistance, namely a short-circuit
(see figure 3.4). Using the work developed in chapter 1 on the current
division in parallel circuits we get
/2 =5 x 5/(10 + 5) = 1.6667 A
which flows upwards through the 10 Q resistor.
Complete solution: The superposition theorem states that the current i
flowing downwards through the 10 Q resistor in figure 3.2 is
/= /1 + (-/2) = 0.6667 - 1.6667 = -1 A
5Q
5A
Figure 3.4 Curren! 12 produced by the 5 A source acting alone.
3.4 Thevenin's theorem
This theorem states that any two-terminal active network, no matter how
complex, can be replaced by a practical voltage source of the type de-
scribed in section 1.12 (see figure 3.5). Thevenin's theorem may be summa-
rised as folIows:
L------oB
Figure 3.5 Thevenin's equivalent circuit.
r---------,
I I
I I 20 5V
I
~
A
I
I
I
I 50
load
I
I
I
L Effective source
--------
J B
(a) (b)
11-......-o----1D--L-{) A
50
load
L - - _....._ _ _ _-Cl B
(c) (d)
Figure 3.6 Diagram for worked example 3.4.1.
An active network, having two terminals A and B to which a load may be

connected, behaves as though it contains a single source 0/ e.m.f. Er. 0/
internal resistance R T • The e.m.f. E T is the voltage between A and B
when the load is disconnected, and R T is the resistance measured between
A and B when the load is disconnected and each internal source within
the original network is replaced by its internal resistance
Worked example 3.4.1.
Using Thevenin's theorem, calculate the current l L in the 5 Q load resistor

in figure 3.6(a).
Circuit Theorems 65
Solution
Initially we will calculate E T when the load is disconnected from the circuit
(see figure 3.6(b». Applying KCL to the bottom node in figure 3.6(b), by
observation we note that 1 = 3 A, hence
ET = -41 + 5 = -(3 X 4) + 5 = -7 V
Next, with the load disconnected, we replace each source in the circuit by
its internal resistance, leaving the circuit in diagram (c). Hence
RT = 4Q
That is, Tbevenin's equivalent circuit for this problem consists of a
E T = V AB = -7 V source in se ries with a 4 Q resistor, so that the circuit in
diagram (d) is equivalent to the original circuit in diagram (a). Hence
IL = EAR T + 5) = -7/(4 + 5)
= -0.7778 A
3.5 Norton's theorem
This theorem states that any two-terminal active network, no matter how
complex, can be replaced by a practical current source of the type
described in section 1.12 and illustrated in figure 3.7. Norton's theorem
may be summarised as folIows:
Any active network, having terminals A and B to which a load may be
connected, behaves as though it contains a current source, IN, of internal
conductance G N • The current IN is the current which would flow from
terminal A to B when they are short-circuited, and G N is the conductance
measured between A and B with the load disconnected and each source
within the network replaced by its internal conductance
...---....---oA
L . . - -.....---oB
Figure 3.7 Norton's theorem equivalent circuit.
Using Norton's theorem, calculate the current I L in the 2 Q load in figure

3.8.
"
A
~ ~
2Q
20 V load 'N
B B
(a) (b)
,.---o--c=-.J--o A
'L2Q
or
0.5 S
~-~---e-~~--~B
(c) (d)
Figure 3.8 Circuit diagram for worked example 3.5.1.
Solution
We will first evaluate IN by short-circuiting the load, as shown in diagram

(b). Now IN = 12 - 11 ; we will calculate 11 and 12 separately. Since ter-
minals A and Bare short-circuited together, then
11 = 20/4 = 5 A
and 12 = 10 x 5/(5 + 3) = 6.25 A
hence IN = 12 - 11 = 6.25 - 5 = 1.25 A
which leaves terminal A.
The internal conductance is evaluated when the externaiload is discon-
nected, and each source within the circuit is replaced by its internal
conductance. The circuit is therefore modified as shown in diagram (c).
The internal circuit comprises a 4 Q resistance (0.25 S) in parallel with a
(3 + 5) = 8 Q resistance (0.125 S). That is
GN = 0.25 + 0.125 = 0.375 S
When the original network is replaced by the Norton equivalent circuit, we
arrive at the circuit in diagram (d). Using the work covered in chapter 1 we
calculate
IL = IN X GL/(GN + G L) = 1.25 X 0.5/(0.375 X 0.5)
= 0.714 A
Circuit Theorems 67
3.6 Relationship between Thevenin's and Norton's circuits
We showed in section 1.13 that a practical voltage source and a practical

current source were interrelated. Using the relationships deduced in that
section, we may say that
R T = lIGN
and E T = IN/G N = INR T
3.7 Reciprocity theorem
Up to this point we have discussed two-terminal or one-port networks.

There is a range of four-terminal or two-port networks (see chapter 8 for
details), which includes filters, transformers, semiconductor devices, etc.;
these require more than a single relationship between the terminal voltage
and current to specify their operation. The reciprocity theorem can be
applied to these networks or, more specifically, to linear bilateral single-
source networks, and may be stated in two ways as follows:
1. If the single voltage source V x in branch X produces the current I y in
branch Y then, when the voltage source is removed from branch X and
inserted in branch Y, it produces current I y in branch X.
2. If the single current source Ix connected between nodes X and X'
produces the voltage V y between nodes Y and Y ' then, when the current
source is removed from between nodes X and X' and inserted between
nodes Y and Y ' , it produces voltage V y between nodes X and X'.
The reader should note that the current and voltage at other points in the
network change when the single voltage (or current) source changes posi-
tion.
Calculate I y in the single-source linear bilateral network in figure 3.9(a).

Remove the source V x and replace it in the branch in which I y ftows, and
verify the prediction of the reciprocity theorem.
Solution
The mesh currents 11> 12 and Iyare shown in diagram (a), and the corre-
sponding mesh equations are
20 30 40
(al
20 30 40
(bI
Figure 3.9 Diagram for worked example 3. 7.1.
10 = 8/1 - 6/2
o= -6/1 + 16/2 - 71 y
o= - 7/2 + 161y
Solving for I y by one of the methods outlined earlier gives
I y = 0.389 A
We now demonstrate the reciprocity theorem by interchanging the excita-
tion voltage and the response current, as shown in figure 3.9(b). The
reader should note that V x is connected so that it acts in the same direction
as I y in figure 3.9(a). The mesh equations for the latter circuit are
o= 81 y - 6/3
o= -61 y + 16/3 - 714
10= -7/3 +16/4
and solving for I y (see section 2.5) yields
I y = 0.389 A
Since the value of I y is the same in both cases, the reciprocity theorem is
demonstrated to be correct.
Circuit Theorems 69
3.8 The maximum power transfer theorem
A practical source of electricity has internal resistance, and when a load is

connected to its terminals, the p.d. in its internal resistance causes the
terminal voltage to fall. Clearly, if the load resistance is zero, no power is
dissipated in the load (even though there may be a large current ftowing in
it). If the load resistance is very high, very little current ftows in the load,
and very little power is consumed by the load. Between these two extreme
values of load resistance, there will be a particular value of load resistance
which consumed maximum power from the supply source. It is this we look
at here.
The maximum power transfer theorem states that, if the supply source
can be described in terms of a Thevenin or of a Norton equivalent circuit,
maximum power is absorbed by a resistive load, Ru when the resistance of
the load is equal to the internal resistance of the source.
Consider the circuit in figure 3.10. The current in the load is IL = Vs/(R s
+ R L ), and the power absorbed by the load is
P = PR L = v;,RL/(R s + RL)2
When the conditions for maximum power are investigated, that is, when
dP/dR L = 0, we find that maximum power transfer occurs when R L = Rs .
The reader will note that when R L = R s , the same amount of power is
absorbed in the load as in the source. That is, the efficiency of power
transfer (when maximum power transfer occurs) is only 50 per cent.
Figure 3.10 Maximum power transfer theorem (d.c. circuits).
The conditions for maximum power in an a.c. circuit are somewhat

more complex, and are described in chapter 6.
If, in figure 3.8(a), the 2 Q load resistor (see also worked example 3.5.1) is
replaced by a variable resistor Ru wh at value of R L absorbs maximum
power, and wh at is the value of this power?
Solution
In worked example 3.5.1 we showed that the equivalent circuit of the

source was a 1.25 A current source shunted by a 0.375 S conductance (or
2.667 Q resistance).
Maximum power is delivered to the load when the resistance of the load
is
RL = 1/G = 1/0.375 = 2.667 Q
N
and the current in the load at this time is

1.25/2 = 0.625 A
The maximum power absorbed by the load therefore is
(0.625)2 x 2.667 = 1.042 W
3.9 The parallel-generator (Millman's) theorem
This theorem is a special case of nodal analysis and is particulady useful

not only in the case of parallel generators, but also in electronic amplifier
circuits, and in the solution of unbalanced three-phase three-wire a.c.
circuits (see chapter 7 for details). We will concentrate here on the general
principles involved so far as d.c. circuits are concerned.
Consider the case of the three paralle1-connected generators EI> E 2 and
E 3 in figure 3.11(a), each having its own internal resistance R J , R 2 and R 3 ,
respectively. Since each generator is represented by its Thevenin equiv-
alent circuit we can, alternatively, represent each one by its Norton equiv-
alent circuit, as shown in diagram (b).
--------
..-VBA -----tl
...
_ VBA
E,_
E,---
A l--__--t_J-e B
A B
E3 ----
(a)
(b)
Figure 3.11 The paral/el-generator theorem or Millman's theorem.

Circuit Theorems 71
The current, 1, leaving the idealised Norton generators at node B is

1 = EIG I + E 2G 2 + E 3 G 3
where G N = lIR N • By KCL, this is equal to the current entering the three
parallel-connected conductances, that is
VBA(G I + G2 + G3 )
hence
EIG I + E 2 G 2 + E 3 G3
V BA = GI + G 2 + G 3
If there are n practical voltage sources in parallel with one another, we may
say
n
L EkGk f Ek
k~1 Rk
VBA = k~1
n n
L Gk
k~1
L
k~1
lIR k
or, more simply, though not quite as comprehensively
If, in figure 3.1l(a), EI = 10 V, E 2 = 20 V, E 3 = -25 V, R I = 20 Q,

R 2 = 15 Q and R 3 = 10 Q, calculate V BA and the current in each generator.
Solution
From the above theory

10 20 25
-+---
20 15 10
111
-+-+-
20 15 10
= -0.6667 = -3.077 V
0.2167
If 11> 12 and 13 ftow towards node B (see figure 3 .1O( a», then
11 = (- EI + VBA)/R I = (-10 - 3.077)/20
= -0.654 A
/2 = (- E 2 + V BA )/R 2 = (- 20 - 3.077)/15
= -1.538 A
/3 = (- E 3 + V BA )/R 3 = (25 - 3.077)/10
= 2.192 A
3.10 Rosen's theorem or the general star-mesh transformation
It is possible to transform a network of N eonduetanees whieh are eon-

neeted to a eommon star point, S, as shown in figure 3.12(a), to a mesh of
eonduetanees whieh are eonneeted between the N terminals as shown in
dia gram (b).
The relationship between the two sets of eonduetanees ean be deter-
mined as follows. Consider the ease where node 1 is eonneeted to ground
(zero potential), and other terminals are energised. Applying Millman's
theorem to figure 3.12(a) gives
n
L ElkGk
= -n---
k~1
VSI
L Gkk~1
and the eurrent entering node 1 is

/ = V G - E 12 G I G 2 + E G G + .. + EwG,G n
13 1 3
1 SI 1 - n
L Gk
k~1
= E G I G2 + E G I G3 +
12 LG 13 LG
This is the same value of eurrent that would flow into node 1 if it were
eonneeted to node 2 by a eonduetanee G I2 = GIG/LG, and to node 3 by a
eonduetanee G 13 = GIG/LG, ete. That is, the star and mesh cireuits in
___"",-_ _z:-N
3
(a) (b)
Figure 3.12 Rosen's theorem or the general star-mesh transformation.

Circuit Theorems 73
diagrams (a) and (b), respectively, of figure 3.12 are equivalent if the
conductance connected between, say, nodes fand g is
GfG
G fg = - n - g-
L Gk
k~l
3.11 The star-ilelta, tee-wye or tee-pi transformation
This is aversion of the general star-mesh transformation, and is restricted

to three elements - see figure 3.13.
2 2
(a) (b)
Figure 3.13 The star-delta and delta-star transformations.
Using the results of the general star-mesh transformation, we see that
G = G;Gj
I] ~G
that is
or
G = G 2 G3 R R
= R2 + R3 + - -
2 3
or R 23 (3.2)
23 ~G R1
G = G I G3 R IR
= R I + R3 + - -
3
or R\3 (3.3)
\3 ~G Rz
3.12 The delta-star, wye-delta or pi-tee transformation
While it is not possible to obtain a general mesh-star transformation, a

delta-star transformation for three elements of the type in figure 3.13 can
be deduced as folIows.
From equation (3.1)
(3.4)
From equation (3.2)
G 23 ( GI + G2 + G3 ) = G 2 G3 (3.5)
and from equation (3.3)
G 13 (G I + G2 + G3 ) = G IG 3 (3.6)
Dividing (3.4) by (3.5) yields
GI = G 3 G 12/G 23
and dividing (3.4) and (3.6) gives
G 2 = G 3 G 12/G 13
Substituting for GI and G 2 in equation (3.3) shows that
R = R 13R 23
3 "LR
where "LR = R I2 + R23 + R 13 • Similarly it may be shown that
R = R I2 R 13
1 "LR
and
R = R 12R 23
2 "LR
3.13 Summary of star-delta and delta-star transformations
For the star-delta transformation
G = Gpj
1J "LG
and for the delta-star transformation

Circuit Theorems 75
R = RjjR jk
1 'LR
where i =1= j and i =1= k.
Unworked problems
3.1. Using the principle of superposition, calculate /1 in figure 3.14.

[-2.667 A]
3.2. Using the principle of superposition, calculate the value of /1> /2 and
/3 in figure 3.15. Calculate the power consumed in the 4 Q resistor,
and show that the principle of superposition does not hold for
power.
[- 2.4 A; 0.553 A; - 2.953 A; 1.22 W]
3.3. Determine Thevenin's equivalent circuit with respect to terminals A
md B for the circuit in figure 3.16. What current would ftow in a
resistor of 10 Q resistance connected between A and B?
[ET = -13 V, RT = 10 Q; 0.65 A (B to A)]
3.4. Determine Thevenin's equivalent circuit with respect to terminals A
and B of figure 3.17. Hence calculate the power which would be
developed in an 8 Q resistot connected between terminals A and B.
[ET . = 1.17 V, RT = 4.44 Q; 70.7 mW]
3.5. Determine Norton's equivalent circuit with respect to terminals A
and B of figure 3.15. A resistance of 8 Q is connected between these
terminals; calculate the current in the resistor.
[IN = 2.833 A, G N = 1.283 S; 0.2515 A]
3.6. Deduce Norton's equivalent circuit with respect to terminals A and
B of figure 3.18 if element Xis (a) a 4 A current source with the
current ftowing upwards, (b) a 10 V voltage source with the positive
pole connected to terminal B.
[Ca) 4 A, 0.5833 S; (b) 5 A, 1.083 S]

3.7. In problems 3.3, 3.4 and 3.5, determine the resistance of the load
resistor connected to terminals A and B which dissipates maximum
power, together with the power in each case.
[10 Q, 4.225 W; 4.44 Q, 0.077 W; 0.779 Q, 1.57 W]
3.8. Calculate the maximum power wh ich may be delivered to a load
connected between A and B in problem 3.6.
[Ca) 6.86 W; (b) 5.77 W]
2Q
Figure 3.14
-
10 V
5V~
Figure 3.15
---
5V
3A
Figure 3.16
...-----oB
Figure 3.17
3.9. For the circuit in figure 3.19, ca1culate 11> 12 , 13 and V BA •

[1.091 A; - 1.455 A; 0.364 A; 4.54 V]
3.10. Convert the star network in figure 3.20 into its equivalent general-
ised mesh network.
Circuit Theorems 77
1-.....- ....--oA
L . . - _ -......- .......---O B
Figure 3.18
-- 10 V
50
A
--10 V
"
100
B
--10 V "
'3
150
Figure 3.19
o A
c B
Figure 3.20
20
A~--'-~-~DB
Figure 3.21
78· Electrical Circuit Analysis and Design
A~B
c~Qc
Figure 3.22
[GAO = 0.0952 S; GAB = 0.0476 S; G AC = 0.119 S;

G BO = 0.019 S; G BC = 0.024 S; Gco = 0.048 S]
3.11. Convert the star-connected set of resistors in figure 3.21 into their
equivalent delta network, and hence calculate the resistance be-
tween A and B.
[1.3 Q]
3.12. Evaluate the equivalent (a) delta network, (b) star network for the
circuit in figure 3.22.
[(a) R Ac = 1.65 Q; R Bc = 4.34 Q; R AB = 11.75 Q;
(b) RAs = 1.09 Q; R Bs = 2.87 Q; R cs = 0.403 Q]
4
Energy Storage Elements
4.1 Introduction
So far, our discussions have covered elements which are either energy
sources or energy dissipators. However, elements such as capacitors and
inductors have the property of being able to store energy, whose V-I
relationships contain either time integrals or derivatives of voltage or
current. As one would suspect, this means that the response of these
elements is not instantaneous.
4.2 Capacitors
A simple capacitor comprises parallel conducting plates separated by a

dielectric. In an ideal capacitor, the charge q stored in the dielectric is
q = Cv
where v is the voltage across the capacitor, and C is the capacitance of the
capacitor in farads (F). It is of interest to point out that the abbreviation for
the unit of charge (the coulomb) is also C; the reader should be careful not
to confuse the symbol for capacitance with that for the unit of charge. The
current, i, wh ich charges the capacitor is
i = dq = C dv
dt dt
Consider tbe circuit in figure 4.1, in which the capacitor has been
connected to a d.c. supply long enough to be fully charge. The differential
relationship for the capacitor at this time is
i = C dv = 3d [1O] = 0A
dt dt
that is, an ideal capacitor is an open-circuit to a d.c. source. Using this
relationship in the circuit in figure 4.2(a), and assuming that the capacitors
79
Figure 4.1 Current in a capacitor in a d. c. circuit.
2 F
(a) (b)
Figure 4.2 Capacitors in a d.c. network.
are fully charged, the circuit can be reduced to that in figure 4.2(b) for the
purpose of the calculation of the steady-state current, I, in the 4 Q resistor.
That is
I = 5 x 6/(4 + 6) = 3A
The voltage waveform, v, applied to the circuit in figure 4.3(b) is described

by
o for t < 0
1.5t for 0 ::::; t < 2
v -
(6 - 1.5t) for 2 ::::; t < 4
o for 4::::; t< 00
and is illustrated in figure 4.3(b). Sketch the waveform of the current
through the capacitor.
Solution
The current in the circuit is

o for t < 0
4.5 for 0 ::::; t < 2
- 4.5 for 2::::; t< 4
o for 4::::; t< 00
and is illustrated in figure 4.3(c).
Energy Storage Elements 81
v
3vi J\
(b)~t(S)
o 2 4
i
4.5
(a)
(C)-+-+--r--t (5)
-4.5
4.3 Energy stored in capacitor
Energy is stored in the electric field of the capacitor, and the instantaneous
energy supplied to a capacitor of capacitance C in time dt is
dW = P dt = vi dt = vC dv dt = Cv dv
dt
The total energy supplied to the capacitor is the time integral of this
expression, as follows
w = fV Cv dv = l.cv 2
o 2
For worked example 4.2.1, sketch to a base of time the graph of energy
stored in the capacitor.
Solution
The energy stored is

o for t < 0
1 3.375t 2 for 0 ~ t < 2
W = -Cv 2 = 1.5v2 =
2 1.5(6 - 1.5ty for 2 ~ t <4
o for 4 ~ t< 00
The resulting graph is shown in figure 4.4.
....,~---r-_.....::::!.,...-t (s)
o 2 4
4.4 Capacitors in parallel
When n ideal capacitors are connected in parallel with one another, each
supports the supply voltage, V s , between its terminals - see figure 4.5. That
is, the charge stored by each capacitor is
QI = CIVs , Q2 = C2V S , Qn = CnVs
and the total charge, Q, stored by the parallel-connected capacitors is
Q = QI + Q2 + ... + Qn = V s ( Cl + C2 + ... + Cn)
If the parallel-connected capacitors are replaced by an equivalent capaci-
tor, CE' then
Q = CEVS = (Cl + C2 + ... + Cn)Vs
That is, the effective capacitance of the parallel circuit is
CE = Cl + C2 + ... + Cn
For a parallel circuit, the effective capacitance is always greater than the
largest value 0/ capacitance in the circuit.
Vs t.L
T
Figure 4.5 Capacitors in parallel.
4.5 Capacitors in series
When capacitors are connected in series with one another (known as a

string of capacitors), the same value of charging current flows through each
capacitor for the same length of time (see figure 4.6). That is, each
capacitor supports the same value 0/ charge, Q, hence
Q = QI = Q2 = ... = Qn
Figure 4.6 Capacitors in series.
where Q = Cl VI = C2 V 2 = ... = Cn V n • Applying KVL to the circuit in

figure 4.6 yields
Vs = VI + V 2 + ... + Vn = -Q + -Q + ... + -Q
Cl C2 Cn
= Q [.1..- + .1..- + ... + .1..-]

Cl C2 Cn
If CE is the effective capacitance of the series circuit, then
V =-
Q
S CE
or
1 1 1 1
-=-+-+ ... +-
CE Cl C2 Cn
For the special case of two capacitors in parallel
C - C t C2
E - Cl + C2
In the case 01 series-connected capacitors, the effective capacitance is always

less than the lowest value 01 capacitance in the circuit.
4.6 Potential division in series-connected capacitors
As stated earlier for series-connected capacitors, the total charge stored by

the string is equal to the charge stored by each capacitor in the string, that
is
or
Figure 4.7 Diagram [ar worked example 4.6.1.
Consequent upon this piece of theory, the capacitor with the smallest value
o[ capacitance supports the largest voltage!
Calculate the effective capacitance of the circuit in figure 4.7 and the value
of Vt.
Solution
The capacitance, Cp , of the parallel section of the circuit is

Cp = 40 + 60 = 100 F
The complete circuit effectively consists of astring of capacitors of 20 F,

50 Fand Cp F. The reciprocal of the effective capacitance of the circuit
therefore is
~ = ~ + ~ + _1_= 0.05 + 0.02 + 0.01

CE 20 50 100
= 0.08 F- t
or
CE = 1/0.08 = 12.5 F
The voltage across the 50 F capacitor can be calculated as follows
V = V CE = 10 12.5 = 2.5 V
n Scn 50
4.7 Inductance
When current flows in a wire, it produces a magnetic field around the wire;
when the wire is wound into a coi! or inductor, the resulting magnetic field
is strengthened. The self-inductance, L henrys (unit symbol H), of the coil

is defined as the ratio of the magnetic Bux, cP weber (unit symbol Wb), to
the current I amperes (also known as the excitation current) which prod-
uces the Bux. That is
cP
L= - H
I
When the current in the inductor changes, the resulting change in magnetic
Bux associated with the circuit produces a self-induced e.m.f, e, in the coil,
that is
di
e = -L - V
dt
where di/dt is the rate of change of current in the circuit. Since electrical
engineers regard an inductor as a passive element rather than a source of
e.m.f., we write the p.d. across the inductor as
di
v = L - V
L dt
That is to say, e and VL act in opposite directions, as shown in figure 4.8.
Let us look for the moment at the circuit in figure 4.9 in which a constant
current of 2 A has been Bowing long enough to allow steady-state operat-
ing conditions to be reached. The volt-ampere relationship for the circuit
at this time is
e = L di = 3 d[2] = 0 V
dt dt
____ •
._--~~~--_L----_
e
Figure 4.8 e.m.f. and p.d. associated with an inductor.
Pure inductor
Figure 4.9 The steady-state self-induced e.m.[. in an ideal inductor in a d.c. circuit.
5F
3Q
10 Q
5H 3Q
2F
v = 15 vt
4Q
(al (bl
Figure 4.10 A more camplex d.c. circuit.
That is, under steady-state conditions in a d.c. circuit, an ideal inductor acts
as though it were a short-circuit.
Looking now at the so me wh at more complex d.c. circuit in figure
4.1O(a) involving both capacitors and inductors, we will calculate the
steady-state value of the current I. For this calculation we replace inductors
by short-circuits and capacitors by open-circuits, leaving the 'steady-state'
d.c. circuit in figure 4. 9(b). Clearly, the steady-state value of I is
I = 15/(10 + 3 + 2) = 1 A
While the above discussion is in order for steady-state d.c. conditions,
there may be other factors operating in the circuit because we have two
types of energy storage elements in the circuit. We will discuss these factors
in chapter 10.
The current in the circuit in figure 4.11(a) is described as follows
(al
(cl -+-+---r-- t (5)
o
-6
Figure 4.11 Diagram for worked example 4.7.1.

o for t < 0
t for 0 ~ t< 3
(6 - t) for 3 ~ t < 6
o for 6 ~ t< 00
Determine the waveshape of the voltage across the inductor.
Solution
The current waveform is shown in figure 4.11(b), and the self-induced

e.m.f. is defined by
o for t < 0
di 6 di 6 for 0 ~ t< 3
L-= - =
dt dt -6 for 3 ~ t< 6
o for 6 ~ t< 00
and is shown in figure 4.11(c).
4.8 Energy stored in an inductor
The instantaneous energy supplied to an inductor is

di
dW = P dt = vLi dt =L dt x i dt = Li di
and the total energy supplied is the time integral of this expression as
follows
W = fI Li di = l.Li 1
o 2
4.9 Inductors in series
For the series circuit in figure 4.12

VI = LI di/dt, V1 = L 1 di/dt, Vn = Ln dildt
Applying KVL to the circuit shows that
Vs = VI + V 1 + ... + V n = (LI + L 1 + ... + Ln) di/dt
If L E is the effective inductance of the circuit, then
Vs = LE di/dt
hence the effective inductance of the circuit is
Figure 4.12 lnductors in series.
LE = LI + L 2 + ... + Ln
That is, the effective inductance 01 aseries circuit is greater than the largest
individual inductance in the circuit.
4.10 Inductors in parallel
Applying KCL to the parallel-connected inductors in figure 4.13 yields

is = il + i2 + . . . +' in
hence
dis di di di
-=-+
l
-+
2
... +-n
dt dt dt dt
and, since the voltage across the circuit is
v = L di l = L di2 =
1 dt 2 dt
that is
or
dis v V
-=-+-+
dt LI L2
Figure 4.13 lnductors in parallel.

If the effective inductance of the circuit is L E , then

dis dis v
v = L - or -=-
E dt dt LE
hence
or
1 1 1 1
-=-+-+ ... +-
LE LI L2 Ln
In the case of parallel-connected inductors, the effective inductance is
always less than the lowest value 01 inductance in the circuit.
For the special case of two inductors in parallel
L _ L IL 2
E - LI + L z
Calculate the effective inductance between terminals A and B in figure

4.14.
Solution
We must begin to analyse this circuit at the point which is most remote
from the input terminals, and then work towards the input. For the parallel
combination of 5 Hand 6 Hinductors
L pl = 5 x 6/(5 + 6) = 2.727 H
The effective inductance of 4 Hand L pl in se ries is

L SI = 4 + 2.727 = 6.727 H
The parallel combination of 3 Hand L SI results in an inductance of
= 3 x 6.727 = 2.075 H
L p2 3 + 6.727
and the effective inductance between A and B is
LE = 2 +L p2 = 4.075 H
4.11 Duality between inductors and capacitors
Two circuits are said to be the dual of one another if the mesh current
equations of one circuit have the same mathematical form as the node
voltage equations of the other.
The differential equations for inductors and capacitors are
Inductors Capacitors
di . C -dv
v = L- 1=
dt dt
1 ft 1 ft
i =- v dt v=- idt
L 0 C 0
That is, inductors and capacitors are dual quantities. An inductor of 3 H is
the exact dual of a 3 F capacitor (see also seetion 2.12).
4.12 Relationship between inductance, reluctance and the number of

turns on a coil
For a magnetic circuit having a constant reluctance, that is, an air-co re
L = NiP
I
Now, the magnetic flux produced is
NI NI
iP = BA = IlHA = 1l- A = -
I S
where B is the magnetic flux density, A is the cross-sectional area of the
magnetic circuit, His the magnetic field intensity, N is the number of turns
on the coil, I is the current in the coil, II is the permeability of the magnetic
path, I is the length of the magnetic circuit and S is the reluctance of the
magnetic path. Hence
L = ~/S
That is, if the reluctance of the magnetic path is constant then doubling the
number of turns on the coil quadrupies the inductance.
~
M
_ _ _ _ _ _~A~_ _ _ _~
- v, ---
v,
Figure 4.15 Magnetically coupled circuits; 11>1 is the ftux leaving coill and 11>2 is the ftux
reaching coil 2.
4.13 Mutual inductance
When current ftows in a coil, a magnetic ftux is established and energy is

stored in the magnetic field. If two coils are wound on the same magnetic
circuit (see figure 4.15) and current ftows, say, in coil1, then some of the
magnetic ftux produced by that coillinks with coil 2. If the current in coil 1
is altered, the ftux entering coil 2 also changes, resulting in a mutually-
induced e.m.f., V 2 , in coil 2. If M 12 is the mutual inductance existing
between the coils, then
M 12 = lP2/i 1
where lP2 is the flux reaching coil2. Similarly, if coil 2 is excited and coil 1 is
the magnetically coupled coil, then the mutual inductance M21 between coil
2 and coil 1 is
M 21 = lP/i2
where i2 is the current in coil 2 and lP 1 is the magnetic ftux entering coil 1.
Since the two coils are wo und on the same magnetic circuit
M 12 = M 21 = M
where M is the mutual inductance between the two coils.
When only one of the coils is excited, that coil is known as the primary
winding, and the coil which has the mutually induced e.m.f. in it is known
as the secondary winding. The two circuits are described as magnetically
coupled circuits.
4.14 Direction of the mutually induced e.m.f. - the dot notation
In figure 4.15, the current i 1 is produced by the applied voltage VI> which
has the polarity shown. The direction of the current induced in coil 2 is
deduced as folIows.
Lenz's law states that the direction of the induced e.m.f. (whether self
or mutually induced) opposes the change producing it. Consequently, the
current induced in coil 2 must produce a magnetic flux which opposes the
flux developed by coill. Since, in figure 4.15, the flux enters coil2 from
the left, the current induced in coil 2 must produce a magnetic flux which
leaves the left-hand side of coil 2. The result is that the mutually induced
e.m.f., V 2 , has the polarity shown in the figure.
Engineers have developed the concept of the dot notation which allows
us to specify ends of coils having similar polarity as folIows. One end (any
end) of the primary coil is marked with a dot, and the end of each
secondary coil (there may be several of these) having the same instan-
taneous polarity as the 'dotted' end of the primary coil is also marked with
a dot.
Thus, in figure 4.15, if we mark the left-hand end of the left-hand coil
with a dot, then we must mark the left-hand end of the right-hand coil with
a dot since they both have the same instantaneous polarity (one polarity
arising from the forcing voltage, and the other being the induced polarity).
If we apply this reasoning to the coupled circuit in figure 4.16(a), we will
quickly be able to deduce a set of mesh equations to solve the circuit.
Lenz's law enables us to say that, for each mutually coupled coil, there is a
mutually induced e.m.f. in that coil. Since each winding carries a current,
there is a mutually induced e.m.f. (a current-controlled voltage source) in
each coupled coil. Thus we can separate the two magnetically coupled
circuits in figure 4.16(a) into two separate circuits, as shown in diagram (b).
We now investigate a method of deducing the direction of the mutually
induced e.m.f.s. When drawing diagram (b), we transfer a copy of the 'dots'
Mdi,-
dt
t t di,
M-
dt
(a) (b)
Figure 4.16 (a) Coupled circuits energised by independent sourees, (b) an equivalent
circuit.
from diagram (a), and insert adependent voltage source in each winding
for each mutually coupled coil (there is one such voltage source Jor each
mutually coupled coi/). The possibility always exists for the introduction of
errors at this point; from experience, the author suggests that the depen-
dent voltage source(s) should be drawn at the opposite end to the dot on
the coil- see figure 4.15(b). The value of the mutually induced e.m.f. is
M din/dt
where in is, in this case, either i1 or i2 • At this point the reader should note
that a mutually induced e.m.f. only exists so long as the current is changing
in the exciting coil.
Since we have assumed that i 1 enters the dotted end of LI, that is, the
dotted end of LI is assumed to be connected to the positive pole of the
forcing voltage, the magnitude of the mutually induced e.m.f. in coil 2 is
M di/dt, and this acts to make the dotted end of coil 2 positive.
By applying a similar reasoning to the mutually induced e.m.f. in coil1,
the reader will confirm that the direction of the e.m.f. is as shown in figure
4.16(b), and that its value is M di/dt. In addition to the mutually induced
e.m.f. in each coil, there is also a self induced e.m.f. of value Ln din/dt in
each coil. The mesh equations for the two circuit are therefore
VI = L di 1 + M di2
1 dt dt
V = M di 1 + L di 2
2 dt 2 dt
4.15 Coefficient of coupling
Suppose that coil1 (of inductance LI) in figure 4.15 produces ftux <PI' and
that a proportion k<P 1 links with coil2 (of inductance L 2 ). The parameter k
is known as the magnetic coupling coefficient, where 0 :::::; k :::::; 1. Also, if a
current ftows in L 2 and produced ftux <P2 , then a ftux k<P 2 links with LI.
Now
Multipyling the two equations gives
M2 __ N 1k<P 2 N 2k<P 1 __ k 2 NI <PI N 2 <P2

= k 2 L 1L 2
i2 i1 i1 i2
Coil2
M
.~.
~
---
M~
dt
M~
dt
(al
- v
(bI
Figure 4.17 Series-connected mutually coupled coils.
4.16 Mutually coupled coils in series and in parallel
Series-connected coils
Using the work on the dot notation, the circuit in figure 4.17(a) can be
re-drawn as shown in figure 4.17(b), the equation for wh ich is
di di di di
v = L I- + M-+ L z - + M -
dt dt dt dt
di
= (LI + L z + 2M) -
dt
That is, the effective inductance of the circuit is LI + L z + 2M H. In the
connection shown, the coils are said to be series-aiding since the flux from
one coil assists or aids the flux produced by the second coil.
If the coils are re-connected so that the flux produced by the coils
oppose one another, they are said to be series-opposing, and the effective
inductance of the circuit is LI + L z - 2M H.
Parallel-connected coils
Consider the two parallel-aiding magnetically coupled coils in figure

4.18(a). The corresponding equivalent circuit is drawn in diagram (b) (the
reader will find it an interesting exercise to verify the circuit), and the
corresponding equations for the circuits are
di l diz
v=LI-+M-
dt dt
di l diz
v=M-+L -
dt z dt
di,
L, dt
vt vf di,
M- M~
dt dt
(a) (b)
Figure 4.18 Parallel-connected mutually coupled coi/s.
Eliminating dizfdt between the two equations gives
or
L L 2 - M 2 di
j di j j
v = - - - - - = L'-
L 2 - M dt dt j
where L; is the equivalent inductance of the branch containing L j • Simi-

lady, eliminating dizldt between the equations gives
where L~ is the effective inductance in the branch containing L 2 • Hence
and
For example, if L j = 1 H, L 2 = 4 Hand M = 0.2 H, then

L; = (1 x 4 - 0.2 2 )/(4 - 0.2) = 1.042 H
and
L~ = (1 x 4 - 0.2 2 )/(1 - 0.2) = 4.95 H
The effective inductance, L E , of the complete circuit is
L'L' 1.042 X 4.95
L = j 2
= 0.861 H
E L' j
+ L'2 1.042 + 4.95
The reader will observe that when L 2 = M then L; ~ 00, and when L = M j
then L; ~ 00. If, for example, in the above calculation M = 1 H, then

L; = 1 H, L~ = 00 and L E = 1 H.
Unworked problems
4.1. A pulse of direct current of amplitude 30 mA is applied for 10 ms to

an ideal capacitor of 5 IlF capacitance. What is the change of voltage
between its terminals at the end of the pulse?
[60 V]
4.2. A practical capacitor can be represented by an ideal capacitor
shunted by a leakage resistor. If a 5 IlF capacitor has a leakage
resistance of 1 MQ, determine an expression for (a) the current
through the leakage resistor, (b) the current through the 'ideal'
capacitor and (c) the energy stored in the capacitor when a voltage
of 200 sin 120:n:t V is applied.
[(a) 2 x 10- 4 sin 120:n:t A; (b) 0.12:n: cos 120:n:t A;
(c) 0.1 sin 2 120:n:t J]
4.3. The operation al amplifier circuit in figure 4.19(a) has the equivalent
circuit in diagram (b). If VA is a voltage which rises steadily from
zero volts at the rate of 1 V/s for 1 s, and then falls at the rate of 1 V/s
until it reaches zero, thereafter remaining at zero, deduce the wave-
form for VB. The voltage gain, A of the op-amp is very large, and it
may be assumed that the op-amp does not saturate.
[VB = - 1 V for 1 s, when it becomes 1 V for 1 s, remaining at zero
thereafter. ]
(a) (b)
Figure 4.19
4.4. The voltage across a 20 IlF capacitor during the time interval
o ~ t ~ 3.333 s is given by v(t) = 50t 2 (5 - t) V. Deduce an expres-
sion for i(t), and calculate the maximum current and the instant at
which it occurs.
[t(lO - 3t) mA; 8.333 mA; 1.6667 s]
Figure 4.20
Figure 4.21
4.5. Calculate the effective capacitance between terminals A and B in

figure 4.20 when C and 0 are (a) open-circuited, (b) short-circuited.
[(a) 6.275 IlF; (b) 6.5 IlF]
4.6. Figure 4.21 is a simplified diagram of astring of suspension insula-
tors. If the maximum voltage per unit (unit capacitance C) is 18 kV,
calculate the maximum line voltage.
[33 kV]
4.7. A current i(t) = 100(2 + 10 sin 100 Jtt) A ftows in a pure inductor of
0.1 H inductance during the time interval 0 :s:; t :s:; 0.02 s. Calculate
the maximum voltage across the inductor.
[104 Jt V]
4.8. If, in figure 4.22(a), (i) element A is a pure inductor and element B
is a pure resistor, (ii) element A is a pure resistor and element B is a
pure inductor, deduce an expression for the voltage vu(t) for the case
where the voltage gain, A, of the op-amp (see diagram (b» is
infinity.
[(i) - ~J VA dt; (ii) - ~ d;tA ]

B B
VA t ! VB
VAt V;t
(a) (b)
Figure 4.22
2H
A C
Figure 4.23
__ D <---
3
( -p <_D
2" 4
Figure 4.24
4.9. A solenoid 1 m in length and 15 cm diameter has 4000 turns of wire

on it; the coil has an air core (110 = 43t x 10- 7 Hirn). Calculate (a) its
approximate inductance and (b) the energy stored when a current of
3 A ftows in the coil.
[(a) 0.355 H; (b) 1.6 J]
4.10. Calculate the equivalent inductance between terminals A and B of
figure 4.23 when C and D are (a) open-circuited, (b) short-circuited.
[(a) 0.774 H; (b) 0.758 H]
4H
4H~8H
• •
'--'"
• ~.
5H 6H
Figure 4.25
4.11. What are the possible alternative locations for the dot notation for
the coupled circuit in figure 4.24?
[1 and 4 or 2 and 3]
4.12. Calculate the effective inductance of the circuit in figure 4.25.
[12 H]
5
The Sinewave, Phasors
and Power
5.1 Introduction
Advantages of an alternating supply over a direct current supply include

not only the simplicity by which it can be 'transformed' or altered in
magnitude, but also the way in which it can be transmitted and controlled.
When a given amount of power is transmitted over a great distance from
a generating source to a load, the voltage is transformed to a high value so
that the current is correspondingly low. The net result is that the power loss
(PR) in the transmission system is redueed when eompared with the case
when the power is transmitted at a lower voltage (and a higher current). At
the receiving end, the voltage is transformed to a lower value before being
conneeted to the consumer.
While there is an infinite variety of alternating waveshapes, such as,
reetangular , triangular , ete., the sine and eosine waves are the best for
eleetrieal power transmission.
When describing alternating quantities, the letters 'a.e.' should simply
be interpreted as meaning 'alternating'. For example 'a.c. eurrent' means
'alternating eurrent', 'a.c. voltage' means 'alternating voltage' , etc.
5.2 Mean or average value of an alternating quantity
When the alternating voltage waveform in figure 5.1(a) is applied to a

resistanee, R ohms, by Ohm's law the current is proportional to the
voltage. The eurrent waveform therefore has the sinusoidal waveform in
diagram (b). The instantaneous voltage at time t is
v = V m sin wt
100
The Sinewave, Phasors and Power 101
(a) 0 II------>,------/------l..-
(b) 0 f----->r-------r------->-
I~R
(c) 0 "'------~-----""-'<.------"'-
Figure 5.1 Sinusoidal (a) voltage and (b) current waveforms associated with a resistive
element. (c) Power waveform in a resistive element.
where w is the angular frequency in radis. The current, from Ohm's law is
. v rnV •
l = - = -- sm wt
R R
Since both the voltage and the current waves are sinusoidal, the area under
the positive half-cycle is equal to the area under the negative half-cycle.
That is, the mathematical average value of both of them over one complete
cycle is zero!
In cases such as the sinewave, electrical engineers take a special view-
point about the average or mean value of the wave. It is, in fact, taken to
be the average value of the rectified wave; that is, the negative part of the
wave is inverted so that it appears in the positive half. The average value of
the sinewave is then taken to be the average value of the rectified sinewave
over the complete cycle; since both positive and 'negative' half-cycles in
this case have the same shape, we may either calculate the average value
over the complete cyc1e, or over the first half-cyc1e! We will do the latter. If
a = A sin wt, then the average value (A av ) or mean value (A mean ) is
1 fT'2
A av or A mean = -- A sin wt dt
TI2 0
where T (= 2nlw) is the periodic time of the wave, hence

A
A av or A mean = - -
[1 cos
- - wt
]T12
TI2 w 0
A [1 2n ]TI2 A[ 2n ]T12
- T/2 2nlT cos T
t 0 = - -; cos t 0 T
= 2Aln = 0.637A
The corresponding value for a sinewave of voltage and current, respect-

ively, are
Vav = 2Vmln = 0.637 V m
I av = 2Im ln = 0.637 Im
The frequency of the wave in hertz (Hz) is given by the number of cyc1es
per second, and is
f = lITHz
The corresponding angular frequency or radian frequency is

w = 2nf radis
5.3 The etfective value or r .m.s. value of a periodic wave
The effective value of an alternating current is equal to the value of direct

current which, when flowing in a resistor, delivers the same power to the
resistor. The waveform of the instantaneous power in a resistor when a
sinusoidal voltage is applied is shown in figure 5.1(c). The instantaneous
power consumed by a resistor R is
p = i 2R W
and the average power consumed during one cyc1e is
P =-
1 fT
i 2R dt = -
RfT F dt
ToT 0
The power delivered by the corresponding direct current, I eff , is

p = RI~ff
henee
and similarly
The values I eff and V eff , respeetively, are the root-mean-square value (or
r.m.s. value, meaning the square root of the mean of the sum of the
instantaneous square values) of the eurrent and the voltage waves. When
dealing with periodie waves, the subseript 'eff' is usually dropped, and the
symbols I and V are taken to me an the r.m.s. or effeetive value of the
eurrent and voltage wave, respeetively.
r.m.s. value 01 a sinewave

The instantaneous eurrent in a sinewave is
i = Im sin wt
whieh has a periodie time of T = 2rrJw. The r.m.s. value of the wave is
I = V[~ [ I~ sin 2 wt dt]
= Im V[2: [ ~ (1 - eos 2wt) dt]
Similarly, for a voltage sinusoid

V = V m tv'2 = 0.707 V m
5.4 Phase angle
The general equation for a sinusoidal wave is

a = A sin (wt + 8)
whieh includes aphase angle, 8, in its argument (wt + 8). A sinewave ean
be thought of as the plot of the vertieal displaeement of a line of length A
whieh rotates in a counterclockwise direction at a eonstant speed w radIs
(see figure 5.2). If the radialline has moved through angle 8 at t = 0, then
w radIs
// I1[' --- / '{

~ .
_ _ A - /" -... A sm wt
I ~ \ I
I 8 \ / \
I I
I 0 7[\
\
/ \
\
',//
.......... _-,./
A sin (wt + 8)
(a) (b)
Figure 5.2 The sinewaves A sin (wt + 0) and A sin wt.
we get the waveform shown in full li ne in diagram (b); this corresponds to

the general sinusoid a = A sin (wt + 8). If the radialline is horizontal at t =
0, that is, 8 = 0, we get the sinewave shown in the broken li ne in diagram
(b); this corresponds to the curve a = A sin wt. The two sinewaves in figure
5.2 are said to be out of phase with one another.
The sinewave A sin wt can be though of as a reference sinewave, since its
value is zero when t = 0, and increases when t is greater than zero. That is,
a sinewave represents the displacement of the tip of a line of length A,
wh ich rotates in a counterclockwise direction. When the line is horizon-
tal, engineers say that it is in the reference direction.
Since sinewaves are drawn out by the tip of a line rotating in a
counterclockwise direction, then the line which traces the curve A sin (wt
+ 8) passes a given position (say the horizontal) before the li ne which
traces the curve A sin wt. That is, the waveform A sin (wt + 8) leads the
waveform A sin wt by 8, the phase angle being measured relative to the
reference direction.
In electrical engineering it is customary (but not mandatory) to give
the phase angle in degrees, and no confusion should arise if we express a
current in any of the following ways
I = 10 sin (wt - n13) A
10 sin (wt - 60°) A

10 sin (2n x 100t - n13)A
10 sin (628.3t - 60°) A
In the case of the latter two expressions, the supply frequency is 100 Hz (or
628.3 radis).
L
0.9
Figure 5.3 Phasor representation o[ 1.273 sin (wt + 8).
S.S Phasors and phasor diagrams
At a given frequency, a sinusoidal quantity is characterised by two para-

meters, namely its magnitude or amplitude and its phase angle. In the UK
we usually me an the Lm.S. value of a wave when we refer to the magnitude
(in so me US texts it may mean the maximum value). We can therefore
describe the waveform 1.273 sin (wt + 8) as the phasor quantity
0.9L8
where the value 0.9 represents the Lm.S. value of the sinewave of maxi-
mum value 0.9\12 = 1.273, and we can represent it as shown in figure 5.3.
Thus the voltage v = 141.4 sin (wt + lt/4) V can be represented as
i = Im sin (wt - 3D")
B,
-I j
O~~--~~---r----7-----+-wt
A,
----
B,
0<
5n/6
(a)
(b)
~
60°
30° 3D"
I
V
(c) (d)
Figure 5.4 Phasor representation o[ a voltage and a current.

100Ln/4 V or as 100L45° V
Suppose that we represent the current and voltage waveforrns in figure
5.4(b) by the pair of rotating lines in diagram (a). If we choose to 'freeze'
or 'fix' the rotating lines at wt = 0, the voltage is represented by line AI and
the current by BI. Since these lines represent the maximum value of the
respective quantities, the phasor diagram at t = 0 is obtained by scaling the
length of each line by a factor of 1/Y2 to give the corresponding r.m.s.
phasors, shown in diagram (c).
Alternatively, we can select so me other angle at wh ich to 'freeze' or
look at the phasors. If we choose wt = 5n/6 rad or 150°, we get the phasor
diagram in figure 5.4(d). The reader will note that in both cases I leads V
by 90° (or V lags behind I by 90°).
5.6 Representation of a sinusoidal quantity as a complex value
Sinusoidal quantities are time-varying and, strictly speaking, we cannot

state the value of, say, a current or voltage that will be correct for all values
of time. Moreover, we need a technique which will allow us to apply laws
and theorems to alternating current circuits. We therefore introduce a
technique here, namely the complex value, which is of vital importance to
electrical and electronic engineers. The concept of complex values is,
without doubt, the most powerful tool in the armoury of the engineer.
The time-domain sinusoidal current
i(t) = Im sin (wt + <jJ)
is expressed as the imaginary part of a complex quantity (see chapter 15
for information on complex numbers) by Euler's identity
i(t) = Im(Im ej(wl + <1»)
We represent the current as a complex quantity by dropping the imaginary

(Im) wording from the expression, and a further simplification is achieved
by suppressing the factor eiw1 as folIows.
I = Im ei<l>
Since phasors are complex quantities, that is, they embody magnitude and
phase angle, they are printed in boldface type.
In addition, the maximum value is converted to its r.m.s. value, I, and
the whole is written in its polar form as follows
1= IL<jJ
The quantity i(t) varies sinusoidally, and contains amplitude and phase
information as a function of time and, accordingly, it is known as the
time-domain representation of the current. The phasor representation, I, is

the frequency-domain representation of the current, and does not explicitly
include the frequency. (The 'frequency' aspect of the frequency-domain
representation is emphasised by the absence of frequency in the ex-
pression!)
Similarly for voltages, we may represent the time-domain sinusoidal
voltage
v(t) = V rn sin (rot + 8)
as the complex phasor quantity
V= VL8
where V is the Lm.S. value of the voltage.
Transform the current i(t) 150 sin (300t - rr/6) A into its frequency-
domain complex polar value.
Solution
The general steps to be followed are:

1. Convert the maximum amplitude to its Lm.S. value.
2. If necessary, convert the phase angle to degrees.
3. Write down the frequency-domain representation.
Since rr/6 rad = 30°, then
1= ~~ L-30° = 106.07 L-30° A
To convert from the frequency-domain representation to the time-domain

representation, it is merely necessary to reverse the steps. If the frequency
is unknown, then it is simply necessary to write rot where 30üt appears in
the equation in the worked example.
5.7 Impedance of elements
The impedance, Z, of an electrical element is its total opposition to flow of

(sinusoidal) current. If a complex voltage V gives rise to the complex
current I in an element then, by Ohm's law, the impedance of the element
is
V
Z =- Q
I
5.7.1 Resistance
A resistance does not store energy, and a change in the voltage across the
resistor gives rise to an instantaneous change in current in the resistor.
That is the sinusoidal waveform of the voltage across the resistor is in phase
with the current in the resistor. The impedance of the resistor is, therefore
v = RL0 = R + jO Q
0
Calculate the complex polar expression for the current in a 5 Q resistor

when a voltage of v(t) = 10 sin (50t - 50°) V is applied to it.
Solution
The frequency-domain representation of the voltage is
V = ~2 L-50° = 7.071L-50° V
hence
1 = 7.071 L-50° = 1.414L-50° A

5
5.7.2 Pure inductance
Since an inductor stores energy in its magnetic field, a change in current

through the inductor causes the stored energy to change and, when this has
occurred, the voltage across the inductor changes. That is, the current
through the inductor and the voltage across it are not in phase with one
another. The phase relationship between the voltage and current associ-
ated with the inductor can be deduced as folIows.
The v-i relationship for an inductor is VL = L di/dt, and if the current is
of the form i = Im sin rot, then
di d(Im sin rot)
VL =L - =L = roLIm cos rot
dt dt
= roLIm cos rot = roLIm sin (rot + 90°)
= V m sin(rot + 90°)
where V m is the maximum voltage across the inductor.

Hence
Vrn = wL/rn = XL/rn
where XL = wL is known as the inductive reactance at frequency w.
Moreover, since vL has a cosine waveform, it leads the current waveform by
90°. Engineers te nd to think of the voltage across the inductor as being in
the reference direction, and we prefer to say that the current in the inductor
lags behind the voltage across it by 90°.
The complex expression for the impedance of the inductor is, therefore
VL VL
Z = - I= /
L - L90° = J'wL = J"XL
A current of i(t) = 14.14 sin(100t+ 15°) A ftows in a 0.5 Hinductor.

Determine the phasor voltage across the inductor.
Solution
The polar complex current in the inductor is

14.14
1= - - L15° = lOL15° A
v'2
and XL = wL = 100 x 0.5 = 50 Q
hence
From Ohm's law
v = IZ = 10L15°
L x 50L90° = 500L105° V
5.7.3 Pure capacitance
Since a capacitor stores energy in its electric field, a change in the voltage
between the capacitor terminals causes the stored energy to change; when
the energy changes, the current ftowing through the capacitor changes.
That is, the voltage between the capacitor terminals and the current
ftowing through it are not in phase with one another. The phase re la-
tionship between the voltage and the current can be deduced as folIows.
The v-i relationship for a capacitor is i = C dvc/dt, and if the capacitor
voltage is vc = V rn sin wt, then
i = C d(Vm sin wt)/dt = wCVm COS wt

= wCVm sin (wt + 90°) = Im sin (wt + 90°)
where Im = wCVm, and is the maximum current in the capacitor. From
Ohm's law
Vm Vm
1=-
m Xc 1/wC
where Xc is the reactance of the capacitor at frequency w. That is
X =_1_
c wC
Morever , since the current has a eosine form, it leads the voltage across the
capacitor by 90° (alternatively, we may say that the voltage lags behind the
current by 90°).
The impedance of the capacitor is given by Ohm's law as follows
Z = V = V L-900 = _ J·X = XL-900 = _1_

c I l e e jwC
The manipulation of complex numbers is fully explained in chapter 15.
If a voltage 3 L - 60° V is applied to an ideal 2 IlF capacitor at a frequency

of 60 radis, calculate the current flowing through the capacitor.
Solution
The capacitive reactance is

Xc = 1/wC = 1/(60 X 2 x 10- 6 ) = 8.333 kQ
and the current in the capacitor is
V 3L-60°
I = - = = 0.36L30° mA
Zc 8.333 x 103 L - 90°
5.7.4 CIVIL - an a.c. mnemonic
In the above context, the mnemonic CIVIL is very useful, as follows

in C. I leads V
~I
CIVIL
T
V leads I in L
5.8 The susceptance of elements
The susceptance of an element is the reciprocal of its reactance, That is

B = 1/X
For an inductance
and for a capacitance

1
= wC
1/wC
5.9 The admittance of elements
The admittance, Y, of an element is the reciprocal of its complex im-

pedance; since it is a complex value, it is written in hold type/ace. That is
Y =~ = G + jB
Z
where G is the conductance and B is admittance.
For a resistance
Y =_1_=!+J'O=G +J'B
R R + jO R R R
That is, G R = 1/R (as it is in d,c. circuits) and B R = 0, The process of

division of complex numbers is fully described in chapter 15,
For an inductance
Y = 1 = 0 + _1_ = 0 - J'/X = G - J' B

L 0
+ J'XL J
·x L
L L L
That is G L = 0 and B L = lIX = lIwL.

L
For a capacitance
1
y =0
c . - j/Xe
=0+ j/Xe = Ge + jBe
or Ge = 0 and Be = lIXe = wc.

The corresponding representation of Ohm's law is
I = VY = V( G ± jB)
5.10 The impedance of elements in series
If n elements Zl' Zz, ... , Zn are connected in series, and current I flows
through them, by KVL
Vs = Z.I + ZJ + ... + ZJ = I(ZI + Zz + ... + Zn)
where Vs is the supply voltage. The effective complex impedance of the
circuit is
n
ZE = Zl + Zz + ... + Zn = L Zk Q
and the complex voltage Vn across the nth element is
A voltage of 10L20° V is applied to aseries circuit containing the foHowing

elements
Zl = lOL45° Q, Zz = 10LO° Q, Z3 = 15L- 80° Q
Calculate the current in the circuit and the voltage across Zz.
Solution
This example illustrates the advantage of having complex numbers in

rectangular form when adding (or subtracting) complex values, and of
having them in polar form when multiplying (or dividing) complex values.
In order to add complex impedance together, it is necessary to convert
them to their rectangular complex values; this is described fuHy in
chapter 15.
Z, = lOL45° = 7.071 + j7.071 Q
Z2 = lOLO° = 10 + jO Q
Z, = 15L-80° = 2.605 - j14.772 Q
and ZE =~ Z = 19.676 - j7.071 = 21.13 L - 21.37° Q
From Ohm's law
V lOL20°
1= - = = 0.473L41.37°A
ZE 21.13 L - 21.37°
and the voltage across Z2 is
v = V _Z_2 = 10 L200 _ _ lO_L_O_O_ _

2 s ZE 21.13 L - 21.37°
= 4.73L41.37° V
The reader should note that, since Z2 is a resistor, the voltage, V2 , across it
is in phase with the current.
5.11 The admittance and impedance of elements in parallel
If n admittances Y" Y 2 , ••• , Yn are connected to the voltage V., then by

KCL
I s = VsY, + V Y + ... + VsY n =
S 2 Vs(Y, + Y + ... + 2 Yn)
where I s is the supply current. The effective admittance of the parallel
circuit is
n
Y E = Y, + Y2 + ... + Yn L Yk
k~'
and the corresponding effective impedance, ZE' is calculated from

1 1 1 1 1
= - = - + -+ ... +- = L-
n
YE
ZE Z, Z2 Zn k~1 Zk
The current, In, in the nth branch is
Yn ZE
In = I s - = Is
YE Zn
In the special case of a two-branch parallel circuit
YE = Y, +Y 2
and
Z.Zl
ZE =
Z. + Zl
The current in each of the two branches is
Y. Zl
I. = I. = I. ---=--
Y. +Y 1 Z. + Zl
Y1 Z.
11 = I. = I. ---
Y. +Y 1 Z. + Zl
Impedances of lOL20° g, lSLO° g and (17.32 - jlO) gare connected in

parallel to a 100L30° V a.c. supply.
Calculate (a) the effective admittance and impedance of the circuit, (b)
the current drawn by the circuit and (c) the current in the branch contain-
ing the capacitor.
Solution
(a) We can solve the problem using either admittances or impedances.

Selecting the former gives
1 = 0.lL-20 = 0.094 0
- jO.0342 S
lOL20°
1
Y2 = -l-S-L-O-o = 0.0667 - jO S
1 1
Y = = 0.OSL30°
3 17.32 - jlO 20L-30°
= 0.0433 + jO.02S S
and YE = ~Y = 0.204 - jO.OO92 = 0.2042L-2.S8° S
hence
ZE = lIYE = 4.9L2.S8° = 4.89 + jO.22 g
(b) From Ohm's law
I s = VSYE = 100L30° + 0.2042L-2.S8° = 20.42L27.42° V
(c) Since the third branch has a negative value of reactance, it must contain
the capacitance. From the theory outlined in section S.10
Y O.05L30°
13 = I s _ 3 = 20.42L27.420-----
YE O.2042L-2.58°
= 5L60° A
Alternatively, 13 could have been worked out as follows
13 = VS /Z 3 = lOOL30oI20L-30° = 5L60° A
5.12 Impedance and admittance of series-parallel circuits
The impedance (or the admittance) of an a.c. series-parallel circuit is

worked out in a similar manner to that for an otherwise similar circuit
containing pure resistances (or conductances). The only exception is that it
is necessary to not only add the camp/ex impedances of series elements, but
also add the camp/ex admittances of parallel elements.
5.13 Power and power factor
If the voltage across an element is

v(t) = V m sin (wt + (}I)
and the current through the element is

i(t) = Im sin (wt + (}2)
then the instantaneous power consumed by the element is

pet) = [Vm sin (wt + (}1)][Im sin (wt + (}2)]
= V mImsin (wt + (}I) . sin (wt + (}2)
The trigonometric identity

sin A . sin B = ~ [cos (A - B) - cos (A + B)]
leads to the following expression for the instantaneous power consumed by
the element
pet) = ! V mIm[cos «(}I - (}2) - COS (2wt + (}I + (}2)]
The average power consumed is obtained by integrating this expression

over one cycle. Since the second term in the expression is a sinusoid of
frequency 2w, its average value is zero, and it may be ignored. The average
power, P, is therefore
P= ivmIm cos «(}I - (}2) = VI cos </> W
2::Ja-
P = VI cas <p
VI"o_
Figure 5.5 The power triangle.
where V and I are the r.m.s. value of the voltage across and the current in
the element, respectively. The angle cp is the phase angle between the
voltage and the current.
The situation in an a.c. circuit is illustrated by the power triangle in
figure 5.5. The average power (or actual power), P, absorbed by the circuit
or element is
P = VI cos cp W
As measured on an ammeter and a voltmeter, the apparent power
absorbed by the circuit is
S = VI volt-amperes (VA)
In the above equation for power, cos cp is known as the power Jactor of the
circuit, where
P power
power factor = cos cp = - = -----
VI volt amperes
power
apparent power
Since cos cp has a positive mathematical sign for either +cp or -cp, it is
customary to define the power factor as follows. In engineering it is usual
to think of the voltage as being the reference phasor, and to state whether I
lags or leads V. If I lags behind V, we say that the circuit has a lagging
power Jactor and, if I leads V we say that the circuit has a leading power
Jactor.
The third side of the power triangle is the reactive VA (V Ar) or reactive
power, Q, as follows
Q = VI sin cp volt-amperes reactive (VAr)
and, finally
5.14 Power, VA and VAr absorbed by ideal elements
In the following we will assume that the r.m.s. voltage across an element is
V, the current in it is I, and the phase angle quoted assume that V lies in the
reference direction.
Resistors
In this case (jJ = 0°, hence

PR = VI cos 0° = VI
SR = VI VA
QR = VI sin 0° = 0
or PR = SR' and an ideal resistor does not consume reactive VA. Also
V = IR so that
PR = VI = V x V/R = V 2/R
and
PR = VI = IR x R = PR
Inductors
In this case (jJ = 90° (lagging) so that

PL = VIcos 90° = 0
SL = VI
QL = VI sin 90° = VI(lagging)
Also V = IXL so that
and
SL = VI = IXL x I = PXL
hence SL = Qu and an ideal inductor does not absorb power.
Capacitors
In a capacitor (jJ = 90° (leading), and

Pe = VIcos 90° = 0
Se = VI
Qc = VI sin 90° = VI(leading)

bence Sc = Qc, and an ideal capacitor does not absorb power. Also V = IXc
so tb at
and
Sc = VI = IXc X 1= J2Xc
5.15 v-i waveforms
From work in earlier sections we know tbat, in a resistor, tbe voltage and
current are in pb ase witb one anotber. Tbese waveforms, togetber witb
tbat of tbe vi product are sbown in diagrams (a) and (b), respectively, of
figure 5.6. Tbe instantaneous value of tbe vi product wave is eitber positive
or zero at all times, and tbe average power consumed is finite. Tbe reader
will note tbat tbe v-i wave is a sinusoid of twice tbe supply frequency (as it
is for tbe otber cases considered bere).
In tbe case of a pure inductor, tbe current lags bebind tbe voltage by 90°
as sbown in figure 5.6(c). Tbe corresponding v-i product wave is drawn in
diagram (d), and its average value over tbe complete cycle is zero! Tbis
confirms tbe work in section 5.14, wbere we sbowed tbat tbe average
power consumed by an inductor is zero. Tbe pbysical reason is tb at tbe
energy absorbed by tbe inductor during tbe period wben tbe magnetic field
is being built up is returned to tbe supply wben tbe magnetic field collapses.
Similarly for a pure capacitor, tbe average power consumed is zero (see
diagrams (e) and (f) of figure 5.6).
Wben tbere is a pbase difference of <p between tbe v and i waveforms, as
sbown in figure 5.7 (a), tbe corresponding v-i product wave is as sbown in
diagram (b). In tbis case, more energy is consumed by tbe circuit tban is
returned to tbe supply, resulting in a net average power consumption by
tbe circuit, wbicb is
P = VI cos <p W
- see also section 5.13.
5.16 Power consumed in an a.c. circuit
We will use tbe information in tbe previous sections to evaluate tbe

apparent power, tbe power and tbe reactive power absorbed in eacb
element in figure 5.8.
Tbe impedance of tbe parallel section of tbe circuit is
(a)
1- T - - . - . l a1
VV~i a~sorbed
Y/A //;/A /
1 2 4 - - . IEnergy
by reslstor
(b) ~~~L-~~~~~ __L-~~L-__ t
(c)
1- T ~I
~ 'o"gy "",,d by L
(d)
\SJ~ t
Energy released by L
v
(e)
~ ~ Energy stored by C
(f)
'&W Ene:gy released by C
Figure 5.6 v, i and p wave[orms (a) and (b) tor a pure resistor, (e) and (d) tor a pure
induetor and (e) and (f) tor a pure eapaeitor.
(al
f};;.; ~E","JY "00"" by clrcult
(bI
~~ t
Energy released
by circuit
Figure 5.7 WaveJorms Jor v, i and v-i product when i tags behind v by <p.
Figure 5.8 Figure Jor the exampte in section 5.16.
Z = 5(-jlO) = 50L-90° = 4.472L-26.570

p 5 - jlO 11.18L-63.43°
= 4 - j2 Q
Since the parallel circuit is in se ries with the (2 + j4) Q section, the total
impedance of the circuit is
ZE = (2 + j4) + (4 - j2) = 6 + j2 = 6.325L18.44° Q
By Ohm's law, the current drawn by the circuit is

Vs 20LO°
I =-= =3.162L-18.44°= 3 - j1 A
1 ZE 6.325L18.44°
This current flows into the parallel section of the circuit, and divides
between the two branches as folIows.
The current in the capacitor is
I = I Zp = 3.162L -18.440 4.472L - 26.57°

3 1 Zc 10 L - 90°
= 1.414L45° = 1 + j1 A
Applying KCL to the parallel circuit gives

12 = I 1 - 13 = (3 - j1) - (1 + j1) = 2 - j2
= 2.828L-45° A
The power consumed by the 2 Q resistor is
T:R = 3.1622 X 2 = 20 W
and the power consumed by the 5 Q resistor is

T:R = 2.828 2 X 5 = 40 W
Neither of the resistors consumes reactive VA. Also, since the inductor
and the capacitor are both pure elements, they do not consume any power.
The reactive VA consumed by the inductor is
QL = T:X L = 3.1622 X 4 = 40 V Ar lagging
and that consumed by the capacitor is

Qc = J2jXc = 1.4142 X 10 = 20 V Ar leading
It should be pointed out he re that the current through the inductor lags
behind the voltage ac ross it by 90°, and that the current through the
capacitor leads the voltage across it by 90°. That is, QL and Qc have
opposite mathematical signs! If we assign a positive sign to Qu then we
must assign a negative sign to Qc. and the total V Ar consumed by the
circuit is
Q = 40 - 20 = 20 V Ar (lagging)
The mathematical sign associated with QL and Qc is discussed further in

section 5.17.
5.17 Complex power
If a voltage V = VLa results in a current 1 = 1 Lß in a circuit, then the

average power P absorbed is
P = VI cos (a - ß)
Using compiex number nomenciature, we may say that
P = VI Re[ej(a - ß)] = Re[Ve ia Ie- iß ]
The first term within the final set of brackets can be recognised as the
phasor voltage. However, since the current in the circuit is
1 = ILß = Ie iß
the second term in the square brackets is not the current phasor. To correct
this, we make use of conjugate notation (see chapter 15 for details) as
follows
1* = Ie- iß
where 1* is the complex conjugate of the current. That is
P = Re[VI*]
and we can let the apparent power become complex by defining it as
S = VI* = Vlej(a - ß) = Vle i</> = P + jQ
where P is the power absorbed by the circuit, Q is the reactive power, and
the apparent power is ISI. The power factor is given by cos 4>.
If the current lags behind the voltage, 4> has a negative sign, with the
result that the complex part of VI* has a positive value. By international
agreement, complex lagging VA is given a positive sign and leading VA is
given a negative sign.
We will now apply this method of calculation to the problem in section
5.16 in which V s = 20LO° V, / 1 = 3.162L-18.44° A, 12 = 2.828L-45° A
and 13 = 1.414L45° A. The complex power absorbed by the circuit is
S = Vs/; = 20LO° x 3.162LI8.44° = 63.24LI8.44°
= 60 + j20 VA
That is, the circuit absorbs 63.24 VA, 60 Wand 20 VAr (lagging). Refer-
ring to the solution in section 5.16, the reader will note that the 2 Q resistor
consumes 20 W, and the 5 Q resistor consumes 40 W, giving a total power
consumption of 60 W. Moreover, the inductor receives 40 VAr (Iagging),
and the capacitor receives 20 VAr (Ieading). In the notation adopted in this
method of calculation these respectively correspond tO + 40 V Ar and - 20
VAr, giving a total reactive power consumption of + 20 VAr, as calculated
above.
Unworked problems
5.1. A sinusoidal voltage waveform has a value of - 45 V at t = 0 and is

rising; it reaches zero volts at t = 0.104 ms, and has its first positive
peak at t = 0.4165 ms. Calculate the frequency and the periodic time
of the wave, and also its maximum value. Write down an expression
for the voltage waveform. If this voltage produces a current of 10 cos
(rot - 135°) A, calculate the phase angle of i(t); with respect to v(t).
[800 Hz; 1.25 x 10-3 s; 90 V; 90 sin (1600Jtt - 30°) V; v(t) leads i(t)
by 15°]
5.2. The voltage indueed in a eoil of an eleetrieal generator inereases at a

eonstant rate from zero volts at zero radians to V m at a radians, then
remains eonstant from a to (n - a), after whieh it deereases at a
eonstant rate until it reaehes zero volts at n radians. This is repeated
with the reverse polarity in the seeond half of the wave. Calculate
the r.m.s. value and the average value of the wave for a value of a of
!
0, nl6 and n12.
ra.(mra.sd.) 0
Vm
nl6
V mv'(7/9)
nl2
V mv'(l/3)
average Vm 5Vm/6 V m/2
5.3. A sinusoidal eurrent of 20 A (r.m.s.) at a frequeney of 50 Hz is
added to a eurrent of the same magnitude whose frequeney is 60 Hz.
Write down an expression for the instantaneous eurrent in the
eireuit, and ealculate its value at t = 10 ms.
[28.28(sin 100 nt + sin 120 nt); -16.62 A]
5.4. Convert the following values into their reet angular eomplex form:
(a) 20L150°,
(b) 9 LlO° - lOLllO°,
(e) (5 - j3) x (6 + j2),
(d) 5L2000/7 L90°,
(e) 8.7 LlO° + (8 - j12).
[Ca) -17.32 + jlO; (b) 12.28 - j 7.83; (e) 36 - j8;
(d) - 0.244 + j 0.671; (e) 16.57 - j 10.49]
5.5. Convert the following eomplex values into polar form: (a) -6 - j8,
(b) (7 + j6) + (-8 - jlO), (e) (-7 + j9)/(5 + j3),
(d) 4L-170° + (2 - j3).
[(a)lOL -126.9°; (b)4.123L -104°; (e)1.96L -96.9°;
(d)4.173 L -117.7°]
5.6. A two-terminal blaek box eontains (a) a 50 Q resistor, (b) a 50 IlF
eapaeitor, (e) a 0.5 H induetor. If the eurrent flowing into the box is
5ej(IOOOt - 30°) A, ealculate the voltage between the terminals of the box
in eaeh ease.
[( a) 250ej(1000t - 30°) V; (b) 250ej(1000t - 12(0) V; 2500ej(1000t + 60°) V]
5.7.. Three wires are eonneeted to anode in a eireuit; the wires earry
eurrents i}> i 2 and i 3 , respeetively. (a) If i 1(t) = 10 sin (300t + 40°) A
and i 2 (t) = 15 eos (300t - 60°) A, ealculate i 3 (t); (b) if 11 =
15L-20° A and 13 = 20L40° A, ealculate 12 ,
«a) 24.9 sin (300t - 146°) A; (b) 30.42L-165.3° A]
5.8. + 40°) mV and
If v AB(t) = 50 sin (50t - 30°) mV, vBC(t) = 25 sin (50t
VAD(t) = 90 sin (SOt - 10°) mV, ealculate VCD and VBD •
100 f-IF 100 f-IH

O----j A
o------..----OB
Figure 5.9
[19.1L165.6° mV (note: Voc = 19.1L-14.4° mV);

32.73L-168.3° mV]
5.9. If the supply frequeney to the eireuit in figure 5.9 is 400 radis,
ealculate the eomplex input impedanee Zin if terminals AB are (a)
open-eireuited, (b) short-eireuited, (e) eonneeted by a 50 Q resistor.
[Ca) 47.17L-32° Q; (b) 20.62L-14° Q; (e) 31.17L-36.2° Q]
5.10. A 1 H induetanee, a 1 !lF eapaeitanee and a resistor Rare eonneeted
in parallel to a supply voltage of frequeney 500 rad/so Express the
real and imaginary parts of the impedanee of the parallel eireuit in
terms of R.
[Real part = R x 106/(106 + 2.25R 2 ); imaginary part =
1500R/(106 + 2.25R 2 )]
S.ll. Wh at value of induetance must be eonneeted in se ries with a deviee
whose internal eireuit is equivalent to a resistanee of 5000 Q in series
with a eapaeitanee of 0.1 !lF, to give an overall eireuit admittanee of
(51.42 + j87.4) !lS. The supply frequeney is 1000 rad/so
[1.5 H]
5.12. Aseries eireuit eontains a resistanee of 5 Q, an induetor of reaetanee
10 Q and a eapaeitor of reaetanee 6 Q. If the r.m.s. voltage of the
supply is 100 V, ealculate (a) the eomplex impedanee and admit-
tanee of the eireuit, (b) the eurrent drawn by the eireuit, (e) the
phase angle and power faetor of the eireuit and (d) the voltage aeross
eaeh element in the eireuit.
[Ca) 6.4L38.66° Q, 0.156L-38.66° S; (b) 15.63L-38.66° A; (e) I
lags V by 38.66°, power faetor = 0.781 (lagging); (d) VR =
78.13L-38.66° V; VL = 156.3°L51.3° V, Vc = 93.75L-128.7° V]
5.13. In problem 5.12, ealculate the apparent power, the power and the
VAr eonsumed. Calculate also the VAr eonsumed by the induetor
and the eapaeitor.
[1563 VA; 1220 W; 976.4 VAr (lagging); 2443 VAr (lagging); 1466.6
VAr (leading)]
5.14. A resistor and a coil are connected in series to a 240 V supply. If the
current drawn from the supply lags the supply voitage by 37°, and its
r.m.s. value is 3 A, calculate (given that the r.m.s. voitage across the
coil is 171.4 V) the resistance and reactance of the inductor, and the
resistance of the resistor. Calculate also the power and V Ar con-
sumed by the circuit.
[30.74 Q, 48.15 Q; 33.26 Q; 575 W; 433 VAr (lagging)]
5.15. A 3-branch parallel circuit contains a 4 kQ resistor in one branch, a 5
kQ inductive reactance in the second branch and a 7 kQ capacitive
reactance in the third branch. Calculate the effective admittance and
impedance of the circuit. If it is energised by a 200 V r.m.s. supply,
calculate the current in each branch, the total current drawn
by the circuit and the phase angle of the circuit. Determine also the
apparent power, the power and the VAr consumed, together with
the power factor of the circuit.
[0.2564 L -12.87° mS, 3.9 L12.87° kQ; I. = 0.05 LO° A; 12 =
0.04L-90° A; 13 = 0.0286L90° A; -12.87°; 10.26 VA; 10 W; 2.29
VAr (lagging); 0.975 (lagging)]
5.16. A two-branch parallel circuit contains impedances of (8 - j7) Q and
(5 + j6) Q, respectively. If the current in the (8 - j7) Q branch is
9.41 L41.2° A, calculate the current in the other branch, and the
supply voitage. Also evaluate the complex admittance of the circuit.
What is the phase angle of the circuit? What apparent power, power
and VAr are consumed?
[12.8L-50.1° A; 100LO° V; 0.157 L-13.41° S; 13.41° (lagging);
1570 VA; 1527.2 W; 364.1 VAr (lagging)]
6
Sinusoidal Steady-state
Analysis
6.1 Introduction
This chapter not only deals with a.c. applications of techniques dealt with
in the work on d.c. circuits, but deals with other circuit theorems particu-
larly appropriate to a.c. circuit analysis. In particular, the application of
complex numbers to electric circuits; the theory of complex numbers is
covered in chapter 15, where the reader can study this and other math-
ematical techniques appropriate to electrical and electronic engineering.
We commence by analysing the circuit in figure 6.1(a), which is specified
in terms of time-domain data. Following the concepts developed in earlier
chapters, the frequency-domain equivalent circuit is drawn in diagram (b).
We shall retain ohmic values in kQ, so that the resulting currents are in
mA. The impedance of the series branch is
Zs = 2 + j1 kQ
and the impedance of the parallel branch of the circuit is
1 kQ
i.(t) I.
-j1 kQ
56.57 sin (5OOOt + SO·) V 4OLSO· V

(a) (b)
Figure 6.1 The circuit in diagram (a) contains time-domain tiata which, in diagram (b), is
converted into its co"esponding frequency-domain equivalent.
126
Sinusoidal Steady-state Analysis 127
j2(1 - j1) j2(1 - j1) 2L90° x 1.414L-45°

Z =
+ (1 - j1) + j1
p
j2 1 1.414L45°
= 2LO° kQ
Since the two branches are in se ries with one another, the effective impe-
dance of the complete circuit is
ZE = Zs + Zp = (2 + j1) + (2 + jO) = 4 + j1
= 4.123L14.04° kQ
By Ohm's law, the current 11 drawn by the circuit is
11 = VS/Z E = 40L900/4.123L14.04° = 9.7 L 75.96° mA
The current in each branch of the parallel circuit may be evaluated by any
of the methods described earlier, typical calculations being
1 = 1
Z2
= 9.7 L 75.96°
J2
x -----''------
3 1 Z2 + Z3 j2 + (1 - j1)
2L90°
= 9.7 L 75.96° x = 13.72L120.96° mA
1.414L45°
and
Z3
1 = 1
2 1 Z2 + Z =9.7L-14.04° mA
3
The power , VA and V Ar in the various sections of the circuit can be

calculated by any one of the methods outlined in chapter 5.
Since types of a.c. analysis deal with circuit equations involving complex
numbers, we will be concerned with the solution of simultaneous equations
involving complex quantities. Generally speaking, their solution is no more
difficult than solving conventional simultaneous equations, with the excep-
tion that we also need to deal with phase angles as weIl as magnitude. That
is, the solution takes a little longer, and needs a little more care. Consider
the solution of the following simultaneous equations
50LO° = 11.18L - 26.56°X + 5L900y (6.1)
o = 5L900X + 3.162L-18.43°Y (6.2)
where X and Yare unknowns. To eliminate variable Y from the equations,
we need to make the coefficient of Y have the same value in both equa-
tions; that is it must have the same magnitude and phase in both equations.
We then add (or subtract) the equations, leaving only the variable X,
whose value can then be determined. Full details of the theory of complex
number manipulation is given in chapter 15.
If we multiply equation (6.2) by the complex value

5L90° = 1.581L108.43°, it makes the coefficient Y in the new
3.162L -18.43°
equation 5L90°. When the new equation is subtracted from equation (6.1),
Y is eliminated as follows
50LO° = 11.18L-26.56°X + 5L900y «61) repeated)
o 7.91LI98.43°X + 5L900y «6.2) x 1.581 L108.43°)
50LO° = 17.68L-8.21°X (SUBTRACT the equations)
To perform the subtraction it is first necessary to convert both equations

into rectangular form, as described in chapter 15. The value of X is
X = 50LO° = 2.83L8.12°
17.68L-8.21°
Substituting this value into equation (6.2) gives
Y= -5L90° x 2.83L8.12° = -14.15L98.12°
3.162L -18.43° 3.162L-18.43°
=-4.47 L116.55° = 4.47 L-63.45°
Alternatively, the equations can be solved by determinants which, in the
opinion of the author, is an easier and more reliable routine. As with other
mathematical processes, this is dealt with in chapter 15.
Yet another method of solving a pair of simultaneous equations with
complex values in polar form is given by the computer program in listing 6.1.
The main program is in lines 10 to 610, inclusive; the input of data is
complete by line 290, and lines 310 to 590, inclusive, make computations
on each of the determinants. Two subroutines are used; the one commenc-
ing at line 1000 calculates the magnitude and phase angle of the determi-
nants, and that commencing at line 2000 keeps the final phase angle within
the limits ± 180°.
Listing 6.1
Solution of two simultaneous equations having complex values.
10 CLS
20 PRINT TAB (3) ; "Solution of two CCIlplex siIrultaneous equations"
30 PRINT TAB(18); "of the form"
40 PRINT TAB(15); "V1 = A*X + B*Y"
50 PRINT TAB(15); "V2 = c*x + D*Y"
60 PRINT TAB(3); "Where V1 and V2 are corcplex values,"
70 PRINT TAB(3); "A,B,C and D are corcplex coefficients,"
80 PRINT TAB(3); "and X and Y are the variables."
90 PRINT
100 INPUT "Magnitude of VI = ", VI
110 INPUT "Phase of VI (degrees) = ", PI
120 REM ** Convert degrees to radians **
130 PI = PI * .01745
140 INPUT "Magnitude of A = ", A
150 INPUT "Phase of A (degrees) = ", PA
160 PA = PA * .01745
170 INPUT "Magnitude of B = ", B
180 INPUT "Phase of B (degrees) = ", PB
190 PB = PB * .01745
200 PRINT
210 INPUT "Magnitude of V2 = ", V2
220 INPUT "Phase of V2 (degrees) ", P2
230 P2 = P2 * .01745
240 INPUT "Magnitude of C = ", C
250 INPUT "Phase of C (degrees) = ", PC
260 PC = PC * .01745
270 INPUT "Magnitude of D = ", D
280 INPUT "Phase of D (degrees) = ", PD
290 PD = PD * .01745
300 PRINT
310 REM ** Ca1culations for main determinant **
320 Mag = A * D: PH = PA + PD
330 Re1 Mag * COS(PH): Iml = Mag * SIN(PH)
340 Mag C * B: PH = PC + PB
350 Re2 Mag * COS(PH): 1m2 = Mag * SIN(PH)
360 GOSUB 1000
370 Det = Mag: DetP = PH * 57.3066
380 REM ** Calculations for determinant X **
390 Mag VI * D: PH = PI + PD
400 Re1 Mag * COS(PH): Iml = Mag * SIN(PH)
410 Mag V2 * B: PH = P2 + PB
420 Re2 Mag * COS(PH): 1m2 = Mag * SIN(PH)
430 GOSUB 1000
440 DetX = Mag: DetXP = PH * 57.3066
450 REM ** Calculations for determinant 'Y **
460 Mag = A * V2: PH = PA + P2
470 Re1 = Mag * COS(PH): Iml = Mag * SIN(PH)
480 Mag VI * C: PH = PI + PC
490 Re2 = Mag * COS(PH): 1m2 = Mag * SIN(PH)
500 GOSUB 1000
510 DetY = Mag: DetYP = PH * 57.3066
520 REM ** There is no solution of Det = 0 **
530 IF Det = 0 'IHEN PRINr TAB (3) ; '''Ihe equations cannot be solved": END
540 REM ** Calculate and print the value of the variables **
550 PRINT TAB(3); "X = "; DetX / Det
560 Phi = DetXP - DetP: GOSUB 2000
570 PRINT TAB(9); "at angle "; Phi; " degrees."

580 PRINT TAB(3); "Y = "; DetY / Det
590 Phi = DetYP - DetP: GOSUB 2000
600 PRINT TAB(9); "at angle "; Phi; " degrees."
610 END
1000 REM ** Calculate magnitude and phase of determinant **
1010 Re = Re1 - Re2: Im = Im! - 1m2
1020 Mag = SQR(Re A 2 + Im A 2): pi = 3.141593
1030 IF Re = 0 AND Im > 0 THEN PH = pi / 2: RETURN
1040 IF Re = 0 AND Im < 0 THEN PH = -pi / 2: RETURN
1050 PH = ATN(Im / Re)
1060 REM ** Compensation for angular ca1cu1ation **
1070 IF Re < 0 AND Im < 0 THEN PH = PH - pi
1080 IF Re < 0 AND Im > 0 THEN PH = PH + pi
1090 IF Re < 0 AND Im = 0 THEN PH = pi
1100 RETURN
2000 REM ** Modify phase angle to within 180 deg **
2010 IF Phi> 180 THEN Phi = Phi - 360: RETURN
2020 IF Phi< -180 THEN Phi = Phi + 360: RETURN
2030 RETURN
6.2 Nodal, mesh and loop analysis
The technique involved for each of these methods of analysis is generally

similar to that described in chapter 2 for resistive circuits, with the excep-
tion that complex values are used.
We will illustrate nodal and mesh analysis using simple problems. When
a cireuit eontains a mixture of voltage and eurrent sourees it may be
necessary to modify the eireuit equations using supermeshes (when using
mesh analysis) or supernodes (when using nodal analysis). Where depen-
dent sourees are involved the teehniques are, onee again, similar to those
outlined for resistive cireuits.
As with d.c. eireuits, loop analysis ean be used to evaluate loop currents
in any eircuit, but has the dis advantage that it is less systematic than mesh
analysis, and is not generally so convenient for computer solution.
In the first example in this section, we apply nodal analysis to the solution
of the cireuit in figure 6.2.
Solution
In this case we take node 0 to be the reference node. Applying KCL to

node 1 we get
2L20° A 4L40° A
Figure 6.2 Circuit Jor worked example 6.2.1.
2L200= «2 + 3) + j(-2 + 3))V. - (3 + j3)V2

= (5 + j1)V. - (3 + j3)V2
= 5.1 L11.31°V. - 4.24L45°V2 (6.3)
The application of KCL to node 2 yields
4L40o= -(3 + j3)V. + «4 +3) + j(3 + 4))V2
= - (3 + j3)V. + (7 + j7)Vz
= -4.242L45°V. + 9.9L45°V2 (6.4)
V2 can be eliminated between these equations by multiplying equation
(6.3) by 9.9L45°/4.24L45° = 2.335LO°, and adding it to equation (6.4) as
folIows.
4.67 L20° = 1l.91Lll.31°V. - 9.9L45°V2 «6.3) x 2.335)
4L40° = -4.242L45°V. +9.9L45°V2 «6.4) repeated)
8.45L29.22° = 8.7 L -4.37°V. (ADD the equations)
Hence
8.45 L29.22° O.982L33.60 V

V. = 8.7 L-4.370
Substituting this value into equation (6.3) gives
V = (5.1 L11.31° x O.982L33.6°) - 2L20°
z 4.24L45 0
= O.778L14.72° V
This calculation involves several polar-to-rectangular and rectangular-to-
polar conversions, which are not shown. A simpler method may be to use
determinants, or the computer program in listing 6.1.
In.examples which follow, details of the solution of simultaneous equa-
tions is omitted.
Figure 6.3 Mesh current analysis.
In this problem we will use mesh analysis to determine the currents in

figure 6.3.
Solution
Applying KVL to mesh 1 yields

lOL20° = «2 + j2) + (4 - j4) )/1 - (4 - j4)/2
= (6 - j2)/1 - (4 - j4)/2
= 6.325 L -18.44°/1 - 5.657 L - 45°/2
and to mesh 2 gives
-30L50° = -(4 - j4)/1 + «3 + j3) + (4 - j4»/ 2
= -(4 - j4)/1 + (7 - j1)/2
= -5.657 L-45°/1 + 7.071L-8.13°/2
Solving these equations in the manner described earlier gives

I 1 = 2.39L163.4° A
and
12 = 3.69 L -148.6° A
6.3 Principle of superposition
The principle of superposition can be applied to linear circuits, including

a.c. circuits. Consider the calculation of the voltage VI in the circuit in
figure 6.4. The value of this voltage is dependent on the two sources in the
circuit, one being a current source and the other a voltage source.
When the circuit is energised by the current source alone (in the mean-
time, the voltage source is replaced by its internal impedance, that is,
zero), the voltage at the left-hand node is VIA' In this case, 2LO° A flows
V, 4 + j5 Q
Figure 6.4 Principle o[ superposition.
into a two-branch parallel circuit comprising (2 + j3) Q in parallel with

(4 + j5) Q, and VIA is
V = 2LOo~-+ j3)(4 + j5) = 2.68 + j3.76 V

IA (2 + j3) + (4 + j5)
When the circuit is energised by the voltage source alone (in the meantime,
the current source is replaced by its internal impedance, that is, infinity),
the voltage at the left-hand node is now VIB' This voltage is the proportion
of the 5LO° V source appearing across the (2 + j3) Q impedance, which is
VIB = 5L 0 0 2 + j3 = 1.81 + J'0 .1 V

(2 + j3) + (4 + j5)
Hence, by superposition
VI = VIA + VIB = 4.49 + j3.86 = 5.92L40.8° V
6.4 Thevenin's theorem and Norton's theorem
Thevenin's and Norton's theorems (see also chapter 3) may be used for
frequency-domain circuits, the only difference being that we need to re-
place the term resistance by impedance, and conductance by admittance.
We will illustrate their use in a.c. circuits by way of examples.
Determine Thevenin's equivalent circuit with respect to terminals A and B

of the network in figure 6.5. Hence calculate the current which would flow
in a (6 - j5) Q impedance connected between A and B.
Solution
Firstly, we must replace the source by its internal impedance (zero in this
case), and determine the impedance between terminals A and B as folIows.
j40 Q
'---e---o A
Figure 6.5 Thevenin's theorem.
Z = 20 X j40
16 + j8 Q
TH 20 + j40
Next we calculate the voltage VAB , which is the Thevenin source voltage as
follows
VTH = VAB = 50LlO° 20LO° = 22.36L-53.44° V

20 + j40
That is, the Thevenin source between terminals A and B comprises a
voltage VTH = VAB = 22.36L-53.44° V having an internal impedance of
(16 + j8) Q. When the externaiload is connected, the total impedance in
the circuit is
ZE = ZTH + (6 - j5) = 22 + j3 = 22.2L 7.77° Q
and the current in the load is
VTH
1 = - = 22.3L-53.44° = 1.007 L-61.21° A
22.2L 7.77°
Repeat worked example 6.4.1 using Norton's theorem.
Solution
We can avoid some work simply by saying that the Norton internal admit-
tance is equal to the reciprocal of the Thevenin internal impedance, that is
YN = l/ZTH = 1/(16 + j8) = 0.0559 L - 26.56°
= 0.05 - jO.025 S
The value of the Norton internal source current is equal to the current
which ftows from A to B when these terminals are short-circuited. That is
The Norton equivalent source current could, alternatively, have been

deduced by saying
IN = VTH/Z TH = 22.36L-53.44°/(16 + j8)
= 1.25L-80° A
The Norton source current divides between the internal admittance of the
source and the admittance of the load, the latter value being
Y L = 1/(6 - j5) = 0.098 + jO.082 S
The effective admittance of the Norton generator and the load in parallel is
YE = YN + YL = (0.05 - jO.025) + (0.098 + jO.082)
= 0.148 + jO.057 = 0.159L21.06° S
I = IN YL = 1.25L-80° 0.128L39.8° = 1L-61.26° A

YE 0.159L21.06°
When comparing the values of I obtained by the two methods of calcula-
tion, minor discrepancies are due to rounding errors.
6.5 Millman's theorem
Other circuit theorems, subject to any special conditions discussed earlier,

apply equally weIl in the frequency domain; Millman's theorem (also
known as the Parallel Generator theorem) is no exception. This theorem is
widely used in electronics, and is weH suited to the solution of unbalanced
3-phase, 3-wire power systems (see also chapter 7) as is illustrated below.
Consider the system in figure 6.6, in which an unbalanced load (see chapter
7 for a description of the term unbalanced) is supplied by an unbalanced
three-phase supply. We need to calculate the current in each line, together
with the voltage VSN '
Solution
Using the equation developed in chapter 3, the voltage VSN is

VANYAS + VBNYBS + VCNYCS
VSN = - ' - - - - - - - - - - -
YAS + YBS + Y cs
---
100LO° V
'A 10LO° g
---
A
110L-130° V
18 20L20° g
---
N S
B
120L 110° V
Figure 6.6 Millman's theorem applied to an unbalanced 3-phase, 3-wire power system.
100LO° 1l0L-130° 120 L 110°

----+ + 40L-4Q0
=----------------------------
1 1 1
---+ +-----
lOLO° 20L20° 40L-40°
2.92L-25.35°
=------ 17.6L-25.71°
0.166LO.36°
= 15.86 - j7.64 V
The current in line A is
100LO° - 17.6L-25.71°
lOLO°
= 8.45L5.19° A
Similarly, it may be shown that
lB = 5.79L-158.4° A and le = 3.33L155.4° A
6.6 Rosen's, star-delta and delta-star theorems
Rosen's theorem or the general star-mesh transformation is as applicable

in the frequency domain as it is to d.c. circuits, as are the star-delta and
delta-star transformations.
Three impedances are connected in delta as folIows. An impedance of

(6 + j3) Q is connected between lines 1 and 2, (7 - j5) Q is connected
between lines 2 and 3, and an impedance of (6 + jO) Q is connected
between lines 3 and 1. Determine the impedances in the equivalent star-
connected network.
Solution
Since each of the star-connected impedances is connected to the star-point

S, then
LZ
where
= (6 + j3) + (7 - j5) + (6 + jO) = 19 - j2
= 19.1L-6° Q
Hence
(6 + j3)(6 + jO) 40.25 L26.57°
19.1L-6° 19.1L-6°
= 2.11L32.57° Q
Also
Zzs = Z21Z,jLZ = 3.02L-2.96° Q
Z3S = Z31Z32/LZ = 2.7 L-29.53° Q
6.7 Maximum power transfer theorem
The maximum power transfer theorem as applied to a.c. circuits is more

involved than is the case in d.c. circuits because of the existence of complex
impedances and complex voltages and currents. There are four general
conditions for maximum power transfer in a.c. networks, and these are
given below without proof:
1. A pure resistance load extracts maximum power from a circuit when the
load resistance is equal to the magnitude of the internal impedance of
the source to which it is connected.
2. A constant reactance, variable resistance load extracts maximum power

when the resistance of the load is equal to the magnitude of the internal
impedance of the network plus the reactance of the load.
3. A constant power factor, variable impedance load extracts maximum
power when the magnitude of the load impedance is equal to the
magnitude of the internal impedance of the network.
4. A variable resistance, variable reactance load extracts maximum power
from a circuit when the impedance of the load is equal to the complex
conjugate of the impedance of the network.
While we cannot illustrate each version of the theorem here, we will take a
look at the final version.
If a load having independently variable resistance and reactance is con-

nected to terminals A and B of the circuit in worked example 6.4.1 (figure
6.5) and extracts maximum power from it, calculate the value of the
impedance of the load and the power absorbed.
Solution
The Thevenin equivalent circuit of worked example 6.4.1 is a voltage

source of 22.36L -53.44° V having an internal impedance of (16 + j8) Q.
Maximum power is absorbed when the impedance of the load is the
complex conjugate of this value, namely
ZL = ZTH * = 16 - j8 Q
The effective impedance of the complete circuit is then
ZE = ZTH + ZL = (16 + j8) + (16 - j8) = 32 Q

1= VTH/Z E = 22.36L-53.44°/32 = 0.7 L-53.44° A
The maximum power absorbed by the load is
P max = IFI R L = 0.72 X 16 = 7.84 W
6.8 a.c. circuits with dependent sources
So far we have only looked at circuits containing independent sources. The

techniques involved in circuits with dependent sources are generally the
same as those in d.c. circuits, with the exception that we are dealing with
complex quantities, as shown in the following worked example.
100 kQ
r----'ooQ"j
I H--'-'~r~~----'---ö
I
110Loot '1.,..
I mV I
I I
I Signal souLrc-e--T,-......- - - - - - - -.....- - O
"--~-±---g~m
:O~im
L _ _ _ _ _ ...J
The circuit in figure 6.7 is a simplified circuit of a transistor amplifier.

Calculate the voltage gain (ViV l ) at a frequency of (a) 1 kHz, (b) 200 kHz.
Solution
One method of solving this problem is by nodal analysis as follows. Initial-

ly, the 10 mV source and its associated 100 Q internal resistor (shown in
the broken box in figure 6.7 are converted into a O.lLO° mA source which
is shunted by a 100 Q resistor, as shown in insert (i). The 100 Q resistor is
seen to be in parallel with the 10 kQ resistor, the two being equivalent to a
99 Q resistor (or a 0.0101 S conductance). The nodal equations for the
circuit are as follows
Node 1:
1+j100X103WC)
0.1 x 10-3LO° = ( 0.0101 + V -
100 X 103 1
= (0.01011 + jWC)Vl - (1 X 10-5 + jwC)V2

Node 2:
1 + j 100 X 103WC)
0= ( 8- V+
100 X 103 I
( 6.667 x 10-3 + 1+j100X103WC) V2

100 X 103
= (7.9999 - jwC)VI + (6.677 X 10-3 + jWC)V2

(a) Whenf = 1 kHz, wC = 2Jt X 1000 x 25 X 10-12 = 1.57 x 10-\ and the
equations are
0.1 x 10-3 + jO = (0.01011 + j1.57 x 10-7 )VI - (1 X 10-5 + j1.57 X 10-7)V2
0= (7.9999 - j1.57 x 10-7 )VI + (6.677 X 10-3+ j1.57 X 10-7)V2
Solving by one of the methods outlined earlier
V2 = 5.423L179S V
hence
voltage gain = V2 1VI = 542.3L179S
(b) Whenf= 200 kHz, wC = 2Jt X 200 X 103 x 25 X 10-12 = 3.142 X 10-5 ,
and the equations are
0.1 x 10-3 + jO = (0.01011 + j3.142 x 10-5 )VI - (1 X 10-5 + j3.142 X 10-5)V2
0= (7.9999-j3.142 x 10-5 )VI + (6.677 x 10-3+j3.142 x 10-5)V2
Solving gives V2 = 2.741L120.4° V
and voltage gain = V/VI = 274.1L120.4°
The reduction in gain (and the change in phase shift) when the frequency is
increased from 1 kHz to 200 kHz is due to the reduction in the reactance of
the capacitor with increase in frequency. Frequency response is dealt with
more fully in chapter 11.
Unworked problems
6.1. Aseries circuit containing a 10 Q resistor and an 8 Hinductor is

energised by a voltage v(t) = 8 cos 2t + 6 sin 2t V. Determine an
expression for the current in the circuit. What is the phase rela-
tionship between the voltage and the current?
[0.53 sin (2t - 4.87°) A; i(t) lags v(t) by 4.87°]
--E,
o
---
E,
.;;.JZ,
o
A ~--Lj-"B
--E,
":;Z,
Figure 6.8
3 Q j4Q
-j3 Q
Figure 6.9
6.2. A voltage of 5 sin (5t - 11:/4) V is applied to aseries circuit containing

a coil of 10 Q resistance and 4 H inductance, and a 0.04 F capacitor.
Calculate the voltage across the coil. [6.2 sin (5t + 74.74°) V]
6.3. For the circuit in figure 6.8, if EI = lOLO° V, E z = 0, E 3 = 6LO° V,
ZI = j4 Q, Zz = 4 Q, Z3 = - j2 Q, use mesh analysis to calculate 1I
and I z • [li = 2.262L-96.4° A; I z = 2.594L-101.3° A]
6.4. If in figure 6.8, EI = (20 + j7) V, E z = (10 + jO) V, E 3 = 0, ZI =
(1 + j2) Q, Zz = (5 - j6) Q and Z3 = j8 Q, calculate 1I and I z using
mesh analysis. [li = 1.21L-50° A; I z = 2.31L-69.9° A]
6.5. Write down, but do not solve, the mesh equations for the circuit in
figure 6.9.
o (5 + j9)/1 jlOlz (2 - j5)/3

-(5 - j2) = -j10/1 + (6 + j10)lz 5/3
(10 + j3) = -(2 - j5)/1 51z + (7 - j8)/3
1001
200 Q -jl00 Q
30 Q 1 kQ tV,
Figure 6.10
j20 Q
L..-.....---o A
50L45° V t 10 Q
L..----~-----~~-~B
Figure 6.11
---
1 LO° V
jl000 Q
Figure 6.12
6.6. Solve the circuit in figure 6.9 for 1.,12 and 13 ,

[I. = 1.05L57° A; 12 = 1.8L63.4° A; 13 = 2.07 L 75.8° A]
6.7. Figure 6.10 shows a simplified equivalent circuit of a common-base
amplifier. Calculate the value of V2 •
[3.09L-178.2° V]
6.8. Using the superposition theorem, calculate the current I in figure
6.1l.
[2.704L-67S A]
6.9. Using superposition, calculate the current I L in figure 6.12.
[4.97 L84.3° A]
6.10. The network in figure 6.13 is the sm all-signal hybrid parameter
equivalent circuit of a transistor. Determine the Thevenin equiva-
h,
, - - - -.....- - - 0 A
~---~--~~---~--uB
Figure 6.13
lent circuit with respect to terminals A and B; parameters h, and h f

are dimensionless, h i is a resistance and h o is a conductance.
[E TH = -hfE/(hoh j - hrh f ); ZTH = l/(h o - hrh/h j ) Q)]
6.11. Deduce Thevenin's equivalent circuit with respect to terminals A
and B for the circuit in problem 6.8.
[ETH = 27.04L-67S V; ZTH = 7.071 L45° Q]
6.12. Deduce the Norton equivalent circuit with respect to terminals A
and B of problem 6.3.
[IN = 0.502L90° A; Y N = 0.354L45° S]
6.13. Use Millman's theorem to solve problems 6.3 and 6.4.
6.14. An impedance of (2 + j4) Q is connected between nodes A and B,
an impedance (3 - j5) Q is connected between nodes Band C, and
an impedance (2 + jO) Q is connected between nodes A and C.
Determine the equivalent star-connected circuit.
[ZAS = 1.26L 71.57° Q; ZBS = 3.69L12.53° Q; Zes = 1.65L-50.9° Q]
6.15. The load in problem 6.14 is energised by a 3-phase, 3-wire system,
whose neutral point is node N. If node A is energised by a voltage of
100L 10° V with respect to node N, node B is energised by a voltage
of 120L -120° V with respect to node N, and node C is energised by a
voltage of 90 L 100° V with respect to node N, use Millman's
theorem to calculate the current flowing into nodes A, Band C,
respectively.
[lA = 111.9L-29.4° A; I B = 16.3L-166.2° A; le = 100.7 L144.2° A]
7
Polyphase Circuits
7.1 Introduction
Polyphase supply systems have a number of advantages over single-phase

systems including the fact that, for a given amount of electrical power
transmitted, the total power loss is lower in the polyphase system; more-
over, the total volume of conductor material needed in the cable is less.
For these and other reasons, power is transmitted across the nation by a
three-phase system. Another feature is that the torque produced by a
single-phase motor is pulsating rather than rotating; that is, an ideal
single-phase motor has no starting torque! In order to cause the rotor of a
single-phase motor to begin to rotate, it is necessary (at starting) to convert
it to a two-phase motor. On the other hand, polyphase motors produce a
smooth torque, and their speed can be controlled using relatively straight-
forward methods.
Industry uses many other types of multi-phase system. For example,
many control systems use two-phase supplies to drive servomotors, and
many rectifier systems use a six-, twelve- or twentyfour-phase supply. In
this chapter we shall concentrate on three-phase systems.
The US notation for phase markings has been adopted, so that the
phases in a three-phase system are called A, Band C rather than R, Y and
B, respectively.
7.2 Three-phase generation
The usual method of producing a three-phase set of voltages is to have

three separate, but identical, windings on the rotor of an alternator , each
winding being physically displaced from the next winding by 120°. Since the
windings are displaced in space by 120°, and are moving through the same
magnetic field system, the net result is three single-phase voltages (referred
144
Polyphase Circuits 145
(a)
(b) (c)
Figure 7.1 (a) Three-phase valtage wavefarms, (b) star-eanneeted valtages and (e) a
typical phasar diagram.
to as the phase voltages) having an angular displacement of 120° between

them.
Alternatively, the voltages can be produced electronically, the phase
displacement being controlled either by R-C circuits and operational am-
plifiers, or by integrated circuits.
While 3-phase voltage sources are widely used, the reader should note
that 3-phase current sources are very uncommon.
7.3 Star-connection or Y-connection
Imagine an alternator with three separate windings with ends marked A

and A', Band B', and C and C', respectively having the time-varying
voltage V AA , , V BB , and VCC' induced in them, the respective waveform
diagrams being shown in figure 7.1(a). The voltages can be written down in
the form
VAN = Vm(AA') sin rot V
VI;IB' = Vm(BB') sin (rot - 120°)
= Vm(BB') sin (rot + 240°) V
vcc = V m(CC) sin (wt - 240°)
= V m(CC) sin (wt + 120°) V

If the ends A', B' and C' (each of which can be thought of as the 'start'
end of the corresponding winding) are connected together at node N, then
the eonneetion diagram is as shown in diagram (b), and the phasor diagram
at wt = 0 is illustrated in diagram (e). Onee again, it is pointed out that the
magnitude of eaeh phasor is the r.m.s. value of the associated sinusoidal
waveform.
Moreover, we use the double-subseript notation for three-phase vol-
tages as follows
V AN = voltage of terminal A with respect to N
VBN = voltage of terminal B with respect to N
V CN = voltage of terminal C with respect to N
In many systems, point N is connected to earth, that is, to a point of neutral
potential, and is known as the neutral point of the supply. Where it is not
earthed, it is sometimes ealled the star point.
The reader should note that, in the above analysis, we have taken VAN to
lie in the referenee direction. Although this will generally be the case, it
may not always be so for any given problem.
7.4 Phase sequence
As explained in chapter 5, a sinusoidal waveform can be thought of as the

vertieal eomponent of the tip of a radial line whieh rotates at eonstant
speed in a counterclockwise direetion. A three-phase set of voltages can
therefore be thought of as the vertieal eomponent of three radial lines
whieh are separated from one another by 120°, and whieh rotate at a
eonstant speed in a eounterclockwise direetion.
The phase sequence of the system in question is the order in which they
pass a fixed point in space. In the ease of the phasors in figure 7.1(e), this is
the sequence ABC or positive phase sequence (PPS). PPS supplies are
generally assoeiated with eleetrical power systems, and are the subject of
the majority of the work in this chapter.
The windings ean be reconnected so that the phase sequenee is reversed,
that is, the sequenee ACB, whieh is known as negative phase sequence
(NPS). Power supply systems normally generate PPS supplies but, under
abnormal conditions such as eertain types of eleetrical fault, NPS compo-
nents may be produeed. NPS systems are discussed more fully in section
7.17.
7.5 Balanced and unbalanced systems
A balanced polyphase supply is one having phase voltages which are equal
in magnitude, and which are phase displaced from one another by an equal
angle; in a three-phase system, this angle is 120°. If either of the conditions
is not met, the supply is said to be unbalanced.
A balanced polyphase load is one in which the impedance of each phase
of the load has the same magnitude and the same phase angle. If either
condition is not met, the load is said to be unbalanced.
In the majority of large industrial loads, both the supply and the load
are balanced, but in domestic situations and sm all industries, we meet with
unbalanced supplies connected to unbalanced loads.
While we have already met with many analytical tools which allow us to
deal with most polyphase circuits, one of the most powerful techniques
(symmetrical components) has yet to be introduced (see section 7.17).
7.6 Phase and line voltages in a star-connected system
As explained earlier, a phase voltage is the voltage induced in a winding or

phase of an alternator . In the case of the phasor diagram of the star-
connected supply in figure 7.2, the phase voltages are V AN , V BN and VCN.
A fine voltage or fine-to-fine voltage is the voltage between a pair of
lines. For example, the line voltage between lines A and Bin figure 7.2 is
VAB = voltage of line A with respect to line B
= V AN - VON
Similarly
V BC = voltage of line B with respect to line C
>---~~A
Figure 7.2 Phase and fine vo/tages.

Figure 7.3 Calculation o[ the Une voltage V AB .
and
VCA = voltage of line C with respect to li ne A
In the case of a balanced system, the phase voltages are equal in magni-
tude, and each is called V p • The way in which the line voltage V AB can be
related to V p is shown below (see also figure 7.3).
In a balanced system
VAB = VpLO° - V pL-120° = Y3 V p L30° V
That is the magnitude, Vu of the voltage between a pair of lines (the Une
voltage) in a balanced star-connected supply system is
V L = Y3 V p
Referring to figures 7.2 and 7.3 we see that
VAB = Y3 V pL30° V
VCA = Y3 V pL150° V
VBC = Y3 V pL-90° V
That is, the fine voltages in a balanced 3-phase supply are equal in magni-
tude, and are phase displaced tram one another by 120".
Calculate complex expressions for phase and line voltages in a three-phase

system in which V p = 250 V.
Solution
V p = 250 V, hence V L = Y3 V p = 433 V

Assuming that V AN lies in the reference direction then

VAN = 250LO° = 250 + jO V
VBN = 250L-120° = -125 - j216.5 V
VCN = 250 L 120° = -125 + j216.5 V
and
VAB = 433L30° = 375 + j216.5 V
VCA = 433L150° = -375 + j216.5 V
V BC = 433L-90° = 0 - j433 V
If VAN = 400 L20° V, VBN = 350 L-130° V and VCN 450 LllO° V,
determine a complex expression for each line voltage.
Solution
In this case we are dealing with an unbalanced set of voltages and, more-
over, V AN is not in the reference direction.
V AB = VAN - VBN = 400L20° - 350L-130°
= (375.88 + j136.8) - (-224.98 - j268.12)
= 600.86 + j404.92 = 724.6L33.98° V
There are two points the reader should note. Firstly, it is necessary to
convert the polar complex values ·to rectangular complex values before
they can be subtracted. Secondly, the answer is given in polar form, which
is more usual form for practical purposes since most engineering instru-
ments give readings in this form rather than in rectangular complex form.
VCA = V CN - VAN = 450LllO° - 400L20°
+ j422.86) - (375.88 + j136.8)
= (-153.9
= -529.78 + j286.06 = 602.1L151.6° V
V = V
BC V = 350L-130° - 450L11O°
BN - CN
= (-224.98 - j268.12) - (-153.9 + j422.9)
= -71.08 - j691.02 = 694.02L-95.9° V

c c
A A
(a) (b)
Figure 7.4 (a) A 3-phase delta-connected generator, and (b) its phasor diagram.
7.7 Delta-connected or mesh-connected three-phase source
If the three separate windings on the rotor of a 3-phase alternator are

interconnected so that the end of one winding is connected to the start of
the next (see figure 7.4(a)), the result is the delta- or mesh-connection,
having the phasor diagram in figure 7.4(b).
Since each phase of the generator is connected between a pair of lines,
then the magnitude of the li ne voltage is equal to the phase voltage. That
is, in a balanced delta-connected system
VL = Vp
or
VAB = V L L30° V
VCA = V L LI50° V
VBC = V L L-90° V
The reader will observe that in a balanced system, the line voltages are
equal in magnitude, and are phase displaced from one another by 120°.
7.8 Three-phase, four-wire, star-star system
A typical system is illustrated in figure 7.5; the neutral point (N) of the
supply is connected to the star point (S) of the load by a line of zero
resistance.
Figure 7.5 A three-phase, four-wire, star-star system.
7.8.1 Balanced supply, balanced load
In this case each phase voltage has the same value, and each is separated
from the next phase voltage by 120°. Also, a balanced load is one in which
the magnitude of the impedance in each phase of the load has the same
value, and each has the same phase angle. That is for each load
Zp = ZpLcJ> Q
The current l BA ftowing from node a of the supply to node A of the load is
VaN V LO° V V
l BA = - = - p - - = - p L cJ> = - p (cos cJ> + j sin cJ»
Zp ZpLcJ> Zp Zp
also
VbN V L-120° V
I bR = - - = P = - p L(-120° + cJ»
Zp ZpLcJ> Zp
= Vp (cos( -120° + cJ» + j sin( -120° + cJ>))

Zp
V
= -p «-0.5 cos cJ> + 0.866 sin cJ» +
Zp
j( -0.866 cos cJ> - 0.5 sin cJ»)

and
V eN V pL120° V
lee =--= = - p L(120° - cJ»
Zp ZpLcJ> Zp
Vp
= « -0.5 cos cp - 0.866 sin cp)
Zp
+ j(0.866 cos cp + 0.5 sin cp»

The current in the neutral wire, I SN (usually described as the neutral wire
current or neutral current, IN) is
IN = I aA + I bB + I ec = 0 + jO A
That is, in a balanced star-connected 4-wire system the current in the neutral
current is zero. In such a system, there is no need for a neutral wire, and a
three-phase three-wire supply can be used (see also section 7.9). This is the
case in most industrial systems (but not in most domestic supply systems).
7.8.2 Three-phase, Jour-wire, star-connected system with an unbalanced

load
With an unbalanced load, it is gene rally the case that IN is non-zero, and is
illustrated in the following example,
W orked example 7.8.1
A 502.3 V, 3-phase, 4-wire supply is connected to a star-connected load

having the following load impedances
ZAS = lOLO° Q, ZBS = lOL20° Q, Zcs = lOL-40° Q
Calculate the current in each line and in the neutral wire.
Solution
The phase voltage is

Vp = VJV3 = 502.3/\/3 = 290 V
The circuit is gene rally as shown in figure 7.5, and the current in line A is
I A = VAN/ZAS = 290Loo/lOLO° = 29LO° = 29 + jO A
in line B is
I B = VBN/ZBS = 290L-120o/lOL20° = 29L-140°
= -22.22 - j18.64 A
and
I c = VCN/ZCS = 290L120o/lOL-40° = 29L160°
= -27.25 + j9.92 A
Figure 7.6 Phasor diagram for worked example 7.8.1.
The current in the neutral wire is
= (29 + jO) + (-22.22 - j18.64) + (-27.25 + j9.92)

= 22.25L-156.9° A
The corresponding phasor diagram is shown in figure 7.6.
7.9 Three-phase, three-wire, star-star system
In many industrial systems, both the supply and the load are balanced, so
that the neutral wire current is zero; in this case the neutral wire may be
removed without any effect on the voltages and currents in the system.
However, if either the supply or the load is unbalanced, or both are
unbalanced, then it is usually the case that the neutral point of the supply is
not at the same potential as the star point of the load. This is iIIustrated in
the following example, in wh ich both the supply and the load are unba-
lanced.
Calculate the star-to-neutral voltage, the voltage across each phase of the
load, and the current in each phase of the load in the following 3-phase,
3-wire, star-star system.
VAN = 200LlO° V, VBN = 220L-140° V, VCN = 180LlOO° V
ZAS = lOLO° Q, ZBS = 15LlO° Q, Zcs = 5L-20° Q
Solution
In this case we can use Millman's theorem to calculate the voltage between
the star point of the load and the neutral point of the supply as folIows.
111
-+-+-
ZAS ZBS V cs
200LlO° 220L-140° 180 L 100°

----+ +
5L-20°
1 1 1
---+ +---
lOLO° 15LlO° 5L-20°
29.44Lll1.9°
----= 82.23 L 102.9°
0.358L9°
= -18.36 + j80.15 V
That is to say, a voltage exists between the neutral point of the supply and
the star point of the load.
Using this value, we can ca1culate the voltage across each phase of the
load as folIows.
VAS = VAN - VSN = 200LlO° - 82.23L102.9°
= 220L-11.9° V
VBS = VBN - VSN = 220L-140° - 82.23L102.9°
= 267.63L-124.1° V
and
VCS = VCN - VSN = 180L100° - 82.23L102.9°
= 97.97 L97.57° V
The current in each line can be calculated as follows.
. VAS 220L-11.9°
I =--
ZAS 22L-11.9° A
A
lOLO°
VBS 267.63L-124.1°
IB =--= 17.84L-134.1° A
ZBS 15LlO°
c
Star point
/
/
/
/
VBS /
/
VB'
B
Figure 7.7 Phasor diagram Jor worked example 7.9.1.
and
V cs 97.97 L97.57°
Ic =-- 19.59L117.6° A
Zcs 5L-20°
The corresponding phasor diagram is shown in figure 7.7.
Note: The problem can, alternatively, be solved either by using mesh
current or node voltage analysis.
7.10 Delta-connected systems
Many three-phase loads are delta-connected rather than Y -connected, the

principal reason being that if the load is unbalanced, it can either be
connected or removed without affecting the voltage distribution at the load
end (in a 3-phase, 3-wire system, the addition or rem oval of an unbalanced
Y-connected load almost invariably produces a change in the individual
phase voltages at the load).
The reader will observe that, in figure 7.8, each phase of the load is
connected to a corresponding phase of the supply. Consequently, we can
(to a large extent) regard a delta-connected system as three single-phase
systems, in wh ich a pair of phases use one line in common to carry the line
current. Arising from these observations, we may say that
VAB
I AB =Z-
AB
Applying KCL at node A shows that

c le C
v. e ~
A IA
f'\., z.e A
I.e
I.
B B
Figure 7.8 A delta-delta connected system.
Applying it to node B yields
and at node C gives
The above equations are gene rally applicable to any delta-connected sys-
tem, whether balanced or unbalanced.
7.10.1 Balanced delta-connected load with a balanced supply
In this case
and
The magnitude of the phase current, I p , flowing in each phase of the load is
Vp VL
1=-
p Zp Zp
That is, each phase current has the same value. Now (see also figure
7.4(b))
and
Z V L150° V
ICA = ~= L =~L(150° - lj»
ZCA ZpLlj> Vp
The line current lA is calculated from
V
lA = lAB - ICA = ZL (L(30° -lj» - L(150° - lj>))
p
~/ VL VL
v3 -(cos lj> + j sin lj» = V3- Llj>
Zp Zp
= V3 IpLlj>
Similarly it may be shown that the magnitude of the other line currents (IB
and le) is V3Ip , and the phase angle between each phase current and the
associated phase voltage at the load (which is one of the line voltages) is lj>.
A balanced 500 V, 3-phase source supplies a balanced 3-phase delta-

connected load of (4 + j5) Q per phase. Calculate the current in each phase
of the load and in each line. Draw the phasor diagram.
Solution
The impedance in each phase of the load is

4 + j5 = 6.4L51.34° Q
and the modulus of the current in ·each phase of the load is

I p = 500/6.4 = 78.125 A
and the magnitude of the line current is
IL = V3Ip = V3 x 78.125 = 135.32 A
Since the load is balanced, each phase current lags behind the associated
phase voltage by 51.34°. The corresponding phasor diagram is shown in
figure 7.9. The reader should note that, since each node of the delta is
separate, it is sometimes convenient (though not strictiy true) to draw all
the phasors from a common cent re point.
7.10.2 Unbalanced delta-connected load with a balanced supply
If the load is unbalanced, it is necessary to calculate individually the phase

currents in the load, and then implement the equations at the beginning of
this seetion to calculate the line currents, as outlined in worked example
7.10.2.
The following impedances are connected in delta to a balanced 3-phase,

500 V supply
ZAB = lOL25° Q, ZBC = 15L-30° Q, ZAC = 20LO° Q
Solution
The phase currents are

VAB 500L30°
I =--= =50L5° = 49.8 + J·4.36 A
AB ZAB lOL25°
500L-90°
- - - = 33.33L-60° = 16.67 - j28.86 A
15L-30°
VCA 500L150°
I CA =Z
-- o - = 25 L 150° = -21.65 + j 12.5 A
-2-0-L-0-
CA
and the corresponding line currents are

IA = lAB -I CA = (49.8 + j4.36) - (-21.65 + j12.5)
= 71.45 - j8.14 = 71.91L--6S A
I B = I Bc- lAD = (16.67 + j28.14) - (49.8 + j4.36)

= -33.13 - j33.22 = 46.92L-134.9° A
I c = I CA - I Bc = (-21.65 + j12.5) - (16.67 - j28.14)

= -38.32 + j41.36 = 56.38L132.8° A
7.10.3 Unbalanced delta-connected load with an unbalanced supply
The solution of this type of circuit is gene rally similar to that outlined in
worked example 7.10.2, with the exception that the unbalanced line vol-
tages are used in the calculation.
7.11 Delta-connected supply and a star-connected load
This situation generally presents no problem if the supply and load are
both balanced, since it can be dealt with as though the load is supplied by a
3-phase, 4-wire supply (even though the neutral wire is absent). The reason
is that, since the load is balanced, the neutral wire current is zero and the
voltage across each phase of the load is VL tv'3.
If either the supply or the load is unbalanced, a convenient method of
dealing with the problem is to convert the load into its equivalent delta
network (see chapter 6), and deal with the circuit as described in section
7.10.
7.12 Star-connected supply and delta-connected load
Once again, the situation is fairly straightforward, because we can calculate

the line voltages (which may either be balanced or unbalanced), and then
treat the problem as outlined in section 7.10.
7.13 Summary of balanced star- and delta-connected systems
Phase Line Phase Line

voltage voltage current current
Star Vp = VJY3 V = Y3V p
L Ip =I L IL =I p
Delta Vp = V V = Vp
L L Ip = I /Y3L IL = Y Jl p
7.14 Power consumed in a three-phase system
The power eonsumed by one phase of a polyphase system is

Pp = Vp/ p eos cp
where eos cp is the power faetor of the load in that phase.
Unbalanced load
The total power eonsumed is the sum of the power in eaeh of the three
phases. That is, the total power PT is
PT=PA+PB+P c
where PA' PB and Pe are the power eonsumed, respeetively, in the A-, B-
and C-phases.
Balanced load, balanced supply
Onee again, the total power is the sum of the individual value of power in
the three phases but, in this ease, an equal value of power is eonsumed by
eaeh phase so that
PT = 3V pI p eos cp
In a Y-eonneeted load, V L = Y3V p and I L = I p and
PT = Y3V L I L eos cp
In a delta-connected load, V L = V p and I L = Y3I p, henee
That is, in any balanced 3-phase load the total power eonsumed is
Moreover, in a balaneed system, eaeh phase eonsumes not only the same
value of VA but also the same value of VAr, so that
ST = V3V L I L
PT = V3V L I L cos <p
QT = V3V L I L sin <p
Calculate the VA, the power and the VAr consumed by the balanced
delta-connected load in worked example 7.10.1.
Solution
We are dealing here with a system with a balanced 500 V supply, whose
line current is 135.32 A and the phase angle of the load is -51.34°.
Hence
ST = V3V L I L = V3 x 500 x 135.32 W = 117.19 kVA
PT = V3V L I L cos <p = 117.19 X 10 3 x 0.6247 W
=73.2 kW
QT = V3V L I L sin <p = 117.19 X 10 3 x 0.781 VAr
= 91.51 kVAr lagging
Alternatively, we may say that in the A-phase of the load
V AB = 500 ,L 30° V and I AB = 78.125 L (30° - 51.34°)
= 78.125 L -21.34° A, so that
S AB = V Auf tB = 500 L 30° x 78.125 L 21.34°
= 39.06 L 51.34° kV A
Hence
ST = 3 x 39.06 L 51.34° = 117.18 L 51.34°
= (73.2 + j91.5) kV A
or
ST = 117.18 kVA
PT =73.2 kW
QT = 91.5 kVAr (lagging)
Calculate the power consumed by the unbalanced 3-phase, 3-wire, star-

connected load in worked example 7.9.1.
Solution
The relevant data are (see also worked example 7.9.1)

V AS = 220 L -11. 9° V, IA = 22 L -11. 9° A
V BS = 267.63L-124.1° V, I B = 17.84L-134.1° A
V es = 97.97L 97.57° V, I B = 19.59L 117.6° A
The power consumed by phase A is
PA = Re(V Jt) = Re(220L -11.9° x 22L11.9°)
= Re( 4840 L 0°) = 4840 W
and by phase B is
PB = Re(VBJ;) = Re(267.63L-124.1° x 17.84L134.1°)
= Re(4774.5LlOO) = 4702 W
and by phase C
Pe = Re(V cJ~) = Re(97.97L 97.57° x 19.57L -117.6°)
= Re(1919.2L-20.03°) = 1803 W
Hence
PT = PA + PB + Pe = 11 345 W
Note: The power in each phase could, alternatively, have been calculated
on the basis of
(phase current) 2 X resistance per phase
7.15 Power measurement in three-phase systems
The majority of power measurement is carried out using analogue watt-

meters, and these have one coil to sense the voltage applied to a circuit,
and another to sense the current flowing in the circuit. The instrument
pro duces a deflecting torque proportional to the average power consumed
by the circuit.
In the special case of a balanced load, it is theoretically only necessary to
w, Cu rrent coi I
v+ u:
~r:-~--ll
'- _~_ J
I
v_ Iv..
IA
3-phase
load
I
Potential coil
v+rt- t
B (star
Potential coil
or
V_ delta)
--I V<'
C
M I_ _ ~_:.JL le
' " / Curren!

W, coil
Figure 7.10 The two-wattmeter method o[ power measurement.
have one instrument, which measures the power consumed by one phase.
The total power is three times the reading of the instrument.
In general (both for balanced and unbalanced loads) it is necessary to
connect correctly (N - 1) wattmeters, where N is the number of lines used
to supply power to the load. That is, a 3-phase, 4-wire system requires
(4 - 1) = 3 wattmeters to measure the total power, and a 3-phase, 3-wire
system needs two wattmeters to measure the total power.
7.16 The two-wattmeter method of power measurement
The total power consumed by a three-phase, three-wire system can be

measured by two wattmeters connected in any two lines. For example, they
can be connected as shown in figure 7.10.
The markings on the wattmeter terminals are as folIows. When current
ftows from terminal M to terminal L, AND terminal V + is positive with
respect to terminal V _, the wattmeter produces a positive deftection_ If the
current ftows in the reverse direction, OR the voltage polarity is reversed,
the wattmeter produces a negative torque. If both the current and polarity
are reversed, the wattmeter produces a positive deftection. For power
measurement, the usual connection between the potential coil and current
coil is as shown in figure 7.10, that is V + is linked to M.
The instantaneous total power consumption by the load in figure 7.10 is
where v AS is the instantaneous volta ge between line A and the star point of
the load (assurning for the moment that the load is star-connected), i A is
the current in line A, etc. Since the load is supplied by a 3-wire system
iB = -(i A + ic)
that is, the instantaneous power consumed is
PT = (v AS - vBs)i A + (v cs - vBs)i c
and the total average power consumed is

ß
PT = V ABI A cos a + V CBI c cos
where ais the phase angle between V AB and I A' and ß is the angle between
V CD and I c; PT is the sum of the readings of the wattmeters.
The reader is asked to note that we have not applied any conditions to
the load, so that it may either be balanced or unbalanced. There are, of
course, two other possible ways in which the wattmeters may be connected
to measure power, and a similar analysis will show that in either case the
sum of the readings of the two wattmeters is equal to the total power
consumed.
Alternatively, using complex quantities for figure 7.10, we may say
that
PT = Re(V AJ~) + Re(V cJ~)
The analysis above assumed that the load was star-connected. It will
prove an interesting exercise for the reader to show that the analysis is also
correct if the load is delta-connected.
7.17 Introduction to symmetrical components
An unbalanced set of 3-phase voltages or currents can be analysed into

three sets of balanced components, namely: .
1. A positive phase sequence (PPS) balanced system having the same phase
sequence as the original unbalanced system (say ABC).
2. A negative phase sequence (NPS) balanced system having a phase
sequence opposite to that of the original unbalanced system (say
ACB).
3. A zero phase sequence (ZPS) system whose elements are aB in phase
with one another, and have the same magnitude.
If either the supply or load is unbalanced, either two or aB three sets of the
symmetrical component elements exist.
If a 'healthy' balanced system has a fault on it (other than a symmetrical
short-circuit), either two or aB three sets of symmetrical components exist.
If only for this fact, the determination of the symmetrical components or
voltage and current is a useful weapon in the armoury of an electrical
engineer.
1__
Ic+~-----I
(c)
10-
(a) (b)
(d)
Figure 7.11 Symmetrical components o[ current.
7.18 Analysis of unbalanced conditions
If I A' I Band I c are a set of 3-phase unbalanced currents, then

I A = I A+ + I A- + I AO
I B = I B+ + I B- + IBo
I c = I c+ + I c- + I co
where I A+' I B+ and I c+ are the PPS components of the currents, I A-' I B-
and I c- are the NPS components of the currents, and I AO' IBo and I co are
the ZPS components of the currents.
Figure 7.11 illustrates the use of these equations. The PPS, NPS and
ZPS components of a set of unbalanced currents are shown in diagrams
(a), (b) and (c), respectively, and the way in which the unbalanced currents
are synthesised from them using the above equations is shown in diagram (d).
For a set of unbalanced voltages, there is a similar set of equations, that is
VA = V A+ + V A- + V AO' etc.
The phase sequence components can be calculated from the original

unbalanced components as folIows. If the complex operator h is defined as
h = 1 L 120°, then
I B+ = h 21 A+ I c+ = hI A+
I B- = hI A_ Ic_ = h 2I A_
I AO = IBo = I co
and
hence
I A = I A+ + I A- + I AO
IB = h 21 B+ + hI B- + IBO
Ic = hl c+ + h 21 c- + I co
or
I A+ = (lA + hl B + h 21 c)/3
I A- = (lA + h 21 B + hl c)/3
I AO = (I + I + I c)/3
A B
The reader should note that in a 3-phase, 4-wire system, the current in the
neutral wire is (lA + I B + I c) = 31 AO; that is to say, the current in the
neutral wire is entirely ZPS current. Similarly, we may conclude that in a
'healthy' 3-phase, 3-wire system, none of the lines carries ZPS current.
However, should a fault occur on the system (such as, for example, an
earth fault when the system has a supply with an earthed neutral point),
then ZPS current may flow in the supply lines.
In an unbalanced but 'healthy' delta-connected system, ZPS current
does not flow in the supply lines, but may flow inside the mesh. This
situation can arise, for example, in the case of a delta-star connected
transformer when current flows in the neutral wire of the secondary circuit.
A corresponding component of ZPS current flows around the closed
delta-connected primary winding in order to maintain m.mJ. balance
between the two windings.
7.18.1 Power consumed by symmetrical components
The total average power consumed in a system is the sum of the individual
powers due to the PPS, NPS and ZPS components. No average power is
associated with the voltage from one phase sequence and the current from
another phase sequence.
A 3-phase, 4-wire balanced supply of line voltage 502.3 V supplies the

following line currents
1A = 29 L 0° A, 1B = 29 L -140° A, 1C = 29 L 160° A
Calculate the PPS, NPS and ZPS components of the line current, and
determine the associated power consumption.
Solution
From the equations derived earlier

1 A+ = (lA + hl B + h 21 c)/3
= (29 L 0° + [1 L 120° x 29 L -140°]
+ [1 L 120° x 29 L 160°])/3
= 29(1 L 0° + 1 L -20° + 1 L 40°)/3 = 26.32 L 6.3° A
1 A- = (I A + h 21 B + hl c)/3
= (29 L 0° + [1 L 240° x 29 L -140°]
+ [1 L -240° x 29 L 160°])/3
= 29(1 L 0° + 1 L 100° + 1 L -80°)/3 = 9.67 L 0° A
1 AO = (I A + 1 B + 1c)/3
= 29(1 L 0° + 1 L -140° + 1 L 160°)/3 = 7.42 L -157° A
Since the supply is balanced, V A- = 0 and V AO = O. The total power is
therefore supplied by the PPS component of the current as folIows. Now
V A+ = V p LO° = (502.3/V3) L 0° = 290 L 0° V
hence
PT = 3 Re(V A+1 A~) = 3 Re(290 L 0° x 26.32 L -6.3°)
= 22788 W
Unworked problems
7.1. The voltage V AB in a balanced 3-phase system is 173.2 L 50° V.

Determine the value of V CN.
[100 L 140° V]
7.2. If the source in problem 7.1 is a negative phase sequence source,

calculate V CN'
[100 L -400 V]
7.3. Calculate the active (in-phase) and reactive current components in
each phase of a Y-connected, 12 kV, 3-phase alternator supplying a
5 MW load with apower factor of 0.7 lagging. If the magnitude of
the line current remains unchanged, and the load power factor is
raised to 0.9 lagging, calculate the new output power from the'
alternator .
[In-phase component = 240.6 A, quadrat ure component = 245.5 A;
6.43 MW]
7.4. A balanced Y-connected load of (6 + j8) Q per phase is connected
to a balanced 3-phase, 500 V supply. Calculate the magnitude of the
line current, the power factor, and the total VA, power and VAr
consumed.
[28.9 A; 0.6; 25 kVA; 15 kW; 20 kV Ar]
7.5. A phase sequence indicator is an instrument which can be used to
determine the phase sequence of a polyphase supply. A simple
indicator comprises the star-connected combination of a 19 !J,F
capacitor connected between phase A and the star point, an electri-
cal lamp of resistance 200 Q connected between phase Band star,
and an identical lamp connected between phase C and star (the star
point of the indicator is not connected to the neutral point 01 the
supply). If the line voltage is 200 V, 50 Hz, calculate the voltage
across the lamp connected to the B-phase when the supply is
(a) PPS, (b) NPS.
[(a) 182.1 L -104'so V; 51.36 L 152.40 V]
7.6. A balanced mesh-connected load of (8 + j6) Q per phase is supplied
by a balanced 500 V, 3-phase supply. Calculate the modulus of the
phase current, its power factor, and determine the total apparent
power, power and reactive volt-amperes.
[50 A; 0.8 (lagging); 43.3 kVA; 34.64 kW; 25.98 kVAr]
7.7. The currents in branches AB, BC and CA of a mesh-connected
system supplied by a symmetrical source of phase sequence ABC are
as folIows:
AB: 50 A at apower factor of 0.8 lagging
BC: 60 A at apower factor of 0.7 leading
CA: 40 A at unity power factor.
Calculate the current in each line.
[I A = 30 L -900 A; 1B = 72.9 L 88.40 A; 1C = 42.9 L -92.60 A]
7.8. An unbalanced set of phase voltages are

V AN = 200 L 0° V, V BN = 200 L - 90° V, V CN = 200 L 90° V
and the corresponding line currents are
I A = 30 L 0° A, I B = 40 L -145° A, I C = 40 L 125° A
Calculate the symmetrical components of voltage and current,
together with the power supplied by each set of symmetrical
components, and the total power consumed.
[V AO = 66.67 L 0° V; V A+ = 182.14 L 0° V; V A- = 48.8 L 180° V; I AO
:;: 9.17 L 159° A; I A+ = 35.65 L -7.2° A; I A- = 3.42 L 20S A;
P us = -1712 W; P pps = 19326 W; P NPS = -469 W; PT = 17 145 W]
7.9. Three delta-connected impedances of (40 - j20) Q per phase are
connected in parallel with three Y -connected impedances of
(20 + j40) Q per phase. If the line voltage is 500 V, calculate the
magnitude and phase angle of the line current, and determine the
total apparent power, the power and the reactive power con-
sumed.
[20.42 A; -8.13°; 5.9 kVA; 5.84 kW; 0.83 kVAr]
7.10. A 300 V, 3-phase supply of sequence ABC is connected to the load
in figure 7.12. Neglecting instrument losses, determine the power
indicated by the wattmeter.
[187.3 W]
40 0 j40 0 M L -j40 0
A~-Ölv_1
v+ L..: __ .J
B
200 200
400
Co------i-....J-----'
Figure 7.12
7.11. The circuit in figure 7.13 is energised by a 200 V, 3-phase supply of

sequence ABC. Calculate the reading of the wattmeter.
[493 W]
7.12. A symmetrical 500 V, 3-phase, 50 Hz supply is connected to the
circuit in figure 7.14. Determine the magnitude of the voltage
between points XY, YZ and ZX for the phase sequence (a) ABC,
(b) ACB.
[Ca) 0 V; (b) 500 V]
M L
A
c
150 -j200
Figure 7.13
A
X
B
Y
C
Z
Figure 7.14
7.13. A balanced star-connected load, each phase contammg a 15 Q

resistor in parallel with an inductive reactance of 5 Q, is supplied by
a balanced 3-phase, 3-wire, 200 V supply. If the line connecting the
supply to the load has a resistance of 0.5 Q per phase, calculate a
complex expression for I A' together with the power consumed by the
load and the total power loss in the transmission line.
[23.47 L -660 A; 2480 W; 826 W]
7.14. Repeat problem 7.13 but for a delta-connected load having an
impedance in each phase consisting of a 15 Q resistance in parallel
with a - j5 Q reactance.
[23.47 L 660 A; 2480 W; 826 W]
8
Two-port Networks
8.1 Introduction
Apart is a pair of terminals in a network where a signal may either enter or

leave. A two-port network is a dass of multi-port networks, which have
more than one port. Figure 8.1 shows a general block diagram of a
two-port network, which has voltage VI between the terminals of one port,
and V 2 between the terminals of the other port. A condition which must be
satisfied in all two-port networks is 1 1 = 1 3 and 1 2 = I •.
Two-port networks are important building blocks in a wide range of
applications, induding electronics, automatic contral systems, communica-
tion circuits, transmission and distribution systems, etc.
There is a wide range of parameters which can be used to define the
operation of a two-port network, and are selected according to the ease
with which they can be applied to a particular situation. In this chapter we
will be looking at admittance parameters, impedance parameters, hybrid
parameters and transmission parameters.
The parameters described in this chapter have particular application to
electrical and electranic engineering; for example, admittance parameters
have special use in high-frequency circuits (transistor amplifiers, RF appli-
cations, etc.), hybrid parameters are particularly useful in describing
transistor characteristics, transmission parameters are important in trans-
mission line calculations, etc.
Figure 8.1 A /Wo-port network.
171
Whenever we deal with most two-terminal networks, it is the usual

convention to assume that current flows into the upper terminal 01 both
ports. However, in the case of the transmission parameters, which are
largely concerned with transmission lines, the current 1 2 ftows out of the
upper terminal of the right-hand port.
In every case, we assume that the two-terminal network is composed of
linear elements and contains no independent sources; dependent sources
are permissible. We may therefore assume that I. and 1 2 are produced by
the superposition of two components, one produced by V. and the other
by V 2 •
It is usual to think of V. and V 2 (or I. and 1 2 ) as being produced by ideal
sources, but this is not necessarily the case. Moreover, a network may be
terminated at either end by another 2-port network, which may further be
connected to another 2-port network, etc.
8.2 Input impedance, output impedance, voltage gain, current gain

and power gain
Input impedance and admittance
When dealing with the input impedance of a 2-port network, we are

concerned only with the V-I relationship at the input port. It is therefore
only necessary to think at this stage about a simple one-port network.
Quite simply, we can define the input impedance of the port as
input voltage V.
Z in = -.-=-___c:..-.
mput current I.
Similarly, we may define the input admittance as
input current I.
y. = =-
m input voltage V.
Consider the 2-port passive network enclosed in the broken lines in

figure 8.2. A load is usually connected to the output terminals of the
network before it can be used, and we will assume that a 2 Q load is
connected. In order to write the mesh equations for the network, we will
assume that a current 13 circulates in a clockwise direction around the mesh
containing the 10 Q, 20 Q and 5 Q resistors. It should be pointed out that
we have adopted the 2-terminal network convention that currentflows into
the output terminals. This is taken account of in the resulting mesh
equations for the circuit (including the load), which are
Two-port Networks 173
r - - - - - - - - .....,
1 2-port network I
lv1
'J Load
20
L ________ ---l
Figure 8.2 A passive two-port network.
VI = 101 1 101 3
o -101 1 + 351 3 + 5(-1 2 )
o - 51 3 - 7( - 12 )
If we let VI = 1 V, and solve the equations for I}> we get 11 = 0.1467 A.
Aeeordingly
Zin = V/lI = 1/0.1467 = 6.82 Q
and
Y in = 1/Zin = 0.1467 S
Output impedance
The output impedanee of a network is simply the Thevenin equivalent

impedanee between the output terminals. There are a number of ways of
ealculating Z out> one of which (after removing the load from the eireuit) is
to replace the input supply souree by its internal impedance, and drive the
output terminals by a 1 L 0° A current. The potential difference between
the output terminals is equal to Zout.
In the case of the network in figure 8.2, we can obtain the output
impedanee by observation as follows. Sinee the network is energised by a
voltage souree, the output impedanee of the network is seen to be a 5 Q
resistor in parallel with a 20 Q resistor (the 10 Q resistor being short-
cireuited by the internal impedanee of the voltage souree), that is
Zout = 5 X 20/(5 + 20) = 4 Q
If, on the other hand, the network had been energised at the input by an
ideal eurrent souree, we would have open-eireuited the input terminals
when evaluating the output impedanee, giving
Zout =5 x (20 + 10)/(5 + [20 + 10]) = 4.29 Q
Yet another method of ealculating the output impedanee is to energise the
input of the network and determine (i) the open-eireuit output voltage
V 20C , (ii) the short-eireuit output eurrent 1 2SC ; the output impedanee is the
ratio of these two values. If 1 2SC is assumed to flow out of the network,
then
and if 1 2SC is assumed to flow into the network, then

Zout = V2oc/(-/2SC)
We illustrate the latter approach using nodal analysis applied to the

network in figure 8.2. If the input is energised by alL 0° A eurrent source,
the nodal equations with the output terminal open-cireuited (1 2 = 0) are
1 1 = 1 = 0.15V 1 - 0.05V 2oc
12 = 0 = -0.05V 1 + 0.25V 20C
Solving gives V 20C = 1.429 V.
The short-eircuit (V 2 = 0) output current when I 1 = 1 L 0° A ean be
evaluated from the following nodal equations
I 1 = 1 = 0.15V 1 - 0 1 2SC = -0.05V 1 +0
giving 1 2SC = -0.3333 A, henee
Zout = V 2OC/( -/ 2SC ) = 4.29 Q
Finally, another method of evaluating Zout is to replaee the input source by

its internal impedance and, with the load disconneeted, drive the output
terminals with a current of 1 L 0° A; the voltage (in volts) between the
output terminals is equal to Zout (in Q).
V oltage gain
The voltage gain, G v , of a network is given under normal operating

eonditions with the load connected by the ratio
G v = V 2 /V.
The reader will note that we use G for gain; it is a convention to use Gin
this way, and the reader should not eonfuse it with G for conduetanee.
To determine its value for the passive network in figure 8.2, let us drive
the input with a current of 1 L 0° A, and determine the voltage from the
nodal equations, which are
I. = 1 = 0.15V. - 0.05V 2
and 0= -0.05V 1 + 0.75V 2
From the second equation we see that
G v = V 2 /V. = 0.05/0.75 = 0.06667
Although, in this case, V 2 is less than V. we, none-the-Iess, refer to the

ratio as the voltage gain.
Current gain
The current gain, G., of a four-terminal network is calculated under

normal operating conditions (with the load connected) from the ratio
G. = /2//'
If we energise the passive network in figure 8.2 by a voltage of 1 L 0° V,
and write down the mesh equations we obtain
V. = 1 = 10/. - 10/ 3
o= - 10/. + 35/ 3 + 5 ( -/ 2)
0= - 5/ 3 - 7(-/2)
where /2 ftows into the network; solving for /. and /2 yields
/. = 0.147 LO° A
/2 = -0.033 L 0° A (ftowing into the network)
and the current gain of the network is
G. = /2//' = -0.033/0.147 = -0.224
The reader will have observed that we could have calculated /. from
/. = V.lZin = 1/6.82 = 0.147 A
Power gain
The power gain, G p, of a network is given by the ratio

G = output power P out
p •
mput power P in
Assuming sinusoidal excitation, the power gain is
G - Re[-V 2 /*]
2
P -
Re V.I.
[ *]
The negative sign in the numerator arises from the fact that the output
current is assumed to ftow into the output terminals of the 2-port net-
work.
There are several ways of calculating the power gain, one of the simplest
being
In the case of the network in figure 8.2, the power gain is

G p = 10.0667 x (-0.224)1 = 0.0149
Although the values above were calculated for a passive network, the
techniques involved are applicable to networks containing dependent
sourees.
8.3 Admittance parameters or y-parameters
In this case the network is defined by the equations

I. = YIIV. + y 12V 2
1 2 = Y 2. V. + Y 22 V 2
The y-parameters are simply constants of proportionality, and have dimen-
sions of siemens (S). We may write the equations in general matrix form as
folIows.
The parameter Y 11 can be evaluated by measuring V. and I. when the

output terminals are short-circuited (V 2 = 0). That is
I.=YII V .+ O
or
Similarly, we may calculate the other y-parameters as follows

I.
Y.2=-
V2
" "
Y12 V, Y21 V,
Figure 8.3 A two-port, y-parameter equivalent circuit.
" '2
200
1005H
Figure 8.4 Calculation 0/ y-parameters.
Because each parameter has the dimensions of admittance, and each

is obtained by short-circuiting either the input or output port, the
y-parameters are also known as the short-circuit admittance parameters.
More specifically, the parameters are known respectively as
Y 11 is the short-circuit input admittance
Y zz is the short-circuit output admittance
Y 12 and Y Zl are the short-circuit transfer admittances
There are several equivalent circuits corresponding to the y-parameter
equations, one circuit being shown in figure 8.3.
It is pointed out here that the parameters y 1\1 Y 12> Y 21 and y 22 are
sometimes known as parameters Yi, Y n Y rand Y 0 ' respectively, corre-
sponding to input, reverse, forward and output parameters.
As a simple example, we will calculate the y-parameters for the passive
2-port network in figure 8.4.
Now, Y 11 is the input admittance with the output terminals shorted, that
is its value is (1110 + 1120) S, or
Y11 = 0.15 S
To calculate Y 12' we short-circuit the input terminals, and apply 1 V to the
output terminals. The current wh ich ftows in the short-circuit in the
direction of 1 1 is equal to the value of Y IZ. Inspecting figure 8.4, we see that
the current whichjlows out of the input terminals is 1120 A. However, we
define 1 1 as the current jlowing into the input terminals, hence
Y12 = -0.05 S
Similarly, to calculateY21' we short-circuit the output terminals and apply

1 V to the input terminals. The current which flows in the short-circuit in
the direction of 1 2 is equal to Y 21. By a similar argument to that above, we
see that
Y21 = -0.05 S
The parameter Y 22 is the admittance between the output terminals with the
input short-circuited which, by observation, is
Y22 = (1/5 + 1/20) = 0.25 S
The y-parameter equations for the 2-port network in figure 8.4 are
therefore
11 = 0.15V 1 - 0.05V 2
1 2 = -0.05V 1 + 0.25V 2
and
0.15 -0.05]
[
[y] = -0.05 0.25
It is not a coincidence that y 12 = Y 21 in the above calculation. Elements
such as resistors, inductors and capacitors (other than, perhaps, electro-
lytic capacitors) can be connected in a circuit in either direction, and the
result is the same. These elements are known as bilateral elements. A
network containing only bilateral elements is known as abilateral network,
and for such a two-port it can be shown that Y 12 = Y 21. Some two-ports
contain non-bilateral elements, such as dependent sources; these networks
also have this property. A two-port for which Y 12 = Y 21 is known as a
reciprocal network.
Determine the y-parameters of the network in figure 8.5, which is a

small-signallinear equivalent circuit of a transistor in the common-emitter
mode with resistive feedback between the collector and base. Calculate the
1.5 kn
" "
V,!600 n

voltage gain and the current gain of the 2-port network when a 2 kQ
resistor is connected to the output terminals.
Solution
Whatever circuit we analyse with the y-parameters, we must stricdy adhere

to the parameter definitions. The parameter Y 11 is the input admittance
measured with the output terminals short-circuited. Since, in this case, the
dependent source in figure 8.5 is shorted, then
Y11 = (11600 + 111500) = 2.333 mS
The parameter Y 12 is equal to the current which ftows in a short-circut
between the input terminals in the direction of I I when 1 V is applied to the
output terminals. Since V I = 0, no current ftows in the dependent source,
and
Y12 = -111500 S = -0.6667 mS
The parameter Y21 is the current ftowing in a short-circuit between the
output terminals in the direction of 1 2 when we apply 1 V to the input
terminals. Since V 2 = 0, the current ftowing through the 1.5 kQ resistor
towards the output is 111500 A. The current in the current source is 0.04V I
= 0.04 A, hence the current ftowing in the short-circuit between the output
terminals gives the value
Y21 = (0.04 - 111500) S = 39,33 mS
Finally, Y 22 is the admittance between the output terminals when the input
terminals are shorted. Since V I = 0, no current ftows through the depen-
dent source, and the 600 Q resistor is short-circuited. Hence
Y22 = 1110 000 + 111500 = 0.767 mS
The y-parameter equations which apply to figure 8.5 are therefore
II = 2.333VI - 0.6667V2 mA
12 = 39.33VI + 0.767V2 mA
In order to calculate the voltage and current gain, we need to connect the
2 kO load, and supply a current to the input; let this be 1 mA. That is
II 1 mA and V2 = -200012 A = -U2 mA
or -0.5V2 mA
Substituting these values into the y-parameter equations for the network
gives
1 X 10- 3 2.333 X 1O- 3 VI 0.6667 X 1O- 3 V2
o 39.33 X 1O- 3 VI + 1.267 X 1O- 3V2
From the second of these equations we get

V2 1.348
Gv = - = - - - = -31.06
VI 0.0434
and from the first of the equations
1 x 10- 3 = 2.333 X 10- 3 X [ -~] - 0.6667 X 1O- 3 V 2

31.06
- 0.472V2
or -1.35 V
hence -0.5V2 = 0.675 mA
therefore GI = I/lI = 0.675/1 = 0.675
8.4 Impedance parameters or z-parameters
Once again, input and output voltages and currents are assigned as in
section 8.1. The impedance parameters are specified by the following
equations
VI zn/l + Zl2 / 2
V2 zu/l + z22 / 2
where
VI
Zn
II
12 = 0
VI
Zl2 =--
12
II = 0
V2
ZU
II
12 = 0
V2
Z22 =--
12
II = 0
Since the parameter values are calculated for zero current at either input or
output, the z-parameters are known as open-circuit impedance parameters,
all having dimensions of ohms. More specifically
I
I, I,
Figure 8.6 A z-parameter equivalent circuit.
Zu is the open-circuit input impedance

Z22 is the open-circuit output impedance
Zu and Z21 are the open-circuit transfer impedances
The above equations establish the equivalent circuit of the network, one
version being shown in figure 8.6.
As with the y-parameters, we often use Zb Zn zfand Zo to represent Zu,
Zl2' Z21 and Z22 respectively.
We will now calculate the z-parameters for the passive 2-port network in
figure 8.4. The parameter Zu is the input impedance with the output
terminals open-circuited, leaving 10 n in parallel with (20 + 5) n, or
Zu = 10 x (20 + 5)/(10 + (20 + 5» = 7.143 n
To calculate Zl2 we open-circuit the input terminals, and cause a current of
1 LO° A to flow into the upper output terminal: Zl2 is equal to the voltage
between the input terminals (the upper terminal being assumed positive).
With the input terminals open-circuited, the current which flows in the
10 n resistor (making the upper input terminal positive) when a current of
1 A enters the upper output terminal is
1 x 5/(5 + (20 + 10» = 0.1428 A
Hence
Zu = 10 x 0.1428 = 1.428 n
Similarly
Z21 =5 x 10/(10 + (20 + 5» = 1.428 n
and
Z22 = (5 x (20 + 10»/(5 + (20 + 10)
= 4.286 n
Once again, it will be seen that Zl2 = Z21 for a reciprocal network.
Figure 8.7 Circuit tor worked example 8.4.1.
Calculate the z-parameters of the 2-port network in figure 8.7. Hence

calculate the current and voltage gain when a 10 n resistor is connected to
the output port.
Solution
Many of the values can, in this case, be obtained by observation. The

parameter Zn is the input impedance when the output terminals are open-
circuited. By observation
Zn = 2 - j4 n
Since ZI2 is calculated with 11 = 0, then VI is equal to the voltage across the
2 n resistor, that is
and
Zu = V.lV2 = 2 n
Similarly, Z21 is calculated with 12 = 0; V2 is equal to the voltage across the
2 n resistor, hence
and
Z21 = V/li = 2 n
Since the circuit contains only bilateral elements, it is a reciprocal network,
and one would expect that Z12 = ZZl"
The parameter ~2 is the impedance between the output temlinals when
the input is open-circuited. That is
Z22 = 2 + j5 n
The z-parameter equations which apply to figure 8.6 are therefore
VI = (2 - j4)/, + 212
V2 = 21 + (2 + j5)/2
1
To determine the current and voltage gain of the complete circuit, we must
connect the 10 n load and apply a voltage of, say, 1 V to the input port. In
this case
VI = 1LO° V V 2 = -10/2
The equations which apply to the complete circuit therefore are
VI 1 = (2 - j4)/, + 212
V2 -101 2 = 211 + (2 + j5)/2
The second equation yields
12 2
G =- - - = 0.154L157.4°
I I 12 + j5
I
and the first gives

(2 - j4)/2
1 + 2/2
0.154L157.4°
or 12 = 0.036L-136S A
Hence V2 = -10/2 = 0.36L43S V
therefore
G V = V/VI = 0.36L43S
8.5 Hybrid parameters or h-parameters
Hybrid parameter or mixed parameter networks are weIl suited to bipolar

transistor circuits, because the parameters are relatively easy to measure
experimentaIly. The hybrid parameter equations for a 2-port network are
written in the form
VI hIlI + h 12 V 2
12 h 21 / 1 +h 22 V2
By the nature of the equations, it is evident that the parameters have

different dimensions, two of them being dimensionless.
The parameters h ll and h 21 are obtained by letting V 2 = 0, while h 12 and
h 22 are determined by letting I I = 0, as follows
hll = ;'1 IV
2 = 0
(ohms)
h 12
= ~: I 11 = 0
(dimensionless)
h 21 (dimensionless)
= ;: I V2 = 0
h 22
= ;21 1
1 = 0
(siemens)
The parameters are respeetively known as

hll short-circuit input impedance
h 12 = open-circuit reverse voltage gain
h 21 short-circuit forward current gain
h 22 open-circuit output admittance
In transistor analysis, the parameters h ll , h 12 , h 21 and h 22 beeome hj, h n h r
and ho , respeetively. In addition, the subseripts b, e and e are added to
indieate whether the transistor is operating in the eommon-base, eommon-
emitter or eommon-eolleetor mode. Thus, in the eommon-emitter mode,
we have the parameters hie , h re , h re and hoe .
Onee again, the equations lead us direetly to the equivalent eireuit of the
network, shown in figure 8.8.
We will analyse the passive 2-port in figure 8.2 to illustrate the appliea-
tion of h-parameters. The value of h l l is equal to the input impedanee of
the network with the output short-eireuited. By observation, this is 10 n in
parallel with 20 n, giving
hll = 10 x 20/(10 + 20) = 6.667 n
The value h 12 is eaIculated from the ratio of VI to V 2 with the input
open-eireuited, and a voltage of 1 LO° V applied to the output terminals.
I, I,
h21 I,
Figure 8.8 A 2-port h-parameter equivalent circuit.

r---------- ---,
E, h,/,
r _ _ _ _ _ _ _ _ _ _ _ _ --.J
L
This is fairly easy to calculate in this case because VI is the voltage across
the 10 n resistor in a chain containing a 10 n and a 20 n resistor between
the output terminals. That is
h 12 = 1 x 10/(10 + 20) = 0.333
h 21 is the short-circuit forward current gain, and can be calculated by
short-circuiting the output terminals, and applying a current of 1 LO° A to
the input terminals. The value of h21 is the ratio of the current in the
short-circuit which ftows in the direction of 12 in figure 8.2 to the input
current. With 1 A at the input, the current in the short-circuit is
- /2 = 1 x 10/(10 + 20) = 0.3333 A
hence
h 21 = 12 / / 1 = -0.3333
Finally, we calculate h22 by open-circuiting the input terminals and evaluate
the admittance between the output terminals, that is
1 1
h = - + 20+10 = 0.2333 S
22 5
For the hybrid parameter equivalent circuit in figure 8.9, determine an

expression for (a) the current gain 12 // 1 , (b) the voltage gain V/VI
Solution
The diagram in figure 8.9 satisfies the small-signal operation of a transistor

in any mode (provided that the parameters for the correct operating mode
are supplied), and the solutions are applicable to any mode of operation.
(a) Current gain
Applying the ruIe for current division in the output circuit we get
hence
(b) V oltage gain

Applying KVL to the input circuit yields
EI - h r V 2
11 = - - - -
Rs + h j
When KCL is applied to the output circuit, we see that
12 = h f / 1 + h OV 2
however 12 = -V2 /Z L
or
-V2
- - = hf / 1 +hVO 2
ZL
Substituting 11 from above gives
or
That is
The negative sign associated with Gv implies that there is a phase shift of
1800 between the input and the output. This value of phase shift will, of
course, be modified at differing values of frequency if ZL contains a reactive
element (see also chapters 11 and 12).
8.6 Transmission parameters
These parameters are also known as the ABCD-parameters or t-para-

meters or a-parameters. As the name implies, they are largely concerned
2-port network
v·DAi~D Dv.
" "
Figure 8.10 Transmission parameters.
with transmission lines (but are also useful in other applications including
control systems). Consequently, we are concerned with current which
leaves the output port (figure 8.10), and we define the parameters as
follows
VI = t 11 V2 + t1212
11 = t21 V 2 + 122 12
A posItIve value of 12 means that current leaves the network. The

t-parameters are related to the ABCD-parameters as follows
The parameters are calculated as follows
t11 = ~: I (dimensionless)
(ohms)
(S)
t22 = -1-
1
I (dimensionless)
12 V2 = 0
The reciprocity theorem can be used to find the relations hip between the
t-parameters of a passive two-port as follows. If we apply voltage V T to the
input terminals of the two-port, as shown in figure 8.11(a) then, since the
output terminals are short-circuited (V2 = 0), the circuit equations are
(8.1)
When V T is applied to the 'input' terminals, as shown in figure 8.11(b), and

the 'output' terminals are short-circuited, then VI = 0 and the direction of
the current reverses. The equations for figure 8.11 (b) are
,...---o--------j t" t'2 ~ __- _ n _ - - - - ,
(a) VT t
L..---o------l t2 , t 22 I-----O----J
I' , I'2
..---0------1 t" t'2 1 - -__-0-----,
(b)
L..---O-----I t2 , t 22 1----o----.J
Figure 8.11 Relationship between the t- or ABCD parameters.
(8.2)
(8.3)
From equation (8.2) we see that
I~ = tu V -rltiZ
and from equation (8.3)
= (tZI - tZZtll / t l2 )VT

The reciprocity theorem states that if the network does not contain voltage
or current sourees, then I; = I zsc . That is
= (tZI - tZZ tll /t l2 )VT
- I zsc
but, from equation (8.1) V = tlzfzsc> or
T
- Izsc (tzi - tzztll/tl2)tl2Izsc

or -1
that is
In terms of the ABCD parameters this becomes

AD - BC = 1
As a simple example, we will calculate the t-parameters of the passive
2-port in figure 8.2.
The parameter 111 is calculated by applying 1 LO° V to the input and

open-circuiting the output (/ 2 = 0), the ratio V/V2 giving the value of 111 as
follows
V2 = 1 x 5/(20 + 5) = 1/5 V
hence
111 = V.1V2 = 1/(1/5) =5
We calculate 112 by short-circuiting the output (V2 = 0) and applying
1 LO° V to the input. The current in the short-circuit (which leaves the top
output terminal) is
12 = V/20 = 1120 = 0.05 A
that is
112 = = 20 n
V/1 2 = 1/0.05
To determine 12 • we open-circuit the output (/ 2 = 0) and apply a current of
1LO° A at the input. By current division, the current in the 5 n resistor is
1 x 10/(10 + (20 + 5)) = 0.286 A
and V 2 = 5 x 0.286 = 1.43 V, giving
12 • = I/V2 = 1/1.43 = 0.7
Finally, we evaluate 122 by short-circuiting the output (V2 = 0) and connect

a 1 LO° A current source to the input. The current in the short-circuit
(leaving the output) is
12 = 1 x 10/(10 + 20) = 0.333 A
hence
122 = 1./1 = 2 1/0.333 =3
also
tl l t22 - t 12 t21 = (5 x 3) - (20 x 0.7) =.1
8.7 Relationships between the y-, Z-, h- and I-parameters
It would be very tedious to deduce the relationship between each set of

parameters, so we will merely indicate how this is done by relating the y-
and z-parameters. The complete relationship between the z-, y-, h- and
t-parameters is listed in table 8.1.
The y-parameter equations for a 2-port network are
I. = YIIV. + y 12 V2
12 = Y2I V • + Y22 V 2
Table 8.1 Transformation between y-, Z-, h- and t-parameters*
~
To
Y Z
-Z12 1
h
-h 12 t
t
-At
Yu YI2 --
Z22
-- -- -- -22- --
Az Az hu hu t 12 t 12
Y
-Z21 Zu h 21 Ah -1 t ll
Y21 Y22 -- -- -- -- -- --
Az Az h ll h ll t 12 t 12
Ah h l2 At
-Y22-
Ay
-Y12
--
Ay
Zll Z 12 --
h 22
--
h 22
-tu- --
t 21
t 21
Z
-Y21 Yu -h 21 1 1 t 22
-- -- Z21 Z22 -- -- -- --
Ay Ay h 22 h 22 t 21 t 21
1 -Y12 Az Z12 t 12 At
-- -- -- -- h ll h 12 -- --
Yll Yu Z22 Z 22 t 22 t 22
h
1 -1
~ - -t 21-
Y21 -Z21
-- - -- h 21 h 22 --
Yu Yu Z22 Z22 t 22 t 22
-Y22 -1 Zu Az -A h -h u
-- -- -- -- -- -- tu t 12
Y21 Y21 Z21 Z21 h 21 h 21
t
-A -Yll 1 -h 22 -1
--y- -- -- -Z22- - - -- t 21 t 22
Y21 Y21 Z21 Z2l h 21 h 21
* For all parameter sets, A p = PUP22 - PI2P21
Solving for VI gives
VI =
where ~y = Y1IY22 - Y12Yll

That is, the equation for VI is
V = Y22 I _ Y12 I
I ~ I ~ 2
Y Y
When we look below at the equivalent z-parameter equation we see that

VI = Z1l1, + zll12
hence
Y22 -Y12
Z =-- and Z =--
li
11 li 12
y
y
similarly
-Y21 Y11
Z
12
=--
li
and Z
22
=--
li
y y
8.8 Interconnection between two-port networks
As mentioned at the outset of this chapter, each parameter set has its own
advantages for a given application. One of these advantages is that each set
of parameters allows other 2-port networks using the same parameter set
to be interconnected in a particular way, thereby simplifying the math-
ematics involved. We look at the interconnections here.
y-parameters
These are particularly useful when ports are connected in parallel both at
the input and at the output, as shown in figure 8.12. For network L
where
and [YLl are the y-parameters for network L. Similarly for network M
Since the networks are in parallel with one another
I,
v,
Figure 8.12 Parallel connection of two 2-port networks.

and
That is
Consequently, each parameter of the resulting parallel netwo-k is the sum

of the corresponding y-parameters of the individual networks, that is
Yu = YUL + YUM' etc.
This can be extended to any number of 2-port networks connected in
parallel at the input and at the output.
z-parameters
The z-parameters are useful where the 2-port networks are connected in
se ries both at the input and at the output, as shown in figure 8.13.
In this case
and
[V) = [VL + V M] = [zL][IL] + [zM][IM]
= ([ZL] + [zM])[I] = [z][l]
so that ZlI = ZlIL + ZlIM' etc.
h-parameters
The primary advantage of h-parameters is the ease with which they may be
measured in transistors. An analysis of 2-port networks similar to that
undertaken above shows that the h-parameters of two 2-ports may be
I, = 11l I, = I"
V1l
I,
t Network L
t V'L
I,
= 11l =1" V,
V,
= I'M = I'M
v,J Network M
tV'M
/, = I'M I, = I'M
Figure 8.13 Series connection of two 2-port networks at the input and at the output.
V, DL-N,_twor_k t
L J--I :,_v-----L~tI_Net_work---,M 0 V,
Figure 8.14 Cascaded 2-port networks.
added together if the two networks are connected in series at the input (see
figure 8.13) and in parallel at the output (see figure 8.12). This is an
unusual configuration, and is little used.
t-parameters
Transmission parameters are useful where 2-port networks are cascaded as

shown in figure 8.14, such as transmission line sections. In this case, the
current leaving the output port of one network is the input current of the
next network. That is
V2 = V3 and /2 = /3
For network L
VI V2 V3
] = [tL] [ ] = [tL] [
/1 /2 /3
and for network M
That is
Consequently the t-parameters of the cascaded network is given by the

matrix product
This product is not the simple mathematical product of the corresponding

elements in the matrix, but is the matrix product, as described in chapter 15.
As a simple example of the above, we will determine the t-parameters of
the half-T network in figure 8.15(a). This can be considered to consist of
the series impedance in diagram (b) cascaded with the shunt admittance in
diagram (c). By inspection, the equations for figure 8.15(b) are
VI
/I =
(al (bI (cl
Figure 8.15 The t-parameters of a half- T network.
or
and the equations for figure 8.15(c) are
or
The reader will observe that, for both circuits
When the two are cascaded, as shown in figure 8.15(a), we may write
[;1] [~~][~ ~][;:]

[1 +y YZ ~][ ~2]
That is t n = 1 + YZ, t 12 = Z, t21 = Y and t22 = 1; also t11 t22 - t 12t21 = 1.
Unworked problems
8.1. Calculate the input impedance of the non-port network in figure

8.16.
[9.1 il]
8.2. Evaluate R in for the circuit in figure 8.17. Assurne the op-amp to be
ideal.
[2 kil]
so 40
20
Figure 8.16
ao
, kO
Figure 8.17
Figure 8.18
8.3. If all the resistance values in figure 8.16 are doubled, calculate the
input admittance of the circuit.
[0.057 S]
8.4. Determine the y-parameters of the circuit in figure 8.18 at a fre-
quency of 50 Hz.
[Y11 = 0.062L-29.8° S; Yll = 0.028L-146.4° S;
Yll = 0.028L-146.4° S; Yll = 0.039L-11.3° S]
8.5. If, in problem 8.4, VI = 10 V at 50 Hz, calculate (a) the voltage gain,
(b) the current gain, (c) Ztn, and (d) the power gain of the network if
a capacitor of reactance 20 0 is connected between the output
terminals.
[(a) 0.486L-14.2°; (b) 0.44L-63S; (c) 18.5L40.7° 0; (d) 0]
-
I 1 1000 5000 12
13
50/3
y,1 500
Figure 8.19
Figure 8.20
8.6. Evaluate the y-parameters of the 2-terminal network in figure 8.19.

[Yll = 9.3 X 10- 3 S; Y12 = -2 X 10- 3 S;
Y21 = -2.7 X 10- 3 S; Y22 = 2 X 10- 3 S]
8.7. Determine the z-parameters of the network in figure 8.20.

[ZI1 = 5.160; Z12 = 2.1 0; Z21 = 2.1 0; Zn = 24.70]
8.8. Repeat problem 8.5, but for the circuit in figure 8.20 and the
associated z-parameters.
[(a) 0.262L-50.1°; (b) 0.067 L-141.1°; (c) 5.05L-1°; (d) 0]
8.9. Calculate the z-parameters for the circuit in figure 8.21 at a fre-
quency of 50 Hz. .
[ZI1 = 5.93L-90° 0; Z12 = 5.07 L57.3° 0; Zn = 4.08L90° 0;
Z22 = 13.63L17.4° 0]
8.10. Repeat problem 8.5 but for the circuit in figure 8.21 and the associ-
ated z-parameters.
[(a) 0.69L131°; (b) 0.2L-39.2°; (c) 5.75L-80.2°; (d) 0]
8.11. If the value of each resistor in figure 8.20 is changed to 10 0,
calculate the h-parameters for the circuit.
[h ll = 160; h 12 = 0.2; h 21 = -0.2; h n = 0.06 S]
8.12. A transistor has the following h-parameters: h ll = 1200 0, h 12 =
5 x 10-\ h21 = 50, h 22 = 20 !lS. If the load connected to its out-
put terminals is 10 kO, calculate the voltage gain and the current
ga in of the transistor amplifier.
[Gy = -420.2; GI = 41.69]
/ 318/-LF 100 /,
~ I--"""--Lj---o
v, v,
Figure 8.21
300
Figure 8.22
8.13. Determine the z-parameters of the lattice network in figure 8.22.

[Zl1 = 35 0; ZI2 = 5 0; Z21 = 5 0; Zn = 35 0]
8.14. A two-port network has the parameters t ll = 0.02; t 12 =

2 0; 121 = 2 mS; 122 = 0.1. If the input source has an internal
resistance of 10 0, and the resistor connected to the output termin-
als has a resistance of 20 0, calculate (a) the voltage gain and (b)
the current gain.
[(a) 14.29; (b) 7.143]
8.15. A communications line is represented by a symmetrical-Jt network,
in which the series element is 120L60° 0, and each shunt admittance
is 2.5 X 10- 3 L90° S. Calculate the value of the t-parameters of the
line.
[t11 = t22 = 0.755L11.4°; t 12 = 120L60° 0; t21 = 4.4 X 10- 3 L95° S]
9
The Transformer
9.1 Introduction
The transformer was briefly intraduced in chapter 4, when mutual induct-

ance was discussed. The majority of transformers have two electrical
circuits, one - known as the primary circuit - contains apower source, and
the second - known as the secandary circuit - is connected to the load.
There are two braad classes of transformer , namely the ideal transformer
and the linear transformer .
An ideal transfarmer is one in wh ich the coils are wound on an iron core,
and the magnetic coupling between the windings is near-perfect; that is,
the magnetic coupling coefficient is practically unity. This class of trans-
former includes the power transformer.
A linear transfarmer is one whose magnetic coupling coefficient is less
than unity, in fact its value can be very low indeed! This class of trans-
former includes the radio-frequency transformer.
Certain types of transformer with an iran core, such as those used in the
output circuit of an electranic amplifier, are rather less than electrically
'perfect'. The reason is that their winding resistance is generally high, and
the magnetic circuit design is such that the magnetic coupling coefficient is
less than unity.
Since the ideal (power) transformer is the type most frequently encoun-
tered, we begin our studies by looking at this type. With this sort of
transformer , we are not particularly interested in its magnetic coupling
coefficient, since we can assume its value to be unity.
9.2 The ideal transformer
An 'ideal' transformer is one in which the coils are wound on an iran co re

and the windings have a very high inductance (ideally infinite), and the
magnetic coupling coefficient is unity. In practice, such an ideal cannot be
198
The Transformer 199
achieved, but is closely approached in a well-designed iron-cored trans-

former.
Assuming that the magnetic flux in the core varies sinusoidally, the
instantaneous flux is given by
cp = CPm sin wt
where CPm is the maximum value of flux in the core, and the instantaneous
value of the induced e.m.f. in a coil of N turns is
e = - N dcp/dt = - NwCPm cos wt
Since the form of the equation for the induced e.m.f. in each winding is the
same, it follows that the e.m.f.s induced in the windings are in phase with
one another. However, this is a philosophical argument because, unless the
windings are electrically connected to one another at a common point we
cannot, at this stage, be absolutely certain wh at the phase relationship iso
Let us assume, therefore, that the induced voltages are in phase with one
another. Moreover, since the winding resistance of apower transformer is
very small, we can say that the voltage across the winding is
v = -e = NwCPm cos wt
The maximum value of the voltage across the coil is
V m = NwCPm
and the r.m.s. voltage in the coil is
V2 x N X 2nfCPm = 4.44NfCPm
Since we are dealing with an ideal transformer whose magnetic coupling
coefficient is unity, then we may say that this voltage is induced in every coi!
on the core. Provided that the winding resistance is low (wh ich is usually
the ca se in apower transformer) then, to a first approximation, the voltage
across the coil is equal to the e.m.f. induced in the coil. That is, in coil 1
VI = IEII
and in coil 2
and we mayaIso say

VI = 4.44NdCPm
V 2 = 4.44N2 fCPm, etc.
Hence
V2 N2
-=-=a (9.1)
VI NI
(a) (b)
Figure 9.1 Simple equivalent circuits af the ideal transformer.
where a is known as the turns ratio of the transformer. An equivalent

circuit corresponding to equation (9.1) is shown in figure 9.1(a). The
equation may be re-written as follows
VI V2
NI N2
That is, each winding in an ideal transformer supports the same number of
volts per turn. This applies even if the transformer has many separate
windings on its magnetic core.
Furthermore, since the transformer is ideal, the windings have little
resistance, and the power loss in them is very low. Assuming that we are
dealing with a two-winding transformer , we may say, to a first approxima-
tion
VIII COS cP = Vzf2 COS cp
where cp is the phase angle between the voltage and current. Hence each
winding in an ideal transformer supports the same number of volt-amperes,
that is
V2 11
-=-=a (9.2)
VI 12
An equivalent circuit corresponding to equation (9.2) is shown in figure
9.1(b). Combining equations (9.1) and (9.2) gives the following rela-
tionship for an ideal transformer
V2 11 N2
-=-=-=a
VI 12 NI
From this we see that
IINI = I 2N 2
That is, each winding in an ideal transformer supports the same number of
ampere-turns.
The Transfarmer 201
9.3 Phasor diagram for an ideal transformer on no-load
It was shown in seetion 9.2 that the voltage aeross eaeh winding on the
transformer were all in phase with one another and, with a sinusoidal ftux
waveform, the voltage has a eosine waveform. That is, the valtage across
each winding leads the flux wavefarm by 90°. Consequently, the phasor
diagram for the primary and seeondary winding for a transformer on
no-load is as shown in figure 9.2.
The phasor diagrams for the primary and seeondary windings are shown
separately in figure 9.2 beeause the two windings are eleetrieally isolated
from one another.
To maintain the magnetie ftux in the eore, the primary winding earries a
na-Iaad current, 10 , This eurrent eonsists of two eomponents:
(1) The magnetising current campanent, I mag , whieh lags behind the eur-
rent by 90° and produces the ftux qJ.
(2) The care lass campanent, le, whieh is in phase with V, and supplies the
power loss dissipated in the magnetie co re (known as the care lass or
iran lass).
The phase angle of 10 is ct>o whieh, in an ideal transformer, approaehes 90°,
that is
and the eore loss is

Po = V,Ie
When an induetive load is eonneeted to the seeondary of the transformer ,
the phasor diagrams are shown in figure 9.3. The seeondary eurrent, 12 ,
lags behind the seeondary terminal voltage, V2, by ct>2' Sinee ampere-turn
v,
90'
"""-_ _ _ _ _ _.1.-_______ <P
Jmag
(a) (b)
Figure 9.2 No-load phasor diagram for an ideal transformer (a) primary winding, and (b)
secondary winding.
------ --,
v, I,
v,
I: _- I I,
/
/
I
/
I
/
'I o
" - - - ' - - - - - - - - - _ - <P

(al (bl
Figure 9.3 Phasor diagram Jor a transJormer with an inductive load (a) primary winding,
(b) secondary winding.
balance is maintained on the two windings, a corresponding current I;

ftows in the primary winding where
I;N I = 12 N 2
or
I; = 12 N 2 /N I
Accordingly, I; is sbown lagging VI by CP2. The current in the primary
winding is the phasorsum of 10 and I; as follows
11 = 10 + I;
Figure 9.4 shows a simplified equivalent circuit of apower transformer
which represents the phasor diagram in figure 9.3. The resistor Re is the
path through which the core loss current ftows, and X m is a pure inductive
reactance through which the magnetising current ftows (no magnetising
current ftows in the ideal transformer itself). The value of Re and X m are
determined by means of a no-load test or open-circuit test on the trans-
former , in which the primary winding is excited at its nominal voltage and
frequency, at which time the no-load current and power loss, Po, are
measured. From these values we can say
CPo = cos-I(PO/VJo)
le = 10 cos CPo
I mag = 10 sin CPo
Re = VI/le
The Transformer 203
I, 1', I,
10
mag
'
V, Re Xm V, Load
--------
'Ideal' transformer
Figure 9.4 Simplijied equivalent circuit 0/ apower trans/ormer taking account 0/ the
no-load current.
9.4 Leakage Dux in a transformer
When a load is connected to the secondary of a transformer , as shown in

figure 9.5, a current 12 flows in the secondary.
Although the coils are wound on an iron core, the coupling coefficient in
a practical power transformer is not quite unity, so that some of the flux
produced by eaeh winding links only with that particular coil, and with no
other coil. Consequently, this flux induces an e.m.f. only in the originating
coil. Moreover, one-half of the magnetic path for this flux lies in the air
surrounding the coil, and the reluctance of this flux path is very high
indeed. This flux, which is produced by each current-carrying winding, is
known as the leakage flux associated with that winding; because of the high
reluctance of the magnetic path, the leakage flux is proportional to the
current in the coil. In figure 9.5, the primary winding leakage flux is cI>u.
and the secondary winding leakage flux is cI>L2'
Since each leakage flux is linked with one winding only, it induces a
'back' e.m.f. in that winding, wh ich opposes the current flow in the wind-
ing. The greater the leakage flux, the greater the voltage drop due to this
cause. Good transformer design aims to reduce the leakage flux to a low
level.
Unfortunately, the simple winding construction in figure 9.5 gives a
large leakage flux, and is not used in practice. The leakage flux is kept to a
minimum either by winding both coils on the same li mb (the shell eonstrue-
tion), or by winding one-half of each winding on each of the verticallimbs
(the eore eonstruction).
Main flux <P
,-- -1
I, I r\
1
(\ I ('\
I 1
t <Pu <P L , ~ tv,

I I
I I
V V V I
I I
'--------_./
Figure 9.5 Leakage flux.
As mentioned above, the leakage flux has the effect of inducing a back
e.m.f. in the winding which produced it. We regard this voltage drop as
being due to a leakage reactance, which can be thought of as being in series
with each winding of the transformer. That is, in figure 9.6, we have a
primary winding leakage reactance Xl> and a secondary winding leakage
reactance X 2 •
A 'practical' transformer can be thought of as an ideal transformer
together with circuit elements which allow for no-load current, winding
resistance, leakage reactance, etc. The sole function of the ideal trans-
former is to carry the main flux and to provide a 'turns ratio'; it neither
stores nor dissipates energy, so that it does not draw magnetising current,
has no hysteresis loss, and has no PR loss in its windings.
A fairly complete form of equivalent circuit for the transformer discus-
sed so far is shown in figure 9.6, in which Xl and X 2 are the respective
primary and secondary winding leakage reactances, R l and R 2 are the
primary and secondary winding resistance, and Re and X m are as described
earlier.
9.5 Inipedance matching with an ideal transformer
An iron-cored transformer is widely used as an interface device between

the output of an electronic amplifier and a loudspeaker; both of these
usually have widely differing values of resistance. The reader will recall
that, for maximum power transmission, the resistance of the load should
have the same resistance as the output resistance of the source (or the
amplifier in this case).
The Transfarmer 205
"
N, V, Load
\...--..J
'Ideal' transformer
Figure 9.6 A complete equivalent circuit o[ a single-phase trans[ormer.
In electronic circuits the load is frequently regarded as a pure resistor,

and the secondary current is
12 = V 2 /R L or RL = V 2 /12
For an ideal transformer
11 = al2 and VI = V 2 /a
The apparent resistance, R 1 , 'seen' by the primary supply source is
Suppose that a loudspeaker of resistance 8 Q is to be matched to an

electronic amplifier of output resistance 100 Q. To obtain maximum power
transfer from the amplifier to the load, we need to satisfy the condition
100 = R L /a 2 = 8/a 2
where a is the turns ratio of the interface transformer. That is
a =N 2 /N1 = 1/v'(lOO/8) = 1/3.54
that is N 1 /N2 = 3.54.
An audio-frequency transformer is generally less 'perfect' than apower
transformer , and it will be found that impedance matching is a little less
perfect than suggested here!
25',
B
500n
'3
t
0.5V
rv
~
"
(b)
Figure 9.7 (a) A dass Apower amplijier and (b) an equivalent circuit.
An equivalent cireuit of a 'dass A' electronie power amplifier is shown in

figure 9.7(a). Calculate the eurrent in the 8 Q load if the r.m.s. input
voltage is 0.5 V. Also determine the power gain of the amplifier.
Solution
The transformer wh ich couples the amplifier to the load has a voltage
step-down ratio of 15:1 (a = 1/15), and is replaced in figure 9.7(b) by an
equivalent circuit (see also figure 9.1(a». The equation for the mesh
containing 11 is
0.5 = 500/ + 40(1 +
1 25)/1 = 1540/1
or 11 = 3.25 X 1O~4 A
and the input power to the circuit is
0.5 x 3.25 X 1O~4 = 1.625 X 1O~4 W
In the mesh in which 12 flows we have
13 /15 = 25/ 1
The Trans/ormer 207
or 13 = 375/1 = 0.122 A
with a corresponding output power of
0.1222 x 8 = 0.119 W
Hence Gp = 0.119/1.625 X 10-4 = 732.3
9.6 The ideal transformer as a two-port network
The ideal transformer can conveniently be described as a 2-port network

using the t-parameters as follows (see chapter 8 for details).
.
Smce a = -N z = -V 2 = -,
II t h e V - I reIatlOns
' f or
NI VI 12
the transformer are VI = V 2 /a and I I = a12 , so that the t-parameter equa-
tions for the ideal transformer are as follows
that is 111 = lIa, 112 = In = 0 and 122 = a.

If the secondary winding of the transformer has a load connected to it,
as shown in figure 9.8, the t-parameters of the load (see also chapter 8) are
111 = 1, 112 = 0, 121 = Y L and 122 = 1. The matrix equation for the trans-
former and load in figure 9.8 is therefore
= [lla
aYL
that is
VI = V4 /a
I. = Y L aV4 + al4
~I
Figure 9.8 A loaded transformer.
Since no other load than Y L is connected to the secondary, then 14 = 0 and

1
where ZL = l/YL • This theme is capable of development, and a loaded

transformer can be analysed by means of a chain of t-parameters or ABCD
parameters.
9.7 Thevenin's equivalent circuit of an ideal transformer and voltage

regulation
We can simplify the solution of many problems involving transformer

circuits by using equations developed earlier, together with TMvenin's
theorem.
If we consider the circuit in figure 9.9(a), in which we have an ideal
transformer with an impedance ZI in the primary circuit (which may, for
example, represent the winding resistance and leakage reactance of the
transformer), and a load ZL in the secondary circuit, we may simplify the
complete circuit by 'reftecting' the load impedance into the primary circuit,
as shown in figure 9. 9(b ).
The 'reftected' load impedance is Z L la 2 (see section 9.6), wh ich is the
input impedance of an ideal transformer with a load ZL connected to its
secondary terminals. Now 1I = Vs/(ZI + Z Lla 2 ) and I z = lila, also
V~ = IIZLla2, hence
V2 = aV~ = IIZLla
Similarly, we can 'reftect' the primary circuit values into the secondary
circuit as folIows. From figure 9. 9(b)
1I = Vs/(ZI + Z Lla 2 )
and since 12 = lila, then
a = ~
v.li ~i~ z,~ v.1

Z
' v L - -- - - "
(a) (b) (c)
Figure 9.9 An ideal transfarmer with an impedance in the primary and in the secandary
circuit.
The Transformer 209
a2Z 1 + ZL
The corresponding equivalent circuit for values reflected into the sec-
ondary circuit is shown in figure 9.9(c).
Power engineers are concerned with maintaining the electrical voltage
at the consumers terminals. The variation in voltage at the terminals of the
consumer is often quoted in terms of the voltage regulation as foBows
modulus of the modulus of the

Per unit no-load voltage fuB-load voltage
voltage
regulation modulus of the no-load voltage
Voc - VL
V oc
If ZL represents an impedance which draws fuB-load current from the
transformer, the equivalent circuits in figure 9.9 can be used to calculate
the voltage regulation of the transformer (in many cases the voltage regula-
tion is given as a per cent figure ).
Referring to figure 9.9, the equation for voltage regulation is
for diagram (b)
Voltage regulation = (Vs - V~)/Vs
for diagram (c)

Voltage regulation = (aV s - V 2 )/aVS
The reader should note that if the dot notation in figure 9.9 is reversed on
one of the windings, there is no change in the magnitude of the voltages
and currents involved, and no change in the voltage regulation. However,
the secondary voltage and current are phase shifted by 180°.
W orked example 9. 7.1
A 100 kV A, 3300/250 V single-phase transformer has a total equivalent

resistance and leakage reactance of 3 Q and 12 Q, respectively, both
referred to the primary winding. What is the fuB-load current at (a) 0.8
power factor lagging, (b) 0.8 power factor leading? Calculate the voltage
regulation in each case.
Solution
The modulus of the full-Ioad secondary current is

12 = 100 000/250 = 400 A
Hence
(a) for a lagging load of 0.8 power factor
11 = 400L-36.9° A
(b) for a leading load of 0.8 power factor
11 = 400L36.9° A
The voltage regulation is calculated as follows. The magnitude of the
impedance connected to the secondary winding to produce full-Ioad cur-
rent is
ZL = 250/400 = 0.625 Q
At this point we have the choice of calculating the voltage regulation based
on values referred to the primary winding (see figure 9.9(b», or to values
referred to the secondary winding (see figure 9.9(c». Selecting the former,
we need to calculate the load impedance referred to the primary winding,
which is
z~ = ZL/a2 = 0.625/(250/3300)2 = 108.9 Q

(a) For a 0.8 power factor lagging load
Z~ = 108.9L36.9° = 87.1 + j65.39 Q
By potential division in figure 9.9(b)
Z'
V' = V
+ Z'L
L
1 S Z1
= 3300 108.9L36.9°
(3 + j12) + (87.1 + j65.39)
= 3026L-3.76° V
Hence
. 3300 - 3026
Voltage regulation = 3300
= 0.083 per unit (p.u.) or 8.3 per cent
That is, the secondary voltage is reduced by 8.3 per cent when full load
(lagging) is connected.
The Trans/ormer 211
(b) For a 0.8 power factor leading load

Z~ = 108.9L-36.9° = 87.1 - j65.39 Q
By potential division in figure 9.7(b)

Z'
V~ = V_---=::.L_
S ZI + Z'L
3300 108.9L-36.9°
(3 + j12) + (87.1 - j65.39)
= 3431L-6.25° V
Hence
. 3300 - 3431
Voltage regulatiOn = 3300 = - 0.04 or - 4 per cent
That is, the secondary voltage rises by 4 per cent when the leading load is
connected.
9.8 The linear transformer
The windings on a linear transformer are wound on an air core, and the
magnetic coupling coefficient between the windings usually has a low
value. The linear transformer is widely used in electronic and communica-
tions circuits, possibly with both primary and secondary circuits being
tuned or resonant.
Figure 9.1O(a) shows a two-winding linear transforrner, having primary
and secondary winding resistance together with a load impedance. Using
the dot notation, outlined in chapter 4, the equivalent circuit of diagram
(a) is shown in diagram (b). The mesh equations are
VI = (R 1 + jwL1)II - jwMI2
o = - jwMII + (R 2 + jwL 2 + ZL)I2
R, R, R, R,
iWL~
I,
Zl
'V ti wM1 '
(a) (b)
Figure 9.10 (a) A linear trans/ormer, (b) an eleetrieal equivalent cireuit.

212 Eleetrieal Cireuit Analysis and Design
Using the notation adopted for mesh analysis, we may say that
Zn = R I + jwL I and Z22 = R 2 + jwL 2 + ZL
The mesh equations can therefore be re-written in the form
VI = Zn I I - jwMI2
o - jwM/ I + Z2J2
hence 12 = - jwM/I/Zn , or
VI = Zn/l + w2~/2/Zn
That is to say, the input impedanee of a linear transformer is

VI w2M 2
Zin = - = Zn + --
II Z22
The input impedance clearly consists of the primary winding self-
impedance, Zn, in addition to an impedance w 2 M 2/Z n , known as the
refleeted impedanee, which is inversely related to the total impedance of the
secondary circuit. Writing
Z22 = R 22 + jX22
then we may rationalise the denominator (see chapter 15 for details) of the
reflected impedance term as follows
That is, if the secondary circuit has a net inductive reactance, it refleets a
capacitive reactance into the primary circuit. However, Zn has a sufficient-
ly large inductive reactance for Zin to remain inductive (this is illustated in
worked example 9.8.1). Similarly, it may be shown that if the secondary
circuit has a net capacitive reactance, it reflects an inductive reactance into
the primary winding.
The reader should note that when an 'ideal' transformer is loaded with a
reactance of any kind, the reflected reactance is of the same kind.
The circuit values for a linear transformer of the type in figure 9.1O(a) are
LI = 75 mH, L 2 = 150 mH, M = 80 mH and R I = R 2 = 10 Q. Calculate
the input impedance of the circuit at a frequency of 1000 radis for a load of
(a) 0.4 H inductance, (b) a 2.5 I-lF capacitance.
The Trans/ormer 213
Solution
The self-impedance of the primary circuit is

Zn = R 1 + jwL 1 = 10 + j(1000 X 7.5 X 10- 3 )
= 10 + j75 Q
Also
(a) Inductive load
The load impedance is
ZL = jwL = j 1000 X 0.4 = j400 Q
and
Z22 = R 2 + jwL 2 + ZL
= 10 + j(1000 X 150 X 10- 3 ) + j400
= 550.1 L88.96° Q
hence
Zin = Zn + (WM)2/Z n
= (10 + j75) + 6400/550.1 L88.96°
= 10.21 + j63.38 Q
(b) Capacitive load
The load impedance is
Zc = lIjwC = 1Ij(1000 X 2.5 X 10- 6 ) =- j400 Q
hence
Zn = (10 + j 150) - j400 = 10 - j250
= 250.2L-87.7° Q
giving
Zin = Zn + (wM)2/Zn
= (10 + j75) + 6400/250.2L-87.7°
= 11.03 + j100.6 Q
The reader will note that, in both cases, the resistance of the secondary
circuit is reflected into the primary circuit as an increase in the resistance
element of the input impedance, that is, the input resistance is higher than
the 10 Q resistance in the primary circuit.
In case (a), the effect of the inductive load is to reduce the input
inductive reactance below the j75 Q reactance of the primary circuit alone,
while in case (b) the capacitive load has the opposite effect.
(c)
(b) L,L,-M'
-M-
" "
L,L,-M' L,L,-M'
Iv,
v'l L,-M L,-M (d)
Figure 9.11 (a) Idealised linear transformer, (b) an equivalent circuit, (c) the electrically
connected T-equivalent circuit and (d) the TT-equivalent circuit.
9.9 T - and :n:-equivalent circuit of a linear transformer
Although a linear transformer has two magnetically coupled, but electrical-

ly isolated windings, it is sometimes convenient to view them as though
they are electrically coupled.
If our coupled circuit has the dot notation in figure 9.11(a), its equiv-
ale nt circuit is shown in figure 9.11(b), whose mesh equations are
VI = jwLl l - jwMI2 = jw([L 1 - M] + M)/I - jwMI2
V2 = - jwM/ I + jWLJ2 = - jwM/I + jw([L 2 - M] + M)/ 2
If we look at these equations as though they refer to electrically coupled
elements, we conclude that the circuit is the one shown in figure 9.11(c).
The corresponding :n:-equivalent circuit is shown in figure 9.11(d).
If the position of the dot on either of the windings in figure 9.11(a) is
reversed, that is, the connections to any one winding are reversed, the
T-equivalent circuit would have an inductance of - M in the common
vertical branch, an inductance of (LI + M) in the top left-hand branch,
and an inductance of (L 2 + M) in the top right-hand branch.
Unworked problems
9.1. Calculate 12 in figure 9.12. The supply frequency is 100 rad/so

[8.19L35° mA]
The Trans/armer 215
Figure 9.12
- v, L R, /,
M~
~
---
L, R, /,
V,
Figure 9.13
Figure 9.14
9.2. Calculate the input impedance of the circuit in figure 9.12.

[296L41° Q]
9.3. Deduce an expression for V, and V2 in figure 9.13.
[(R, + jwL,)I, - jwMI2 ; jw(M - L 2 )I2 ]
9.4. If the source in figure 9.14 is (a) a 5LO° V voltage source (positive
terminal at the top), (b) a 5 L 0° A current source (current ftowing
upwards), calculate V AB if the frequency is 1 rad/so
[(a) 0.588L-101.3° V; (b) 3L-90° V]
9.5. If a load of 20 Q is connected between A and B in figure 9.14,
calculate V AB for both sourees.
[(a) 0.584L -107° V; (b) 2.986L - 95.7° V]
9.6. Calculate the total power consumed in figure 9.15 if / = 100 Hz.
[0.41 W]
Figure 9.15
Figure 9.16
9.7. Calculate the current I in figure 9.16 if the supply frequency is 1000
radis.
[O.03LO° A]
9.8. The primary winding of a coupled circuit has a resistance of 20 Q and
a self-inductance of 6 mH. The corresponding values for the sec-
ondary circuit are 10 Q and 4 mH, respectively. If the mutual
inductance between the two windings is 2 mH, and a pure resistive
load of 40 Q is connected to the terminals of the secondary winding,
calculate th~ input impedance if the supply frequency is 5000 radis.
[36.49 L53S Q]
9.9. Calculate the load current in problem 9.8 if (a) a voltage source of 10
L 0° V, (b) a current source of 1 L 0° A is connected to the primary
winding.
[Ca) 5.09L14.74° mA; (b) 0.186L68.2° A]
9.10. The primary inductance of a coupled circuit is 0.5 H, that of the
secondary is 0.1 H, and the mutual inductance is 60 mH; the primary
winding resistance is 12 Q and that of the secondary is 20 Q. If a
capacitor of capacitance 1 IlF is connected in series with the primary
winding, and a capacitor of 0.2 IlF is connected between the ter-
minals of the secondary winding, calculate the input impedance at
(a) 5000 radis, (b) 800 radis.
[Ca) 2478L89.2° Q; (b) 850L-89.2° Q]
The Transformer 217
9.11. A 50 kVA, 3300/250 V single-phase transformer has a primary

winding resistance of 5 Q and a secondary winding resistance of
0.03 Q. If the total leakage reactance referred to the secondary
winding is 0.13 Q, calculate the voltage regulation of the transformer
for full load at apower factor of (a) 0.8 lagging, (b) unity power
factor.
[(a) 10 per cent; (b) 4.7 per cent]
10
Transient Solution of
Electrical Circuits
10.1 Introduction
In electrical circuits containing one or more energy storage elements, a

transient state exists whenever the circuit conditions change. A circuit
containing one storage element is described as a first-order circuit or a
single-energy circuit, and one containing a capacitor and an inductor is
known as a second-order circuit or double-energy circuit, which is character-
ised by a linear differential equation containing a second-order derivative.
Transient analysis of circuits is probably one of the most difficult areas
of work covered in a course and, by its nature, involves a higher mathema-
tical conte nt than many other topics. In many cases, transient analysis is first
approached via c1assical mathematical methods and, later, by the D oper-
ator method. However, using these methods, many solutions are obtained
via a 'back door approach', and one needs to know many special techni-
ques (or should they be called 'tricks of the trade'?) to obtain the correct
answer.
Moreover, with some forcing functions, solutions cannot be obtained
using c1assical methods. Also, if initial conditions exist in the circuit, such
as a charge on a capacitor or an initial current in an inductor, obtaining a
solution using c1assical methods can be very tedious.
If only to highlight some of the difficuIties wh ich may be encountered,
we will take a brief look in sections 10.2 and 10.3 at the solution of first-
and second-order circuits, respectively, using classical methods. Following
this we turn our attention to solution by the Laplace transform method,
since this has a number of advantages from an engineering viewpoint.
However, it should be pointed out that there is a small 'drawback' ofthe
Laplace transform method, and this is that one has to traverse a basic
mathematical 'fog' before one enters the calmer waters of electrical cir-
218
Transient Solution 0/ Electrical Circuits 219
cuits. The reader is encouraged to understand the basic mathematics of

Laplace transforms; in any event, he can refer back to the mathematics
when studying the worked examples. Generally speaking the mathematics
involves integral and differential calculus, but no more than is found in a
first year subject covering these topics.
10.2 C1assical solution of first-order systems
The circuits in figures 10.1 and 10.2 contain a single energy storage element
(see chapter 4), in which Vs is the supply voltage and i is the circuit current
at time t seconds after closing the switch. At this time voltages VR , VL and Vc
are across the components.
t = 0 ---Vs
Figure 10.1 Simple first-order R-L system. Figure 10.2 Simple first-order R-C system.
Inductive circuit
For Figure 10.1, by KVL
But
di
V = L-
L dt
and therefore
di
Vs = iR + L
dt
or
Vs . L di
-=1+-- (10.1)
R R dt
Capacitive circuit
For Figure 10.2, by KVL
= iR + Vc
Now, since
dt
then
dv c
V s = RC--+ v (10.2)
dt c
In general
dx
F=x+r- (10.3)
dt
where F = driving function = steady-state value of x, in which x = circuit
variable, r = circuit time constant and dx/dt is the first derivative.
Using the 0 operator, where 0 = d/dt, then F = x + rOx.
The solution of linear, first-order equations of this type, where the
variables can be se para ted on the two sides of the equation consists of two
parts
where x t is the transient solution obtained by making F = 0, and x ss is the

steady-state solution, that is, the value of x as t ~ 00.
For the transient solution of the equation,
dx t
Xt + T--= 0
dt
X tis always of the form Ae -tl" where A is a constant to be determined
from the circuit's initial conditions, that is, the value of x at t = O.
In general, the complete solution is of the form
x = Ae- tl, + X ss (10.4)
Transient Solution 01 Electrical Circuits 221
Constant "Voltage source, v s = E volts, and the switch is closed at t = 0
For the inductive circuit shown in figure 10.1, from equation (10.1)
L di E
i+--=-
R dt R
or i + rDi = final current where r = UR
Since the inductor has no voltage across it when the CUTTent is steady,
E
iss = -
R
hence the complete solution for the current in the circuit in figure 10.1 is
E
i = Ae-t/~ +- (10.5)
R
For the capacitive circuit shown in figure 13.2, from equation (10.2)
dv c
Vc + RC-= E
dt
so
vc + rDv c = final voltage
Since the capacitor is fully charged in the steady state, no current ftows
in it and V ss = E.
The complete solution for the capacitor voltage in the circuit in figure
10.2 is
Vc = Ae- tiT + E (10.6)
Initial conditions
Substitute the values of i and v at the instant the switches are closed in the
two circuits. Let these be zero and VI volts respectively.
Since the current through the inductor and the voltage across a capacitor
cannot be changed instantly, then, at t = 0, in equation 10.5 for the
inductive circuit in figure 10.1
E
0= A +
R
that is
E
A=--
R
Hence
E
i = -(1 - e- RtIL ) (10.7)
R
In equation 10.6 for the capacitive circuit in figure 10.2
Vl =A +E
then
Hence
v =E - (E - Vl)e-tIRC (10.8)
In both cases the solution is of the form:
instantaneous value = final value - step size x e -tiT
which is illustrated graphically for two cases in figure 10.3. In diagram (a),
final value > initial value, while in diagram (b), final value < initial
value.
Final value
Initial value
Time
(a)
Initial
value--"
Circuit
variable Step
size
Time
- - - - - _.=-::=-=.--_ _-1
Final value
(b)
Figure 10.3 Transients in a first-order system with an initial condition.

Transient Solution oJ Electrical Circuits 223
While these solutions are relatively straightforward, circuit analysis

becomes increasingly difficult if the network is 'excited by, say, a sinewave
or a complex function, particularly when initial conditions are involved.
10.2.1 The time constant 1: of a circuit
In many first-order circuits, a quantity known as the time constant, r, gives

an indication of the transient performance of the circuit, particularly when
it is excited by a d.c. source.
Where the response of a circuit is described by the expression
J(t) = Ae -tiT
the value of J(t) falls from a value equal to A at t = 0, to zero when t
approaches infinity. More interestingly, J(t) reaches a value less than 0.01
A when t = Sr. The period Sr is often described as the settling time of the
transient and, beyond this time, we can assurne that steady-state conditions
have been reached. Similarly, if the response is described by
J(t) = A(l - e -tIT)
the value of J(t) rises from zero when t = 0 to a final value of A when t
approaches infinity. In fact, it can be shown to rise to a value greater than
0.99 A when t = Sr. Once again, we can think of Sr as being the settling
time of the transient.
It is also useful to think of a time of 0.7r as the 'half-life' of the transient
since, in both the above cases,J(t) changes by 50 per cent for each 0.7rtime
change. Also, J(t) changes by 63.2 per cent (either a reduction or an
increase) when t = r.
In the case of a simple RC circuit
r = lIRC s (R in Q, C in F)
and in an RL circuit
r = LIR s (R in Q, L in H)
10.2.2 Circuits reducible to first-order format
Circuits which have a single energy storage element may be solved by first
obtaining Thevenin's equivalent circuit. The circuit time constants then
become LlR Th or CR Th •
10.3 Classical solution of second-order systems
The voltage equation for the series RLC circuit in figure 10.4 is given by
KVL
that is
iR + L - +-
di
dt
1
C
f idt = v s
Differentiating this equation with respect to time to give

di d 2i i dv s
R-+L-+-=-
dt dt 2 C dt
or
d 2i R di i 1 dv s
-+--+-=--
dt 2 L dt LC L dt
R L C
t =
':J
0 - Vs
Figure 10.4 Second-order series RLC circuit.
The highest derivative in the output variable, i, is second order and the
solution is again made up of two parts: namely
The steady-state solution iss is determined by single-phase theory. For

example, when the source is a constant voltage, iss is zero because of the
capacitor.
The transient solution, i t is obtained by equating the left-hand side of the
equation to zero. For convenience in interpretation, the coefficient of the
highest derivative is reduced to unity. Hence, using the D operator for
clarity
Transient Solution of Electrical Circuits 225
(10.9)
The 'standard form' of this linear second-order differential equation is

(10.10)
where W n is the undamped natural frequency of the oscillations and ~ is the
damping factor.
Comparing equations 10.9 and 10.10, shows for the series circuit that
W n =_1_ and ~ =_R_= ~dE

VLC 2w n L 2 VL
The condition ~ = 0 implies zero circuit resistance.
The solutions to equation 10.10 are determined by the value of ~ as
folIows:
1. ~ < 1, damped oscillations
i l = Ae-~wn' sin(wot + 8)
2. ~ = 1, critical damping
i l = (A + Bt)e -Wn'
3. ~ > 1, overdamped response
4. ~ = 0, no damping, continuous oscillation

i l = A sin(wnt + 8)
where A, B, and 8 are constants determined by the initial conditions of the
circuit, and Wo is the frequency of the damped oscillations. This is given
by
Initial conditions
With constant voltage e volts applied to the circuit in figure 10.4, by

KVL
Ri + LDi + v c = E
Let the voltage across the capacitor be V land the circuit current be zero
before closing the switch, hence at t = 0
LDi + VI = E
. (E - VI)
Dl=~-~
L
Steady-state solution
iss = 0, since the capacitor is fully charged to E.

Hence the complete solution is given by i = i t . Substituting these values
into 1, 2, 3 and 4 above gives: .
(E - V)
i= 'te- Wnl
L
. (E - V,) .
1 = sm wnt
wnL
As with first-order systems, the solution can be very involved if the
circuit is excited by a sinusoidal signal or complex wave, particularly when
initial conditions are involved.
10.4 The Laplace transform
The Laplace transform, F(s), of a function of time f(t) is defined as
F(s) =.!l'[f(t)] = [ f(t)e-S1dt
The Laplace transform is a one-sided transform, which assumes that f(t)

only exists for t ~ O. The variable s = a + jw (see also chapter 11) is also
complex, and a restriction on its value is that its real part must be large
enough to make the integral convergent.
The lower limit on the integral is t = 0 but, because of possible
discontinuities at t = 0 we should, strictly speaking, use a value of t which is
just less than zero. We generally denote this as t = 0 -. The integral then
becomes
F(s) = L~ f(t)e-S'dt
Thus the operator2'[f(t)] transforms a functionf(t) in the time domain into
a function F(s) in the s-domain or complex frequency domain. Also, given
F(s), we can obtain f(t) by the inverse process as follows
f(t) =2'-1 [f(t)]
However, because the direct use of this method needs results from com-
plex-variable theory, we shall (at alm ost any cost!) avoid employing it.
Instead, we will develop a table of Laplace transforms together with the
functions from which they are derived, and use it whenever we need to find
the inverse of a transform. Trust an engineer to find an easier solution!
After investigating a number of special functions associated with electri-
cal circuits, we will devote so me time to the derivation of a few basic
Laplace transforms before we move on to circuit analysis.
10.5 Step, impulse and ramp functions
Any one of many forms of signal may be applied to an electrical circuit,

and we take a look here at three important forms, namely the step, the
impulse and the ramp.
The unit step signal, u( t)
The simplest step signal is the unit step (u(t», which is illustrated in figure
1O.5(a), and is mathematically described as follows
0 for - 00 < t< 0
u(t) = (
1 for 0 ~ t< 00
(a) (b)
Figure 10.5 (a) The unit step function, (b) a shifted unit step.
The function is not continuous at t = 0 and, in rigorous mathematical

terms, the derivative of u(t) does not ex ist for t = O.
We can time-shift the unit step function to cause the sud den transition
from zero to unity to occur at some time T as folIows. We define the shifted
unit step as follows
u(t - T) = jo for
1 for
-00
T
< t< T
~ t< 00
This is illustrated in figure 1O.5(b) and, in this case, the function is not
continuous at t = T.
The unit step function can also be amplitude-scaled to represent a step
of any magnitude merely by multiplying the unit step by a factor K, that is
Ku(t) or Ku(t - t I).
A forcing function frequently encountered in electrical circuits is the
unit voltage pulse; this can be represented by two shifted unit steps as
folIows. Consider the shifted unit step functions u(t - t l ) and -u(t - t 2 );
these are illustrated in figure 1O.6(a), and their sum (the unit pulse)
»
(u(t - t l ) - u(t - t 2 is shown in figure 1O.6(b). The equivalent electrical
circuit wh ich produces the unit pulse is illustrated in figure 1O.6(c).
o I--...L---.--- j,-----,--O----L.-
t, t2
-1 Vu(t-t2 )
-u(t-t2 )
(a) (b) (c)
Figure 10.6 (a) The shifted unit step junetions u(t - t l ) and u(t - t 2 ) together generate the
unit pulse jorcing funetion in (b). The eorresponding equivalent eleetrieal
eireuit is shown in (e).
The unit pulse is often used in electrical circuits as a sampling pulse. For
example, if we wish to apply a burst of high-frequency oscillations to an
electronic system for, say 0.2 !!S, we can represent the resulting wave as
v(t) = [u(t - t l ) - u(t - t 2 )]V m sin rot V
where t 2 = t l + 0.2 x 10 -6. A typical resulting waveform may be as shown
in figure 10.7.
We can also use the delayed step function to describe the operation of a
switch which closes at time t = t l as folIows. If the voltage source KV 1
shown in figure 1O.8(a) is connected to an external circuit when t = t l , the
t, t, + 0.2x 10-<
Figure 10.7 A sinusoidal high-frequency burst.
KV,t
~'
t = t,
K"(~,)I(
(a) (b)
Figure 10.8 (a) Actual switching circuit, (b) its equivalent circuit.
complete circuit can be replaced by its mathematical equivalent in diagram

(b).
The unit impulse function or delta function, ö( t)
If we differentiate the unit step function, u(t), with respect to time, its
value is zero for all time except at t = O. At the latter time u(t) is
mathematically indeterminate, and its differential is infinity, that is, the
differential is a pulse of infinite value and of zero width. However, all is not
lost, for we can look more closely at the unit step function.
A practical unit step cannot change its value from zero to unity in zero
time, so let us assurne that its change in value occurs over the period -D./2
to D./2 (see figure 1O.9(a». The derivative of the practical unit step function
is shown in figure 1O.9(b), and we see that its amplitude is lID.. Clearly, as
D. approaches zero, the amplitude of the pulse approaches infinity, thereby
approximating to the ideal impulse. Moreover, the area of the impulse in
diagram (b) is unity, hence mathematically
+
f_~ D(t) dt = J:- D(t) dt = 1

and
d
[u(t)] = D(t)
dt
m, A
(a) (b) (c)
~ o ~ _~
-
0 ~
- -~ +~
+-
2 2 2 2
Figure 10.9 (a) A praetieal unit step funetion, (b) the derivative of the praetieal unit step,
(e) a triangular pulse with similar eharaeteristies to pulse (b).
(a) (b)
Figure 10.10 (a) The unit ramp funetion r(t), (b) the magnitude-sealed and time-shifted
ramp funetion Kr(t - t 1).
The reader should note that, in many texts, it is assumed thafthe slope of
the practical unit step function commences at t = 0, and completes the
slope at t = ~. The overall result is the same as that outlined above, other
than the unit step function is time shifted by ~/2.
As with the unit step, we can scale the unit impulse to any magnitude
simply by multiplying by the factor K, that is, it becomes Kc5(t).
The unit ramp function, r( t)
The unit ramp function r(t) in figure 1O.1O(a) is defined methamatically by

0 for - 00 <t<0
r(t) = (
t for 0 ~ t< 00
As with the other 'unit' functions, we can both magnitude-scale and time-
shift the ramp function, as shown in figure lO.lO(b).
Relationship between the unit step, unit impulse and unit ramp functions
The mathematical relationship between the three functions is

Transient Solution o[ Electrical Circuits 231
d
dt
[r(t)] = u(t) r(t) = f~o u(t) dt
d
- [u(t)] = ö(t)
dt
u(t) = f~o ö(t) dt
The lower limit associated with the integral may be, in fact, any value less
than zero.
The voltage
v(t) = -2u(t) + 4r(t) - 4r(t - 8) - 2u(t - 8) V
is applied to a capacitor of 1.5 F capacitance. Determine an expression for
the current in the circuit, and sketch its waveshape.
Solution
At this stage it is helpful to sketch the voltage waveform in order to

appreciate wh at type of function we are dealing with. The step function
-2u(t) corresponds to a step of -2 V which occurs at t = 0, as shown in
figure 1O.11(a). Next, the function 4r(t) is a ramp of slope 0.5, commencing
at t = 0 (see figure 1O.11(b)). The ramp function -4r(t - 8) is a ramp of
slope -0.5 (that is, negative-going), commencing at t = 8 s (see figure
10. l1(c)). Finally, the step function -2u(t - 8) is a step of -2 V,
commencing at t = 8 s (figure 10. 11 (d)).
The forcing function is the sum of these components, and is shown in
figure 1O.11(e). The capacitor current is
dv dv
i = C - = 1.5 - A
dt dt
I
hence
d d 1
i = 1.5 [ -[ -2u(t)] +- br(t)] + -d [-2r(t
1 - 8)] + -d [-2u(t - 8)]
dt dt dt dt
and, using the relationships deduced earlier
= 1.5 [-2ö(t) + 4u(t) - 4u(t - 8) - 2ö(t - 8)]
= -3ö(t) + O.75u(t) - 0.75u(t - 8) - 3ö(t - 8)
0 2 4 6 8 10 12
I
I I I I l........L.-t
(al
-2 2u(tl
.-~
2-~
(bI
0 8
0 8 12
(cl I I~t
-2- - - - - - - - : = : = t t i t - 8 1
0 8
I
(dl I
-2- 2u(t-81
(eI i
1x-1
_2 0 4 8
Figure 10.11 The build-up ofv(t).
)"
0 2 4 6 8 10
(al I I
~36(tl
0.7~
(bI I
0 8
0 8
(cl I
-0.75----- f
0 8
(dl I
l--3.\ft-8f
(eI 07:1 ,I
Figure 10.12 Construction of the current waveform through the 1.5 F capaeitor.
The corresponding current waveform is constructed in figure 10.12. The

sudden negative-going transition of current at the beginning and end of
figure 10.12(e) would (theoretically at least) result in a current pulse of
amplitude -oo! In practice, of course, this could never happen because the
output resistance of the source would limit its magnitude.
10.6 Laplace transforms of some useful functions
In the following we develop Laplace transforms of a number of the more

useful functions in electrical engineering. These and other Laplace trans-
form pairs are given in table 10.1.
Table 10.1 Short table of Laplace transforms

f(t) =!i'-l [F(s)] F(s) =!i'[f(t)]
1 or u(t) (unit step) 1/s
A A/s
e- at lI(s + a)
ö(t) (unit impulse) 1
t u(t) (ramp) lIs 2
u(t - 1) (delayed step) e -sT/S
Rectangular pulse (1 - e -sT)/S
Lf(t) dt F(s )/s
df(t)
sF(s) - f(O)
dt
d 2f(t)
s2F(s) - sf'(O) - f(O)
dt 2
sin rot ro/(S2 + r( 2)
cos rot S/(S2 + r( 2 )
e -at ( cos rot - : sin rot )
sin (rot ± lJ) (ro cos lJ ± s sin lJ)/(S2 + r( 2)

cos (rot ± lJ) (s cos lJ ± ro sin lJ)/(S2 + r( 2)
t 1/s 2
t n (n is a positive integer) nt/sn + 1
t sin rot 2roS/(s2 + r( 2)2
t cos rot (S2 _ r( 2)/(S2 + r( 2)2
e -at(' n!/(s + a)n + 1
sinh at a/(s2 - a 2)
cosh at S/(S2 - a 2)
e- at cosh ßt (s + a)/«s + a)2 - ß2)
e -at sinh ßt ß/«s + a)2 - ß2)
e -at cos rot (s + a)/«s + a)2 + r( 2 )
e- at sin rot ro/«s + a)2 + ( 2)
Derive the Laplace transform of the unit step function u(t).

Solution
The step function is defined as follows

0 for - 00 ::::; t ::::; 0
u(t) = (
1 for 0::::; t::::; 00
hence
If the step has amplitude A, then

A
2' [Au(t)] =
s
Deduce the Laplace transform of f(t) = t n •
Solution
This expression can be integrated by parts (f u dv = uv - Jv du), so that if

u = t n and dv = e -S'dt, then v = -e -s'/s and du = nt n - ldt as follows
= 0
ntn -
+ [ - - - e-s ,
1 ]00
-
f 00 n(n - l)t n - 2
e-s , dt
S2 0 0 -S2
00 ,
n.
= 0 + 0 + ... - f --n e-s ' dt
o -s
n!
=--
Note: n is a positive integer. It also follows that

!i' [
tn - 1] - -1
(n - I)! sn
and
!i' [t] = l/s 2
The above example allows us to write down the Laplace transform of the
ramp voltage v(t) = kt. Its Laplace transform is
!i' [kt] = k/s 2
Evaluate the Laplace transform of /(t) = e at .

Solution
= [ - (s ~ a) e-(s-a)t 1: -~-a
=-s
This assumes, of course that the real part of s is greater than a. It also
follows from the above that
!i' [e- at] = 1/(s + a)
The function e -at occurs frequently in electrical circuits where l/a is the
time constant, 1', of the circuit. The function is a decaying exponential
which falls to slightly less than 1 per cent of its initial value when t = 51'
= 5/a. Similarly, the function (1 - e -at) is the equation of a rising
exponential curve, which rises to slightly more than 99 per cent of its final
value when t = 51' = 5/a.
For exponential curves of this kind, the period 51' is known as the settling
time, which is the time taken for the transient period to have practically
settled out.
Determine !i' [sin rot] and !i' [cos rot].
Solution
1
Writing sin rot as - (e irut - e-iwl) we get
2j
5f' [sin rot]
(e JWI + e - JWI) then

. . 1· .
and smce cos rot IS "2
2' [cos rot] = S/(S2 + ro 2)
Evaluate 5f' [t sin rot).
Solution
If we differentiate e -sI sin rot with respect to s, we get

d
- (sin rot e -SI) = -t sin rot e -SI
ds .
and, using the result from worked example 10.6.4
f oo
o
d d ro
- (sin rot e- SI ) dt = - - - - -
ds ds (S2 + ro 2)
or
- f OO
o
t
-2ros
sin rot e- sl dt = - - - -
(S2 + ro 2 )2
hence
Determine the Laplace transform of f(t) = sinh at.
Solution
Since sinh at = ~(e al - e -al), then
5f' [t sinh at] = ~ f~ (e at - e -al) e -SI dt
="2
1( 1
s- a- s+a
1) = S2 -
a
a2
Transient Solution oJ Electrical Circuits 237
Worked example 100607
Determine 2 [sin (wt + e)] 0
Solution
Expanding sin (wt + e) as sin eocos wt + cos eosin wt, then

2 [sin (wt + e)] = 2 [sin eocos wt + cos eosin wt]
s sin e w cos e
= +---
S2 + w 2 S2 + w 2
s sin e + w cos e
10.7 Properties of the Laplace transform
The Laplace transform has several properties of particular interest to

electrical and electronics engineers, and we will look at a number of
themo
Scalar multiplication
2 [k J(t)] = kF(s)
For ex am pie
lOs
2 [10 cos wt] = 10 2[cos wt] = - - -
S2 + w2
Addition
For example
2 [lOt + 3e -21 - 5 sin 4t]
= 102[t] + 32[e -21] - 52 [sin 4t]
10 3 20
=-+------
S2 s + 2 S2 + 16
Time differentiation
fi' [ dt 1=
df(t)
sF(s) - f(O)
where f(O) is the initial condition which exists in the circuit.
Time integration
ff [ [I\t) dt 1 ~ : F(s)
Time shift
The time-shift theorem states that if a time function is delayed by time a in

the time domain, then the result in the frequency domain is multiplied by
e _·as.
fi'[f(t - a)u(t - a)] = e-asF(s)

For example, the rectangular voltage pulse in figure 1O.13(a) can be
described by the following expression.
00 00
24684
-1
v(t)
Figure 10.13 (a) The rectangular pulse shown can be considered to be the sum of the unit
step functions in (b) and (c).
v(t) = u(t - 4) - u(t - 6)

Since the Laplace transform of u(t) is 1Is then, since u(t - 4) is u(t) delayed
by 4 s, the transform of the first delayed function is e -4o/S • Similarly, the
transform of the second delayed step function is e -6s/S , hence the overall
Laplace transform is
e -40 - e- 6s
fi'[v(t)] = V(s) = ---
s
v~"1 I 0 [,
OrT T+r 2T
Figure 10.14 The Laplace transform of a train of rectangular pulses.
This theorem can also be used to determine the Laplace transform of a

periodic time function as follows. If Jet) is a function with aperiod of T for
positive values of T, then
F(s) = F I (s)/(l - e- ST )
where FI(s) = ([(t)e- ST dt, is the transform of the period of the time
function. The (1 - e -ST) in the denominator accounts for the periodicity of
the function.
Consider the rectangular pulse train in figure 10.14. The Laplace trans-
form of the first period of the function is
That is, the transform for the pulse train is

V m (1 - e- ST )
V(s)-----
s (1 - e- ST )
Frequency shi[t
~ [e -at Jet)] = F(s + a)

This theorem states that replacing s by (s + a) in the frequency-domain
corresponds to multiplying Jet) by e -at in the time domain. Thus, making
reference to table 10.1, we have
s+a
~ [e- at cos wt] = - - - - - -
(s - a)2 + w 2
By inference we can also say that
~[eat Jet)] = F(s - a)
Frequency differentiation
dF(s)
~[-tf(t)] = -
ds
This theorem states that differentiation with respect to s in the frequency

domain is equivalent to multiplication by -t in the time domain. By
implication, we may say that
ff [tnf(t)] = (-l)n dnf(s)

ds n
For example
ff [t sin rot]
2ros
r
Frequency integration
ff[ f~)] = F(s) ds
For example
sin rot]
ff [ - - =
t
Frequency scaling
For k ~ 0
ff [f(ft)] = 1F ( ~ )
This theorem allows us to obtain the frequency-scaled version of a function
of time whose Laplace transform exists. For example, since the transform
of a 1 radis sinewave is
ff [sin t] = lI(s2 + 1)
then the transform of a 2 kHz (4000n radis) sinewave is
1 1
4000n (_S_)2
ff [sin 4000t] = - - - - - - - - - : : - - -
+1
400ün
400ün
Initial value theorem
The initial value of a function is given by

f(O+) = lim sF(s)
s~'"
where f(O+) is the value of f(t) when t is zero (or, strictly speaking,
marginally greater than zero!). This theorem states that the initial value of
the function f(t) is obtained by putting s = 00 in the equation for sF(s).
We shall illustrate this by using a function for wh ich the initial value of
f(t) is known. One example is the cosine wave, where f(O+) = 1. In this
case
and its initial value is

f(O+) = s---+
lim sF(s) = s--+
(Xl
lim (s x 00
S/(S2 + »= 1
w2
Final value theorem
Provided that all the poles of sF(s) are completely within the left-hand part
of the complex frequency plane (see chapter 11 for details), then the final
value of the function is given by
f( 00) = s~o
lim sF(s)
where f( 00) is the value off(t) when t approached infinity. That is, the final
value of f(t) is obtained by letting s be zero in sF(s).
If f(t) is a pure sinusoid, F(s) has poles on the jw axis, and use of the
final value theorem would lead us to the conclusion that the final value is
zero! However, we know that the final value of a sinusoid is indeterminate,
so beware of poles on the jw axis. The only exception is a simple pole at
s = O.
Once again, to illustrate this theorem, we will consider a function of time
whose final value we know. Such a function is f(t) = 1 - e-ar, whose final
value is unity. In this ca se
sF(s) = (1s -s+a

s - -1)
- =--
s+a
a
hence
a
f(oo) = s~o
lim sF(s) = lim - - = 1
s~o s + a
10.8 Representation of circuit elements in the s domain
So far, we only considered circuit elements which do not have initial

conditions imposed on them, that is, an initial charge on a capacitor, or an
initial current in an inductor. The Laplace transform allows us a fairly easy
method of dealing with these conditions. However, we look at basic circuit
elements before dealing with complete circuits.
Resistors
The v-i equation in the time domain for a resistor is

v(t) = Ri(t)
and the Laplace transform of this equation is
.!l' [v(t)] = .!l' [Ri(t)] = R.!l' [i(t)]
that is we may say
V(s) = RI(s)
The transfer function V(s)/l(s) = Z(s) = R is the s-domain impedance of the
element, and the ratio l(s)/V(s) = Y(s) = 1/R is the s-domain admittance of
the element. Clearly, the s-domain impedance and admittance are identical
to the time-domain values.
Inductance
The time-domain v-i relationship for the circuit in figure 1O.15(a) is

di(t)
v(t) = L -
dt
Hence
L 5L
i(t) 1(5) 1(5)
~')1 i(O)
~')1 ~')1 5L
i(O)
5
(a) (b) (c)
Figure 10.15 (a) Induetor L with initial eurrent i(O) in the time-domain, (b) s-domain
Thevenin equivalent cireuit, and (e) the Norton equivalent cireuit.
di(t)]
V(s) = 2' [ L dt = sLI(s) - Li(O)
The resulting equivalent circuit for the inductor is shown in figure 1O.15(b).
Re-writing the equation in the form
V(s) i(O)
I(s) = -+-
sL s
gives the s-domain equivalent circuit in figure 1O.15(c).
If no initial current ftows in the inductor then, in the case of figure
1O.15(b), the initial condition generator is short-circuited and, in the case
of figure 1O.15(c), it is open-circuited.
Capacitance
The time-domain v-i equation for the circuit in figure 1O.16(a) is

dv(t)
i(t) = C - -
dt
-
i(t) C
o---l"--il
~"l
v(o)
(al
Figure 10.16 (a) Time-domain circuit [or a capacitor with an initial voltage v(O) between
its terminals, (b) the Norton s-domain equivalent circuit and (e) the s-domain
Thevenin cireuit.
that is
I(s) = sCV(s) - Cv(O)
from which the equivalent circuit in figure 1O.16(b) is drawn. Re-writing
the equation in the form
I(s) v(O)
V(s) = -+-
sC s
from which we may deduce the equivalent circuit in figure 1O.16(c).
10.9 Introduction to analysis of first-order circuits using the Laplace

transform
We will approach transient analysis of networks by considering systems

containing a single energy storage element. The solutions provided are
intended to give an indication of typical solutions, but are by no means
exhaustive.
Determine for figure 1O.17(a) an expression for i(t) and vL(t) if (a) EI = 0,
(b) EI = -20 V. In both cases, calculate i(t) when t = 0.01 s. The blade of
the switch changes from A to B when t = 0; in both cases it can be assumed
that the switch is ideal, and the changeover occurs instantaneously.
(a) (b)
Figure 10.17 (a) Circuit [or worked example 10.9.1, (b) a s-domain equivalent circuit.
Solution
This example will illustrate how easy it is with Laplace transforms to insert
initial conditions into the circuit.
Assuming that steady-state conditions have been reached before the
switch blade changes from A to B, the initial current in the circuit is
i(O) = E/R. The circuit equation (see figure 1O.17(b)), when t ~ 0 in the
s-domain, is
lOu(t) + Li(O) = (R + sL)/(s)
or
lOu(t) = (20 + O.ls)/(s) - O.li(O)
(a) In this case 1(0) = 0/20 = 0 and the s-domain equation for the circuit is
10
-; = (20 + O.ls)/(s)
hence
10 100 0.5 0.5

I(s)
s(20 + 0.1s) s(200 + s) s 200 + s
= 0 5 ( -.l _ _
1 _)
s 200+s
The penultimate step in the above equation is obtained by the method of

partial fractions (see chapter 15 for details), but any other method can be
used.
Using the table of Laplace transforms we see that
i(t) =2'-1 [/(s)] = 0.5(1 - e- 2oo,) A
so that when t = 0.01 s, the current is
i ool = 0.5(1 - e- 200 x 001) = 0.432 A
We can evaluate the initial and final value of the current as follows
100
initial value = lim s/(s) = = 0
HOO 200 + s
100
final value lim s/(s) = 0.5 A
HOO 200 + s
Also vL(t) = L di/dt, then
100 10
VL(s) = L(s/(s) - f(O» = L 200 + s = -20-0-+-s
hence from the table of Laplace transforms

vL(t) = lOe- 2OO! V
The corresponding waveforms for i(t) and vL(t) are shown in figure
1O.18(a) and (b), respectively.
(b) In this case, the initial current (in the direction shown in the figure) is
i(O) = -20/20 = -1 A. The s-domain equation for the circuit is
-10 = (20 + 0.1s)/(s) - 0.1( -1)

s
or
10 _ 0.1
s 10 0.1
I(s) = - - -
20 + 0.1s s(20 + 0.1s) 20 + 0.1s
ill)
0.5
0.4 8
0.3 6
0.2 4
0.1 2
0 +----.-----.-----.----.---.-- 1 Ims) 0 +----.----.--...,..::::::::;::==r-- 1 Ims)

0 5 10 15 20 25 o 5 10 15 20 25
la) Ib)
Figure 10.18 (a) i(t) and (h) vL(t) [ar part (a) o[ worked example 10.9.1.
100 1
s(200 + s) 200 + s
0.5 0.5 1 0.5 1.5

------
s 200 + s 200 + s s 200 + s
henee
i(t) =g>-l [fes)] = 0.5 - 1.5e-2OOt A
and the eurrent when t = 0.01 s is
i O.01 = 0.5 - 1.5e-2 = 0.297 A
Onee again we ean ealculate the initial and final value of the eurrent as
follows.
1.5s
initial value lim sF(s) = 0.5 - -1 A
...... 00 200 + s
1.5s
final value = lim sF(s) = 0.5 - = 0.5 A
...... 0 200 + s
The s-domain expression for the voltage aeross the induetor is
1.5s )
VL(s) = L(sF(s) - t(O» = 0.1 ( 0.5 - - (-1)
200 + s
= 0.1 ( 1.5 - 1.5s ) = 0.15 ( 1 _ __

200+s
s_)
200+s
30
200 +s
hence
VL(t) =!i'-l [VL(s)] = 30e-2OOt V
A diagram showing i(t) and vL(t) for the initial condition in the problem is
given in figure 1O.19(a) and (b), respectively.
i(tl
0.5
O+--...J7fI.:.....-...1..-_...1..--....L..._ _. t (msl 20
-0.5 10
-1
--+--,---,---,--=-r--t (msl
5 10 15 20
(al (bi
Figure 10.19 (a) i(t) and (b) vL(t) for part (b) of worked example 10.9.1.
Determine an expression for the voltage across the capacitor in figure

1O.20(a) if the capacitor is initially charged to 40 V with the polarity shown.
Also evaluate an expression for i(t).
Solution
In this example we will illustrate the use of the parallel-connected initial

condition source. The s-domain circuit is drawn in figure 1O.20(b), in which
""'I[]40y
i(t) 10
10
s
100
(a) (b)

the initial. voltage across the capacitor is converted to its initial current
source. The latter corresponds to a current source of Cv(O) = 80 A. The
nodal equation for diagram (b) is
5L' [lOu(t)] + 80 = ( lo + 2S)V((S)

or
10 + 80 __ 1 + 20s
- - - V((s)
s 10
therefore
100 800
V((s) = s(1 + 20s) + 1 + 20s
5 40
-----+----
s(0.05 + s) 0.05 + s
=100(~-
s
1
0.05 + s
)+ __ 4_0_
0.05 + s
Hence the voltage across the capacitor is

vl(t) = 100(1 - e-{)05') + 40e-{) 05, V
= 100 - 60e-{) 05, V
The resulting graph is shown in figure 1O.2l.
Since i(t) = C dvc(t)/dt, the current in the capacitor can be calculated
using the time-differentiation property
I(s) = C(svtCs) - t(O))
v(t) v
,,-/ .....
/<100(1-e- 005t )
- v
/
/
---
40 'f. / 40e- 005' V
20 I '-....c
-+----r-----:;:=-=--.---..,......----I_ t(5)
o 20 40 60 80
Figure 10.21 Time variation of the vo/tage across C for t > O.

5 40s )
=2 ( + -40
0.05 + s 0.05 + s
2
= (5 + 40s - 40(0.05 + s»
0.05 +s
6
0.05 +s
that is
i(t) = 6e-D05t A
10.10 Sinusoidal excitation of first-order systems
Here we open our account with systems excited by an alternating current

source. We illustrate this by means of an RC circuit in wh ich the capacitor
carries an initial charge.
Determine an expression for i(t) in the circuit of figure 1O.22(a)
Solution
The equivalent s-domain circuit is drawn in figure 1O.22(b) in which the

alternating voltage source is expressed in the form
2' [20 sin 2t] = 20 x 2/(S2 + 22 ) = 40/(S2 + 4)
and the initial charge on the capacitor has the value
10
20 sin 2t.u(t) t
15 V 40
5'+4
(a) (b)
Figure 10.22 (a) Electrical circuit, (b) s-domain circuit.

v(O)/s = 15/s
The loop current equation for the circuit is
I(s) ( 10 + 10 ) = ~ _ 15
s r +4 s
where (10 + lO/s) is the s-domain impedance of the circuit. The first part of
the right-hand side of the above equation is the result of the a.c. source,
and the second part is produced by the initial charge on the capacitor. We
will keep the two parts separate during the early section of the solution so
that we can study the individual results.
s 40 s 15
I(s) = - - - - - - - -
(lOs + 10) (S2 + 4) (lOs + 10) s
4s 1.5
---------
(s + 1)(r + 4) s+1
= [_~+ 0.8s 3.2]_~ +

s+1 r+4 s+1
The bracketed part of the solution was obtained by partial fractions (see
chapter 15 for details), and relates to the effect ofthe a.c. source. From the
table of Laplace transforms, the solution is
i(t) = (-0.8e- t + 0.8 cos 2t + 1.6 sin 2t) - 1.5 e-t
= (-0.8e- t + 1.789 sin(2t + 26.57°» - 1.5 e-t
At this point, the reader will observe that the action of connecting the a.c.
source appears to produce a 'd.c.' transient current (-0.8e- t ) with a time
constant of 1 s, which will decay to zero in about 51' = 5 s. The time
constant is the natural time constant of the circuit (liRe), and the initial
'd.c.' current of -0.8 A just compensates for the value of 1.789 sin(2t +
26.57°) A when t = 0; the net result is an initial a.c. current of zero! The
complete equation is therefore
i(t) = -2.3e-t + 1.789 sin(2t + 26.57°) A
The reader should note that the steady-state a.c. component of i(t) can be
predicted quickly using conventional theory as folIows. The impedance of
the circuit is
Z(jw) = R - j/wC = 10 - j/(2 x 0.1) = 10 - j5
= 11.18L-26.57° Q
and
l(jw) = V(jw)/Z(jw) = 20/11.18L-26.57°
= 1.789L26.57° A
The value of current calculated here is the peak value because we have
used the peak value of the alternating voltage.
10.11 Solution of series second-order circuits using Laplace transforms
A second-order or double-energy circuit is one containing two different

types of energy storage element, such as Land C. The series circuit in
figure 10.23 is typical of this kind; initial conditions may, of course, exist in
the circuit. The s-domain impedance of the circuit is
---
Figure 10.23 Double-energy series circuit.
V,(t)
Z(s) = R + sL + 1/sC
and the current in the circuit is determined from the expression
l(s) = !l' (Vl(t)]/Z(s)
The s-domain expression for the current depends, of course, on the nature
of the supply voltage. For the moment, we will assume that a step function
of voltage is applied to the circuit, that is !l'(Vl(t)] = V/s; in this cast;, the
equation for l(s) is
V/s V/L
l(s) = ---=---- = ---=----
R + sL + s~ r + f + lc
s
V/L
(10.11)
The denominator of the final expression represents the characteristic equa-

tion in what is known as the standard form, in which wn is the undamped
natural frequency, and , is the damping factor of the circuit, which may be
zero or any positive value. From the denominator of the above equations
Wo = l/\I(LC)
and 2'wo = R/ L
or ,= ~ V[~ ]
The response of the circuit is generally described in terms of the value of ,
namely
1. ,= ° (undamped response)
° < ,< 1 (underdamped response)
,=
2.
,>
3. 1 (critically damped response)
4. 1 (overdamped response)
For a step voltage applied at t = 0, the response is as shown in figure 10.24.

The reader will note that the undamped and underdamped response are
oscillatory, implying that the characteristic equation of the system has
complex poles in the left-hand half of the s-plane (see chapter 11).
°
The underdamped response is of particular interest to engineers because
it corresponds to many practical systems. In this case < , < 1, that is
~ V[ ~ ]< 1 or R < 2 \I(LlC). In this case, the coefficient of s in
the equation for I(s) is finite. Completing the square in the donominator
of equation (10.11) gives
V/L
I(s) = - - - - - - ' - - - -
(s + 'woY + (w~ - '2W~)
If w~ = w~ (1 - '2) or Wo = Wo \1(1 - '2) where Wo is the damped
oscillatory frequency or the oscillatory frequency, then
VJL woVI/L
I(s) = - - - - -
(s + 'w
o )2 + w~ wo[(s + 'woY + w~]
and, from the table of Laplace transforms, we see that
VI .
sm w
i(t) = -- e-~wnt t
woL 0
Also, from the equation for Wo, we see that when , = 1 then Wo = 0, that is,
sin wot = 0. That is, when the damping factor is unity, the output variable
i(t) i(t)
(a) (b)
(c) (d)
Figure 10.24 Response of the circuit in figure 10.23 to a step change in applied voltage for
(a) undamped response, (b) underdamped response, (c) critical response,
(d) overdamped response.
(i(t) in this case) just reaches its steady-state value without either over-
shoot or undershoot (see also figure 1O.24(c)).
We will look at each of the four types of response in worked exarnple
10.11.1.
For aseries circuit of the type in figure 10.23, in which L = 2 H, C = 2 F

and V 1(t) = 10 V, determine expressions for i(t) if the resistance R is (a)
zero, (b) 1 Q, (c) 2 Q and (d) 3 Q.
Solution
(a) R = 0 (undamped response)

In this case
C= ~ V[ Z] = 0 and Wn = lIV(LC) = 0.5 radis

hence
V/L 10/2
I(s) = - - ----=
r + w~ r + 0.25 0.5 (S2 + w~)
From the table of Laplace transforms we see that

i(t) = 10 sin 0.5t A
That is, the current has a peak value of 10 A, and it oscillates at a
frequency of 0.5 radis.
(b) R = 1 Q (underdamped response)
~= ; V[ ; ] = 0.5 and wn = 0.5 radis
consequently
10/2 5
I(s) =
r + (2 x 0.5 x 0.5s) + 0.5 2 r + O.5s + 0.25
Completing the square in the denominator gives
5
I(s) - - - - - - - - - -
(s + 0.25)2 + (0.25 - 0.25 2 )
5
(s + 0.25)2 + 0.1875
5 0.433
=-- --------
0.433 [(s + 0.25)2 + 0.433 2]
= 11.55
[(s + a)2 + w~]
where Wo = 0.433 and a = 0.25. The table of transforms gives

i(t) = 11.55 e~·25r sin 0.433t A
(c) R = 2 Q (critically damped response)
~= ; V[ ; ] = 1, wn = 0.5 radis and Wo = 0, that is
no oscillations. The s-domain expression for the current is

V/L V/L 10/2
I(s) = =---=---
r + 2wns + w~ (s + W n )2 (s + 0.5)2
I!
=5
(s + a)l+l
and from the table of Laplace transforms we see that

i(t) = 5te -D5t A
The general form of response is shown in figure 1O.24(c). This is a case
where we should be rather careful when using the concept that 'the tran-
sient has decayed in about 5 time constants'. In this case, apparently 'l' =
1/0.5 = 2 s; if we use this length of time in the solution, we will find that the
current is 0.34 A, which is not insignificant when compared with the peak
current of about 3.7 A (which occurs after about 2 s). In fact, it takes about
16 s for the transient to decay to an insignificant value.
(d) R = 3 Q (overdamped response)
In this case
~= ~ V[; ]= 1.5 and Wn = 0.5 radis

and the s-domain equation for the circuit current is
10/2 5
fes) = =-----
S2 + (2 x 1.5 x 0.5s) + 0.5 2 S2 + 1.5s + 0.25
Completing the square in the denominator gives
5 5 [ ß ]
fes) = (s + 0.75Y - 0.3125 = 0.559 (s + aY - fP
and, from the table of Laplace transforms

i(t) = 8.945 e-D75t sinh0559t A
While the solution has some mathematical merit, it does not present an
engineer with amental picture of what is happening. More information can
be obtained by converting the result into its exponential form as follows
i(t) = 8.945 e-D75t ~ [(e0559t _ e-D559t)]
= 4.4725 (e-D·191t - e-l. 309t) A

The longest of the two time constants is 1/0.191 = 5.24 s, so that it will take
5 x 5.24 = 26.2 s for the current to have fallen to zero (or nearly so).
The reader should note that it takes much longer for the overdamped
circuit to settle than does the critically damped circuit.
We will look in this worked example at the operation of an electro-

mechanical system, namely a conventional contact-breaker ignition system
of an automobile.
t= 0
The basic circuit is shown in figure 10.25, in which a resistor, a capacitor

and an inductor are connected to the 12 V battery in the vehicIe. The
resistor is merely a current-limiting device; the capacitor is connected
across a switch (known as the 'points'), and the voltage developed across
the ignition coil is applied to the sparking plugs.
The points open and dose periodically (depending on the speed of the
engine) and, in this problem, we need to calculate the maximum voltage
across the coil (assuming, that is, the sparking plugs have not broken
down).
Solution
We can assume that the points are dosed long enough for the current to
have reached its steady-state value of 12 V/2 Q = 6 A, before the points are
opened.
The s-domain equivalent circuit is shown in figure 10.26, in which the
initial condition source has a value of U(O) = 1 x 6 = 6 V (assisting the
battery) and the corresponding mesh current equation is
-12
s
+ 6 = l(s)(2 + s + 106 /s)
or
12
-+6
s 12 + 6s
I(s)
106
s+2+-
s
s+2
6---------
(s + 1)2 + [(103)2 - 1]
s+1
6 ------ - 6 X 10-3 - - - - - -
(s + 1)2 + (10 3 )2 (s + 1)2 + (10 3 )2
~ Li(O) = 6
t VL(s)
Figure 10.26 Automobile ignition circuit in the s-domain.
= 6
s+a - 6 X 10-3 - _ _ __
w
(s + a)2 + w 2 (s + a)2 + w 2
Taking the inverse transform gives
i(t) = 6e-t cos loo0t - 6 x 1O-3e-t sin l000t
For all practical purposes we may ignore the second term, hence
i(t) = 6e-t cos loo0t
The voltage across the ignition coil is, therefore
di(t)
vL(t) = L - - = 6 x (-1000) x 6e-t sin 1OO0t
dt
= -6000 e- sin 1000t V
t
The periodic time of the sinusoid is T = 2rrJw = 2nll000 s or 2n ms. The

maximum voltage occurs when the sinusoid is maximum, or when t = n/2
ms, at which point sin 1000t = 1. Hence
vL(max) = -60ooe-ltI2OOO = -5990 V
10.12 s-Domain transfer functions
A transfer function is the ratio of the response of the circuit to the forcing
function applied to the circuit. Having obtained the transfer function of the
circuit or system, it is a matter of applying circuit analysis to obtain the
output.
The technique for obtaining the s-domain transfer function follows the
same general lines as in the frequency domain, and worked example
10.12.1 illustrates the general method.
Determine the s-domain transfer function H(s) = Vo(s)IV)(s) for the circuit
in figure 10.27. Assume that the operational amplifier is ideal.
I
i,(t)
R,
V,(t)
Solution
Since the op-amp is ideal we may assurne, for the purpose of analysis, that
vx(t) = O. That is
i1(t) = vl(t)/R I or I.(s) = V.(s)/R I
and or Iz(s) = I.(s)
also
Now, since Iz(s) = I.(s), then

Vo(s) -Iz(s)Ri(l + R Cs)
H(s) = - - = --'--'-----'-----'-
2
V.(s) I.(s)R 2
10.13 Transients in magnetically coupled circuits
The time-domain equations for the coupled circuit in figure 1O.28(a) are
di1(t) di2 (t)
v1(t) = LI d(+ Md(
L,i,(O)
1,(s)
-- + Mi,(O)
---
L,i,(O) + Mi,(O)
1,(s)
V,(s) V,(s)
(a) (b)
Figure 10.28 (a) Basic two-winding magnetically coupled circuit, (h) s-domain equivalent
circuit with isolated windings.
Taking Laplace transforms of these equations gives

V 1(s) = L 1[sI1(s) - il(O)] + M[slz(s) - i2 (0)]
VZ<s) = M[sll(s) - il(O)] + L 2 [slz(s) - i2 (0)]
or
V1(s) = L 1sl1(s) + Mslz(s) - [L1il(0) + Miz(O)]
Vz(s) = Msl 1(s) + L 2 s1z<s) - [L 2i 2 (0) + Mil(O)]
where i l (0) and i 2 (0) are initial currents in LI and L 2 , respectively. The
s-domain equivalent circuit in figure 1O.28(b) is modelled on these equa-
tions. Worked example 10.13.1 illustrates how these equations can be used
to solve for the transient solution in a magnetically coupled circuit.
Derive expressions for the current in the primary and secondary winding of
the transformer in figure 1O.29(a). There are no initial conditions in the
circuit.
Solution
We should note in figure 10.29(a) that (1) theprimary winding is excited by

a step function of direct voltage and (2) the current is assumed to leave the
dotted terminal of the secondary winding. The second point means that the
s-domain equivalent circuit appears as shown in figure 10.29(b). The mesh
equation for the primary circuit is
10
- = (20 + O.ls)ll(s) - 0.051z(s)
s
(a) (b)
and for the secondary is

0= -0.05sI1 (s) + (20 + [0.1 + O.Ol]s)Iis)
= -0.05sI1 (s) + (20 + O.11s)Iis)
The equation for the secondary circuit shows that
20(20 + O.11s)
Il(s) = 12 (s)
S
Substituting this expression into the primary circuit equation yields

10 [(20 + 0.lS)20s(20 + O.11s) ]
S = Iis) - 0.05s
0.17s2 + 84S + 8000]

12 (s) [ s
or
10
0.17s2 + 84s + 8000
58.82
S2 + 494s + 47 059
whence W n = V47 059 = 216.9 radis, and 2Cw n = 494 or C= 1.14. That is,
i 2 (t) has an overdamped response. Completing the square in the denomina-
tor of the equation gives
58.82 0.498 X 118.2
I(s) - =------
2 - (s + 247)2 - 118.2 2 (s + 247)2 - 118.22
= 0.498 ß
(s + a)2 - {f
The table of transforms gives
iit) = 0.498e-247t sinh 118.2t
= 0.498e-247t(e118.21 - e- 118 •21)/2
= 0.249(e- I28 .8t - e- 36S ·2t)
The longest time constant is 1/128.8 = 7.76 ms, andthe transient will have
died away in 5 x 7.76 = 38.8 ms (say 40 ms).
The transform of the primary current is obtained by inserting the ex-
pression for 11 (s) into the mesh equation of the secondary as follows
I 1 ()
S = 20(20 + O.l1s) 11 ()
S
S
20(20 + O.l1s)
= -'------'- - - - - - - -
58.82
S S2 + 494s + 216.92
= 1176.4 [ 20 + 0.11 ]
s(r + 494s + 216.92) r + 494s + 216.92
The first expression inside the brackets is reduced into partial fraction form
(see chapter 15 for details) to give the following
494 x 20 20s
--+----
1764[ 20 216.92 216.92 0.11]
Ms) = 1 . 216.92s - s(r + 494s + 216.92) + r + 494s + 216.92
= 0.5 _ 0.5 s + 494 + 1.095 118.2

s (s + 274)2 - 118.22 (s + 274)2 - 118.22
For simplicity, we will write (s + 274)2 - 118.22 = [D], giving
1 (s) = 0.5 _ 0.5 s + 247 _ 0.5 247 + 1.095 118.2
1 s [D] [D] [D]
= 0.5 _ 0.5 s + 247 + 0.05 118.2
s [D] [D]
= 0.5 _ 0.5 s+a + 0.05 ß
s (s + a)2 - W (s + a)2 - W
The inverse transformation gives
i1(t) = 0.5 - 0.5e- 247t cosh 118.2t + 0.05e- 247t sinh 118.2t
= 0.5 - O.225e-128.8t - O.275e-365.2t A
The primary steady-state current is 0.5 A (as could be observed from the
d.c. conditions in the Circuit) and, once again, the longest time constant in
the expression for i1(t) is 1/128.8 = 7.76 ms so that the transient part of i1(t)
will have died away after 38.8 ms (say 40 ms).
Unworked Problems
10.1. A voltage given by

v(t) = u(t) - 4u(t - 2) + 3u(t - 4) V
is applied to a 2 Hinductor. Deduce an expression for the current
in the inductor.
[0.5r(t) - 2r(t - 2) + 1.5r(t - 4) A]
10.2. Deduce the Laplace transform of (1 - e- at ) , where ais a constant.
[a/(s(s + a))]
10.3. Evaluate the Laplace transform of f(t) = u(t) - u(t - a).
[(1 - e-sa)/s]
10.4. A cyclic sawtooth waveform whose first cycle is described by
f(t) = (t) for 0 :s;; t:s;; 1, after which it instantly becomes zero again.
Determine the Laplace transform of the wave.
[(1 - (s + l)e-S)/( (1 - e-S)s2)]
10.5. Determine the Laplace transform of the function
(a) (t - a)e-a(t - a)u(t - a)
(b) ö(t) + (a - b)e-btu(t).
[(a) e-as/(s + a)2; (b) (s + a)/(s + b)]
10.6. A voltage E is applied at t = 0 to aseries circuit containing a
resistor R, an inductor Land a capacitor C, with no initial condi-
tions in the circuit. Derive an expression for the current, i(t), if the
circuit is (a) underdamped, (b) critically damped, (c) overdamped.
(a) ~ e-\;Wnt sin w t· (b) E te- wnt .
woL 0' L '
(c) ~ e-\;Wnt sinh at, where a = WO V({; - 1)

aL
10.7. A capacitor is charged through a pure resistor by a battery of
constant voltage E; derive an expression for the instantaneous
charge on the capacitor if it is initially uncharged. If C = 10 IlF and
R = 1 MQ, calculate the time taken for the capacitor to receive 85
per cent of its final charge.
[CE(l - e-tIRC); 18.97 s]
i(t)
20u(t) t
Figure 10.30
10.8. Calculate the time taken for i(t) in figure 10.30 to fall to 4 mA.
[0.035 s]
10.9. A direct voltage, VI' is applied to a coil of resistance 10 Q and
inductance 1 H, and is maintained constant until all transients have
settled out. At this time the voltage is changed to V 2 • (a) Derive an
expression for the current in the coil after the voltage has changed
to V2. If (b) VI = 200 V and V2 = 100 V, (c) VI = 100 V and
V2 = 200 V, calculate the current in the coil at (i) 0.1 s, (ii) 0.3 s
after the voltage has changed.
[(a) (V2 + (VI - V2)e-Rt'L)/R; (b) (i) 16.32 A, (ii) 19.5 A;
(c) (i) 13.68 A, (ii) 10.5 A]
10.10. A surge generator used for testing high-voltage electrical apparatus
produces its voltage by initially charging a number of parallel-
connected capacitors to a very high voltage. To produce the voltage
surge, the capacitors are connected in se ries giving, in this case, a
total voltage across the series-connected capacitors of 300 kV ..
The equivalent circuit of the surge generator is shown in figure
10.31; the spark gap merely acts as a switch and, for the purpose of
analysis, can be regarded as a short-circuit. The apparatus under
test, connected between terminals A and B, can be regarded as
having an infinite impedance. Determine an expression for the
voltage across the apparatus under test.
[242(e-O· 02t - e-498t) kV (t in IlS)]
Spark
1 mH
___-oA
~----------~---oB
Figure 10.31
10.11. In problem 10.10, calculate the maximum voltage between A and

B, and the time taken to reach the maximum value.
[235.7 kV; 1.11 lAs]
10.12. Aseries circuit contains a 10 Q resistor and a 0.1 Hinductor. If a
voltage of lOe-50l u(t) is applied to the circuit, deduce an expression
for the current in the circuit.
[2( e- lOOI - e-50I)]
10.13. A circuit containing a resistance of 20 Q in series with an induct-
ance of 0.2 H is energised by a voltage v(t) = 250 sin(300t + cf» V.
Deduce an expression for the current in the circuit if the supply is
connected when (a) cf> = 0°, (b) cf> = 90°. There are no initial
conditions in the circuit.
[Ca) 3.95 sin(300t - 71.57°) + 3.75e-lOOI A;
(b) 3.95 sin(300t + 18.43°) - 1.25e-lOOI A]
10.14. A magnetically coupled circuit similar to that in figure 10.29 has the
following values: d.c. supply = 20 V, resistance in the primary
circuit = 10 Q, L l = 0.2 H, L 2 = 0.4 H, M = 0.1 H. The load is a
pure resistor of 10 Q resistance, and there are no initial conditions
in the circuit. Derive an expression for the current in the primary
and in the secondary windings.
[il(t) = 2 - 0.292e- 22 66/ - 1.708e-63061 A; i2 (t) = 0.707(e- 22 66/ - e-63(61)
A]
10.15. Calculate, for problem 10.14, the maximum value of i 2 (t) and the
time when it occurs.
[0.255 A; 0.025 s]
10.16. Derive the transfer function v2 (S)/V l (s) for the RC network in figure
10.32.
[ (s + 1/(RlCl) )(s + 1/(R2 C2 » ]
s2RlR2ClC2 + s(RlCl + R lC2 + R 2 C 2 ) + 1
R,
Figure 10.32
R, R,
Figure 10.33
t= 0 t= 0
an
Figure 10.34
10.17. Derive v2 (S)/V 1 (s) for figure 10.33.

[1I(S2R 1R 2C1C2 + s(R1C1 + R 1C2 + R 2C2) + 1)]
10.18. If both switches in figure 10.34 are closed when t ~ 0, and both
capacitors are fully charged when t = 0, determine an expression
for i(t) for t > 0 if both switches are simultaneously opened when
t = O.
[e- 035t A]
11
Complex Frequency, the
s- Plane and Bode Diagrams
11.1 Introduction
This chapter introduces an important aspect of circuit analysis, namely

complex frequency. This is a unifying concept which ties together many
analytical techniques into a ne at package.
Complex frequency is introduced by looking at an exponentially
damped sinusoidal function, for example the voltage
(11.1)
where ais areal quantity, and usually has a negative value. While the
reality of the equation (11.1) is not easy to grasp we have, in fact, looked at
it in other forms on several occasions.
For example, in the case of a direct voltage, both a and ware zero,
giving
v(t) = Vm cos (J = Vdc
In the general sinusoidal case, ais zero, giving
v(t) = V m COS (wt + (J)
If we allow w to be zero, then we have the exponential case as follows

v(t) = V meot cos (J = Vdce ot
Thus the general expression in equation (U.1) includes the d.c. case, the
general sinusoidal case and the exponential function.
Initially in this chapter we look at the mathematical aspects of complex
frequency and the s-plane, and then we transform the impedance of ele-
ments into the s-domain. This is particularly important when we study
topics such as the frequency response of circuits and Bode diagrams. Once
266
Complex Frequency, the s-Plane and Bode Diagrams 267
we have mastered the concept of the s-domain 'impedance' of simple

elements, it is a relatively simple step to determine the s-domain im-
pedance of complete circuits.
11.2 The exponential form of a complex number
Expanding eiß as a Maclaurin series we get

( . ()2 ( . ()3
eiß = 1 + j() + _J_+_J_ + ...
2! 3!
= (1 _!!.-+!!.--
2! 4!
... ) + j (() -~+~- ... )
3! 5!
= cos () + j sin ()
It mayaiso be shown that
e- iß = cos () - j sin ()
That is to say, eiß is a complex number, and can be represented by a point
which is unit distance from the origin and subtending an angle () with the
real axis. Hence any point v, at distance V from the origin and subtending
angle () with the real axis can be represented in the form
v = Veiß = V(cos () + j sin ()
The reader should note that for any point v, there is an infinite number of
other corresponding points, but differing from one another by an angle
which is an integral number of 2Jt radians, that is their angle is «() + 2Jtn),
where n is an integer.
11.3 Complex frequency
If () = rot, we may write

ej9 = cos rot + j sin rot
or if s = a + jro, where s is known as the complex frequency, then
eS' = e(a + iw), = eoteiw'
= eot( cos rot + j sin rot)
If the magnitude of the expression is V, then
ve' = Veot(cos rot + j sin rot)
Let us consider the effect of various values of a and ro in the above
expression:
(1) v(t) = 10 S = 0 + jO
(2) v(t) = 8e-51 S = -5 + jO
(3) v(t) = 4 cos 20t SI = 0 + j20
S2 = sr= 0 - j20
(4) v(t) = 7e-2tcos 40t SI = -2 + j40
S2 = sr = - 2 - j40
(5) v(t) = 6e71 cos (9t + 30°) = 7 + j9
sr=
SI
S2 = 7 - j9
Expression (1) corresponds to a 'd.c.' term in which there is no growth or
decay of the function, nor is there any sinusoidal oscillation. Expression (2)
is an exponentially decaying function in which u = -5; the larger the
negative value of u, the more rapid the decay of the function. A positive
value of u corresponds to a function which grows exponentially.
The third expression is that of a sinusoid with no growth or decay in its
magnitude. In this case, there is a pair of values of s, one being the
conjugate of the other. Since each value of S possesses both real and
imaginary parts, the sum of the two values identifies areal sinusoidal
frequency; in this case the angular frequency is 20 radis.
In a similar manner there are, in expression (4), a pair of conjugate
values; in this case there is an exponential decay associated with the
u = - 2 component. Finally, expression (5) is related to a sinusoid which
grows exponentially with time.
It is interesting to note that, in each conjugate pair of values, one has a
negative value of w! Such a value does, in fact, exist. For example, if we
look at the expression for the compound angle
cos (- 40t + 4J) = cos (-40t).cos 4J - sin (-40t).sin 4J
= cos 40t. cos 4J + sin 40t. sin 4J
= cos (40t - 4J)
This implies that it has the same frequency as the function cos (40t - 4J),
but it has a different numerical value at time t.
Where there is a product term such as 9 sin 8t.cos 5t, the roots contain
the sum and difference of the frequencies. In this case the roots are SI =
j(8 + 5) = j13, S2 = -j (8 + 5) = -j13, S3 = j(8 - 5) = j3, S4 = -j(8 - 5) =
-j3.
Generally speaking, the larger the value of u, the more rapid the rate of
growth or decay of the quantity; the greater the value of w, the higher is
the frequency.
11.4 The s-plane
We can map complex frequencies on the s-plane, as shown in figure 11.1,

where U is plotted in the horizontal or 'real' direction, and w is plotted
c ___ \ _
s=:+ja>, V"
-0,
"
-o-.----~~-------- __----~------------._---------4~+a
-ja> ---.r-
Figure 11.1 Waveforms for complex frequencies in the s-plane.
in the vertical or 'imaginary' direction. At the origin of the graph,

where s = 0 + jO, we get a unit step or 'd.c.' condition. At point A, the
function decays with time constant of l/a; at point B, the function
oscillates at a frequency of W t radis (note: the conjugate point is not
shown in the figure). If 1- al > 1 wl, the rate of decay will is fairly smalI.
The reader is reminded that a pair of conjugate points is necessary to
give a sinusoidal oscillation. A complex pair of points in the right-hand half
of the s-plane (see point D - note: the conjugate point is not shown in the
figure ) gives rise to an oscillatory signal which grows exponentially to
infinity.
In general, a conjugate pair of points in the right-hand half of the
s-plane represents an unstable condition.
11.5 Transformation of impedance into the s-domain
Just as we have associated the sinusoid I cos (wt + </J) with phasor I = lei</>,
we can relate an impedance with its s-domain equivalent as folIows.
Resistance
The relation between the voltage applied to a resistor and the current
through it is given by Ohm's law as follows
v = Ri
If i = Ieutei(rut + 4» = st
Iei4>e , then the voltage aeross the resistor is
v = Veutei(rut + 4» = Vei4>est
That is
or Vei 4> = Rlei4>

therefore
VL4> = RIL4>
henee
Z(s)=VL4>=RLO°
R IL 4>
and YR(s) = lIR
Inductance
In a pure induetor VL = L(dildt); if the eurrent is = Ie i4>est and

VL = Ve i8est , the v-i relationship is
Ve i8est = L~(Iei4>est)
dt
= Lslei8est
or Vei8 Lslei8
henee VL8 LsIL 8
That is v= LsI
The s-domain impedanee of a pure induetor therefore is
V
ZL(S) =-= Ls
I
and YL(s) = l/Ls
Capacitance
Here the relationship is i = C(dvc/dt), and
Ie i4>et = C~ (Ve i8est )

dt
Camp lex Frequency, the s-Plane and Bode Diagrams 271
or CsVe j9
hence IL cp CsV L cP
that is I CsV
and the s-domain impedance of a pure capacitance is
V 1
Z (s) = - = -
C I Cs
and
Series and parallel combination 0/ similar elements

The reader should note that the usual rules apply to series and parallel
combinations of similar elements. For example, if two resistors of resist-
ance Rare connected in series, then ZR(S) = 2R, and if connected in
parallel then ZR(S) = R/2. If two inductors of inductance L are connected in
series, then ZL(S) = 2Ls, and if connected in parallel then ZL(S) = Ls/2, etc.
11.6 Frequency response as a function of w
We now consider the forced response of circuits as the supply frequency w

(or f) changes; during this section we assurne that a = 0, or S = jw. We can,
therefore, either develop circuit equations in terms of S or in terms of jw.
Generally speaking, the former is more useful since it is difficult, when
terms have been multiplied, to convert an equation from the jw form to the
s form beeause it is not always possible to know how many 'j' terms were
involved (remember: j2 = -1, j3 = - j, j4 = 1, ete).
Initially we consider the frequeney response of a fairly simple eircuit -
the RC parallel eircuit in figure 11.2 - in order to define some basic terms.
We look at the way in whieh both the magnitude and the phase shift vary
with frequeney. The impedanee of the circuit is
z-3j
Figure 11.2 A simple Re parallel circuit.
-w -----j---->t<~-t---- w
1
Re -450
-w --+---t---+,- __ w -90 0 -------
1 o 1
-q,
Re Re
(a) (b)
Figure 11.3 (a) Magnitude-trequency plot tor +00 and -w ot the impedance tor figure
11.2, and (b) the phase-trequency plot.
R(lICs) R
R + lICs 1 + RCs
Alternatively, we can evaluate Z(jw) simply by replacing s by jw as
folIows:
') R(lIjwC) R
Z(Jw = =----
R + lIjwC 1 + jwRC
= y' R L-tan-1(wRC) = IZILljJ

(1 + [WRC]2)
where both Izi and ljJ are functions of w.
The curve showing how the magnitude of the impedance changes with
frequency with both + wand - w is drawn in figure 11.3(a), and the
corresponding phase-frequency plot is in figure 11.3(b). The two curves
comprise the jrequency response of the impedance Z. It can be shown that
the magnitude curve is symmetrical about the w = 0 axis; the phase angle
has its sign reversed for w < O. For these reasons, we generally need only
interest ourselves in the frequency response for positive values of w.
Consider now the power consumed by the circuit at the frequencies
w = 0 and w = lIRC; we williater refer to the latter frequency as w\. At
w = 0, the capacitive reactance is Xc = lIwC = 00, and no current flows in
it at this frequency. If the circuit is energised by a current source I, then all
the current flows through the resistor, and the power consumed at zero
frequency is
Po = RI/12
and is the maximum power consumed by the circuit. At frequency w\, the
current in the resistor is
1 1
I --------
R + lIjw1 C 1 + jw1CR 1+j
1 =_1_ L _ 45o
V2L45° V2
and the power consumed at frequency w 1 is
P = 1~12 R = 1l12R = Po
1 V2 2 2
That is, the power consumed when w 1 = lIRC is one-half the maximum
power. Consequently, we say that w 1 is the half-power frequency, or the
half-power point, or (for reasons given later) the cut-off frequency. In
general, a frequency (and there may be more than one of these) which
gives a power response which is one-half of the maximum response is
known by one of the above names.
In the above case, the phase shift at the half-power frequency is -45°.
In general, depending on the circuit, the phase shift at a half-power point is
±45no, where n is an integer.
We will now look at two simple examples which are of particular interest
when we study resonance (see chapter 12).
Sketch the frequency response curve for the impedance of a two-branch
parallel circuit which has a pure inductor, L, in one branch, and a pure
capacitor, C, in the other.
Solution
Although this is an idealised situation (strictly speaking, there is no such
thing as a 'pure' Land a 'pure' C), it will be very helpful in giving an
overall view of the frequency response of this type of circuit. The complex
frequency impedance is
Ls(lICs) Ls
Z(s)
Ls + lICs 1 + LCs2
or, replacing s by jw
(. ) jwL wLL90°
Z )w = 1 + LC(jwY = 1 - w2 LC
when w = 0, then Z(jw) = OL90°. As w increases in a positive direction,

the magnitude of Z(jw) increases and the phase angle remains constant at
90°. However, at some frequency Wo, the denominator has zero value; this
is when olöLC = 1, or
-w~----,-----~'-----,---~~W
o
(b)
- 90 0 -1------.....,
-W _--+-------+------+-____ (j)
'---____-I - -90 0
-rp
(c)
Figure 11.4 (a) Parallel LC eireuit, (b) the magnitude response and (e) the phase response.
Wo = l/V(LC)
This is shown in figure 11.4.
At this frequency, which is known as the resonant [requency, Wo, the
denominator of the impedance equation is zero, and the magnitude of the
impedance is infinity. When W > Wo, then 01 LC > 1 and the magnitude of
the impedance begins to diminish until, at W = 00, IZ(jw)1 = o.
Also, when w2 LC > 1 the sign of the denominator of Z(jw) becomes
negative, resulting in the phase shift becoming - 90°.
The corresponding values for - ware shown in figure 11.4.
Poles and zeros
At this point we will define two critical values of complex frequency s, one
being known as a zero and the other as a pole, which determine both the
transient and the steady-state behaviour of a circuit. A zero is a value of s
which makes the value o[ a complex function equal to zero, and a pole is a
value of s which makes the value o[ a complex function equal to infinity.
z_~
121
~
(a)
-w w
-Wo 0 Wo
(b)
q,
-900 ;---
-W_-+------il-----+-_w
o
_ -90 -I--_...J
0
(c)
Figure 11.5 (a) Series resistaneeless LC cireuit, (b) the magnitude response o[ the
impedanee o[ the cireuit and (e) the phase response.
Zeros occur in the numerator of the expression, and poles occur in the
den ominator . Since the complex impedance in this problem is
Ls
Z(s) - - - -
1 + LC~
there is a zero at s = 0, and a pair of poles at s = Wo = lIv'(LC) and
s = - Wo = -lIv'(LC). The two poles form conjugate points on the
s-plane (see section 11.4).
Repeat worked example 11.6.1 but for aseries circuit containing a pure
inductor and a pure capacitor.
Solution
The circuit is shown in figure 11.5(a), and the circuit impedance is

1
Z(s) =Ls+-
Cs
or, replacing s by jw
Z(jw) = jwL --j-= j (WL __1_) = _j(_1__ WL)

wC wC wC
= (~C - wL ) L - 90° Q
Clearly, when w = 0, IZ(jw)1 = 00 (see diagram (b», and the phase shift is
-90°, as shown in diagram (e) of figure 11.5. As the frequeney (+ w)
inereases, IZ(jw)1 deereases in value and the phase angle remains eonstant
at -90° until some frequeney Wo, when woL = lIwoC. At this frequeney
(wo = lIV(LC», whieh is the resonant frequency of the series cireuit, the
modulus of the eireuit impedanee is zero.
When w > wo, Z(jw)1 inereases with w, finally approaehing infinity as w
approaehes infinity. At the same time, the phase shift beeomes +90°, as
shown in figure 10.5.
Corresponding values for -ware also shown in the figure.
11.7 Transfer functions
Often we need to look at the frequeney response of the ratio of the foreed
funetion, for example, the output voltage or eurrent, to the foreing fune-
tion whieh produees it, whieh mayaiso be a voltage or eurrent. The ratio of
the foreed funetion to the foreing funetion is known as the transfer function
H(s) or H(jw). That is
H(s) = response (s) = output (s)
foreing funetion (s) input (s)
or
. ) response (jw) output (jw)
H ( JW = ---=---";';;""-'---
foreing funetion (jw) input (jw)
We ean, for example, look on impedanee as a transfer funetion in which
V
H(s) = Z(s) = - (s)
I
or
H(jw) = Z(jw) = ~ (jw)

I
where V is the response and I the forcing function. In general, a transfer

function may take on one of many dimensions. For example, it may be
dimensionless, as is the case when we are considering the voltage gain of an
operation al amplifier; it may have the dimensions of impedance, as in the
case considered above; it may have the dimensions of, say, radians per volt
where we are considering the operation of a voltage-driven position control
system, etc.
Evaluate the transfer function V1(s) for the circuit in figure 11.6.
VI
Solution
Bearing in mind that the two 1 F capacitors are connected in series (giving
an effective capacitance of 0.5 F), the transfer function is
H(s)
Vb)
=- - -+-lIs- -
- = - -0.8
VI(s) 1 + 8s + 1I0.5s
0.125(1 + 0.8s)
S2 + 0.125s + 0.25
0.125(1 + 0.8s)
=---------'---"""'"-----
(s + 0.0625 + jO.5)(s + 0.0625 - jO.5)
Derive an expression for H(s) = VO(s)/VI(s) for the operational amplifier

circuit in figure 11.7. The operation al amplifier is ideal. This circuit is the
basis of many active filter networks.
Solution
The circuit is that of an active filter, and is applicable to a low-, a high- or a

band-pass filter, according to the position of resistors and capacitors in the
circuit (see also problem 11.11).
At node 2
11 = 1 + 1 + I. = (VI - V2 )YI
2 3
The individual currents are

12 = V2Y2 13 = (V2 - V S )Y3
I. = (V2 - Vo)Y. I s = (Vs - Vo)Ys
Since the op-amp is ideal, V s is practically zero, so that no current ftows
into the amplifier, and
13 = V 2Y 3
I s = - VoYs
Inserting the above expressions for current into thefirst equation gives
V 2Y 2 + V 2 Y 3 + V 2Y. - VoY. = (VI - V 2 )YI
and, after a little more algebraic manipulation we get
H(s) = V o = - YI Y3
VI YS(YI + Y2 + Y3 + Y.) + Y 3 Y.
At this stage we are dealing with an academic exercise, and the reader
should note that the admittance values should be of an electrically accept-
able form. The reader should refer to problem 11.11 for typical values for a
practical circuit.
V, V3 V. I. V.
V,
I, Z, 13 Z3 Z. I.
Z.
W orked example 11.7.3
Derive an expression for the input impedance, ZIn(s), for the generalised
impedance convertor (GIC) in figure 11.8. The op-amps are ideal.
Solution
The purpose of a GIC is to convert an impedance of one kind into an

impedance of another kind, that is, it makes a capacitor 'look' to the
extern al circuit as though it were an inductor. This is particularly useful in
integrated circuit applications where inductors are difficult to manufacture.
The circuit layout in figure 11.8 looks fairly involved, but it is drawn in this
way to make the analysis as simple as possible.
Since the op-amps are ideal, the potential difference between the am-
plifier input terminals is practically zero. We may therefore say that VI =
V3 and V 3 = V s ; hence, we can use VI to represent both V 3 and V s . Also,
since the op-amp input current is practically zero, we may say that 14 = I s ,
therefore
or, using the above simplification

V4 - VI = VI
or
Z4 Zs
Also 12 = 13 , or
V2 - VI
Z2
that is
v = V ( 1 _ Z;Z4 )
2 I Z;Zs
The input impedance of the circuit is
VI VI
Zin(S) =1= (V - V)/Z
I I 2 I
ZI
ZI Z3ZS
Z2 Z4
If ZI = Z2 = Z3 = Zs = Rand Z4 = lIsC, then
RxRxR
Zin(s) = R/sC = (R 2 C)s = Ls
That is, the capacitor C in the Z4 position appears as though it were an

inductor L at the input of the circuit, where L = R 2 C H. If R = 1 kQ and C
= 1 IlF, then L = 10002 X 10-6 = 1 H.
11.8 Bode diagrams
In order to understand more fully the way in which the magnitude and
phase shift of a transfer function of a given circuit changes with frequency,
we need a simple method of estimating the shape of the curves. Accurate
curves can be plotted after so me more-or-Iess frantic manipulations with a
calculator (or a programmable calculator), or they can be obtained directly
by an all-out assault using a computer (see also chapter 14, where the
SPIeE package is described). However, in the short term, we need a fairly
quick method of obtaining the response curves; one method has been
provided by Hendrik Bode of the Bell Telephone Laboratories, and is
described here.
We have already met with frequency response in section 11.6, where
simple frequency response diagrams were described. Here we look in a
little detail at the frequency response of more general forms of transfer
function in electrical circuits.
Many transfer functions combine elements having an equation similar to
the following
H(s) = K (s + ZI)(S + ZI) .. .

s"(s + PI)(S + PI) .. .
where K is a constant gain (or attenuation) factor which is independent of

frequency, z is a frequency at which a zero occurs in the complex frequency
plane, p is a frequency at which a pole occurs in the complex frequency
plane and n is an integer.
Finally: there may be quadratic terms in the denominator , similar to the
one shown below, which can give rise to aresonant effect.
First, we must look at the way in which we can estimate the overall
magnitude and phase response from the many parts of the transfer function
of the system.
A Bode diagram is an asymptotic plot of the magnitude and phase of the
transfer function to a base of frequency, so that we can write the above
equation as a function of w in the following form
H(jw) = IH(jw)1 LH(jw)
IHI(jw)1 LHI(jw) X IHz(jw) I LHz(jw) .. .
IH3 (jw)1 LH3 (jw) X IHljw)1 LHljw) .. .
IHII X IHzl X . . . (
= IH I X IH I X arg LHI +LHz + ... - {LH3 +LH4 + ... })
3 4 ...
The final expression teIls us that the overall magnitude is obtained by

multiplying (or dividing) individual magnitudes of each part of the transfer
function; the overall phase shift is obtained by adding (or subtracting) the
individual phase angles of the different parts.
This may see m a relatively complex procedure, but it can be reduced to
aseries of simple steps when we look at the basic techniques for expres-
sions commonly found in electrical and electronic circuits.
To simplify the process of drawing the diagrams, both the magnitude
and phase curves are drawn to a logarithmic scale for the abscissa. The
phase scale is drawn linearly; to simplify the multiplication and division of
the magnitude values, the magnitude scale is in logarithmic units called
decibels (dB), where the decibel is defined as follows
H dB = 20 log IH(jw)1
and logarithms to base 10 are used.
The Bode diagram yields two graphs. One is the magnitude or gain
expressed in dB plotted to a base of frequency (or to a function of
frequency, which may be dimensionless) on a logarithmic scale. The other
is a graph of phase shift (usually in degrees) plotted to the same logarithmic
frequency base.
40
LH(jw)
20 log 18 = 25.1 dB
20
-t---r---.-__ W (log) -t--'T"""-"T'""- w (radIs)

0.1 1 10 0.1 10
(a) (b)
Figure 11.9 Bode diagram (a) magnitude plot, (b) phase plot tor H(joo) = K.
Normally both graphs are drawn on the same sheet of log-linear paper.
In this way it is possible, at a glance, to review the frequency response of
the system.
11.9 Bode diagram of H(joo) = K
If K = 18, the decibel magnitude is

H dB = 20 log 18 = 25.1 dB
for all 00, and results in the graph in figure 1O.9(a).
The phase shift for a simple numeric gain factor is zero for a11 00, as
shown in figure 11. 9(b).
11.10 Bode diagram of H(jw) = (jWT)"
In this case the dB gain is

H dB = 20 log IjwTln = 20n logY(wT)2
= 20n log WT
If n = 1, the way in which the magnitude changes with WT is shown in table
11.1. We see from the table that the graph is a straight line having a slope
016 dB/octave (strictly speaking, it is 6.02 dB/octave), meaning that there
is a change of 6 dB in the magnitude each time the frequency changes by a
factor of 2. Altematively, the slope is 20 dB/decade (that is, a change of 20
dB for each ten-fold change in frequency), as shown in figure l1.lO(a). A
look at the mathematics shows that the slope of the magnitude plot, for any
integer n, is 6n dB/octave (approx.) or 20n dB/decade.
The phase angle associated with (jWT)" is
L(jWT)" = L(j)" = 9On°
Table 11.1
roT H dB Slope o[ graph
0.1 -20
0.2 -14 (approx.) } 6 dB/octave 20 dB/decade
1.0 0
2.0 6 (approx.) } 6 dB/octave 20 dB/decade
)
10.0 20
20 dB/decade
100.0 40
H•• n=2 LHUw)
40 180· n=2
20 90- n=l
0 w~ 0 w~
0.01 0.1 10 100

-20 _90· n =-1
-40
(b)
n =-1
(a)
Figure 11.10 (a) Magnitude plot ofH(joo) = (jOO't)D for n = -1, 1 and 2,
(b) corresponding phase plot.
This is illustrated in figure 1l.10(b) for values of n of - 1, 1 and 2.
11.11 Bode diagram of H(jw) = (1 + jWT)

Gain-frequency plot
The expression for the gain in dB of this expression is

H dB = 20 log 11 + jWTI = 20 log V (1 + ah2 )
We will look at the value of this expression over two ranges of frequency,
namely (1) when (WTY « 1 and (2) when (WT)2 » 1.
High-frequency
asymptote
40
20 Low-frequency
~ asymptote
o +"':"""'T'"-"'T'"-+
0.01 0.1 /1
/
-20 / Corner
/ frequency
/
(a)
Low-frequency
asymptote (b)
Figure 11.11 (a) Asymptotic magnitude-frequency plot lor H(jw) = (1 + jw'C),

(b) a means 01 sketching the curve in the region 01 the corner frequency.
1. (wr)2 « 1
In this case H dB = 20 log v' (1 + w2 r 2 ) :::= 20 log 1 = 0 dB for all w within
this frequency range. This results in the low-frequency asymptote in
figure 11.11(a), which lies on the 0 dB axis.
2. (wr)2» 1
Here H dB = 20 log v' (1 + w2 r 2 ) :::= 20 log wr, which is a straight line of
slope 20 dB/decade cutting the 0 dB axis when wr = 1 (see also section
10.11). This gives the high-frequency asymptote (see figure 11.11(a))
which cuts the Wi axis at Wi = 1 or where W = 1Ii.
Clearly, the two asymptotes intersect when the high-frequency asymp-
tote has the value H dB = 0, that is, when wr = 1. This frequency is known
as a corner frequency, wC '
Occasionally the asymptotes do not give sufficiently accurate informa-
tion to sketch the curve in the region of the corner frequency, and it is
necessary to plot a few more points. In particular, values at the corner
frequency together with frequencies which are one-half and double the
corner frequency are usually sufficient to allow a fairly accurate curve to be
sketched, as folIows.
Magnitude at w = W c
The actual gain at this frequency is
H dB = 20 log v' (1 + 1) = 3 dB (strictly 3.01 dB)
Magnitude at w = w c /2 and w = 2wc
The dB gain at wc /2 is 1 dB (wh ich is 1 dB away from the low-frequency
asymptote), and at 2wc is 7 dB (which is 1 dB away from the high-

frequency asymptote).
Magnitude at W ~ (wc /10) and W ~ 10wc
The dB gain at these frequencies almost lie on the low- and high-frequency
asymptotes, respectively.
Taking these values into account, a plot of the gain curve in the region
of the corner frequency is shown in figure ll.ll(b).
Phase-frequency plot
The phase-frequency response is calculated from the

LH (jw) = L(1 + jWT) = tan-1wT
This is a smoother curve than the magnitude-frequency asymptotic curve
but, nevertheless, we can draw a simple asymptotic response as follows
1. When WT is small
Here, very approximately, we take this to mean 0 ~ WT < 0.1.
LH (jw) : : : : tan- 1 0 = 0°
2. When WT is large
Here we can take this to mean 10 ~ WT < 00.
LH (jw) : : : : tan- 1 00 = 90°

3. When 0.1 < WT < 10
The plot approximates in this region to a straight line which joins the low-
and high-frequency phase asymptotes. This is a line of slope 45°/decade.
The resulting asymptotic phase-frequency curve is shown in full line
figure 11.12. The actual phase curve is approximately 5° above the junction
of the low- and mid-frequency asymptotes, and is 5° below the junction of
the mid- and high-frequency asymptotes. The actual curve and the mid-
frequency asymptotic line intersect when WT = 1, whereLH(jw) = 45°.
Aseries circuit contains a resistor of 25 n resistance and an inductor of

0.25 H inductance. Write down an expression for the s-domain impedance
of the circuit, and sketch the frequency response curve showing how the
impedance varies with frequency.
LH(jw)
90°
Actual ~/
curve I
I ~-- Asymptotic
I plot
45°
o~~~-~~-r--~-~_wr
0.01 0.1 10 100
Figure 11.12 The simplijied asymptotic phase angle response tor H(joo) = (1 + joJ't). The
actual curve is shown dotted.
Solution
The expression for the s-domain impedance is

Z(s) = H(s) = R + Ls = 25 + 0.25s = 25(1 + O.Ols)
that is
Z(jw) = H(jw) = R + jwL = 25 + jwO.25
= 25(1 + jwO.01)
The final step in the impedance equations has been taken in order to
convert the expression into the standard form developed in section 11.11.
Initially, the equation is reduced to one containing the factors described
earlier, namely a constant (= 25) and a term of the form (1 + jW'l'), that is,
(1 + jwO.01). The magnitude in dB of each is drawn separately in figure
1l.13(a) and, finally, the two are added to give the complete asymptotic
magnitude plot, which is drawn in fulliine in figure 11.13(b). The K = 25
factor gives a constant gain of 28 dB, and the (1 + jwO.01) term gives a
break point at O.Olwc = 1 or W c = 100 radis. In this case, the Bode diagram
has been plotted to a logarithmic base of frequency. The factor (1 +
jwO.01) indicates that there is a simple zero at s = -100 on the s-plane.
There is a correction of +3 dB to be made at the corner frequency on
the magnitude plot, and the corrected curve is shown in dotted line in
figure l1.13(b).
An inspection of the magnitude plot indicates that the impedance of the
circuit changes from about 25 n (corresponding to 28 dB) at 10 radis (1.592
Hz) to about 250 n (or 48 dB) at 1000 radis (159.2 Hz).
Next we look at the phase plot. The constant term of 25 in the H(jw)
Actual
plot
40
20 log 25 = 28 dB
---------~---
~/
/ 40
28dB
\A
-'" Linearised
plot
20 / 20
(1 + 0.01 jw) /
~/
~r---I""--f_-r__-~w (radis) -t-,---,---,---r__-~ w (radis)
10 100 1000 10000 10 100 1000 10000
(a) (b)
Figure 11.13 (a) Plot of the magnitude factors ofH(jw) = 25(1 + jwO.01), (b) the
complete Bode magnitude plot.
LH(jw)
90°
Linearised
plot
-+-r--='T---,---,,---"--W (radis)
10 100 1000 10000
Figure 11.14 The Bode phase plot for 25(1 + jwO.01).
expression imparts zero phase shift, so that the overall phase shift is equal
to that of the (1 + jroO.01) term. This produces corner frequency on the
asymptotic phase plot (see figure 11.14) at 100 radis. If necessary, the
asymptotic phase plot can be corrected to give an accurate curve, as shown
in broken line in the figure.
Normally, both the magnitude and phase plots are shown on one dia-
gram.
11.12 Bode diagram of H(jm) = (1 + jmT)"

As with the case of the H(jm) = (1 + jmT) response, the low-frequency
gain is approximately zero dB up to a frequency of mT = 1, that is, m = 1fT.
At this point the high-frequency asymptote assumes a slope of 20n
dB/decade (or approximately 6n dB/octave), as shown in figure 11.15(a); the
reader will find it an interesting exercise to verify the value of the slopes.
Once again, the actual magnitude curve will deviate from the associated
asymptotes, but in this case the deviation will be n times greater than in the
case for the H(jm) = (1 + jmT) curve. That is, at the corner frequency, the
deviation is 3n, dB, and at twice and one-half of the corner frequency the
deviation is n dB from the associated asymptote.
n=3 n=2
LHUro)
80
60 270° ~--n=3
40 180° ._--n=2
20 90°
ror 0
-20 _90° 0.01 100
-40 -180° --n=-2

-60 -270° '----n=-3
-80
n =-2 (b)
n =-3
(a)
Figure 11.15 (a) The magnitude-frequency and (b) the phase-frequency asymptotic
response for H(jw) = (1 + jw"t)n.
The phase shift (see figure 11.15(b)) is, as might be expeeted, n times
greater at eaeh frequeney than it was for the H(jw) = (1 + jwr) eurve.
Onee again, the angular deviation of the aetual eurve from the start of the
low-frequeney phase asymptote, and the end of the high-frequeney asymp-
tote in figure 11.15 is 5n°.
A reasonably aeeurate phase eurve (to within about 10 aeeuraey) ean be
obtained with a little more trouble, by drawing straight-line segments
between the points listed in table 11.2.
Table 11.2
(Ur Phase angle
0.01 00
0.1 5n°
0.5 25n°
2.0 65n o
10.0 85n°
100.0 90n°
The voltage gain, Vou/V1m of a transistor amplifier is

Vout -2.5s
H(s) = V in (s) = (1 + sIlOO)(1 + silO 000)
Draw the Bode diagram for the amplifier.
Solution
The expression for the voltage gain can be written in the form
. Vout . -0.5jw
H(Jw) = Vin (Jw) = (1 + O.Oljw)(l + lO-Sjw)
We can look at the solution in terms of the factors -0.5, JW, (1 +
O.Oljw)-\ and (1 + 1O- 5 jw)-I, as follows:
1. The gain associated with the constant factor is
H dB = 20 log 1-0.51 = -6 dB
and is shown as the broken li ne (1) in figure 11.16(a). The reader should
note that the negative sign associated with the numerator is a 'phase'
factor of 1800 rather than a 'gain' factor.
2. The factor jw results in a gain which increases at a constant rate of 20 dB
per decade, and passes through 0 dB at w = 1 radis, shown by broken
line (2) in figure 11.16(a).
3. The factor (1 + O.Oljw)-1 gives a corner frequency at w = 110.01 = 100
radis, shown in broken line (3) in figure 11.16(a).
4. The broken line (4) in figure 1l.16(a) corresponds to the factor
(1 + 1O- 5 jw)-\ and has a corner frequency of 1110- 5 = lOS radis.
HdB /'
40 /
(a) //(2)
20 /'
/'
/'
1 /' 10 100 103 10' 10' 10· 10'
o / w (radIs)
(1)· - - - - - - , - - - - - - - - - . . . , . - -(1)
,
,
/ ~
~
(3)~
'
(4)~
34dB
/-- .......
(b) 20 ;.-", Actual curve / "
Figure 11.16 The gain curve Jor worked example 11.12.1.

10 100 103 10' 10' 10· 107

O::--lf-L-----I,.,.........
'-- .......-L----L..(2-)-~.......
- ...........L.-/--+-(-3)-....Leo- w (radIs)
-90°-+-_ _...::-- !]1':~ ~ ____ . . . '--~ j __ _
1
Total asymptotic
phase curve
-270°
Actual curve ~
.....,----
LH(jw)
Figure 11.17 The phase curve for worked example 11.12.1.
The overall magnitude response is the sum of the four separate elements,
and is shown in figure 11.16(b). The corrections to the corner frequencies
are shown in broken lines in figure 1l.16(b). The mid-band gain is seen to
be 34 dB, corresponding to a numerical gain of 50.
When dealing with the phase shift, we can divide the transfer function
into three factors, namely (-0.5jw), (1 + O.Oljw)-l and (1 + lO-Sjw)-l, as
folIows:
a. The phase shift associated with the factor (-0.5jw) isL-j = -90°,
shown as broken line (1) in figure 11.17.
b. Tbe term (1 + O.Oljw)-l gives a phase lag which increases from zero (or
thereabouts) to -90° (or thereabouts) over the range 10 radis to 1()3
radis, shown as broken line (2) in figure 11.17.
c. This is similar to (2), but the phase shift occurs over the frequency range
1()4 to 1()6 radis, shown as broken line (3) in the figure.
The total phase shift is the sum of the three phase shifts, and is drawn in
full line in figure 11.17. The linearised phase plot can be corrected at the
corner frequencies, as outlined earlier, and is shown in broken line in the
figure.
11.13 Plot of H(jw) = 1 + 2~ [ ~~o] + [~~O]2

<AI <AI
This quadratic expression represents a conjugate pair of zeros (if it is in the

numerator of a transfer function) or poles (if in the denominator) on the
s-plane; these frequently occur in electrical and electronic circuits. The
quantity ~ is the damping factor, imd Wo (as we shall see later) is the corner
frequency of the asymptotic response for this'function; when ~ = 0, Wo is
the resonant frequency of the system.
40 40
20
o~ ______+-____ ~_w
0.1 10
(a) (b)
1
Figure 11.18 (a) Bode asymptotic magnitude plot for H dB = 20 log 1 + j2'r" [~ _ [~21
(b) Details at the corner frequency for different values of damping factor.
Magnitude Response
The expression for the magnitude is
1. When w/wo « 1
H dB "'" 20 log 1 = 0 dB
This gives the low-frequency asymptote, which is a straight line on the 0

dB axis.
2. When w/wo » 1
H dB "'" 20 log 1- [ :0 ]21 = 40 log :
The resulting high-frequency asymptote is a line of slope 40 dB/decade.

The two asymptotes intersect (which is the corner frequency of the plot)
»
when (40 log (w/wo = 0, that is when w/wo = 1 or w = Wo, as shown in
figure 11.18(a).
In order to get an accurate impression of the magnitude curve, we must
make corrections at the corner frequency. To a first approximation we may
say that, whatever the value of ein the range 0.1 to 1 (which is the range in
which most engineers are interested), the actual magnitude curve begins to
leave the low-frequency asymptote at about w/wo = 0.2, and rejoins the
high-frequency asymptote at about w/wo = 2 (see figure 11. 18(b». We
need one other cardinal point at a given frequency in order to sketch the
Table 11.3
~ H dB
0.1 -14
0.25 -6
0.5 0
1 6
magnitude curve for a given value of ~; an appropriate frequency is w/wo

= 1, or w = WO' The value of H dB at this frequency is calculated from
H dB = 20 log 11 + j2~ - 11 = 20 log Inl
A list of values for various damping factors is given in table 11.3, and the
associated magnitude curves are shown in figure 11.18(b).
The reader should note that, when there is negative peak in the magni-
tude curve, it occurs at a frequency which is a little less than w/wo = 1.
Phase response
The phase shift associated with H(jw) = 1 + 2~ [ Wo

jw ] + [ ~~o
LV ] 2 is
2~ w/wo
LH(jw) = tan- 1 ---~
1 - (w/WO)2
1. When w/wo < < 1
LH(jw) = tan- 1 (2~(w/wo» = 0°
This gives the low-frequency asymptote, which lies on the 0° axis. Within
reasonable accuracy, this can be through to span from a very low
frequency up to about w/w o = 0.1.
2. When w/wo > > 1
LH(jw) = tan- 1 (-2~/(w/wo» = 180°
That is, the high-frequency asymptote is a straight line at a phase shift of
180°. This spans from about w/w o = 10 up to infinite frequency.
3. When w/wo = 1
LH(jw) = Lj2~ = 90°
The asymptotic phase curve therefore consists of three straight lines, as
shown in figure 11.19(a); the mid-frequency line which joins the low- and
high-frequency asymptotes has a slope of 900 /decade, and has a phase shift
of 90° when w/wo = 1.
Finally, we need to calculate a few values of w/wo which will allow us,
LH(jw)
180'
90'
O'-tT---'~---,r-----,r-----,_ !!!.
Wo
0.01 0.1 10 100
(a) LH(jw)
180'
,.~---
--;
/,-;
I/Ift,
1/
~ ~ 1
90'
~e,o;,o."
.-,~~-0.1
".-e
0' '::-e/ !!!.
0·.01 0.1 w,
10 100
(b)
[:J
Figure 11.19 (a) Linearised approximation to the phase response ofH(jw) =
1+ 2~ [:::) + 2
and (b) the estimated curves for a range of
selected values of damping factor.
with a reasonable degree of accuracy, to sketch the phase curve for any
selected value of ;;. We will seleet the values w/wo = 0.5, 1 and 2.
1. When w/wo = 0.5, the overall phase shift is
LH(jw) = tan- 1 (1.3333;;)
2. When w/wo = 1, the overall phase shift is 90°.
3. When w/wo = 2, the overall phase shift is
LH(jw) = tan- 1 (-1.3333~)
The corresponding values for various values of ~ are listed in table 11.4.
The corresponding phase curves are drawn in broken Hne in figure
11.19(b); the points in table 11.4 are shown on the curves as large dots.
If the quadratic expression is in the denominator of the transfer function
(see, for example, worked example 11.13.1), both the gain and phase
curves are inverted or reversed. That is, the high-frequency asymptotic
gain curve has a slope of -40 dB/decade, commencing at w/wo = 1; the
asymptotic phase plot has a slope of -900 /decade, commencing at w/wo
= 0.1, causing the phase to change from 0° to -180° by w/wo = 10.
Table 11.4
Damping factar Phase shift (deg.) at
w/w o = 0.5 w/w o = 2

0.1 7.6 180 - 7.6 = 172.4
0.25 18.4 180 - 18.4 = 161.6
0.5 33.7 180 - 33.7 = 146.3
1 53.1 180 - 53.1 = 126.9
The transfer function of an operation al amplifier circuit is

Vout -2000s
H(s) - Vin (s) - 10 000 + 20s + S2
Draw the Bode diagram for the amplifier.
Solution
This expression can be re-written in the form

-0.2s
H(s) - - - - - - - -
1 + 2 X 1O-3s + 1O- 4s 2
hence
-0.2jw
H(jw) = ------=-----
1 + 2 X 1O- 3 jw + 1O- 4 (jw 2 )
The numerator contains a constant together with a simple linear function
of w, and the denominator is a quadratic expression. The linearised gain
curve is plotted as folIows.
1. The decibel gain of the numerator constant is
H dB = 20 log 0.2 = -14 dB
which is shown in figure 11.20(a) as the broken line (1). The reader should
note that the negative sign associated with the numerator term is a 'phase'
factor of 180°, rather than a 'gain' factor.
2. The gain associated with the numerator jw term is a line of slope 20
dB/decade, crossing the 0 dB axis at w = 1 radis. It is shown as broken line
(2) in figure 11.20(a).
3. The quadratic term in the denominator has a corner frequency at w =
\1(1110- 4 ) = 100 radis. The damping factor is determined from the fact that
2Clwo = 2 x 10- 3 , or C= 0.1. The resulting curve is shown as broken curve
(2)
20 dBldecade /
40 ,~V/
/
20 /
/ / (3) ,-,14 dB
/ I
o -Fr=.=.;;t'/==r=/:::.,./--,-1r----r- w (radis)
0.1 / 1 10 100\ 1000
-14- -L-------- L _ -(1)
-20 / \
\
......-,,
\
\
-40
-40 dBldecade \
\
Figure 11.20 (a) Gain plots for the individual factors of the transfer [unction in worked
example 11.13.1 and (b) the overall gain plot.
(3) in figure 11.20(a). The maximum gain for a damping factor ofO.1 is + 14
dB in the region of the corner frequency of 100 radis.
The three asymptotic plots in figure 11.20(a) are combined to give the
complete magnitude plot in figure 11.20(b). The maximum is 40 dB at the
corner frequency of 100 rad/so
The phase curve is plotted in two parts. The - j (= 1/j) term in the
numerator contributes a constant phase shift of -90° (see broken line (1) in
figure 11.21), and the quadratic term in the denominator gives a phase shift
which increases from a low value (about _5°) to about -180° between the
frequencies of 100 radis and 1Q4 radis (see broken line (2)). The overall
phase shift is the sum of the two; the asymptotic phase curve is shown in
fulliine in figure 11.21, and the actual phase curve is in broken line.
LH(jw)
0° -F"F"==t-----,r---,---,- w (radis)
10 100', 103 10' 10'
,
-90° +---""'tO<:",-~------(l)
" ,
-180°
\ " ' -_____ (2)
\ . .
\ _ _ Total asymptotlc phase
-270° /'_::"---
Total phase
Figure 11.21 Phase plot for worked example 11.13.1.

Unworked problems
11.1. State the complex frequencies in the following time functions:

(a) 5 cos 8t, (b) (5 + 4e- 6t cos 8t, (c) 5 + 4e- 6t cos 8t,
(d) 4e- 6t cos (8t - 20°), (e) 4 cos 3t - 5 sin 2t, (f) 8 cos 2t. sin 3t
[(a) j8, -j8; (b) j8, -j8, -6 + j8, -6 - j8; (c) -6 + j8, -6 - j8;
(d) -6 + j8, -6 - j8; (e) j3, -j3, j2, -j2; (f) j1, -j1, j5, -j5]
11.2. Write down the function of time having the following complex
frequencies: (a) -6, (b) 7, (c) 2, -3, (d) 0, -6 + j4, -6 - j4; (e)
-5, 6, 7 + j2, 7 - j2.
[(a) Ae- 6t ; (b) Ae7t ; (c) Ae2t + Be- 3t ; (d) A + Be- 6t cos (4t + 111); (e)
Ae- 5t + Be6t + Ce7t cos (2t + 111)]
11.3. A voltage VI is defined by
(3 + 4s)
200 = (2 + 3s)VI + VI
s
Determine an expression for VI> and calculate its value if s =
-3 + j4. What is its value at t = 0.2 s?
[2oos/(3s 2 + 6s + 3); 16.67 V; 6.37 V]
11.4. At what value of s is Z(s) zero in figure 11.6. At what value of s is
Z(s) = 7.9 = jO.06 O?
[-0.0625 ± jO.348; -5 - j5]
11.5. Determine an expression for Zin of figure 11.22 as a function of a?
Hence determine the frequencies of the poles and zeros. What is
the circuit impedance at (a) 0 = 0, (b) 0 = oo?
[2(0 + 3.732)(0 + 0.268)/(1 + 0); zeros at -3.732 and -0.268, and
a pole at -1; Z(O) = 20; Z(oo) = 00]
Figure 11.22
11.6. The complex-frequency impedance of a circuit has simple zeros at

s = -6 and s = -10, and a pair of poles at s = -4 ± j8. If its
impedance at infinite frequency is 5 0, deduce an expression for
z,
v, I
Figure 11.23
Z(s). If a forcing voltage v(t) = lOe- 21 cos (5t + 60°) V is applied to

the circuit, determine the current i(t) ftowing into the network.
[5(s + 6)(s + 1O)/(s2 + 8s + 80); 1.57e- 21 cos (5t + 1.6°) Al
11.7. If the op-amp in figure 11.23 is ideal, deduce the transfer function
H(s) = Vis)IV(s) as a ratio of two polynomials in s. Determine
H(s) if (a) Z( = RB Zr = Rf in parallel with Cr. (b) Z( = R( in series
with Cl' Zr = R f •
[-ZrCs)/Z(s); (a) -Rtf(Rl(sRfCf + 1)); (b) -sC l Rtf(l + sRlCl )]
11.8. In problem 11.7, given that R f = 10 kil, calculate values of R l and
Cf so that H(s) is (a) -10, (b) -250/(s + 50).
[Cf = 0 (open-circuit), R l = 1 kil; (b) Cf = 2 fJ.F, R l = 2 kil]
11.9. The asymptotic magnitude-frequency Bode diagrams of a number
of systems are described below. Draw the magnitude-frequency
and phase-frequency diagrams, and deduce the transfer function
for each.
(a) Constant gain of 40 dB up to 2 radis, after which the gain falls at
the rate of 20 dB/decade.
(b) Constant gain of 6 dB up to 20 radIs, after which it increases at
the rate of 6 dB/octave.
(c) Constant gain of 10 dB up to 5 radis, after which it falls at the
rate of 6 dB/octave up to 10 radis, after which the gain remains
constant.
(d) The gain falls at 6 dB/octave up to 4 radis, after which it falls at
40 dB/decade. The gain at 1 radis is 60 dB.
(e) The gain falls at 40 dB/decade up to 20 radis, when it falls at 6
dB/octave up to 2000 radis, when it falls again at 12 dB/octave. At 2
radis the gain is 60 dB.
(f) The gain falls at 6 dB/octave up to 5 radis, after which the
reduction in gain becomes 40 dB/decade. At 10 radis the fall in gain
re duces to 6 dB/octave again until, at 100 radis, the gain reduction
be comes 40 dB/decade again. The gain at 1 radis is 60 dB.
[Ca) 10/(1 + 0.5jw); (b) 2(1 + 0.05jw);
v,1
Figure 11.24
(c) 3.162(1 + 0.1jw)/(1 + 0.2jw); (d) 1OO0/(jw(1 + 0.25jw»;

(e) 4000(1 + 0.05jw)/«jW)2(1 + 0.0005jw»;
(f) 1000/(1 + 0.1jw)/(jw(1 + 0.2jw)(1 + 0.01jw»]
11.10. Draw the Bode diagram for each of the following transfer func-
tions: (a) 10(1 + 0.2jw), (b) 20(1 + 0.02jw)Z, (c) 1/(5 (1 + 0.2jw)2),
(d) 1O/(jw(1 + 0.05jw», (e) 100(1 + 0.2jw)/(jw(1 + 0.05jw»,
(f) 20(1 + 0.1jw)/«jW)2(1 + 0.05jw))]
11.11. Deduce the transfer function H(s) = Vis)/V1(s) for the amplifier in
figure 11.24. Plot the Bode diagram for the circuit if R = 1 kn,
C, = 2.388 X 10- 3 F, CZ = 10.613 f.lF.
[1/«R 2C,C2)S2 + 3RC,s + 1); with the values given, the peak
response on the gain curve occurs at a frequency of 1 Hz, when the
phase shift is 90° and the gain is 14 dB. Note: the reader may find it
advantageous to reler to worked example 11.7.2
12
Resonance
12.1 Introduction
It was shown earlier that as the frequency increases, inductive reactance

increases and capacitive reactance reduces. Because of this change in the
magnitude of the reactances, and of the opposing phase angle of the two
forms of reactance, there may be a frequency at which XL and Xc in a given
circuit are equal in magnitude and opposite in phase. When this occurs, we
reach a condition of electrical resonance in the circuit.
Depending on the circuit, this can result in a large current in the circuit
and a large voltage across part of the circuit. This is of particular interest to
electrical engineers.
The resonance phenomenon is not limited to electrical engineering, and
can exist in mechanical systems, as witness high amplitude vibrations at the
natural [requency of mechanical systems. Perhaps the most famous of these
was the destruction of the Tacoma Narrows Bridge in Washington State,
USA, when the bridge was forced into astate of resonance by a pulsating
gale.
12.2 The resonant condition
Resonance in a two-terminal network containing at least one inductor and

one capacitor is defined as the condition which causes the input impedance
at one jrequency to be purely resistive. At this frequency, the voltage across
the circuit is in phase with the current through it. In some circuits there
may be more than one resonant frequency.
A wide variety of resonant conditions occurs in electrical circuits includ-
ing series resonance, parallel resonance, resonance with selected
frequencies, resonance between magnetically coupled circuits, etc,.
299
Resonance may or may not be desirable, depending on the circuit in which

it occurs.
We have briefty looked at resonance in chapter 11 (see worked
examples 11.6.1 and 11.6.2) when we considered the frequency response of
idealised parallel and series circuits. In this chapter we will analyse practi-
cal circuits which include resistance.
12.3 Series resonance
We will look at a number of points of significant interest to electrical and

electronic engineers.
12.3.1 Resonancefrequency andfrequency response
The impedance of the series RLC circuit in figure 12.1 is

Z = R + j(wL - lIwC)
R jwL 1/jwC
~-------4~r-------~
Figure 12.1 Aseries RLC eireuit.
When wL = lIwC, that is, when XL = Xc, the impedance of the circuit is
purely resistive. This occurs at the resonant Irequency, wo, when woL
= lIwoC or
Wo = l/'v'(LC) radis
and
10 = wcl2n: = lI(2n:v'(LC» Hz
At resonance, the circuit impedance is
Zo = R + jO = RLO° n
and is the minimum impedance of the circuit. The current at resonance is
10 = EJR A
Resonance 301
IZI
(lowRI
!'i
"tJ
c:
ca
~ ~------~----------------
w
~
cx:-
(al
~
Cl
c:
ca
~ Or-------~r---------·w·
~
0..
(bI
"tJ
c:
ca
E
~
:;
u
w
(cl
Figure 12.2 The effect 01 frequency change in aseries circuit on (a) reactance and
impedance, (b) phase angle and (c) cu"ent and modulus 01 admittance.
and is therefore the maximum current in the circuit. If R is reduced in

value, the current at resonance is increased. At the re sonant frequency, the
supply voltage is
E. = loR
The frequency response of the series circuit is shown in figure 12.2.
90°
900 Es = VR
Figure 12.3 Phasor diagram for aseries circuit containing ideal elements at resonance.
At resonance XL = Xe, SO that the magnitude of the voltage across each

reactive element is the same at resonance. That is
foXL = foXe
A phasor diagram for aseries circuit containing ideal elements at res-
onance is shown in figure 12.3.
However, if the coil in the circuit has some resistance (as it will in
.practice), then the voltage across the coil at resonance is not equal to the
voltage across the capacitor (see worked example 12.3.2), and is not at 90°
to the current.
12.3.2 The quality laetor, Q., 01 the series eireuit
The quality factor of aseries circuit may be defined in any one of several
ways, including the following
energy stored in a given time
Qs = -----=~------==------
energy dissipated in the same time
power in a reactive element
power in the resistance
A high value of Qs indicates a low level of energy dissipated for a given
energy stored in a reactive element. If we use the inductor as the reactive
element, then
J2X L XL woL
Q =-=-=-
s PR R R
or, using the capacitive element
J2X e Xe 1
Qs = PR = R = woCR
Resonance 303
Also, we may say that

modulus of the voltage across L
_ )(L _ III)(L _ at resonance
Qs - R - IIIR - modulus of the voltage across R
at resonance
and
modulus of the voltage across C
_ )(e _ III)(e ______a_t_r_es_o_n_a_n_ce_ _ _ __
Qs - R - IIIR - modulus of the voltage across R
at resonance
or
Ivoltage across either L or C at resonancel
Qs = IEsl
Now, for the series circuit at resonance
woL 1 L 1
Qs = R = V(LC) R = RV(LlC)
Typically, in a well-designed high-quality (high-Q) communications circuit,
the value of Q may be 50 or greater. It has, of course, a much lower value
in a power-frequency circuit. In power circuits, where resistance values are
much lower than they are in electronic circuits, series resonance can give
rise to dangerously high values of current. Moreover, these currents can
produce very high values of voltage across the reactive elements, possibly
causing damage to the insulation of the elements, and excessive heating in
all the elements.
Sometimes the only effective circuit resistance is in the winding of the
coil, so that we can refer to the Q-factor of the coil, which is
)(L woL
QS=R=R
Aseries circuit contains a resistor of 4 n resistance, a pure inductor of 0.1

mH inductance, and a 111F capacitor. Calculate (a) the resonant frequency
of the circuit, (b) the net impedance of the circuit at resonance, (c) the
current in the circuit at resonance if E. = lOLO° V, (d) the voltage across
each element in the circuit at resonance and (e) the Q-factor of the circuit
at resonance.
Solution
(a) The resonant frequency is

Wo = 1/V(LC) = 1/V(0.1 X 10-3 X 1 X 10-6 )
100000 radis or 15 915.5 Hz
(b) At resonance
XL = wL = 100 000 X 0.1 X 10-3 = 10 n
and
Xc = 1/wC = 1/(100 000 x 1 x 10- 6) = 10 n
hence the impedance of the circuit at resonance is
Zo =R + j(XL - Xc) = 4 + j(10 - 10) = 4 n
(c) At the resonant frequency
10 = Es/Zo = lOLoo/4 = 2.5LO° A
(d) voltage across each element in the circuit is
VR IR = 2.5 x 4 = 10 V = Es
VL = IZ = 2.5 x lOL90° = 25L90° V
L
Vc = IZ c = 2.5 x lOL-90° = 25L-90° V = -VL

(e) We can calculate Qs from any of the equations in section 12.3.2 as
follows
= XLIR = 1014 = 2.5
Qs
Note: IvLI = Qs x IEsl = 2.5 x 10 = 25 V
If, in worked example 12.3.1, the whole resistance of the circuit is in the
coil, calculate the voltage across the coil at resonance.
Solution
Since all the resistance is in the coil, the impedance of the coil is (4 + j 10)
= 10.77 L68.2° Q. At the re sonant frequency the current is 2.5 A, and the
voltage across the coil is
10Zcoil = 2.5LO° x 10.77 L68.2° = 26.93L68.2° V
That is, the voltage across the coil at resonance does not lead the current
by 90°; this is due to the effect of the resistance of the coil. However, the
quadrature voltage across the coil is 25L90° V, wh ich is of opposite phase
angle but equal magnitude to the voltage across the capacitor.
Resonance 305
12.3.3 The peak voltage across R, Land C
One would imagine from the foregoing that, since the current rises to its
peak value at Wo, the p.d. across R, L, and Cwould peak at this frequency.
However, it is shown in the following that this is not quite the case.
Since the voltage across R is proportional to the current then, in fact, VR
does peak at Wo. However, since the capacitive reactance commences at a
high value and reduces with increasing frequency, I rises at a faster rate
than Xc falls. It can be shown that IVcl peaks just before resonance, and
IvLI peaks at a frequency just above resonance (the reader will find it an
interesting exercise to verify this fact). The two voltages are equal in value
at Wo.
For values of Qs greater than about 5, the frequency difference at which
the peak voltage occurs across R, Land Cis generally insignificant.
12.3.4 Bandwidth and selectivity 0/ aseries RLC circuit
If we plot the variation in power consumed in aseries RLC circuit to a base

of frequency, we get a graph of the type in figure 12.4.
At frequencies wl and w 2 , the power consumed is one-half the maximum
power consumed (wh ich occurs at the resonant frequency), and these
frequencies are known as the half-power frequencies or half-power points.
The sharpness of the peak in the resonance curve is described by the
bandwidth, B, and is given by
The smaller the bandwidth, the 'sharper' the amplitude response. The
values of (VI and W 2 can be calculated as folIows. At a half-power frequency
I/ol2R 0 ]2
power consumed
Po
= '2 = -2- = [/
v'2 R = [Rv'2
Es ]2
R
Power
Bandwidth, B
IL-_+-.........-t-_ _ _ w
o
Figure 12.4 Variation in power with frequency in aseries RLC circuit.
That is, the modulus of the impedance of the circuit at a half-power

frequency is
jzj = RV2 Q
The impedance of the circuit at any frequency is

Z = R + j(OlL - 1/wC)
=R+J. [ w LwoL
x - - -1- x -WoC]
-'
woL wC woC
Since jzj = RV2 at a half-power frequency, it is clear that at a half-power

frequency the quadrature component must have a value of unity, or
In the case where its value is -1 then
Qs(w
2
- Wo) = -1 or Qsw + 000 00 - QsWo =0
000 00
Solving the quadratic equation for 00 gives a positive value and a negative
value. Selecting the positive value of 00 gives the half-power frequency
In the case where its value is -1, we get

Qsw - 000 00 - Qs% = 0
and the resulting half-power frequency is
The bandwidth of the series resonant circuit is, therefore

Wo 1 R
B = 002 - Oll =- = -- = -
Qs CR L
The selectivity of a circuit describes the sharpness of the response curve,
Resonance 307
and defines the ability of the cireuit to diseriminate in favour of frequencies

near resonanee. Its value is equal to the Q-factor of the cireuit, and is
seleetivity = Q = wc/B
Calculate the half-power frequencies and the bandwidth of the series

cireuit in worked example 12.3.1. Determine also complex impedanee (a)
10 per cent below and (b) 20 per cent above the resonant frequeney.
Solution
The reader is reminded, in worked example 12.3.1, that Wo = 100 000 radis
and Qs = 2.5. From the foregoing
w\ = w\ [ - 2~s + V([2~J 2 + 1) ]
=100000 [ 1 '\ /( [
+V 1 ] 2 +1 ) ]
2 X 2.5 2 X 2.5
= 100 000 [-0.2 + V1.04] = 81 980 radis
and
= 100 000 [0.2 + V1.04] = 121 980 radis

henee
B = W2 - w\= 40000 radis
or B = wc/Q = 100 000/2.5 = 40 000 radis
The reader should note at this point that the resonant frequeney is not
midway between the two half-power frequencies (in fact, it is the geometrie
mean of the two frequeneies, that is Wo = V(w\w2 ».
. The eomplex impedanee at any frequency ean be ealculated from the
expression
(a) When w = 0.9wo
Z = 4 [1 + j2.5( 0.9w Wo
o - ~)]
0.9wo
= 4[1 + j2.5 X (-0.2111)] = 4.52L-27.82° Q
(b) When W = 1.2wo
Z = 4 [1 . (-
+ )2.5 1.2w
-o - -Wo)
- ] = 5.44L42S Q
Wo 1.2wo
12.4 Parallel resonance
In this case, the attention of the reader is directed towards circuits contain-
ing G, Land C in parallel.
12.4.1 Resonantfrequency andfrequency response of an ideal paraUel

circuit
An ideal parallel circuit contains a pure inductor, a pure capacitor and a

pure conductance in parallel, as shown in figure 12.5. The admittance of
the circuit is
Figure 12.5 Parallel circuit containing ideal elements.
Y G + j(B c - B L )
= G +j ( wC - W~) S
Once again, the circuit is resonant when the reactive element 01 the admitt-
ance is zero, wh ich occurs at the resonant Irequency Wo, when woC = lIwoL,
that is
Wo = 1/V(LC) radis or 10 = 1I(23tV(LC» Hz
Since, at resonance, the reactive element of the admittance is zero, the
current at resonance is
Resonance 309
Clearly, G can have any value, induding zero (corresponding to an infinite

resistance shunting the circuit), so that the current at resonance represents a
minimum value 0/ current.
Since, at resonance, Y o = G, then the circuit impedance at resonance is
Zo = 1/G
The value of Zo is known as the dynamic impedance or dynamic resistance
of the circuit.
At resonance, the parallel LC section of the ideal parallel circuit
(known as a tank circuit) , behaves as though it were an open-circuit, that is,
no current is drawn from the supply. The reason for this is that the
magnitude of the current in both the Land the C branches has the same
value (this is because they both have the same value of reactance at
resonance), but are 1800 out of phase with one another.
The frequency response for the circuit is shown in figure 12.6, and the
phasor diagram at resonance is in figure 12.7
:>-
.,
"0
<:
!Xl"jL------'---------t_
w
r:c:'
6
,. ,,"
..- ...
B,
-
/'
/
I
I
(a)
+90 0
.,
Cl
<:
~ 0 r----~~-----w ..
.c
a..
-90 0
(b)
(c)
Figure 12.6 Frequency response diagram for a parallel GLC circuit.

Figure 12.7 Phasor diagram tor a parallel GLC circuit at resonance.
12.4.2 The quality /aetor, Q,. 0/ an ideal paraUel eireuit

As with the series cireuit, the Q-faetor is defined as the ratio of the energy
stored at resonanee in one of the reaetive elements to the energy dissipated
in the eireuit (whieh is in the shunting eonduetanee). Alternatively, it is the
ratio of the reaetive power in a reaetive element to the power dissipated in
the shunting eonduetanee. Seleeting the induetive braneh gives
E~X
Q [2X X R
=....!:........!=_L_=_=
2 L R
__ 1
= __
P [iR E; R XL woL GwoL
R 2
where R = lIG. Similarly, it may be shown that

V(C/L)
Qp = woCR = RV(C/L) = G
Also at resonanee
modulus of the eurrent in L or C
Qp = -----~-------
supply eurrent
If we seleet the induetor
Similarly, seleeting the eapacitive braneh gives

Qp = woCR
Resonance 311
12.4.3 Bandwidth and seleetivity 0/ a parallel eireuit

As with the se ries circuit, the ideal parallel circuit consumes maximum
power at resonance (the reader will find it an interesting exercise to argue
this point both quantitatively and qualitatively), and the bandwidth is
equal to the band of frequencies between the two half-power frequencies.
The admittance of the parallel circuit at any frequency is
y=.!.R + j (wc-
wL
1 ) = .!. [ 1 + j Qp (~ _ Wo)]
R Wo w
where G = l/R, and at each half-power frequency Iyl = Y2JR, which occurs
when the quadrat ure term has unity value, that is Qp (:0 -:0 )= ± l.
Considering the case for + 1 and - 1 separately, we obtain the two half-
r
power frequencies
Wo [ - 2~pvi([ 2~p +
r
w1 + 1) ]
W2 = Wo [2~p + vi ([ 2~p + 1) ]
The reader should note that W 1 and W 2 are given by the same formulas as
for the series circuit (see section 12.3.4). The bandwidth of the circuit is
B =W 2 - W1 = wo/Qp = woXL/R = woXe/R
also
selectivity = Qp
12.4.4 Resonanee in a praetieal parallel eireuit
A simple practical parallel circuit has two branches (see figure 12.8(a»,
each containing a reactive element and some resistance (the resistance in
the capacitive branch may, alternatively, be shown as a leakage resistance
in parallel with the capacitor). A typical phasor diagram at resonance is
shown in figure 12.8(b).
If we convert the circuit in figure 12.8(a) into the equivalent ideal circuit
in figure 12.8(c), we can use the equations developed earlier to determine
various factors associated with the circuit.
Initially we will convert the capacitive branch of figure 12.8(a) into its
equivalent parallel circuit. The admittance of this branch at frequency w is
1 jwC
_......::...._--= jwC(1 - jwCR e
y =-----
I Re + l/jwC 1 + jwCR e 1 + (WCR e)2
y--
(a) (b)
y- G'
(c)
Figure 12.8 (a) A practical parallel circuit, (b) a typical phasor diagram at resonance and
(c) the equivalent ideal parallel circuit.
__RdwCY
:::..0...._"--_+ j wC = G~ + jwe
1 + (wCRcY 1 + (wCRcY
where e = C/[l + (WCRc)2] and G~ is the equivalent conductance in
parallel with e. In the case of the inductive branch of figure 12.8(a) we get
1 R L - jwL
Y2
RJ + jwL R L 2 + w2L 2
RL wL G' ,1
= - j = J WL'
R~ + w2V R~ + w2L 2 L -
where L' = (R~ + W 2V)/w 2 L, and G~ is the equivalent conductance in

parallel with L'.
The equivalent ideal parallel circuit in figure 12.8(c) comprises the two
above branches in parallel, having an effective conductance of
G' = G' + G' = Rc(wCY + RL

c L 1 + (WCRc)2 R~ + w2L 2
We can therefore say that, for the circuit in figure 12.8(a), the re sonant
frequency is
Resonance 313
WO = 1/v'(L'C') = 1
v'(LC)
V[LlC -
LlC -
R 2L] radis
R~
In the more usual case where Re = 0, the re sonant frequency is
1
v'(LC)
VL- [ L
CR~]
radis
and if, at the same time R~ « (WOL)2 (or QL » 1), the dynamic im-
pedance at resonance is
W orked example 12 .4.1
A transistor amplifier with an output resistance of 50 kQ is connected to a

load comprising a two-branch parallel circuit wh ich is re sonant at 60 kHz;
the parallel circuit contains a coil of inductance 1mH and resistance 15 Q.
If the bandwidth of the complete circuit is to be 3 kHz, calculate the other
components in the circuit (assurne that the capacitor has no leakage resist-
ance).
Solution
The Q-factor of the complete circuit is
Q = frJ R = 60 000/3000 = 20
Since the output impedance of the amplifier is 50 kQ, it is effectively

connected across the resonant circuit as shown in figure 12.9(a). The
idealised equivalent circuit is shown in figure 12.9(b), and we will calculate
the value of L' and C' in the following. From the work in the section above
Raut
= 50 kO
c 50 kO c
~
Parallel
circuit
(a) (b)
Figure 12.9 (a) The equivalent output circuit for worked example 12.4.1 and (b) the
idealised circuit.
L' (R~ + w~U)/aroL)
(15 2 + [2:rt X 60 oooy X 1 X 10- 6 )

(2:rt X 60 000)2 X 1 X 10-3
1.002 X 10-3 H or 1.002 mH
or, at the resonant frequency
X~ = woL' = 377.75 Q
and
G~ RL/(R~ + w~U)
15/(152 + (2:rt X 60 oooy X 1 X 10-6 )
1.0537 X 10- 4 S or 0.10537 mS
The effective conductance shunting the L' C section of the circuit is
G" = G~ + 1150 000 = 1.2537 X 10- 4 S
hence the Q-factor for the curcuit in figure 12.9(b) is
Q 1/(G"X~) = 1/(1.2537 X 10- 4 X 377.75)
= 21.12
This is larger than the required value of 20 (see the beginning of the
solution), and we can reduce the overall Q-factor to the required value by
shunting the circuit with another conductance. The total conductance
needed to reduce the Q-factor to 20 is
G = 1IQX~) = 11(20 X 377.75) = 1.324 x 10- 4 S
hence the extra shunting conductance is
G - G" = (1.324 - 1.2537) x 10- 4 = 0.0703 X 10-4 S
which corresponds to a shunting resistance of 142.25 kQ.
Finally we can calculate the value of the capacitor C from Wo =
1IV(L' C) as follows
C = 1I(%L' )
= 1I([2:rt x 60 000]2 x 1.002 x 10-3 ]
= 7.02 x 10-9 F or 7.02 nF
The resulting circuit is shown in figure 12.10.
Resonance 315
Amplifier
r--- ------,
I
I
I
I 40kO 142.25 kO 7.02 nF
I
I
I
I
L _ _ _ _ _ _ _ _ .J ~
Parallel
circuit
Figure 12.10 The eomplete eireuit for worked example 12.4.1.
12.5 Scaling electrical circuits
When analysing a particular circuit, the values used may be difficult to

handle or to visualise. It may, for example, be easier to think in terms of a
resonant frequency of 1 radis rather than 10367.6 radis, or an impedance
level of 1 Q rather than 579.7 Q. Using standard equations, we can 'scale' a
circuit so that it appears, initially, to have relatively simple values; after
initial calculations, we can return the circuit to its original form by the
inverse of the scaling process.
Scaling is achieved in two stages, namely impedance scaling and fre-
quency scaling as folIows.
12.5.1 Impedance or magnitude scaling
To magnitude-scale an impedance by a factor Km, we need to multiply the

impedance of the element by a scaling factor Km. Thus a resistor of R Q is
scaled by Km to a value KmR Q. An L H inductance of impedance jwL is
scaled to an impedance of jwKmL Q; that is, the scaled value of inductance
is KmL H. A capacitor of C F capacitance has an unscaled impedance of
l/jwC Q, and this is scaled to Km/jwC Q; that is, the scaled capacitance is
CI Km. Thus, the magnitude scaling changes are
R becomes KmR
.L becomes KmL
C becomes CI Km
Similarly, it can be shown that voltage values are multiplied by Km, and
current values are divided by Km. Ratios of similar quantities, such as the
Q-factor of aresonant circuit, are unchanged by magnitude scaling. Other
non-impedance factors, such as resonant frequency, are unchanged by
impedance or magnitude scaling.
Consider the following values from worked example 12.3.1: R = 4 Q,

L = 0.1 mH, C = I!lF. If we magnitude scale these by a factor of Km = 10\
then the scaled values are R' = 4 X 104 Q, L' = 1 H, C' = 1 X 10-10 F. The
original values refer to aseries circuit having aresonant frequency of
100000 radis, and a Q-factor of 2.5. The corresponding values for the
scaled circuit are
w~ = 1/Y(L' ,C') = 1/Y(1 x 1 x 10-10)
= 100 000 radis
and
Q~ = w~L'IR' = 100000 x 1/(4 X 104 ) = 2.5
both of wh ich are unchanged by magnitude scaling.
12.5.2 Frequency scaling
In this case, each frequency-dependent function is multiplied by a scaling

factor K f ; for example, the scaled frequency w becomes KtW. Consequent-
ly, the re sonant frequency and bandwidth of a tuned circuit are increased
by K f •
Clearly, the resistance of a resistor is independent of frequency, and its
value is unchanged by frequency scaling. In the case of an inductor, its
reactance is proportional to frequency, so that the frequency-scaled in duc-
tive impedance is re la ted to the original value by
jwL = jKtWL'
where L' is the frequency-scaled inductance; that is, the effective
frequency-scaled inductance is L' = LIK f • Similarly, a frequency-scaled
capacitance C' is ClKf • That is, the frequency scaling changes are
R be comes R
L becomes LIKf
C becomes CIKf
Since XL = wL = KtW x LlKf = wL (and similarly for a capacitive react-
ance), impedance levels are unchanged by frequency scaling. Following
this line of discussion, it will be seen that the Q-factor of aresonant circuit
is unchanged by frequency scaling.
Suppose we were to frequency scale the values in worked example
12.3.1 by Kf = 10-S, then
L' = LlKf = 0.1 x 10-3/10-5 = 10 H
C' = CIKf = lQ-6/10-5 = 0.1 F
Resonance 317
and
W~ =1/Y(L'C') = l/Y(10 x 0.1) = 1 radis
That is, the re sonant frequency has been multiplied by K f = 10-5 • The
Q-factor of the circuit is
Q' = XUR' = w~L'IR' =1x 10/4 = 2.5
which is unchanged.
12.5.3 Combined scaling
Many problems involve both magnitude and frequency scaling, and the
result involved both scaling factors. For R, Land C the effect is as follows
R becomes KmR
L becomes KmLl K f
C becomes CI KmKf
12.5.4 Universal resonant circuits
Universal resonant circuits are scaled circuits with R = 1 Q and Wo = 1

radis, that is, LC = 1. All conventional passive resonant circuits, that is,
se ries RLC and parallel GLC circuits, can be designed from this type of
circuit.
Calculate the impedance and frequency scaling factors for a parallel res-
onant circuit, in which the resonant frequency is 10 000 radis, and the LC
section is shunted by a 10 kQ resistor.
Solution
The resistance R' which shunts the resonant section is

R' = KmR = Km X 1 = 10 000
hence Km = 10 000
Also w~ = KrWo = 10000 X 1 = 10 000
that is K f = 10 000
Aseries RLC circuit containes the following elements: R = 100 Q, L = 1.6

mH and C = 0.01 IlF. Scale the circuit so that it contains a 2 Hinductor and
a 2 F capacitor.
Solution
From the above

L' = 2 = 1.6 X 10-3 KmlKf
C' = 2 = 0.01 x l~/KmKf
Multiplying the two together gives
L'C' = 4 = 1.6 x lO-11IK;
or Kf = 2 X 10-6
From the expression for L' we get
Km = 2K/1.6 X 10-3 = 2.5 X 10-3
hence
R' = KmR = 100Km = 0.25 Q
L' = KmLlKf = 1.6 X 1O-3 K m1Kf =2H

C' = CIKmKf = 0.01 X 1~1 KmKf =2F
12.6 Passive and active filters
Passive series- and parallel-resonant circuits can be used in both band-pass

and band-stop filters. A band-pass filter is a circuit which will transmit
frequencies within adesignated range known as the pass-band, and attenu-
ates other frequencies. A band-stop filter attenuates frequencies within the
stop-band, and allows other frequencies to be transmitted.
Both passive and active filters are widely used. A passive filter is one
using only R (or G), Land C elements.
12.6.1 Passive band-pass filter
The basic characteristic of a band-pass filter is shown in figure 12.11(a).

This characteristic can be achieved by the series circuit in figure 12. l1(b)
which, at resonance gives Vo = VI (that is if the resistance of the coil is
small compared with the value of R).
Resonance 319
~ R
V,
Pass-band
1.0
1v.
f, f,
(a) (b) (c)
Figure 12.11 (a) A band-pass filter characteristic, (b) and (c) simple passive circuits.
The parallel circuit in figure 12.11(c) gives a similar frequency response

since, at resonance, the dynamic resistance of the parallel circuit is much
higher than the resistance of R.
In fact, at resonance, both circuits have an output resistance of R Q
(approximately). The frequencies 11 and 12 are the cut-off Irequencies of
the filter.
12.6.2 Passive band-stop filter
The general characteristic of this type of filter is shown in figure 12.12(a).

The se ries circuit in figure 12.12(b) can achieve this characteristic since, at
resonance, the net voltage, Vo, across Land Cis zero (or nearly so); this,
of course, assurnes that the resistance of the coil is sm all compared with
that of R.
Similarly, the parallel circuit in figure 12.12(c) gives a similar character-
istic because, at resonance, the impedance of the parallel circuit is very
high (ideally infinite) when compared with R (which is low).
R, L
~
v,
v11lv. v'l
1.0
1/ Y2
o I 0
Iv.
f, f,
(a) (b) (c)
Figure 12.12 (a) Band-stop filter characteristic, (b) and (c) simple passive circuits.
12.6.3 Active RCfilters
These are a dass of filters utilising operational amplifiers as active elements

in conjunetion with passive resistors and eapacitors. They ean be used to
provide almost any filter eharaeteristie, that is, low-pass, high-pass, band-
pass, band-stop, ete. and are the subjeet of many specialised texts.
We will not dwell on their operation here, and it is sufficie,t to say that
we have already analysed a cireuit whieh is the basis of mal:y forms of
aetive filter. The reader should refer to worked example 11.7.2 and prob-
lem 11.11 for this type of eireuit.
12.7 Selective resonance
Aperiodie voltage whieh is non-sinusoidal ean be shown mathematieally to

eontain sinusoids of many frequencies (see ehapter 13 for details). It may
happen that a eireuit produees selective resonance with one of the frequen-
eies, whieh may result in a large value of voltage or eurrent at the seleeted
frequeney in the eireuit.
This is one reason why supply authorities diseourage eonsumers from
eonneeting apparatus whieh generate high values of harmonie voltage or
eurrent as, for example, may occur when high-power maehines are eon-
trolled by eertain types of thyristor-eontrolled equipment.
A voltage of 100(sin wt + sin 2wt + sin 4wt) is applied to aseries eireuit

eomprising a resistanee of 10 Q, a eapaeitor of 10 IJ.F, and a 0.625 H
induetanee. If w = 200 radis, ealculate the eurrent in the cireuit due to
eaeh harmonie frequeney.
Solution
For the 200 radis frequeney

Z:zoo = 10 + j(200 x 0.625 - 1/(200 x 10 x 1ü-6»
= 10 + j(125 - 5(0) == 10 - j375 = 375.1L-88S Q
Z_ = 10 + j(400 x 0.625 - 1/(400 x 10 x 1ü-6»
= 10 + j(250 - 250) = 10 - jO = lOLO° Q
Resonance 321
Z_ = 10 + j(800 x 0.625 - 1/(800 x 10 x 1~»

= 10 + j(500 - 125) = 10 + j375 = 375.1L88S Q
Hence the current for the 200 radis frequency component is
11.00 = 100/375.1L-88S = 0.267 L88S A
The current for the 400 radis frequency component is
1400 = 100/lOLO° = 10LO° A
and for the 800 radis frequency component is
1_ = 100/375.1L88S = 0.267 L-88S A
In this case, the circuit selectively resonates with the second harmonie
frequency component, and passes a much larger current at that frequency.
12.8 Tuned coupled circuits
Many communications circuits contain a tuned band-pass filter comprising

two parallel circuits, both tuned to the same frequency, in whieh the
inductive elements in the two circuits are magnetically coupled.
The general arrangement of the coupled circuit is as shown in figure
12.13(a), and a corresponding equivalent circuit is in diagram (b). The
individual resistance of each circuit is shown but, in practiee, the value of
R 1 and R 2 are very much lower than the circuit reactance at the operating
frequency. We can, to simplify calculations and without significant loss of
accuracy, ignore the two resistance values at this stage. In this case, the
mesh currents are
o = j(wL 1/wC)/
1 - J - jwM/2
o = - jwMI + j(wL2 -
J 1/wC2)/2
c, c,
(a) (b)
Figure 12.13 (a) Inductively coupled tuned circuits, (b) an equivalent circuit.
Substituting the value of 12 from the second equation into the first equation
gives
and since I. is not zero, then

(wL I - l!wCI )(wL2 - l!wC2) - w2M 2 = 0
Since each circuit has the same resonant frequency, it follows that
giving, after some manipulation

1 2 1
- - - + - (1 - k2 ) = 0
ww~ w~
where k = Mtv'(L IL 2), and is the magnetic coupling coefficient between

the two inductors. The above equation is a quadratic in (l!w), and solving
gives
w = w~lv'(l ± k)
That is to say, there are two resonant peaks in the response, one being at
frequency we/V(l + k) and the other at we/V(1 - k).
Critical coupling
When maximum power is transferred in figure 12.13(a) from the primary

winding to the secondary winding, we say that critical coupling has been
achieved between the two circuits. If the mutual inductance needed to give
this coupling is Me then, at this coupling, the effective impedance of the
secondary circuit at resonance is R 2 , and this impedance referred to the
primary winding is (woMe)2/R 2. Maximum power is therefore transferred to
the secondary winding when
that is when
Me = V(R I R 2 )/wO
If the magnetic coupling coefficient which gives critical coupling is k e , then
Me = k e V(L IL 2), hence
ke =MdV(LIL2) = V(~ ~)= l!V(QIQ2)

woLl woL2
Resonance 323
Unworked problems
12.1. Aseries RLC circuit contains R = 20 Q, L = 0.1 Hand C = 0.1

IJ.F. Calculate the resonant frequency, the Q-factor, the bandwidth
and the half-power frequencies.
[10 000 radis; 50; 200 radis; 9900.5 rads/s; 10 100.5 radis]
12.2. In problem 12.1, if the applied voltage is 100 mV, calculate the
current in the circuit and the voltage across C at (a) 0.8wo, (b)
1.3wo.
[(a) 2.22 x 10--4 L87S A; 0.28L-2S V; (b) 1.88 x 10--4 L-87.8°
A; 0.145L-177.8° V]
12.3. If, in the circuit problem in 12.1, the resistor is shunted by a
dependent current source of 0.02Vu where VL is the voltage across
the inductor, calculate the new value of resonant frequency. The
current in the dependent source ftows in the same direction as that
in the original circuit.
[12 910 radis]
12.4. Calculate the impedance at resonance of the circuit in problem
12.3, and determine the current in the circuit at 1.3 times the
re sonant frequency if the r.m.s. supply voltage is 0.1 V.
[(20 + jO) Q; 0.243L-87.2° mA]
12.5. A 5 V a.c. source with an output resistance of 2 Q is connected to a
se ries circuit which resonates at 1 MHz. What values of inductance
and capacitance are connected in the circuit if the resistance of the
inductor is 3 Q, and voltage across the capacitor at resonance is
100 V?
[15.9 IJ.H; 1.59 nF]
12.6. An ideal parallel circuit of the type in figure 12.8(c) contains a
resistance of 10 kQ in one branch, a 10 mH inductance in the
second branch, and a 1 IJ.F capacitance in the third branch. Calcu-
late (a) the re sonant frequency of the circuit, (b) its Q-factor at
resonance, (c) the bandwidth of the circuit and (d) the lower and
upper half-power frequencies.
[(a) 10 000 radis; (b) 100; (c) 100 radis; (d) 9950.1 radis, 10 050.1
radis]
12.7. Calculate the impedance of the circuit in problem 12.6 at (a) Wo,
(b) 0.8wo, (c) 1.2wo•
[(a) 10 kQ; (b) 222.2L88.7° Q; (c) 272.6L-88.4° Q]
12.8. A coil of 10 Q resistance and 0.2 H inductance is connected in
parallel with a capacitor of 100 IJ.F capacitance. Calculate the
resonant frequency of the circuit, and evaluate the current in each

branch at resonance if the supply voltage is 1 V. What current is
drawn from the supply?
[34.7 Hz; I coi' = 0.0224L-77.1° A; I e = 0.0218L90° A; 5 mA]
12.9. Convert the circuit in problem 12.8 into a GLC parallel circuit
containing the ideal elements of the kind in figure 12.5.
[5 mS; 0.211 H; 100 !lF]
12.10. A communications receiver circuit contains a 3-branch parallel cir-
cuit similar to that in figure 12.5. The circuit is tuned by a variable-
capacitance capacitor over a broadcast band from 0.5 MHz to 1.6
MHz. If the value of Qp does not exceed 40 and R is 25 kQ,
calculate the value of L, together with the maximum and minimum
value of C.
[198.9 !lH; 49.74 pF; 509.3 pF]
12.11. If, in the circuit in problem 12.6, adependent voltage source is
inserted in series with the capacitor so that the voltage genera ted
assists the original current through the capacitor, determine an
expression for the input admittance of the circuit if the magnitude
of the dependent source is 1OO01R , where IR is the current in the 10
kQ resistor. Calculate the re sonant frequency of the circuit.
[1Q-4 + 1.1 x 1Q-6 jw + 100/jw; 1517 Hz]
12.12. Scale the se ries RLC circuit in worked example 12.3.1 so that it
contains (a) a 1 H inductance and a 1 F capacitance, (b) it contains
a 1 nF capacitance and is resonant at 1 MHz.
[(a) 0.4 Q, 1 H, 1 F; (b) 4 kQ, 1 nF, 23.3 !lH]
12.13. In a band-pass filter of the type in figure 12.11(b), at a frequency of
20 kHz XL = 2.513 kQ, Xe = 397.9 Q and R = 100 Q; calculate
(a) the centre frequency Uo), (b) the half-power frequencies, (c)
the Q-factor at resonance and (d) the Q-factor of the circuit if a
resistance of 250 Q is connected between the output terminals.
[(a) 7958 Hz; (b) 8366 Hz, 7570 Hz; (c) 10; (d) 14]
12.14. In problem 12.13, calculate the two half-power frequencies and the
pass-band if a load resistance of 100 Q is connected between the
output terminals.
[7761 Hz, 8159 Hz; 398 Hz]
12.15. Two circuits tuned to a frequency of 1 kHz are magnetically cou-
pled as shown in figure 12.13(a). Each circuit has a resistance of 100
Q, and the Q-factor of the primary and secondary circuits are 50
and 100, respectively. Calculate (a) the critical coupling coefficient
and (b) the lower and upper re sonant frequencies.
[(a) 0.01414; (b) 0.993 kHz, 1.007 kHz]
13
Harmonics and Fourier
Analysis
13.1 Introduction
Thus far we have thought almost exclusively of alternating waveforms as

being pure sinusoids. However, there is a range of practical devices includ-
ing inductors, semiconductors, etc., which result either in the flow of
non-sinusoidal currents or in the production of non-sinusoidal voltages. It
is the analysis of these which interests us now.
The Fourier series, developed by Baron Jean Fourier, is aseries of terms
that represent a non-sinusoidal waveform. The simplest form is the trigo-
nometric series, in which the waveform is represented as the sum of a d.c.
term together with a large number (theoretically infinity) of pure sinusoids.
As a eonsequence of the principle of superposition, the effect of the
waveform on circuits can be analysed using standard techniques.
The exponential form of the Fourier se ries is more compact than the
trigonometrie form, but is less easy to understand!
13.2 Harmonics
Sinusoids are, by far, the most frequent form in which periodie waveforms
are met, that is, waveforms in which f(t) = f(T + t), where T is the
periodie time of the wave.
A waveform which is both periodic and non-sinusoidal is said to be a
complex wave. A complex wave may be shown to be built up from a
zero-frequeney term (or d.c. term) and aseries of sinusoids or harmonie
waves whose frequency is an integral multiple of the fundamental frequeney
(or first harmonic frequency). The fundamental frequency is the basic
waveform which establishes the general periodic time of the complex wave.
325
I
Complex waveform
/
Third harmonie
-I
Figure 13.1 A complex waveform which is the sum of the fundamental frequency and the
third harmonic.
Thus the second harmonie has a frequeney whieh is twice the fundamental
frequeney, the third harmonie frequeney is three times the fundamental, ete.
A complex wave whieh is the sum of a fundamental frequency and a third
harmonie frequeney is shown in figure 13.1. In this ease the third harmonie
is in phase with the fundamental, and has a lower amplitude. In general,
harmonies frequeneies are not in phase with the fundamental, and do not
have the same amplitude as the fundamental.
Harmonics are gene rally produeed in eleetrical eireuits by equipment
having a non-linear eharaeteristic. The simplest ease iS,for example, an
iron-eored eoil whose B-H eurve is non-liriear. A more eomplex ease is a
reetifier cireuit, whieh allows ftow of eurrent in one-half of the supply
voltage wave, but prevents it in the other half-eyde. Eleetronic deviees
sueh as thyristors and triaes give rise to even more eomplex waves.
The resulting harmonies produee problems in eleetrieal cireuits ranging
from meter-reading errors to seleetive resonanee.
13.3 Trigonometrie Fourier series
A periodic waveform for whieh f(t) = f(T + t), where T is the periodic
time of the wave, ean be represented by the trigonometrie Fourier series of
f(t) in the form
f(t) = ao + a 1 eos wt + a2 eos 2wt + .. .
+ b1 sin wt + b2 sin 2wt + .. .
or in the generalised form
L
~
f(t) = ao + (an eos nwt + bn sin nwt)

n=l
where w = 2rrJT radis, and nw is the nth harmonie frequeney. The eoef-
Harmonics and Fourier Analysis 327
ficients ao, an and bn are the trigonometric Fourier coefficients of f(t).

Alternatively, the se ries can be written in the form
L
~
f(t) = ao + Cn sin (nwt + CPn)

n=l
or
L
~
f(t) = ao + n cos (nwt + On)

C
n-l
where
cn = V(a~ + b~)
CPn = tan- 1 (an/b n)
On = tan- 1 (-bn/a n)
We are not concerned here with the proof of Fourier's theorem, but we
have a deep interest in evaluating the Fourier coefficients. In order to do
this we need to know the value of certain integrals, which are given below.
Firstly, the average value of a sinusoid over a complete cycle is zero.
[ sin wt dt = 0
[ coswtdt=O
and the value of the following definite integrals is zero
[ sin kwt cos nwt dt = 0
[ sin kwt sin nwt dt = 0
[ cos kwt sin nwt dt = 0
also
[ sin2 nwt dt = ;
[ cos2 nwt dt = ;
The reader should note that the integration period can be any range of T,
for example, the integration could be over the range 0 to T (as it is above),
or from - T/2 to T/2, or from - T/4 to 3T/4, etc.
We now turn our attention to a method of evaluating the Fourier

eoeffieients.
The value of ao
If we integrate the Fourier series over one eomplete eyde we get
[f(t) dt = [ao dt + [ ~1 (an eos nwt + bn sin nwt) dt

and, sinee every term in the seeond expression is either a sine or a eosine,
then
or
ao = ~ [f(t) dt
That is, the eoeffieient ao is the average value of f(t) over one complete
cycle. It is therefore the d.c. component of f(t).
The value of an
To evaluate this term, we multiply both si des of the Fourier se ries by

eos kwt, and then integrate both sides over one eyde, as follows
[f(t) eos kwt dt = [a o eos kwt dt +
[ n~l (an eos kwt eos nwt dt + bn eos kwt sin nwt dt)
Every term on the right-hand side of the equation is zero exeept for the
ease where k = n, when
[f(t) eos nwt dt = ; an

or
an = ~ [f(t) eos nwt dt

That is
an = twiee the average value of (f(t) eos nwt) over one eyde
The value 01 bn
The proeedure for evaluating bn is generally similar to that for an> with the
exeeption that we multiply both si des of the Fourier series by sin kwt
before integration. The final result is
bn = ~ [/(t) sin nwt dt

or
bn = twiee the average value of (f(t) sin nwt) over one eyde
Using these relationships, we will take a look at the Fourier analysis of a
reet angular wave and a reetified sinewave.
Determine the Fourier series for the reetangular wave in figure 13.2.
v(t)
0
ri2 i
- -1 -
Solution
In this ease the graph is plotted to a base of time, and the integral will be
evaluated over a range of T seeonds. The equation of the wave is
V rn for - T/4 < t < T/4
{
v(t) = -Vrn for T/4< t< 3T/4
By observation, the area above the axis in eaeh eyde is equal to the area
below it, henee
The value of an can be calculated as folIows. Since wT = 2n, then

w = 2n/T. Also, since the waveform repeats itself between T/2 and T, we
shall analyse the wave from 0 to T/2 below.
2 fT'2
an = - J(t) cos nwt dt
T/2 0
= i{T fmV
0
m cos 2nnt/T dt +
fm - V
T/4
m cos 2nnt/T dt}
4Vm f T'4 2nnt fT'2 2nnt

= --{ cos--dt - COS--dt}
ToT TI4 T
TI4
= 4V m { [~sin 2nnt] T . 2nnt ]TI2 }
- [ --sm--
T 2nn T 0 2nn T TI4
= 4Vm . nn/2
--sm
nn
If n is even, an = O. For odd values of n, an is finite. The value of b n can be
calculated as folIows.
2 fT'2
bn = - J(t) sin nwt dt
T/2 0
4 JT/4 JTI2
= -{ V m sin (2nJrt/T) d( wt) + - V m sin (2nJrt/T) dt }
T o m
= 0 for all n
That is, there are no sine terms in the series. Hence, the Fourier series is of
the form
4Vm(
v (t) = -
n
1
- cos wt - -cos
3
3wt 1
+ -cos
5
5wt - ... )
Express the rectified current waveform in figure 13.3 as a Fourier series.

Solution
The waveform can be expressed over the range O~t~O.l as

i(t) = {Im sin wt for 0 ~ t ~ 0.05
o forO.05~t~0.1
r-----t------4~---_\. __ t (5)
o 0.05 0.1 0.15
1-- 1 cycle .. I
Since T = 0.1 s, then! = 1rr = 10 Hz and w = 20re rad/so

The d.c. component or zero-frequency component is
ao = -
1 fT !(t) dt
T °
{JO.05
1 Jo.t}
= - Im sin 20ret dt + 0 dt
0.1 ° 0.05
= Im/re
The reader should note that the integration is broken down into intervals
for which the functional form of i(t) is known.
The value of an can be calculated as folIows.
an = -
2 fT!(t) cos nwt dt
T °
= -2 {f0 005 Im sin 20ret cos 20rent dt + foot 0 dt }

0.1 ° 0005
= 101m f:0 5(sin (n + 1)20ret - sin (n - 1)20ret) dt)
We break the calculation at this point to note that

05
at = 101 m
J°
o o
sin (2 X 20ret) dt) =0

The calculation is continued below to include other values of n
an = m[
~
20re
- -1- cos (n
n+l
1
+ 1)20ret + --cos (n - 1)20ret ]~
n-1 °
101 [ - -1( 1 - cos (n + 1):rr) - - -1( 1 - cos (n - 1):rr) ]

=~
20:rr n+1 n-1
Points to note he re are that if n is odd (and =1= 1), then (n ± 1) is even, so
that an = 0; that is a3 = a5 = a7 = ... = O. Ifn is even then (n ± 1) is odd
and
a = m lOI [_2 ___2_]
n 20:rr n+1 n-1 (n + l)(n - 1):rr
The coefficient bn is calculated as follows
bn = -
2 fT f(t) sin nwt dt
T °
~ {f
oo05 fool }
= Im sin 20:rrt sin 20:rrnt dt + 0 dt
0.1 ° 0005
= 201m f°
Oo05
sin 20:rrt sin 20:rrnt dt

Ifn=l
b = 201
1 m
f° o o05
sin 2 20:rrt dt =
I
-2!!.
2
If n =1= 1, the calculation proceeds as follows
bn = 20Im [05 ~(cos(n - 1)20:rrt - cos (n + 1)20:rrt) dt

=0
Hence
i(t) = Im ( 1 + ~ sin 20:rrt - 3. cos 40:rrt

:rr 2 3
- -2 cos 80:rrt - -2 cos 120:rrt )

15 35
13.4 Waveform symmetry
Many waveforms exhibit symmetry either about a particular point or a

particular axis. A knowledge of the type of symmetry involved may tell us
that certain coefficients in the Fourier series may be absent, allowing us (in
certain cases) to simplify the calculations involved. If more than one form
of symmetry exists, there will be more than one factor missing from the
series.
When investigating the symmetry of a waveform, and if a d.c. compo-
nent exists, the reader will find it useful to visualise the wave without its
d.c. component.
The two most readily recognisable forms of symmetry are evenjunction
symmetry (or even symmetry) and odd-function symmetry (or odd sym-
metry).
Even symmetry
An even function is defined as one for which f(t) = f( - t), that is, it is
symmetrical about the y-axis, as shown in figure 13.4(a). The Fourier series
for the function contains only cosine terms (no sine terms exist): ao may
exist, an exists and b n = 0. The series is of the form
f(t) = a o + a t cos wt + a2 cos 2wt + ...
Odd symmetry
An odd function is one in which f(t) = - f( - t), that is it is symmetrical

about the origin (see figure 13.4(b». TheFourier series for this function
contains only sine terms; to summarise, a o = 0, an = 0, b n exists, and the
Fourier series is of the form
f(t) = b t sin wt + b2 sin 2wt + ...
Half-wave repetition
This is a wave in which f(t) = f(t + T/2), as shown in figure 13.4(c). The
Fourier se ries for this type of repetition contains only even terms; to
summarise, ao may exist, and only even terms exist in an and b n as folIows.
If the series has the general form
f(t) = Co + Ct sin( wt + <Pt) + C2 sin(2wt + <P2) + C3 sin(3wt + <P3) + ...
then
f(t + T/2) = Co - C t sin( wt + <Pt) + C2 sin(2wt + <P2) - C3 sin(3wt + <P3)
+ ...
Hence if f(t) = f(t + T/2), then the Fourier series for f(t) contains only
even harmonics.
o TI2 T Time
t- -t "'
(a)
Time
(b)
y y
.--
I f(t) ~ T/2
Time
f(t + T/2)
•I
T Time
t I TI2
(c) (d)
Figure 13.4 Waveform symmetry.
Half-wave inversion
In this type of wave,f(t) = - f(t + T/2), as shown in figure 13.4(d). The

Fourier series contains only odd terms; to summarise, ao = 0 and only odd
terms exist in an and bn .
If the series has the general form described in the Half-wave repetition
section, it is clear that if f(t) = - f(t + T/2), neither the 'd.c.' component
nor even terms exist (see also worked example 13.4.1).
Identify the symmetry of the waveforms in figure 13.5.

v(t) i(t)
-...,f---------'l:-------;t-t- wt
(a)
o T/2 T
(b)
Figure 13.5 Figure [ar worked example 13.4.1.
Solution
The waveform in figure 13.5(a) eorresponds to the idealised magnetie flux

distribution in the air gap of an alternator. The waveform has odd symmet-
ry and half-wave inversion. Consequently ao = 0, an = 0 and beyen = O.
The wave in figure 13.5(b) has a 'd.e.' term whieh, by observation, has
unity value, that is, ao = 1. If we visualise the 'a.c.' component without the
'd.c.' term, we see that the wave is an even function and has half-wave
inversion. That is b n = 0 and a even = O.
13.5 Line spectrum
A popular method of presenting the results of Fourier analysis is in the

form of a line spectrum of the wave. In any given problem, there may be
either one or two line speetrum diagrams. One diagram shows the ampli-
tude of each frequency component in the form of a vertical line drawn at
the corresponding frequency, the length of the line indicating the magni-
tude of the harmonic (this is the most usual form of line spectrum).
In addition, there may be a phase speetrum, the length of each vertical
line representing the phase angle of the partieular harmonie component
(see worked example 13.5.1).
For a partieular harmonie frequency nwt, we can combine the si ne and
cosine terms in the Fourier series as follows
an cos nwt + bn sin nwt = V(a~ + b~) sin (nwt + tan- 1 (a n/b n »
or
an cos nwt + bn sin nwt = V(a~ + b~) cos (nwt + tan- I ( - bn/an»
Using the above relationships, the magnitude and phase spectra can be
drawn. Depending on whether the sine or cosine version is chosen, the
phase shift between the two will differ by 90°.
If we replaee the eosine and sine terms in the Fourier series by their
exponential equivalent, we get the following simplified series (showing
onlyao and the nth harmonie).
f(t) = ao + ... + an eos nwt + ... + bn sin nwt + ...
+ ... + ~j (e jnwt _ e- jnwt) + ...
+ . .. + -1 (an - J·bn) e jnwt + ...

2
+ -1 (an - J·bn) e jnwt + ...

2
This series indieates that there is yet another version of the line speetrum,
with speetrallines at frequeneies +nw and - nw, e'aeh of these lines being
one-half the length of the original amplitude speetrum. The length of the
d.e. term (a o) is unehanged. If, for example, the trigonometrie Fourier
se ries of a wave is as shown in figure 13.6(a), the exponential se ries
indieates that the alternative speetrum in figure 13.6(b) is also true for the
wave.
Amplitude
10 10
8
6
3 3
I---+---,----t-<- w . .~--,---~--~---+---.--~--~+w
o 2 3 -2 -1 o 2 3
(a) (b)
Figure 13.6 Amplitude speetrum 01 a eomplex wave using (a) the trigonometrie Fourier
series, (b) the exponential Fourier series.
Draw the phase and magnitude speetra for a waveform with the following
Fourier analysis
v(t) = 10 - 3 eos t + 2 eos 2t - 5 eos 3t
+ 4 sin t - 6 sin 2t + 2 sin 3t - 12 sin 4t
Solution
The line spectrum values can be calculated as follows; it is assumed that

each term in the series can be expressed in the form c,. cos(ncot + tP,.).
d. c. component
Amplitude = 10
co = 1 radis component
Amplitude = Cl = V(ai + bi) = V(3 2 + 42) = 5
Phase shift = tPl = tan-l(-b/a l) = tan- l(-4/-3) = -127°
Amplitude = C2 = V(ai + bi) = V(22 + 62 ) = 6.3
Phase shift = tP2 = tan- l(-b/a2 ) = tan- l(6/2) = 71S
Amplitude = C3 = V(a; + b;) = V(5 2 + 22 ) = 5.4
Phase shift = tP3 = tan- l(-bja3 ) = tan- l(-2/-5) = -158.2°
Amplitude = C4 = V(a; + b;) = V(Q2 + 122 ) = 12
Phase shift = tP4 = tan- l(-bia4 ) = tan- l(12/0) = 90°
The Fourier series can therefore be expressed in the form
v(t) = 10 + 5 cos(cot - 127°) + 6.3 cos(2cot + 71S)
+ 5.4 cos(3cot - 159.2°) + 12 cos(4cot + 90°)
and the corresponding line spectra are shown in figure 13.7.
12
10
6.3
5 5.4
~
Q)
"tJ :.c., 3
'ae: I---+---+---+----t- w ~--+--+---+---+ __ w
:6'0 2 3 4 ~ 0
Q..
2 4
~
-158.2°

13.6 Circuit response to a non-sinusoidal forcing function
A eomplex foreing voltage ean be thought of as a direet voltage in se ries

with an infinite number of eosine and sine voltages, as shown in figure 13.8.
We ean eonsider the response of a linear network to this voltage by
ealculating the eurrent due to eaeh independent harmonie of the eomplex
voltage. By superposition, the resulting eurrent is the sum of the individual
terms, as shown in worked example 13.6.1.
a, cos wt'
a 2 cos 2wt'
I
I
I
b,Sinwtt$
b 2 sin 2wt'~
Figure 13.8 Circuit response to a non-sinusoidal forcing function.
Determine an expression for the eomplex eurrent, i(t), in the cireuit in

figure 13.9(a), when it is energised by a reetified sinewave of 100 V peak
amplitude (see figure 13.9(b». The fundamental frequeney of the wave is
1000 radis.
Also determine the most signifieant terms in the Fourier series for the
voltage vR(t) and vL(t) for the voltage aeross the resistanee and induetanee,
respeetively.
Solution
Firstly we need the trigonometrie Fourier expression for the eomplex

wave; fortunately we have already dedueed the first few terms in the series
for half-wave reetified eurrent in worked example 13.3.2, and the eorres-
ponding series for a voltage wave is
b
0.01 H
lB'oov
i(t) 100 v(t)
~" ot------'1T'----2.J..P-
. .W-t
.
100
(a) (b) ----

31.53 V 15L 90° V
2000 radis
(c)
v(t) = V m( 1 + ~ sin wt - ~ eos 2wt

:n: 2 3
2
- -eos 4wt - -2 eos 6wt - ... )
15 35
Sinee the terms in the series diminish rapidly as the order of the harmonie
increases, we shall eonsider only terms up to and including the seeond
harmonie. Converting the seeond harmonie cosine term into a sine term,
the expression for v(t) is
v(t) = 31.83 + 50 sin wt - 21.22 sin (2wt + 90°) V
When the maximum a.c. values are converted into r.m.s. values, the eircuit
diagram for the first three terms in the series is as shown in figure 13.9(e).
We now consider each component in turn.
d.c. term
The current, 10 , produced by this term is

10 = aJR = 31.83/10 = 3.183 A
The p.d. across the resistor produced by this voltage is
VRO = RIo = 10/0 = 31.83 V.
and across the inductor (which has no resistance) is zero.
Fundamental frequency term
The r.m.s. voltage is 50/v'2 = 35.36 V, hence VI = 35.36 LO° V. The

cireuit impedance at the fundamental frequency of 1000 radis is
ZI = R + jw.L = 10 + j(l000 x 0.01) = 10 + jlO
= 14.14L45° Q
hence
11 = V';ZI = 35.36Loo/14.14L45° = 2.5L-45° A
and 11max = 3.54 A. Hence
VR1 = RI =
1 10 X 2.5L-45° = 25L-45° V
and
VLl = jX.Il = j10 X 2.5L-45° = 25L45° V
Seeond harmonie term
The r .m.s value of this component is 21.22/V2 = 15 V and, since cp = 90°,

V2 = - 15 L 90° = 15 L - 90° V. The circuit impedance at a frequency of
2000 radis is
Z2 =R+ jw 2 L = 10 + j(2oo0 X 0.01) = 10 + j20
= 22.36L63.4° Q
that is
or 12max = 0.95 A. The voltage across the circuit elements is

VR2 = RI2 = 10 x 0.67 L-153.4° = 6.7 L-153.4° V
and
VL2 = jXJ2 = j20 x 0.67 L-153.4° = 13.4L-63.4° V
Hence, the first three terms in the Fourier series for i(t) are
i(t) = 3.183 + 3.54 sin (lOoot - 45°) + 0.95 sin (20oot - 153.4°) A
The first three terms in the Fourier series for vR(t) are
vR(t) = 31.83 + 35.36 sin (loo0t - 45°) + 9.48 sin (2oo0t - 153.4°) V
and the terms in the Fourier series for vL(t) are
vL(t) = 35.36 sin (lOoot + 45°) + 18.95 sin (2oo0t - 63.4°) V
At this point the reader will find it interesting to verify that v(t) = vR(t)
+ vL(t).
13.7 T.M.S. value of a complex wave and the power supplied
The r.m.s. value of the complex wave

y(t) = ao + (al eos wt + ... + an eos nwt + ... )

+ (bi sin wt + ... + bn sin nwt + ... )
is given by
Yrms = v' (a~ + t(ai + . . . + a~ + . . .) + t( bi + . . . + b~ + . . .))
or, if An is the r.m.s. value of an, and B n is the r.m.s. value of bn, then
Yrms = v'(A~ + (Ai + ... + A~ + ... ) + (Bi + ... + B~ + ... ))
The reader may like to verify that the r.m.s. voltage aeross the resistor in
worked example 13.6.1 is 41 V, and the r.m.s. voltage aeross the induetor
is 28.37 V. Aeeordingly, the r.m.s. supply voltage is
Vs = v'(4F + 28.372 ) = 49.86 V
whieh is in elose agreement with the r.m.s. voltage of 49.88 V ealculated in
worked example 13.7.1.
If v(t) and i(t) are the respeetive voltage across and the eurrent in the
eireuit, then the average power supplied by the nth harmonie is
1 f2Jt
= - VnIn sin nwt sin(nwt - cfJn) d(wt)
2Jt 0
= VnIn eos cfJn

where cfJn is the phase angle between V n and In.
The average power supplied by a voltage of one frequeney (say nwt) and
a eurrent of another frequeney (say mwt) is zero. Henee the total power
supplied by a eomplex voltage (or eurrent) is
PT = Po + PI + ... + Pn + .. .
= VJo + VJI eos cfJI + ... + VnIn eos cfJn + ...
00
= VJo + ~ VnIn eos cfJn

n=l
where V n and In are r.m.s. values.

The power Jactor of a eireuit to which a eomplex wave is applied is
defined as
power eonsumed
power faetor = --........!---------
r.m.s. voltage X r.m.s. eurrent
Calculate, for worked example 13.6.1, (a) the r.m.s. current in the load,
(b) the r.m.s. voltage across the load, (c) the power supplied by each
component of the complex wave, (d) the total power supplied and (e) the
power factor.
Solution
(a) The r.m.s. value of the current is
I = V(I~ + I;m + I;m )

= V (3.183 2 + 3.~42 + 0.~52) = 4.1 A
(b) The r.m.s. value of the voltage across the load is
V = V(V~ + V~m + V~m)

=
..J( 2
V 31.832 + 250 + -21.22
2-
2)
= 49.88 V
(c) We can calculate the power consumed by each component of the

complex wave as follows.
d.e. eomponent
Po = Volo = 31.83 X 3.183 = 101.3 W
or, alternatively Po = I~R = 3.1832 X 10 = 101.3 W
or Po = V~(R/R = 31.8)2/10 = 101.3 W
Fundamental [requency
PI = Re(Vll~) = Re(35.36LO° X 2.5L-45°)
= Re(62.5 - j62.5) = 62.5 W
or PI = IiR = 2.5 2 X 10 = 62.5 W
or PI = V'i(R/R = 25 2/10 = 62.5 W
Second harmonie
P2 = Re(VJ~) = Re(15L-90° X 0.67 L153S) = 4.5 W
or P2 = PzR = 0.672 X 10 = 4.49 W
or P2 = V;(R/R = 6.7 2/10 = 4.49 W
(d) Total Power
PT = Po + PI + Pz = 168.3 W
or PT = PR = 4.F x 10 = 168.1 W
(e) The power faetor of the eireuit ean be ealculated from
power faetor = power/(r.m.s. volt-amperes)
= 168.3/(49.88 x 4.1) = 0.823
13.8 EtJect of harmonics in a.c. systems
Sinee resistanee is independent of frequeney, the voltage and eurrent will,

for eaeh individual harmonie, be in phase with one another in a resistor.
In the ease of an induetor, the reaetanee to the nth harmonie is n times
greater than it is to the fundamental; eonsequently, the eurrent produeed
(for eaeh volt) by that harmonie is n times sm aller than for the fun-
damental.
For a eapaeitor, the reaetanee to the nth harmonie is n times sm aller
than it is to the fundamental; eonsequently, the eurrent produeed (for eaeh
volt) by that harmonie is n times greater than for the fundamental.
Consequently, if measurements of impedanee are made in a eireuit
eontaining reaetive elements using a eomplex supply frequeney, large
errors ean arise unless allowanee is made for the individual harmonies.
Harmonies ean also be produeed in 3-phase systems. It ean be shown
that it is unusual for even harmonies to be present in these systems, but
multiples of the third harmonie (ealled triple-n harmonics) ean produee
serious problems. In some eases, odd harmonies have a positive phase
sequence (see also chapter 7), and others have a negative phase sequence.
13.9 Harmonic analysis
If we are provided with numerical data for a waveform (as we may from an
oseillogram) then, using the theory in section 13.3, we can ealculate the
Fourier eoeffieients of the wave. The eoeffieients are ealculated from ordin-
ates taken at fixed intervals along the wave.
However, the reader should note that this method ean, on oeeasions,
give misleading results; for example, if the spacing of the ordinates is sueh
that they are placed at the nodes of the nth harmonie, then the nth
harmonie will apparently have zero value! Moreover, errors will arise if
one tries to extend the theory to predieting high-order harmonies.
A eurrent waveform has the values in table 13.1 in its first half-eyde; they
are repeated with a negative sign in the seeond half-eyde.
Calculate the maximum value of the fundamental frequeney, together
with the seeond and the third harmonies, and determine the phase angle
between the fundamental and the third harmonie.
Solution
In this type of problem, the solution ean be obtained in tabular form (see
table 13.2) and, in so doing, the risk or error is redueed.
Table 13.1
(f' 0 20 40 60 80 100 120 140 160
i (A) -17.3 16.9 64.3 103.9 115.8 94.5 69.3 47 30.2
Table 13.2
(f' f(t) f(t)cos 0 f(t)sin 0 f(t)cos 20 f(t)sin 20 f(t)cos 30 f(t)sin 30
0 -17.3 -17.3 0 -17.3 0 -17.3 0
20 16.9 15.9 5.8 12.9 10.9 8.5 14.6
40 64.3 49.3 41.3 11.2 63.3 -32.15 55.7
60 103.9 52.0 90.0 -52.0 90.0 -103.9 0
80 115.8 20.1 114.0 -108.8 39.6 -57.9 -100.3
100 94.5 -16.4 93.1 -88.8 -32.3 47.25 -81.8
120 69.3 -34.7 60.0 -34.7 -60.0 69.3 0
140 47.0 -36.0 30.2 8.2 -46.3 23.5 40.7
160 30.2 -28.4 10.3 23.1 -19.4 -15.1 26.2
Sum 524.6 4.5 444.7 -246.2 45.8 -77.8 -44.9
The ealculations for the first half-eyde are shown in table 13.2, and for the
seeond half-eyde the eorresponding total values are
f(t) f(t) eos 8 f(t) sin 8 f(t) eos 28 f(t) sin 28 f(t) eos 38 f(t) sin 38
-524.6 4.5 444.7 246.2 -45.8 -77.8 -44.9
The totals for the eomplete eyde are
f(t) f(t) eos 8 f(t) sin 8 f(t) eos 28 f(t) sin 28 f(t) eos 38 f(t) sin 38
o 9 889.4 0 0 -155.6 - 89.9
Thus ao = mean value of l:.f(t) = 0118 = 0
and an = 2 X mean value of ~f(t) cos n8

bn = 2 X mean value of ~f(t) sin n8
hence
a1 = 2 X 9/18 = 1
a2 2= X 0/18 = 0
a3 =2 X (-155.6)/18 = -17.3
b1 =2 X 889.4/18 = 98.8
b2 = 2 X 0/18 = 0
b3 = 2 X ( - 89.8)/18 = - 9.98
that is, the Fourier series for the wave is

f(t) = cos 8 + 98.8 sin 8 - 17.3 cos 38 - 9.98 sin 38
= 98.8 sin (8 + 0.6°) + 19.97 sin (38 - 120°)
The reader should note that the - 120° phase angle of the third harmonie
corresponds to a phase angle of - 40° at the fundamental frequency, so that
the phase angle between the two sinusoids is 40.6°.
Unworked problems
13.1. A waveform has the following Fourier analysis:

f(t) = 10 + 6 sin (20Jtt + 40°) + 8 sin (40Jtt + 70°) +
2 sin (60Jtt + 90°)
Calculate (a) the mean value off(t) , (b) the r.m.s. value off(t), (c)
the periodic time of the second harmonie and (d) the value of f( t) at
t = 0.1 s.
[(a) 10; (b) 12.33; (c) 0.05 s; (d) 23.38]
13.2. For the waveform in figure 13.10, evaluate (a) ao, (b) at> (c) b2 •
[(a) -1.6; (b) - 0.75; (c) 0.88]
fit)
2
-+--r--+--r-......--+---r-~_ t (s)
L:J
2 4
12
-6
Figure 13.10
o~t - -rr+----:::-2rr--+-- wt
Figure 13.11
Amplitude
1234567
Figure 13.12
,!k. /'
I wt
-rr 0 rr 2rr
Figure 13.13
13.3. Evaluate the trigonometrie Fourier series for the square wave in
figure 13.11 and (b) draw the line speetrum for the wave.
[(a) /(t) = 4~m (sin wt + ~ sin 3wt + ~ sin 5wt + .. .);

(b) see figure 13.12]
13.4. Determine the trigonometrie Fourier series for the tri angular wave-
form in figure 13.13.
[ Vm
2
(1 +.!Jt2 eos wt + _8_
(3Jt) 2
eos 3wt + _8_eos 5wt + ... )
(5Jt) 2
I
13.5. A voltage waveform is defined by the following
:rt :rt
v(t) = 200 eos rot for - - ~ rot ~-
2 2
:rt 3:rt
v(t) = 0 for -~ rot~-
2 2
Deduee the Fourier series for the wave.
I200 (
-
:rt
:rt 2 2
1 +-eos rot + - eos 2rot - - eos 4rot
2 3 15
+ :SCOS6wt- ... l]
13.6. If the periodie frequeney of the wave in problem 13.5 is 300 rad/s,
and is applied to aseries eireuit eontaining a 1 kQ resistanee and a
0.8 IlF eapaeitanee, determine the Fourier se ries up to the 6th
harmonie for the eurrent in the cireuit.
[23.3 eos (rot + 76S) + 18.4 eos (2rot + 64.4°) +
5.9 eos (4rot + 46.2°) + 3 eos (6rot + 34.8°) mAl
13.7. For problem 13.6, ealculate (a) the r.m.s. value of the voltage and
the eurrent, (b) the power supplied by eaeh harmonie, (e) the total
power supplied and (d) the power faetor .
[(a) 99.84 V, 21.5 mA; (b) d.e. power == 0, fundamental power =
0.271 W, 2nd harmonie power = 0.169 W, 4th harmonie power =
17 mW, 6th harmonie power = 4.5 mW; (e) 0.462 W; (d) 0.215]
13.8. Write down the Fourier series for eaeh of the following pe rio die
funetions, and determine the periodie time of the wave in eaeh
ease: (a) 6 sin 2 20t, (b) 5 eos 2 lOt,
(e) 5 + 4 sin 100t + 8 sin 2 100t + 10 eos 2 100t.
[(a) 3(1 - eos 40t), 0.05:rt s; (b) 2.5(1 + eos 20t), 0.1:rt s;
(e) 14 + 4 sin 100t + eos 200t, 0.02:rt s]
13.9. The trigonometrie Fourier se ries for aperiodie wave is

f(t) = 5 - 10 eos 4t + 5 eos 12t - 9 eos 20t + 4 sin 4t
+ 15 sin 12t - 4.5.sin 20t + 8 sin 28t
Produee a table showing the magnitude and phase angle for eaeh
term in the se ries assuming that the expression for the nth harmon-
ie is C n sin (nrot + cf>n). Calculate the r.m.s. value of the wave.
[The table is
Order Modulus <p
0 5 0°
1 10.77 -68.2°
3 15.81 18.4°
5 10.06 -116.6°
7 8 0°
; 17.05]
13.10. A Fourier series expressed in the form
00
J(t) = a o + L C n eos (nwt + <p)

n =1
has magnitude and phase angle speetrum for the first five terms as
folIows.
Order of harmonie o 1 3 5 7
Magnitude 10 8 10 4 3
<po o 0 -45 135 90
Deduce the Fourier se ries for the wave.
[10 + 8 eos wt + 7.07 eos 3wt - 2.83 eos 5wt + 7.07 sin 3wt
- 2.83 sin 5wt - 3 sin 7wt]
13.11. The following values were obtained from measurements on a
voltage wave.
cJ>0 0 30 60 90 120 150 180 210 240 270 300 330
voltage (V) -134-444 -326 -106 -94-144 -66 144 126 306 494 344
Calculate the value of the parameters of the Fourier series up to
and inciuding the third harmonie.
[- 100 eos 2wt - 34 eos 3wt - 300 sin wt - 173 sin 2wt
- 94 sin 3wt]
14
Computer Solution of
Electric Circuits
14.1 Introduction
A range of software packages is available for the analysis of circuits using

mainframe, mini and personal computers. The attention of the reader is
directed here to one of the most popular of them, namely the Simulation
Program with Integrated Circuit Emphasis or SPICE, developed at the
University of California, Berkeley, which is widely available.
With the space available, only an introduction can be given here; for
more details, the reader should study a copy of the User Guide for the
version of SPICE employed in his computer system.
SPICE enables circuit analysis to be performed on d.c; and a.c. circuits
(or a combination of the two) which include elements such as resistors,
inductors, coupled circuits, capacitors, semiconductor devices, dependent
and independent sources, transmission lines, etc. SPICE can also be used
to perform frequency response analysis, Fourier analysis, transient analy-
sis, and can handle problems involving electronic noise, temperature
effects, etc.
The reader should note that some of these features may not be available
in some versions of SPICE for personal computers.
The circuit to be analysed is described using a text edi~or, which allows
the user to create and modify the input file (one program line per circuit
component). The first line of the file normally contains a title statement,
and the file is terminated by a '.END' statement. Such element as voltage
sources, current sources, resistors, inductors, transformers, capacitors,
diodes, transistors, etc., can be included in the input file.
Next, the input file is read by analyser, which checks the file for errors; if
none are present, it performs the required analysis and supplies the results.
On completion, control is returned to the editor.
349
Depending on the type of analysis required, the results may either be in

numerical form, or graphical form, or both.
Abrief introduction to SPICE circuit elements is given in section 14.3,
and greater detail can be obtained from section 14.4, where programming
examples are described.
SPICE provides a faster solution to circuit problems than can be
obtained by hand or calculator, and is less error-prone than conventional
methods. However, the reader is reminded that computer solution is not
an aid to learning! Computer solution comes into its own when the solution
of the problem is either tedious and/or complex, or when a range of
solutions is needed for different circuit parameter values.
All the PLOTs or graphs in this chapter are typical 'text' type displays
produced by a printer. Most versions of SPICE have a graphics post-
processor option, which provides a very high quality output on a printer or
plotter.
We will, at this stage, look at a very simple circuit comprising two
series-connected resistors, R1 and R2, supplied by a battery VI. An input
file describing the circuit may be as follows.
Simple Circuit
V1 1 0 DC 10; 10 V d.c. source
R1 1 2 5', R1 connected between nodes 1 and 2
R2 2 0 15; R2 connected between no des 2 and 0
.END this line MUST be included
Each input file MUST either commence with a 'title line' ('Simple Circuit'
in this case), or a blank line; the purpose of this li ne is to enable the
function of the file to be described in one line. This is followed by aseries
of 'element' lines, which may be written in any order. In this case we have
chosen the order V1, R1, R2; we could have chosen the order R2, R1, V1,
or R1, V1, R1, etc. The finalline must be a '.END' line; the '.' at the
beginning of the line teIls SPICE that this line contains a control function.
Other controllines are introduced in sections 14.3 and 14.4 and, apart from
'.END', controllines may be included at any stage in the file (but before
'.END').
In-line comments can be added following a ';' at the end of each line. A
complete line of comments can be included in the file by commencing the
'comment' line with a ,*, (see worked example 14.1).
Computer Solution 0/ Electric Circuits 351
SPIeE symbolic values

Value SPIeE symbolic form Exponent form
10- 15
F lE-15
10- 12 P lE-12
10- 9 N lE-9
10- 6 U lE-ti
10- 3 M lE-3
10 3 K lE3
10 6 MEG lE6
10 9 G lE9
10 12 T lE12
14.2 Circuit description
Each terminal of each element in the circuit being analysed is connected to

anode, the nodes being numbered in the range 0, 1,2,3 ... n; the value
of n is only limited by the version of SPICE being used. Node 0 is the
reference (or ground) node. A two-terminal element, such as a resistor, an
inductor, a capacitor, etc., is connected between two nodes, and a three-
terminal element such as a transistor is connected between three nodes.
The value of a component, such as the resistance of a resistor, can be
specified either by entering its actual value (usually a maximum seven
figures), or by expressing the value in exponent form, such as 1.5E-6, or by
writing it in symbolic form as shown in the table above. Thus, a resistance
of a 1200 ohm resistor may be specified either as 1200, or as 1.2E3, or as
1.2K.
14.3 Element description
In the following, the SPICE specifications of a number of the more

important circuit elements are described.
Resistors
The general specification is of the form

RXXXXXXX NI N2 VALUE
in which RXXXXXXX is the designation of the resistor in the circuit being
analysed; the XXXXXXX part of the designation is an alphanumeric string
from one to seven characters. For example, this could be R2, R23, or
R3AC4XY, etc. NI and N2 are the nodes between which the resistor is
connected, and V ALUE is the resistance of the resistor in ohms. A typical

specification may be
R2 1 2 3.5
The reader should note that SPICE does not accept conductance values,
and that a resistance of zero ohms is not allowed.
Capacitors
These are specified in the form

CXXXXXXX N+ N- VALUE (IC = ICOND)
where N + is the positive node, N - is the negative node, and VALUE is
the capacitance in farads. The ( ) brackets indicate that the contents are
optional and, where present, give the initial charge on the capacitor
between N + and N - in volts.
Examples include
C3 1 0 O.5E-6
C29 4 5 O.7U IC = - 2.3
In the case of C29, node 4 is charged to -2.3 V with respect to node 5.
Inductors
The general specification is

LXXXXXXX N+ N- VALUE (IC = ICOND)
where N + is the positive node, N - is the negative node, and VALUE is
the inductance in henrys. Where present, the initial current (in A) flows
inside the inductor from N + and N - .
Examples include
L4 4 5 5.6M
L6 2 3 4 IC = 1.5
In the case of L6, an initial current of 1.5 A flows from node 2 to node 3
through the inductor.
Where an inductor is part of a coupled circuit (see the following
section), node N+ is the end marked by a 'dot' using the dot notation (see
chapter 9).
Coupled circuits
A coupled circuit is presented as separate inductors with a specified

magnetic coupling coefficient between the windings in the form
KXXXXXXX LYYYYYYY LZZZZZZZ VALUE

where XXXXXXX, YYYYYYY and ZZZZZZZ are alphanumeric strings
from one to seven characters in length. KXXXXXXX is the designation of
the coupled circuit, such as K8, while LYYYYYYY and LZZZZZZZ are
the designation of the windings of the coupled circuit; the designation of
each of the windings is given separately. If the transformer has two
windings of self-inductance LI and L 2 , and a mutual inductance M, the
coupling coefficient between the windings is k = Mtv'(L I L 2 ) , where 0 < k
~ 1; the value of the coupling coefficient is given in the V ALUE section of
the transformer specification. A complete specification of a two-winding
coupled circuit may be
K3 L1 L2 0.65
L1 4 0 2
L2 5 0 4
This represents a transformer with a coupling coefficient of 0.65, having
a coil LI (of inductance 2 H) connected between no des 4 and 0, and coil L 2
(of inductance 4 H) connected between nodes 5 and O. The 'dotted' end of
coil LI is connected to node 4, and the 'dotted' end of L 2 to node 5.
Independent voltage source
The general specification of an independent constant (d.c.) voltage source

has the form
VXXXXXXX N+ N- (DC)VALUE
for example V4 1 0 DC 7.5
and for a sinusoidal a.c. source it has the form
VXXXXXXX N+ N- (AC (ACMAG (ACPHASE) »
for example VS 3 4 AC 6.7
where VXXXXXXX designates the voltage source in the circuit, N + is the
positive node, and N- the negative node. In the case ofthe constant (d.c.)
source, VALUE is the direct voltage produced by the source; in the a.c.
case the voltage is given by ACMAG, and ACPHASE is the phase angle of
the source. If ACMAG is omitted, a value of one volt is assumed; if

ACPHASE is omitted, a phase angle of zero is assumed.
SPICE also allows a selected range of other independent sources such as
a pulse, an exponential wave, etc., to be used.
Linear dependent voltage sources
The general form of a linear voltage-controlled (or voltage-dependent)

voltage source is
E~ N+ N- NC+ NC- VALUE
where N + and N - are the positive and negative nodes, respectively, ofthe
controlled source, while NC+ and NC- are the respective positive and
negative nodes of the dependent (or controlling) voltage source. VALUE
is the voltage gain from the controlling to the controlled source.
The general form for a linear current-controlled (or current-dependent)
voltage source is
H~ N+ N- VNAM VALUE
where N + and N - have the meaning given above, and VNAM is the name
of the voltage source through which the controlling current flows. The
direction of positive controlling current flow is from the positive node,
through the source, to the negative node of VNAM. VALUE is the
transresistance (ohms) between the controlling current and the controlled
voltage.
Independent cu"ent source
For a constant (d.c.) source the general specification has the form
IYYYYYYY N+ N- (OC) VALUE
for example 15 1 0 OC 5.3
and for an a.c. source it has the form
IYYYYYYY N+ N- (AC (ACMAG (ACPHASE) »
for example 142 5 3 AC 5.6
IYYYYYYY designates the current source within the circuit, and
the arrow designating the direction of current flow through the source is
from node N + to N -. VALUE is the value of the current. In the case of an
a.c. souree, if ACMAG is omitted a eurrent of 1 A is assumed, and if
ACPHASE is omitted a phase angle of zero is assumed.
Computer Solution o[ Electric Circuits 355
Linear dependent current sources
The general form for a linear voltage-dependent (or voltage-controlled)

current source is
GXXXXXXX N+ N- NC+ NC- VALUE
where N + and N - are the positive and negative nodes, respectively, of
the controlled source, NC+ and NC- are the respective positive and
negative nodes of the controlling source, and VALUE is the transconduc-
tance (in S) between the two. Current flow is from N +, through the source,
to N-.
The general form for a linear current-dependent (or current-controlled)
current source is
FXXXXXXX N + N- VNAM VALUE
where N+ and N- are as defined above, VNAM is the name ofthe voltage
source through which the controlling current flows, and V ALUE is the
current gain between the two.
Print-out analysis
The general form of statement is

.PRINT PRTYPE VARl (V AR2 ... V AR8)
in which PRTYPE is the type of analysis to be performed, such as d.c.,
a.c., transient, noise, etc. For example
.PRINT DC V(l) V(3, 4)
In the above example, V(l) is the voltage at node 1 (with respect to
node 0), and V(3, 4) is the voltage at node 3 with respect to node 4.
In the case of an a.c. analysis, five forms of output can be obtained as
folIows:
VR - real part
VI - imaginary part
VM - magnitude
VP - phase shift
VDB - 20 X loglO (magnitude)
As with the d.c. case, differential voltage can be printed, for example
.PRINT AC VP(l, 2) VR(4,5)
The current flowing in an independent voltage source can be specified as
follows
I (VXXXXXXX)
and the current in the voltage source can be printed, for example
.PRINT AC IM(VS) IP(VS)
A voltage source must be present in a branch before the current in the
branch can be printed; it is sometimes useful to insert an independent
voltage source of zero volts in a branch in order to determine the current in
that branch.
Plotting
The general form of plot is produced by the .PLOT statement described

below
.PLOT PLTYPE OVl(PLOl, PHIl) (OV2) (PL02, PHI2) ...
(OV8)
Up to eight output variables (OVl to OV8) may be plotted; the type of
analysis is specified by PLTYPE (wh ich may be DC, AC, or TRAN
(SIENT». The low and high boundaries for the independent variables may
be specified by the user (PLO, PHI), but if the plot limits are not specified,
the SPICE program will automatically determine suitable minimum and
maximum sc ale values.
Table 14.22 shows a typical plot of two variables (gain in dB and phase)
plotted against frequency. Both of the vertical seal es were chosen auto-
matically. Note also that the value of the first specified output variable
(gain) has its values printed alongside the independent variable (frequency)
values. There is no limit to the number of .PLOT lines specified for each
type of analysis.
Transient analysis
The general form of statement is

TRAN TSTEP TSTOP (TSTART) (TMAX) (UIC)
where TSTEP is the plotting andJor printing increment of time for the
output. TSTOP is the final value of time, and TSTART is the initial value
of time used in computing the transient response; the default value of the
latter is zero. Using TMAX enables the computing time step to be smaller
than TSTEP. The option al parameter UIC means 'Use Initial Conditions',
whose values are specified using IC = ... with inductors or capacitors, or
a .IC line to specify initial node voltages.
.PRINT TRAN and .PLOT TRAN are two lines which are respectively
used to print out a table of node voltages, and to plot a graph of these as a
function of time. For example, the lines
.TRAN O.lM 5M
.PLOT TRAN V(4) V(l)
produce a graph of the voltages at no des 4 and 1 in a network, starting at
zero time and incrementing in steps of 0.1 ms for a duration of 5 ms. The
graph abscissa points will be those of time, with the corresponding values
of node 4 voltages at these times. The scales of both voltages will be
calculated and marked automatically.
14.4 Programming examples
In the following we will consider aseries of worked examples associated

with each chapter in the book as folIows:
Worked example number Chapter number
14.1 1
14.2 2
14.3 3
14.4 4
14.5 5
14.6 6
14.7 7
14.8 8
14.9 9
14.10 10
14.11 and 14.12 11
14.13 12
14.14 13
The worked ex am pIes in this chapter solve examples which have been
worked out manually in other chapters, and illustrate typical techniques for
the solution of circuits. The reader should make reference to the circuits
mentioned in the text.
Worked example 14.1
Calculate the value of 11> 12 , 13 and V p in figure 1.23 using the SPICE
circuit analysis package (see also worked example 1.17.1, page 23).
Solution
Since this is the first example of the use of SPICE, we will describe the
program in a little more detail than usual.
The input file in table 14.1 contains a description of the circuit; the first
line in the input file is a title Une, which does not contain program data.
Following the title line are aseries of comment lines which commence with
a '*'; the purpose of these lines is to make the program readable. The
reader should note that SPICE assurnes that current ftows inside any source
from a '+' node to a '-' node.
The reader should refer to figure 1.23 when studying the input file.
SPICE does not include an 'ammeter' as a circuit element, but it does
evaluate the current ftowing through each voltage source. We can there-
fore insert a zero-value voltage source at any point in the circuit where an
ammeter is needed; voltage sources VI, V2 and V3 are used as 'ammeters'
in this example. The reader will recall that the positive direction of current
through such a source is from the '+' node to the '-' node inside the
source.
Each element line describes one element in the circuit; circuit elements
can be given in any order. In this case, the circuit elements are taken from
left to right as figure 1.23 is viewed. A current source is an 'I' element, a
resistor is an 'R' element, and a voltage source a 'V' element.
At the end of the input file, it is usual to specify the type of analysis
required. Each analysis type is preceded by a 'dot', for example '.DC' or
'.AC', etc. In this case we have not specified an analysis type, since the
information provided in the input file will be sufficient to allow SPICE to
give the required output, namely the voltage at each node and the current
in the voltage sources VI, V2 and V3.
The '.OPTIONS' controlline enables the user to specify one or more
options which will apply to his circuit and/or input file. The .OPTIONS
controlline is itself optional, and can be omitted. In this case (as in other
cases in the book), we have included the NOPAGE option, which causes
page ejects to be suppressed, concatinating the print-out, and saving
considerable amount of waste paper being produced. Finally, the input file
is terminated by a '.END' command line.
The reader will find it an interesting exercise to draw the circuit using
the nodes listed in the input file because, in fact, we have added the three
voltage sources VI, V2 and V3 for the reason given earlier.
Each independent current source is described as an 'I' element, and
characters are added to the 'I' to describe the source, that is, 11 and 12.
Each current source line is as follows
IXXXXXX NI N2 DC VALUE
where IXXXXXX is the circuit definition ofthe source, such as 11, etc. The
Tab1e 14.1
*** Worked Examp1e 14.1 ***

*** 14 A d.c. current source ***
*source name
* 1 (+) node
*1 1 (-) node
*1 1 1 type of source
*1 1 1 1 va1ue
*1 1 1 1
I1 0 1 oe 14
*** 2 ohm and 4 ohm resistors ***
*e1ement
*1 (+) node
*1 1 (-) node
*1 1 1 va1ue
*1 1 1 1
Rl 2 0 2
R2 3 0 4
*** 4 A d.c. current source ***
I2 1 0 oe 4
*** 5 ohm resistor ***
R3 4 o 5
* The fo11owing ZERO VALUE vo1tage sources act as ammeters *
*source name
*1 (+) node
*1 1 (-) node
*1 1 1 type of source
*1 1 1 1 va1ue
*1 1 1 1
VI 1 2 oe 0
V2 1 3 oe 0
V3 1 4 oe 0
* The fo11owing 1ine saves paper
.OPTIONS NOPAGE
• END
current ftows within the source from node Nt to node N2. The 'De'
statement is optional, but is included in this case to remind us that it is a
d.c. source. If it is an a.c. current source, it would be defined as an 'AC'
source, the 'AC' being mandatory. Finally, we include the VALUE of the
current.
Resistors are defined as 'R' elements, the two numbers following the
element definition are the nodes between which the resistor is connected.
For example, the line
R2 3 0 4
teUs us that resistor R2 is connected between no des 3 and 0, and its value is
4 Q. SPICE uses node 0 as the reference or zero-voltage node.
Most versions of SPIeE do not have 'ammeters' , but they can give
details of the current flowing in an independent voltage source. In this
case, we insert zero-value voltage sources VI, V2 and V3 in series with
resistors Rl, R2 and R3, respectively, in order to determine the current in
each of the resistors. The independent voltage sources are defined as 'V'
sources as foUows
VXXXXXX N+ N- oe VALUE
where VXXXXXX is the name given in the circuit to the source. Node N +
is the positive terminal of the source, and N - is the negative terminal.
Once again, 'oC' is optional and, in each ca se here, the value of the source
is zero. When SPIeE was written, it was decided that positive current enters
the positive terminal of an independent voltage source. If the reader draws
the circuit using the node numbering in the input file, he will see that the
current in each voltage source enters the positive node. Should the node
numbering be reversed, the corresponding value of current printed in the
output file has a negative sign!
The output file in table 14.2 shows that the voltage at each node is
10.526 V, which corresponds to the result in worked example 1.17.1 and,
likewise, the current in resistors Rl, R2 and R3, corresponding to the
current in voltage sources VI, V2 and V3.
Table 14.2
N:DE VOLTAGE N:DE VOLTAGE N:DE VOLTAGE N:DE VOLTAGE

( 1) 10.5260 ( 2) 10.5260 ( 3) 10.5260 ( 4) 10.5260
VOLTAGE SOURCE CURRENTS

NAME CURRENT
VI 5.263E+00
V2 2.632E+00
V3 2.105E+00
Table 14.3
.PARAM current = 10
Il 0 1 DC {1.4*current}
Rl 2 0 2
R2 3 0 4
I2 1 0 DC {0.4*current}
R3 4 0 5
In many applications it is convenient to use a parameter instead of a

numeric value, and the usefulness of parameters is greatly extended by
allowing them to be combined into arithmetic expressions. The early part of
the input file in table 14.1 is modified as shown in table 14.3 to include a
parameter called 'current' , the parameter first being defined in a
'.PARAM' control line; the value of 'current' is arbitrarily set at 10 A.
Next, the parameter is included in an arithmetic expression enclosed in
curly brackets { }, in the lines for 11 and 12. The value of 11 is
'current' x 1.4 = 14 A
and in the case of 12 it is
'current' x 0.4 = 4 A.
A salutary note must be given here about the use of computer software
for circuit solution. While the software may give the solution quickly, it
does not necessarily help the reader to understand the engineering pro-
cesses involved.
Worked example 14.2
Calculate the voltage at each node in figure 2.14 using the SPICE pro gram-
ming language (see also worked example 2.9.3 on page 46).
Solution
The input file describing the circuit is given in table 14.4 SPICE cannot
accept conductance values, and each is converted into its equivalent
resistance value. To specify each resistance in this case, each has been
numbered by the nodes it is connected to. For example, the 4 S conduc-
tance in figure 2.14 is described as resistance RB of value a 0.25 Q (SPICE
only accepts resistance values), connected between node 1 and 3. Similarly
with the current sources; the 3 A source which drives current from node 2
to node 3 is described as 123, etc.
This problem introduces us to the SPICE version of the voltage-
controlled current source, which is a 'G' source. Using the above conven-
tion, we call it G 13, wh ich drives current from node 1 to node 3. The
voltage-controlled current source is defined as follows
GXXXXXX N+ N- NC+ NC- VALUE
where GXXXXXX is the circuit 'name' of the source, and current is driven
within the source from N + to N -, that is, current leaves terminal N -.
Node NC+ is the positive controlling node, NC- is the negative control-
ling node and VALUE is the transconductance in Siemens of the
source.
Table 14.4
*** Worked Example 14.2 ***

*** Resistance va lues ***
R01 0 1 0.1667
R12 1 2 0.5
R23 2 3 0.3333
R03 0 3 0.2
Rl3 1 3 0.25
** Independent current sources **
102 0 2 DC 2
123 2 3 DC 3
** Voltage-controlled current source **
*source name
*1 N(+) N(-) NC (+) NC(-) transconductance
*1 1 1 1 1 1
G13 1 3 3 2 1.5
.OPTIONS NOPAGE
. END
Table 14.5
NODE VOLTAGE NODE VOLTAGE NODE VOLTAGE
( 1) .0707 ( 2) .0174 ( 3) .3151
The relevant part of the output file is given in table 14.5, and the voltage
at each node is seen to agree with the results of worked example 2.9.3
Worked example 14.3
Determine V AB and the current in each generator in figure 3.11 (a ) (see also
worked example 3.9.1 on page 71).
Solution
The circuit in figure 3.11 is described in the SPICE input file in table 14.6.
The first three element lines specify the three independent direct voltage
sources and the nodes to which they are connected. In this case we have
chosen to enter the 'positive' direction of the voltage in the input file. That
is, the positive terminal of V3 is listed in table 14.6 as being connected to
node O. The next three element lines describe the value and circuit
connections of each resistor. The reader should draw the corresponding
circuit diagram, and note that the common point to which the resistors are
connected is node 4 (corresponding to node Bin figure 3.11(a».
Once again, it is not necessary to specify the type of analysis needed,
because SPICE will perform an operating point analysis and will calculate
the voltage at eaeh node in the output file, whieh is given in table 14.7.
The reader will note that the voltage at nodes 1, 2 and 3 eorrespond to
the voltage of the respeetive independent sourees, and the voltage at node
4 eorresponds to V BA in worked example 3.9.1. SPI CE also outputs the
eurrent in eaeh voltage souree but, as deseribed in worked example 14.1, it
is the eurrent jiowing into the positive terminal 0/ the corresponding voltage
source. That is, we need to multiply eaeh eurrent in table 14.7 by -1 to get
the eonventional eurrent.
The eurrent ftowing in the voltage souree V3 is also given a negative
value but, sinee the positive node of V3 is eonneeted to node 0, the eurrent
ftows through this souree from node 0 to node 1. Onee again, we see that
an engineering vision is neeessary when interpreting the results of
SPICE.
Table 14.6

*** Independent Voltage Sources ***
VI 1 0 DC 10
V2 2 0 DC 20
V3 0 3 DC 25
*** Resistors ***
R1 1 4 20
R2 2 4 15
R3 3 4 10
.OPTIONS NOPAGE
• END
Table 14.7
**** EM\LL SIGUlli BIAS SOllJrI(N 'IEM'ERA'IURE = 27.000 !)EX; C

1) 10.0000 ( 2) 20.0000 3) -25.0000 ( 4) -3.0769

NAME CURRENT
VI -6.538E-01
V2 -1.538E+00
V3 -2.192E+00
TOTAL POWER DISSIPATION 9.21E+01 WATTS
Worked example 14.4
Using SPICE software, plot the waveforms of voltage aeross and eurrent
through the induetor in figure 4.11(a) (see also worked example 4.7.1 on
page 86).
Solution
In this example we will introduce some additional features of the SPICE

language, the input file being shown in table 14.8. In order to plot a graph
of a variable, we must first perform a transient analysis on the circuit and,
in order to do this, we must excite the system with a time-dependent
source. That is, we must describe a source to SPICE in terms of the value
of the signal and the time at which it occurs.
In this case the current source is described by me ans of a Piece-Wise
Linear (PWL) function as follows
IXXXXXX N+ N- PWL(Tl, VI TI, V2 ... Tn, Vn)
where IXXXXXX is the name used to describe the independent current
source, and the current flows inside the source from node N + to N -, that
is, the current leaves node N -. Pairs of values inside the PWL brackets
describe pairs of coordinates of the current waveform, Tn, Vn correspond-
ing to the nth value of time (seconds) and the nth value of current
(amperes). Pairs of points are corrected linearly (the comma between the
time and value for each point is optional). Thus the three points in the
PWL wave describe a triangular current wave.
Table 14.8

*** The following current source is described by a ***
*** Piece-Wise Linear function (PWL) ***
*source name
*1 N(+) N (-) T1,V1 T2,V2 T3, V3
*1 1 1 1 1 1
I1 0 1 PWL(O,O 3,3 6,0)
* VZERO is used as an anmeter
VZERO 1 2 0
* 6 HInductor
*inductor name
*1 N (+) N(-) value
*1 1 1 1
LI 2 0 6
.OPTIONS NOPAGE
*transient analysis
*1 Tstep Tstop
*1 1 1
.TRAN 0.3 6.6
* PLOT the TRANsient analysis
. PLOT TRAN V(2) I (VZERO)
• END
Computer Solution 01 Electric Circuits 365
As stated earlier, current ftows into the 'positive' pole of an independent

voltage source, and SPIeE can evaluate the current ftowing in an inde-
pendent voltage source. Therefore, in order to measure the current in the
circuit, we introduce the concept of a zero-value voltage source, VZERO,
in this circuit. The purpose is to use the source as an 'ammeter' . Finally, a
6 Hinductor is connected to the PWL current source and the 'am-
meter'.
The '.TRAN' command line causes SPIeE to compute the transient
response of the circuit every 0.3 s for aperiod of 6.6 s. Finally, the '.PLOT
TRAN' line causes SPIeE to plot the TRANsient response showing not
only the voltage across the inductor (V(2)), but also the current in the
circuit (I(VZERO)). The reason that we ask for I(VZERO) to be plotted
is that, in order either to print or plot a current, it must ftow through a
voltage source.
Table 14.9
ux;;m:
*: V(2)
+: I(VZER»
TnE V(2)
(*)-- -l.OOOOE+Ol -5.0000E+OO O.OOOOE+OO 5.0000E+OO 1.OOOOE+Ol
(+)-- -4.121IE-30 1.OOOOE+OO 2.0000E+OO 3.0000E+OO 4.0000E+OO
O.OOOE+OO O.OOOE+OO + *
3.000E-Ol 6.000E+OO • + *
6.000E-Ol 6.000E+OO + *
9.000E-Ql 6.000E+OO• +. *
1.200E+OO 6.000E+OO• + *
1.500E+OO 6.000E+OO• + *
1.800E+OO 6.000E+OO• + *
2.100E+OO 6.000E+OO• .+ *
2.400E+OO 6.000E+OO• + *
2.700E+OO 6.000E+OO • + *
3.000E+OO 6.000E+OO • + *
3.300E+OO -6.000E+OO • * +
3.60OE+OO -6.000E+OO • * +
3.90OE+OO -6.000E+OO • * .+
4.200E+OO -6.000E+OO • * +
4.500E+OO -6.000E+OO • * +
4.800E+OO -6.000E+OO • * +
5.100E+OO -6.000E+OO • * +.
5.400E+OO -6.000E+OO • + *
5.700E+OO -6.000E+OO • + *
6.000E+OO -6.000E+OO + *
6.300E+OO 5.673E-20 + *
6.60OE+OO O.OOOE+OO + *
The corresponding section of the output file is shown in table 14.9. The
reader will note that we have not specified the range over which the results
are to be plotted, and we have left it to SPICE to decide. With the form of
print-out used here, it is not possible to show the sudden transition in
voltage across the inductor when the current changes in value, and the
reader should not think that the current changes linearly between the
points on the graph. Several suppliers of versions of SPICE offer a graphics
post-processor to provide an accurate, high-quality graphics output.
Worked example 14.5
Solve the a.c. series circuit in worked example 5.10.1 (see page 112) using
SPICE.
Solution
The input file is given in table 14.10, and it is left as an exercise for the
reader to draw the corresponding circuit diagram. SPICE can accept
voltages and currents in polar complex form, and can output data in either
polar or rectangular form. The first element line in table 14.10 describes
the voltage source as an a.c. source connected between node 1 and node 0,
and having a magnitude of 10 V and a phase angle of 20°.
Unfortunately, SPICE can only accept information about inductance
and capacitance, and cannot accept reactance data. However, there are
several ways round this problem, and one is as follows. If we use a
frequency of 1 radis (or 0.1592 Hz), then we get the correct result ifwe let
L = XL> and C = 1/Xc . This is done here.
Consider impedance Zl, which comprises a resistance of 7.071 Q in
series with an inductor Li == XLl = 7.071 H. We therefore show a
resistance of 7.071 Q connected between nodes 1 and 2, and an inductor of
7.071 H connected between nodes 2 and 3. The latter has a reactance of
7.071 Q at the excitation frequency of 1 radis or 0.1592 Hz. Similarly for
impedance Z3, where we have a resistance of 2.605 Q in se ries with
a capacitive reactance of 14.77 Q. The corresponding 'capacitance' is
C 3 = 1/XC3 = 1/14.77 = 0.0677 F = 67.7 mF. Finally we use an indepen-
dent zero-value voltage source, Vam, as an ammeter.
The method of dealing with the frequency of the source is shown in the
'.AC' controlline, and is specified in the following line .
.AC DIST NOPOINTS STARTFREQ ENDFREQ
This control line is, strictly speaking, intended to deal with a range of
frequencies for frequency response calculations; in this case we only need
one frequency. Where the word DIST appears we enter the way in which
the frequencies are distributed between the STARTFREQ frequency and

the ENDFREQ frequency (both in Hz). There are three options, namely
LIN (LINear), ocr (OCTave) and DEC (DECade); we choose LIN.
Where NOPOINTS appears we insert the number of points to be calcu-
lated. Thus the line
.AC LIN 1 0.1592 0.1592
implies that we need one frequency between 1 radis and 1 radis!
There are two '.PRINT' command lines, each asking for the current
through the 'ammeter' and the voltage of node 3 with respect to node 4 to
be printed. The two lines do, however, ask for the data in a different form.
Table 14.10

*** Voltage Source ***
*source name
*1 N (+) N(-) type of source
*1 1 1 1 magnitude phase (deg)
*1 1 1 1 1 1
VS 1 0 AC 10 20
*** Inpedance Zl ***
R1 1 2 7.071
* Let LI = XLI
LI 2 3 7.071
*** Inpedance Z2 ***
R2 3 4 10
*** Inpedance Z3 ***
R3 4 5 2.605
* Let C3 = 1/XC3
C3 5 6 67. 7M
*** ArmIeter ***
Vam 6 o AC o
*analysis type
*1 LINear number of frequencies
*1 I I start freq end freq
*1 1 1 1
.AC LIN 1 0.1592 0.1592
.OPTIONS NOPAGE
* magnitude of I (Vam) phase of I (Vam)
* 1 1
.PRINT AC IM (Vam) IP (Vam) VM(3,4) VP(3,4)
* real part of I (Vam) imaginary part of I (Vam)
* 1 1
.PRINT AC IR (Vam) II (Vam) ~(3,4) VI(3,4)
• END
The first of these lines asks for the data in polar form as follows
IM(Vam) = current Magnitude through Vam
IP(Vam) = current Phase angle, etc.
and the second line asks for
IR(Vam) = Real part of the current through Vam
II(Vam) = Imaginary part of the current through Vam
The relevant parts of the output file are given in table 14.11.
Table 14.11
FREQ IM (Vam) IP(Vam) VM(3,4) VP(3,4)

1.592E-01 4.733E-01 4.136E+01 4.733E+OO 4.136E+01
FREQ IR (Vam) 11 (Vam) VR(3,4) VI(3,4)
1.592E-01 3.553E-01 3.128E-01 3.553E+OO 3. 128E+OO
Worked example 14.6
Use SPICE computer software to calculate the voltage gain of the transis-
tor amplifier equivalent circuit in figure 6.7(a) at a frequency of (a) 1 kHz
and (b) 200 kHz (see worked example 6.8.1 on page 139).
Solution
The circuit is fairly straightforward, with the exception that the transistor is
simulated by a voltage-controlled current source. The input file is shown in
table 14.12, and the appropriate section of the output file in table 14.13.
The reader will note that the results agree with those of worked example
6.8.!.
The reader is advised at this point of a difficulty which may arise in some
a.c. problems, although it does not occur here, and that is SPICE is
organised (for many types of solution) only to deal with angles which lie in
the range ±180°. If the angle is (for example) less than -180°, it will
be shown as the corresponding positive angle; similarly, positive angles
greater than + 180° are shown as negative angles. That is, once again,
engineering judgement must be applied to the results.
Also, this analysis (and that of worked example 6.8.1) is based on
sm all-signal analysis, and does not allow for practical features such as the
effect of 'saturation' of the transistor if too large a voltage is applied to the
input.
Table 14.12

*** 10 mV a.c. voltage source ***
VS 3 0 AC 10M
*** Circuit components ***
R31 3 1 100
R10 1 0 lOK
R12 1 2 lOOK
R20 2 0 150
C12 1 2 25P
*** Voltage-controlled current source ***
*name of source
*1 (+) output node
*1 1 (-) output node
*1 1 1 (+) controlling node
*1 1 1 1 (-) controlling node
*1 1 1 1 1 transconductance
*1 1 1 1 1
G20 2 0 1 0 8
.OPTIONS NOPAGE
.AC LIN 2 1k 200k
.PRINT AC VM(2) VP(2)
• END
Table 14.13
FREQ VM(2) VP (2)

1.000E+03 5.423E+OO 1.795E+02
2.000E+05 2.741E+OO 1.204E+02
Worked example 14.7
Solve the unbalanced three-phase system in worked example 7.9.1 using

SPIeE software (see page 153).
Solution
In this case, using the techniques described earlier in the book, we specify a
set of unbalanced voltages and unbalanced loads. The input file is given in
table 14.14, and the first three lines specify the unbalanced phase voltages
as follows
V an = 200 L 10 V 0
V bn = 220 L - 140 V 0
V cn = 180 L 100° V
where node 0 is the neutral point and lines a, band c are connected to
SPICE nodes 1, 2 and 3, respectively. Three zero-value voltage sources
Va, Vb and Vc are used as 'arnrneters' to measure the three line currents.
As with worked example 14.5, we use a frequency of 1 radis (0.1592
Hz), so that the 'inductance' in phase B is equal to the value of the
inductive reactance in phase B, and the capacitance in phase Cis equal to
lIX c. The '.AC' line specifies that we are using a single frequency of 0.1592
Hz or 1 radis.
Finally there is a set of '.PRINT' control lines causing the results
obtained to be output; these are listed in table 14.15. Each '.PRINT' line in
the input file is seen to produce its own set of results and, in the absence of
the '.OPTIONS NOPAGE' line, each will generate a separate page of
Tab1e 14.14

*** Unba1anced 3-phase supp1y ***
Van 1 0 AC 200 10
Vbn 2 0 AC 220 -140
Vcn 3 0 AC 180 100
*** Ammeters in 1ines a, band e ***
Va 1 7 AC 0
Vb 2 8 AC 0
Ve 3 9 AC 0
*** Z in phase A = Zas ***
Ras 7 6 10
*** Z in phase B = Zbs ***
Rbs 8 4 14.77
Lbs 4 6 2.605
*** Z in phase C = Zcs ***
Res 9 5 4.698
Ces 5 6 0.5848
.OPTIONS NOPAGE
.AC LIN 1 0.1592 0.1592
*** V(6) = Vsn ***
.PRINT AC VM(6) VP (6)
*** V(7,6) = Vas, V(8,6) = Vbs ***
.PRINT AC VM(7,6) VP(7,6) VM(8,6) VP (8, 6)
*** V(9,6) = Ves ***
.PRINT AC VM(9,6) VP(9,6)
*** I (Va) = Ia, I (Vb) = Ib ***
.PRINT AC IM (Va) IP (Va) IM (Vb) IP(Vb)
*** I(Ve) = Ie ***
.PRINT AC IM (Ve) IP(Ve)
• END
Tab1e 14.15
FREQ VM(6) VP (6)
l.592E-01 8.224E+01 1.028E+02
FREQ VM(7,6) VP(7,6) VM(8,6) VP (8, 6)
1.592E-01 2.200E+02 -l.193E+01 2. 678E+02 -l.241E+02
FREQ VM(9,6) VP(9,6)
l.592E-01 9.794E+01 9.764E+01
FREQ IM (Va) IP(Va) IM (Vb) IP(Vb)
1.592E-Ol 2.200E+Ol -l.193E+Ol 1.785E+Ol -1.341E+02
FREQ IM (Vc) IP(Vc)
1. 592E-Ol 1.959E+Ol 1.176E+02
paper! When the results are compared with those ofworked example 7.9.1,
the reader will note that they are the same and will, of course, appreciate
that while the computer analysis is straightforward, it does not necessarily
give a dear understanding either of the processes involved or of the
practicability of the results. Only an experienced 'engineering eye' can give
the latter.
Worked example 14.8
Using SPIeE software, caIculate the value of the parameter Y2t in the
circuit in worked example 8.3.1 (see page 178).
Solution
Strictly speaking, SPIeE was not developed for this type of problem, but it
is so versatile that it can be used to solve most problems. The only
limitation to its use is in our own mind! The reader will recall that
V2 =0
All we need to do is to apply a 1 V source to the input, and measure 1 2 with
the output short-circuited. This is a simple matter to implement with
SPIeE.
The input file is shown in table 14.16, in which node 0 is the node
common to the input and the output, node 1 being the input node, and
node 2 is the output node. In this case, V2 acts as an 'ammeter' , with
'positive' current ftowing into node 2. The appropriate part of the output
file is given in table 14.17.
As with earlier SPIeE files, we do not need to specify the type of
Table 14.16

*** Calculation of y21 = I2/V1 with v2 o ***
*** Input source ***
V1 1 0 DC 1
*** passive circuit elements ***
R1 1 0 600
R2 1 2 1.5K
R3 2 0 lOK
*** Voltage-controlled current source ***
G 2 0 1 0 0.04
*** V2 = 0 is used as an ammeter ***
V2 0 2 DC 0
*** Note: y21 = current in short-circuiting "ammeter" V2 ***
.OPTIONS NOPAGE
• END
Table 14.17

NAME CURRENT
V1 -2.333E-03
V2 3.933E-02
analysis to be performed, since we only need the current flowing into the
output node, which is 0.03933 A, hence Y21 = 1 2 /V 1 = 39.33 mS.
Worked example 14.9
Using SPICE software, solve the linear transformer problem in worked

example 9.8.1 (see page 212 and figure 9.1O(a» for the inductive load.
Solution
In this ca se we will drive the circuit with a 1 L 0° A a.c. source (11), whose
current flows from node 0 to node 1; the magnitude and the phase angle of
the source are given following the 'AC' expression. The default phase
angle is zero and can be omitted, but is included in the specification of 11
for completeness.
The reason for selecting a 1 A current source is that the input impedance
of the circuit is equal to the voltage at the input terminals. The SPICE
primary circuit consists of a 10 Q res ist an ce R1, connected between nodes
1 and 2, and a 75 mH inductance L1, connected between nodes 2 and O.
The secondary circuit contains the following: a 10 Q resistance R2 between
nodes 4 a.nd 5, a 150 mH inductance L2 between nodes 4 and 3 together
Computer Solution of Electric Circuits 373
**** CIRCUIT DESCRIPTION ****

* The following is an a.c. CURRENT source of 1 A
11 0 1 AC 1 o
* See figure 9.1O(a) for the circuit elements
Rl 1 2 10
R2 4 5 10
* PRIMARY winding
L1 2 o 75M
* SECONDARY winding
L2 4 3 150M
* MAGNETIC COUPLING
* name of coupled circuit
*: name of first inductor
*: name of second inductor
*: magnetic coupling coefficient
* I
I
K1 L1 L2 0.75425
* Extra resistance to 'electrically' link the * primary winding to the
secondary winding.
Rextra 0 3 1MEG
* Inductive load
Llqad 5 3 0.4
.OPTIONS NOPAGE
* Calculate values at 500, 1000 and 1500 radis
.AC LIN 3 79.58 238.7
* Print the rectangular form of the input voltage
.PRINT AC VR(l) VI(l)
.END
with the load. The latter is a 0.4 H inductance Lload connected between
nodes 3 and 5. Coils L1 and L2 are magnetically coupled with a coupling
coefficient of
k = M/V(Ll x L2) = 80 x 1O- 3/V (75 x 10- 3 x 150 x 10- 3 )
= 0.75425
Since SPICE demands that all nodes must have a 'd.c. link' to node 0, a
large value of resistance, Rextra, is used to connect the primary and
secondary windings together. Since current does not flow in this resistor, it
has no significant effect on the result of the analysis.
The '.AC' controlline calls for calculations at three frequencies, namely
500,1000 and 1500 radis, whose values are converted into Hz in the .AC line.
In worked example 9.8.1, the results were presented in rectangular
form, and this is wh at we have done here. The .PRINT control line
requests the 'real' part of the voltage at node 1, i.e., (VR(l)), and the
'imaginary' part (VI(l)) to be printed. The reader should compare the
value of the input impedance at 159.2 Hz (1000 radIs) with that obtained
above. It is, of course, a simple process to repeat the calculation for a
capacitive load.
The relevant part of the output file is as folIows.
FREQ VR(l) VI(l)
7.958E + 01 l.021E + 01 3.169E + 01
l.592E + 02 l.021E + 01 6.337E + 01
2.387E + 02 l.021E + 01 9.505E + 01
Worked example 14.10
Plot graphs of it(t) and ilt) in figure 1O.29(a) - see also worked example
10.13.1 on page 259.
Solution
In this case we will choose to excite the circuit with a PULSE, as described
in the V1line in the input file (see table 14.18). A reason for selecting this
form of driving function is that we need to do a transient response (see the
'. TRAN' controlline) on the output, and, for this purpose it is necessary to
energise the system by a time-dependent function, the repetitive PULSE
being one example. The PULSE function has the following variables
PULSE (IV PV TD TR TF PW PP)
where
IV is the initial value = 0

PV is the pulsed value = 10 V
TD is the delay time before the pulse is applied = 0
TR is the rise time of the pulse = 1 Ils
TF is the fall time of the pulse = 1 Ils
PW is the pulse width = 50 ms
pp is the pulse period = 51 ms
In this case, it is merely necessary to make the pulse width longer than the
transient period involved in the problem, and the pulse period longer than
the sum of TR, TF and PW.
Perhaps, at this time, we should look at the '.TRAN' controlline. This
calls for SPICE to evaluate the transient solution for 40 ms in steps of 2 ms.
The reader will recall that we estimated that the secondary current would
reach its steady-state value (zero) in about 40 ms. Consequently, a transi-
ent period of 40 ms is allowed for in the input file; a step period of 2 ms
provides (40/2) + 1 = 21 results in this period.
Returning to the circuit description, 'Vpri' and 'Vsec' act, respectively,
as primary and secondary ammeters. The reader will note the ';' separator
Table 14.18

*** The input source is a PULSE ***
*output nodes pu1sed va1ue fall time
* I I initial va1ue I time delay pulse width
* I I I I I rise time I I period
* I I I I I I I I I
V1 1 0 PULSE (0 10 0 1U lU 50M 51M)
*** Ammeters ***
Vpri 1 2 0 ;primary ammeter
Vsec 4 7 0 ;secondary ammeter
*** Primary circuit resistance ***
Rpri 2 3 20
*** Load ***
Rload 7 6 20
Lload 6 5 0.01
*** Coupled circuit ***
L1 3 0 0.1
L2 4 5 0.1
K L1 L2 0.5
*** d.c. linking resistor ***
Rlink 0 5 lMEG
.OPTIONS NOPAGE
• TRAN 2M 40M
. PLOT TRAN I (Vsec) (0,0.1) I (Vpri) (0,0.5)
• END
Tc:Ible 14.19
IEGID:
*: I(Vsec)
+: I(Vpri)
TnE I(Vsec)
(*)-- O. ()()()(E+OO 2 .5OQ(E-<l2 5. ()()()(E-()2 7 •5OQ(E-<l2 1. ()()()(E-<l1
(+)-- O.()()()(E+OO 1.2SO<E-<l1 2.5OQ(E-<l1 3.7SO<E-<l1 5.()()()(E-<l1
O. ()()()E+OO O.OOCE+OO X
2.00CE-<l3 7.193E-<l2 • + *
4.()()()E-<l3 9.129E-<l2 • + *
6. ()()()E-<l3 8.735E-<l2 • +. *
8.00CE-<l3 7.56lE-<l2 • * +
1.00CE-<l2 6.229E-<l2 • * +
1.200E-<l2 4.996E-<l2 • * +
1.400E-<l2 3.95lE-<l2 • * +
1.6OOE-<l2 3.093E-<l2 • * +
1.BO<E-<l2 2.41lE-<l2 • * + .
2.00CE-<l2 1. 87lE-<l2 • * + .
2.200E-<l2 1. 452E-<l2 • * +.
2.400E-<l2 1.!23E-<l2 • * +.
2.6OOE-<l2 8.693E-<l3 • * +.
2.BO<E-<l2 6.717E-<l3 • * +.
3.()()()E-<l2 5.196E-<l3 • * +
3.200E-<l2 4.013E-<l3 • * +
3.400E-<l2 3.103E-<l3 • * +
3.6OOE-<l2 2.396E-<l3 •* +
3.8OOE-<l2 1. 853E-<l3 •* +
4.00CE-<l2 1.43OE-<l3 •* +
used between the circuit element field and the comment field. The electri-
cal circuit and the magnetically coupled circuit are defined in the input file
in the manner described earlier in the book. Once again, all nodes must
have a 'd.c.' link to node 0, so that a large value of linking resistor, Rlink,
connects one point in the primary circuit to one point in the secondary circuit.
Finally, we will take a look at the method of displaying the output data
(see also table 14.19). The '.PLOT' controlline calls on SPICE to PLOT
the current through the voltage source Vsec and through Vpri, in that
order. However, the .PLOT controlline used here differs from that used
hitherto, since we use it to set limiting values to the range over which the
current is plotted. As a general rule, if the user does not know the range
over which the results will appear, it is advisable to let SPICE itself make
the decision. A dis advantage which sometimes occurs with this arrange-
ment is that the resulting graphs do not quite appear in the position that
one would like!
In this case, we'have a general idea of the values involved, and we can
specify them in the .PLOT line. The limits are quoted in parenthesis, with
the minimum value first and the maximum value second. That is, the
secondary current is to be plotted between values of zero and 0.1 A, and
the primary current between zero and 0.5 A. The output from the printer is
shown in table 14.19, with each point on the graph of the first named
variable, namely I(Vsec), being plotted with a '*', and each of the points
on the graph of the second named variable, I(Vpri), being plotted with
a '+'.
When I(Vsec) and I(Vpri) have the same value (as they both do
initially), the point is marked with an upper-case 'X'.
The table of results on the left of the graphs shows, in the first column
the time in ms and, in the second column the value of the first named
variable, namely I(Vsec). A print-out of all the values of all variables could
have been obtained had we included a '.PRINT' control line in the input
file. Once again, we see that the transients have practically reached their
final value in about 40 ms.
Plot the frequency response diagram of the circuit in worked example

11.11.1 (see page 285) using SPIeE software.
Solution
The circuit is described in the input file in table 14.20 in standard SPIeE
format. In this case we have chosen to drive the circuit using an a.c. current
source of 1 L 0° A, so that the voltage at node 1 (the input) is equal to the
input impedance of the circuit. The '.AC' controlline requests the compu-
ter to calculate the results in frequency decades, five points per decade,
from 0.1592 Hz (1 radis) to 1592 Hz (10 000 radis), that is, over 4
decades.
The '.PRINT' and '.PLOT' controllines request the computer to output
the magnitude in dB and the phase shift (in degrees). The relevant sections
of the output file are shown in tables 14.21 and 14.22.
The .PRINT controlline causes both the decibel value of the voltage at
node 1 and the corresponding phase angle to be printed for each frequency
(see also table 14.11).
The .PLOT line does not specify the range over which the results are to
be plotted and (see table 14.22) SPIeE chose the following
VDB(1): 20 dB to 100 dB, and marks points with a ,*,
VP(1): -50° to 150°, and marks points with a '+'.
However, the .PLOT control line only results in the value of the first
named variable in the line to be printed, that is, VDB(l). The first column
of results in table 14.22 corresponds to the frequencies at wh ich the
calculations occur, and the second column contains the corresponding
values of VDB(1).
Tab1e 14.20

I 0 1 AC 1
R 1 2 25
L 2 0 0.25
.OPTIONS NOPAGE
.AC DEC 5 0.1592 1592
* The vo1tage V(1) is equa1 to the circuit impedance *
.PRINT AC VDB (1) VP (1)
. PLOT AC VDB (1) VP (1)
• END
Tab1e 14.21
FREQ VDB (1) VP (1)
1. 592E-01 2.796E+01 5.731E-01

2.523E-01 2.796E+01 9.083E-01
3.999E-01 2.796E+01 1. 439E+00
6.338E-01 2.797E+01 2.280E+00
1.004E+00 2.798E+01 3. 611E+00
1.592E+00 2.800E+01 5.712E+00
2.523E+00 2.807E+01 9.008E+00
3. 999E+00 2.822E+01 1. 410E+01
6.338E+OO 2.860E+01 2. 171E+01
1.OO4E+01 2.941E+01 3.226E+01
1. 592E+01 3.097E+01 4.501E+01
2.523E+01 3.342E+01 5.776E+01
3.999E+01 3.660E+01 6.830E+01
6.338E+01 4.023E+01 7.590E+01
1.004E+02 4.407E+01 8.100E+01
1.592E+02 4.800E+01 8.429E+01
2.523E+02 5.198E+01 8. 639E+01
3. 999E+02 5.597E+01 8.772E+01
6.338E+02 5.996E+01 8.856E+01
1.004E+03 6.396E+01 8.909E+01
1.592E+03 6.796E+01 8. 943E+01
Computer Solution 01 Electric Circuits 379
Table 14.22
ImID:
*: vrB(l)
+: W(l)
(*)-- 2 •()()()(E+Ol 4.0000E+01 6.0000E+01 8. ()()()(E+01 1. ()()()(E+02

(+)-- -5. ()()()(E+Ol O.OOOOE+OO 5. ()()()(E+Ol 1. ()()()(E+02 1.5OCXE+02
1.592E-Dl 2.796E+Ol • * +
2.523E:-Dl 2.796E+Ol • * +
3. 999E-Dl 2.796E+Ol • * +
6.338E-D1 2.797E+01 • * .+
1.004E+OO 2.798E+Ol • * .+
1. 592E+OO 2.8OCE+Ol • * .+
2.523E:+OO 2.807E+Ol • * . +
3. 999E+OO 2.822E+01 • * +
6.338E+OO 2.86OE+Ol • * +
1.004E+Ol 2.94lE+01 • * +
1. 592E+01 3.097E+01 • * +.
2.523E:+Ol 3.342E+Ol • * . +
3. 999E+Ol 3.66OE+01 • * +
6. 338E+01 4.023E+01 • * +
1.004E+02 4.407E+01 • * +
1. 592E+02 4.8OOE+01 • * +
2.523E+02 5.l98E+Ol • * +
3. 999E+02 5.597E+Ol • * +
6. 338E+02 5.996E+Ol • * +
1.004E+03 6.396E+01 • * +
1. 592E+03 6.796E+Ol . * +
Using PSpice software, plot the gain and phase response of the generalised
quadratic transfer function
HUw) ~ I + n[~1 + [~r for, ~ 0.1 (see also section 11.13).

Solution
PSpice is aversion of SPIeE which can handle complex transfer functions

which are expressed in the s-domain, that is as a Laplace transformation
(see also chapter 10). For the purpose of the program, we allow the
complex operator jw to be replaced by s, that is, s == jw. The input file in
table 14.23 gives a solution to the problem.
PSpice deals with an s-domain transfer funetion as though it were

the 'gain' of a voltage-dependent voltage source, or an 'E' source. This (the
s-domain transfer funetion) has been given the name 'Equadratie' in the
program. The output from the souree is between nodes 2 and 0, and the
input is the voltage at node 1, that is, V(l). In the line defining the transfer
funetion, the expression LAPLACE informs the computer that it must
expeet a transfer funetion as a Laplaee transform for C = 0.1, whieh it
expeets to find within a pair of { } braekets.
As with most SPICE programs, the software expeets to find a 'd.e.' link
between eaeh node and node 0, and a one megohm resistor is eonneeted
between the input souree no des (node 1 and 0) and between the output
nodes (nodes 2 and 0).
The .PLOT eontrol line requests a plot of the output magnitude re-
sponse in dB, together with phase shift in degrees to a base of frequeney
plotted to a deeade seale of frequency (10 points per deeade) - see also the
'.AC' eontrolline. The resulting output is shown in table 14.24. The points
Table 14.23
*** Generalised Quadratic Transfer Function ***

* H(s) = 1 + (2*zeta*s/omega 0) + (s*s/(omega O*omega 0»*
*** where s replaces jw ***
V 1 0 AC 1
*** ZETA = 0.1, omega 0 = 1 ***
*name of source
* (+) output node
* I (-) output node
* I I LAPLACE trans form
* I I I signal source
* I I I I LAPLACE transform expression
* I I I I I
Equadratic 2 0 LAPLACE {V(l)} {1 + (2*0.1*s/1) + (s*s/(1*1)}
* A high value resistor must be connected between node 1 and 0 *
* and between node 2 and 0 *
R1 1 0 1MEG
R2 2 0 1MEG
.OPTIONS NOPAGE
*** Logarithmic plot of frequencies between 0.1 and 10 radis ***
*AC sweep
*1 sweep in DECades
*1 1 10 points per decade
*1 I I start freq end freq
*1 1 1 1 1
.AC DEC 10 15.92M 1.592
. PLOT AC VDB(2) VP(2)
• END
corresponding to the gain in dB are plotted using '*', and the phase angles
are plotted using '+'. The values on the left oftable 14.24 are, firstly, a list
of frequencies (in Hz) and, secondly, corresponding gain values (in dB) of
the transfer function.
When evaluating phase angles using SPICE, the reader should be aware
that some versions output angles for certain solutions only in the range
±180°. That is, if the phase angle associated with a given transfer function
slowly changes from, say, -90° to -270°, then SPICE may convert phase
angles in excess of -180° into their 'positive' phase angle equivalent, that
is, - 190° becomes + 170°.
The graph in table 14.24 should be compared with those in section
11.13.
'Iable 14.24
IEJ:»l):
*: Vffi(2)
+: VP(2)
rnm Vffi(2)
(*)-- -2 •()()()()E+<l1 O. OOOOE+OO 2. ()()()()E+<l1 4 • ()()()()E+<l1 6 •0000E+<l1
(+)-- O. OOOOE+OO 5. 0000E+<l1 1. ()()()()E+<l2 1. SOOOE+<l2 2. 0000E+<l2
1.592E-02 -8.557E-02 + *
2. 004E-02 -1. 36<E-01 + *
2.523E-02 -2.165E-01 + *
3.176E-02 -3.456E-01 .+ *
3.999E-02 -5. 54OE-01 •+ *
5.034E-02 -8. 943E-Ql .+ *
6.338E-Q2 -1.461E+OO .+ *
7.979E-02 -2.437E+OO • + *
1.004E-01 -4.226E+OO. + *
1.265E-01 -7.927E+OO. + *
1.592E-01 -1.398E+<l1 • * +
2. 004E-01 -3. 909E+OO • * • +
2.523E-01 3.785E+OO • * +
3. 176E-01 9.57lE+OO • * +
3. 999E-01 1. 455E+<l1 * +
5.034E-01 1. 91lE+<l1 . * +
6. 338E-01 2.345E+<l1 • * +
7. 979E-01 2.766E+<l1 • * +
1.004E+OO 3. 179E+<l1 * +
1.265E+OO 3. 587E+<l1 * +
1. 592E+OO 3. 992E+<l1 * +
W orked example 14.13
Using SPIeE software, write a program which calculates the impedance of

the se ries RLC circuit in worked example 12.3.1 (see page 303), in 10 per
cent steps, from 20 per cent below to 20 per cent above resonance.
Solution
Not every implementation of SPIeE provides simple facilities for determi-

nation of impedance, but we can get around this by driving the circuit with
an alternating current of 1LO° A. The complex value of voltage between
the current source terminals is then equal to the complex input impedance
of the circuit. An input file for the circuit in worked example 12.3.1 is given
in table 14.25.
To remind the reader, the series circuit contains R = 4 Q, L = 0.1 mH
and C = 1 IlF. The resonant frequency is 100 000 radis, which we take to
be 15 920 Hz (to an accuracy of four decimal places).
SPIeE detects the capacitor as an open-circuit, and it is necessary to
include an additional resistor, 'Rextra', to shunt the current source in order
to provide a 'd.c.' path between node 1 and node O. The value of 'Rextra' is
sufficiently high to have no effect on our calculation.
The '.AC' control line causes the frequency to sweep from 12.73 kHz
to 19.1 kHz, in five linear steps. This provides us with data at (approx.)
80 krad/s to 100 krad/s in steps of 10 krad/s, which will include the resonant
frequency of 100 krad/s.
Table 14.25

*** Input impedance calculation ***
*** Circuit is driven by a constant current ***
I o 1 AC 1
R 1 2 4
L 2 3 O.lM
C 3 0 1U
** Rextra shunts the current source **
Rext ra 0 1 1MEG
.OPTIONS NOPAGE
** Calculate values in 10 per cent steps around resonance **
* Five LINEAR steps between 80 k radis to 100 k radis *
.AC LIN 5 12.73K 19.1K
** Input impedance is equal to V(l) **
.PRINT AC VM(l) VP (1)
• END
Tab1e 14.26
FREQ VM(1) VP (1)
1. 273E+04 6.024E+OO -4.839E+01
1.432E+04 4.524E+OO -2.785E+01
1. 592E+04 4.000E+OO -8.898E-03
1.751E+04 4.432E+OO 2.552E+01
1. 910E+04 5.427E+OO 4.252E+01
The '.PRINT' control line requests the magnitude (VM(l» and the
phase (VP(l» of the voltage across the current source to be output. These
values are, incidentally, the magnitude and the phase of impedance of the
circuit at the respective frequencies.
The reader will note that the results in table 14.26 (wh ich is the relevant
part of the output file) at 1.432E+04 Hz (approx. 80000 radis) and
1.91OE+04 Hz (approx. 120 000 radis) generally agree with the results
obtained in worked example 12.3.3. The impedance at the re sonant fre-
quency of 1592 Hz (100000 radis) is seen to be 4 Q.
Solve the Fourier analysis problem in worked example 13.3.1 (see page
329) using SPIeE software.
Solution
Fourier analysis (that is, a '.FOUR' controlline appears in the input file in
table 14.27) can be performed by SPIeE in conjunction with a transient
('.TRAN') analysis of the waveshape; that is, a '.TRAN' controlline must
appear in the same input file as a '.FOUR' control line.
SPIeE performs FOURier analysis on waveforms which exist for t > 0,
that is, the va lues exist only in positive time. The rectangular waveform is
therefore described in this case for the time interval 0 < t< 1 s by means of
a piece-wise linear stimulus 'Vpwl', which is applied to a resistor R of value
1 Q. Each point on the waveform is defined by two values, namely by its
time in seconds and its value, as follows (see also worked example 14.4)
PWL(O,l 0.25,1 0.25001, -1 0.75, -1 0.75001,1 1,1)
That is, at t = 0 the voltage has a value 1 V, and remains at that value until
t = 0.25 s. SPIeE only accepts a practical waveform, that is it must have
a practical value of fall time from 1 V to -1 V. A fall time of 0.01 ms
has been assumed; similarly, a rise time of 0.01 ms has been adopted at
t = 0.75 ms. The points specified on the waveform are joined linearly. That
is to say, the waveform is not an ideal square wave, but is trapezoidal.

The transient response is computed by the .TRAN controlline, which
requests the computer to evaluate the transient response every millisecond
for aperiod of 1 s (this period is also referred to later).
The .FOUR control line causes the computer to evaluate the d.c.
component together with the amplitude and phase of the fundamental
frequency and all harmonics up to the 9th harmonic. The fundamental
frequency is specified by the user, and is the first value in the .FOUR
controlline, that is, it is 1 Hz. Also written in this line is the output variable
on which the Fourier analysis is to be performed, namely V(I). There can
be several output variables in this line, but we only need one. The reader
is asked to note that the transient analysis period must be at least
1/(fundamental frequency) s long; in our case this is 1/1 = 1 s.
The relevant part of the output file is listed in table 14.28. The reader
will note that SPIeE reports a sm all d.c. component, which is due to the
fact that the PWL waveform used in the program is not a true square wave,
together with any mathematical 'noise' produced by the program. The
table produced by the program lists not only each Fourier component and
its associated phase angle, but also the 'normalized' values. The latter
assurnes that the magnitude of the fundamental frequency is 1.000, and
that its phase shift is 0.000°.
One needs to look carefully at the table in order to decide wh at values
are significant; in practice we can ignore any harmonie with a magnitude
less than about 1 per cent of the normalised fundamental frequency
component. Applying this as a general rule, we can ignore all the even
harmonics. Looking at the PHASE (DEG) column, we see that the phase
shift of the fundamental frequency is 90°, that is, it is a cosine term in the
Fourier series. Similarly, the third harmonic is a -cos term, the fifth
harmonic is a cos term, etc. The results in table 14.18 are in general
agreement with the results of worked example 13.3.1.
Once again, it is pointed out that a computer package may not provide
all the answers needed for a particular problem. This example can be used
to demonstrate that engineering judgement is often needed to make a final
decision. If the .TRAN control line had been written '. TRAN 10M 1',
that is, transient values to be calculated at 10 ms intervals rather than the
1 ms period used here, the phase angle values would have differed by many
degrees from the values in table 14.28, and the result would not have
agreed very weB with the theoretical analysis. The result would, of course,
be produced much more quickly, but not so accurately!
Table 14.27
Worked Example 14.14

*** Fourier analysis of a square wave ***
*** Voltage is a Piece-Wise Linear (PWL) source ***
Vpwl 1 0 PWL(0,1 0.25,1 0.25001,-1 0.75,-1 0.75001,1 1,1)
R 1 0 1
.OPTIONS NOPAGE
*TRANsient analysis
* I 1 InS steps
* I I up to 1 s
* I I I
.TRAN 1Ms 1s
*FOURier analysis
* 11Hz fundamental frequency
* I analyse the voltage at node 1
* I I
.FOUR 1 V(l)
• END
Table 14.28
roJRIER ~ CF 'lRI\NSIENl' RESPCNSE V(l)

OC a:M'CNENI' = -l.OOlOOlE-03
HAIMNIC ~ roJRIER NCFMALIZED PHASE NCFMALIZED
~ (HZ) a:M'CNENI' a:M'CNENI' (Dm) PHASE (Dm)
1 1.000E+OO 1. 273E+OO 1.000E+OO 9.000E+Ol O.OOOE+OO

2 2.000E+OO 2.002E-D3 1. 572E-D3 9.000E+01 -1.40lE-DS
3 3.000E+OO 4.244E-01 3. 333E-D1 -9.000E+01 -1.800E+02
4 4.000E+OO 2.002E-D3 1. 572E-D3 -9.000E+01 -1.800E+02
5 5.000E+OO 2.547E-D1 2.000E-01 9.000E+01 1.868E-DS
6 6.000E+OO 2.002E-D3 1. 572E-D3 9.000E+01 -3.268E-D8
7 7.000E+OO 1. 819E-01 1. 429E-D1 -9.000E+01 -1.800E+02
8 8.000E+OO 2.002E-03 1. 572E-03 -9.000E+01 -1.800E+02
9 9.000E+OO 1. 415E-D1 1. lllE-D1 9.000E+01 3. 735E-D8
'IUl'AL HARM:mC DIS'I\:RI'ICN = 4.288100E+01 PERCENT
15
Complex Numbers,
Matrices, Determinants and
Partial Fractions
15.1 Imaginary numbers
The concept of imaginary numbers was introduced by mathematicians to

allow them to express the square root of a negative number. For example,
if x 2 = -9 the solution is given by saying
x = imaginary operator x \19
The imaginary number is useful to electrical engineers, who gave the
imaginary operator the symbol j so that, in the above case
x = j3
(Mathematicians use the symbol i to represent the imaginary operator, but
as this symbol is used by electrical engineers to represent electrical current,
the symbol j is used.)
Thus
= V(-l)
j2= (V(-1»2=-1
p = j X j2 = -j
j 4 = j 2 X j 2 = 1, etc.
The reader should note that
imaginary number = imaginary operator (j) X real number
386
Complex Numbers, Matrices, Determinants and Partial Fractions 387
15.2 Complex numbers
A complex number is the sum of areal number and an imaginary number,

and is either written in bold Roman type (as in this book), or has a bar
drawn over it (as is often the ease in hand-written material). Thus
V = a + jb
or
v = a + jb
It is important to note that both a and b are real numbers, but the
eomponent jb is an imaginary number sinee b is multiplied by the im-
aginary operator j.
15.3 Representation of complex numbers
There are four ways of representing a eomplex number, namely

reet angular or eartesian form V = a + jb
polar form V = r L (J
exponential form V = r X eiß
trigonometrie form V = r(eos (J + j sin (J)

The last three forms above are, from Euler's identity, generally the
same form. The relationship between them is
r = V(a 2 + b 2 ) (J = tan -l(b/a)
The reetangular and polar forms are most widely used in eleetrieal
engineering. The seven eomplex numbers V 1 to V, represented in figure
15.1 are written in polar and reetangular form as follows.
Vl = 3 = 3 + jO = 3LO°
V z = 4 + j2 = 4.47L26.57°
V 3 = j3 = 0 + j3 = 3L 90°
V 4 = -2 + j4 = 4.47L 116.57°
V s = -2 = -2 + jO = 2L180° (or 2L-1800)
V 6 = -4 - j3 = 5L216.87° (or 5L-143.13°)
V, =1 - j3 = 3.16L-71.57°
V2
Figure 15.1 Representatian a[ camp/ex numbers.
15.4 Conjugate of a complex number
The conjugate, V*, of the complex number V = a + jb = r L8 is
V*=a-jb=rL-8
where r = V(a 2 + b 2 ) and 8 = tan ~I(b/a).
If V = 3 + j4 = 5L53.13°, then V* = 3 - j4 = 5 L - 53.13°. Also
(V*)* = V. A complex number and its conjugate are said to form a
conjugate complex pair of numbers. Other useful praperties are
(V,V 2 )* = VW~
(V, ± V 2 )* = V~± V~
V*1
V*2
15.5 Sum and difference of complex numbers
To perform addition (or subtraction) of complex numbers, the numbers

must first be converted into rectangular form, and the real parts must be
added together (or subtracted fram one another), and then the imaginary
parts added together (or subtracted fram one another) as folIows.
IfV, = 5 + j6 and V 2 = -3 - j8, then
VI + V 2 = (5 + j6) + (-3 - j8) = 2 - j2
V, - V 2 = (5 + j6) - (-3 - j8) = 8 + j14
15.6 Multiplication of complex numbers
When the numbers are expressed in polar form, the multiplieation pro-
eeeds as follows
r 1 L8 1 X r 2 L8 2 = r 1r 2 L(8 1 + ( 2 )
In general, it is far more eonvenient to earry out multiplication of
eomplex numbers using the polar form than it is using rectangular form.
However, when numbers are expressed in reet angular form, the multipliea-
tion ean be earried out as follows.
= (ac + j 2bd) + j(ad + bc)
(a + jb)(c + jd)
= (ac - bd) + j(ad + bc)
If VI = 3 + j4 = 5L53.13° and V 2 = -5 - j5 = 7.07L -135°, then

VI x V 2 = 5L53.13° x 7.07L-135° = 35.35L-81.87°
or
VI X V 2 = (3 + j4)(-5 - j5) = (-15 + 20) + j(-15 - 20) = 5 - j35
The product of a conjugate complex pair of reet angular numbers is
(a + jb)(a - jb) = a 2 + b 2
and the product of a conjugate complex pair of polar numbers is
rL8 x rL-8 = r 2L(8 -- 8) = r 2
both giving areal number and no imaginary term. For example, if V
= 3 + j4 = 5L53.13°, then
V x V* = (3 + j4)(3 - j4) = 9 + 16 = 25
or
V x V* = 5L53.13° x 5L-53.13° = 25LO°
15.7 Division of complex numbers
If V and I are eomplex numbers then, using polar values, division is earried
out as follows
V r 1 L8 r1
- = = - L (8 - cj»
I r 2 Lcj> r2
In general, it is far more eonvenient to use polar values rather than

rectangular values when performing division.
Using rectangular complex values, division is carried out as follows

V V X 1*
1 1 X 1*
The product 1 X 1* gives areal number with no imaginary part (see
section 15.6), and the process of dividing by (I X 1*) is known as rationalis-
ing the denominator. For example
if + jb and 1 = c + jd, then
V = a
V (a + jb)(c - jd) (ac + bd) + j(bc - ad)
1 (c + jd)(c - jd) c2 + d 2
SlIppose V = 4 + j4 = 5.66L45° and 1 = 3 + j4 = 5L53.13°, then
V 5.66L45°
- - - - = 1.132L-8.13°
1 5L53.13°
or
V (4 + j4)(3 - j4) (12 + 16) + j(12 - 16)
1 (3 + j4)(3 - j4) 9 + 16
28 - j4 .
----'-- = 1.12 - JO.16
25
15.8 Powers and roots of complex numbers
The Nth power of a complex number is calculated as follows

(RL8Y = RN L N8
For example
(2L600)3 = 8L180°
There are N roots of the complex number RL8 as follows
(RL 8) lIN = R lIN I 8 + [nN X 360]
where n has a value in the range 0, 1,2, .. , (N - 1).

For example -{I(8L600) has the values 2L20°, 2L140° and
2L260°.
15.9 Matrix representation
Consider the equations

VI = Z 1.1 1 + Z 12/l + Z 13/3
V l = ZU/ l + Zn/l + Z13 / 3
V 3 = Z 31/1 + Z 3l/l + Z 33 1 3
These can be written in a matrix form of ordered rows and columns of
elements of the same kind as follows
or, alternatively, in the even more ordered form

[V] = [Z][/]
where [V] is a voltage matrix, [Z] is an impedance matrix and [I] is a
current matrix. The formalised matrix representation is weIl suited to
calculator and computer solution of equations.
The double subscript notation is used to identify the position of an
element within a matrix. The first subscript denotes the row in which the
element is found, and the second denotes the column. Thus Zl3 is the
element in the second row of the third column. A simple mnemonic to
remember the order of the subscripts is Roman Catholic (Rows, Columns).
Both the voltage and the current matrix, above, are written in the form
of a column, and each is described as a column matrix or vector. These
have only one column, but may contain any number of rows (in the case
considered, both have the same length). Since there is only one column in a
column matrix, only one subscript is necessary to define the position of an
element within it.
The impedance matrix is a square matrix, in which the number of rows is
equal to the number of columns.
The major diagonal of a square matrix is the diagonal line of elements
going from the top leftmost element to the bottom rightmost element, that-
is, the elements Z 11' Z ll' Z 33 in the impedance matrix above lie on the
major diagonal.
A diagonal matrix is a rectangular matrix in which all the elements are
zero except those on the major diagonal.
An identity matrix or unit matrix is a diagonal matrix in which the value
of each of the elements on the major diagonal is unity.
Other forms include the following. A row matrix has only one row
containing number of columns as follows
A rectangular matrix is one having M rows and N columns (M";: N), and
is described as an M by N or M x N matrix, as shown below.
all a 12 a lN
a 21 a 22 a 2N
aMI a M2 a MN
A null matrix is one in wh ich every element is zero.
15.10 Matrix addition and subtraction
Two matrices can either be added together or subtracted from one another
if they are of the same order. If A = [aij] and B = [biJ are two M x N
matrices, their sum (or difference) is the matrix C = [c ij]' where each
element of C is the sum (or difference) of the corresponding elements of A
and B. That is
A ± B = [a ij ± b ij]
~ ] and B = [: -:]
[ -3 4 4
IfA = 0 then
-6 -7
1~
[-3 + 3 4+4 8
A+B= 5-5]=[0
0+6 -6 -7 7+ 8 6 -13 ]
A-B=
[-3 - 3 4-4 5 - (-5) ] = [ -6 0
10]
0-6 -6 - (-7) 7- 8 -6 1 -1
15.11 Matrix multiplication
The matrix product AB (which must be carried out in that order) can be
computed only if the number of columns in A is equal to the number of
rows in B. B is not necessarily conformable to A for multiplication, that is,
the product BA may not be defined. The following should be observed:
1. AB ,,;: BA generally .
2. AB = 0 does not always imply A = 0 or B = O.
3. AB = AC does not always imply B = C.
If matrices are conformable, multiplication is performed on a row by

column basis; each element in a row is multiplied by the corresponding
element of a column, and the products are summed.
If A is a 1 x M row matrix, and B is a M x 1 column matrix, then
= [ k ~ 1 a Ikb kl ]
Note: BA is not defined.

-6
Suppose that A = [-1 2 3] and B = 0 then
7
AB ~ [-1 2 3] -~]~ [-1(-6) + 2(0) + 3(7)] ~ [27]

a ll a 12 b l1 b 12
If A = a 21 a 22 and B = then
a 31 a 32
a ll b 11 + a l2 b 21 a l1 b 12 + a 12b 22
AB = a 2l b 11 + a 22b 21 a 21 b l2 + a 22b 22
a 3l b 11 + a 32b 21 a 31 b l2 + a 32 b 22
Note: BA is not defined.
-1 0 4 5
Suppose that A = 3 -4 andB = then
5 6 -6 7
AB =
-1(4) + 0(-6) -1(5) + 0(7)
3(4) + (-4)(-6) 3(5) + (-4)(7) =
-4
36 -5]
-13
5(4) + 6(-6) 5(5) + 6(7) -16 67
5 -3 0 I.
Also, if R = -3 12 -5 andl = 12 then
0 -5 11 13
5/ 1 -3/z -0/ 3
RI = -3/ 1 12/ 2 -5/ 3
-011 -5/z 111 3
A matrix may be multiplied by a scalar k (which should not be confused

with the 1 x 1 matrix [k]), to give
kA = Ak = [ka;j]
that is, each element in the matrix is multiplied by k.
15.12 The determinant of a square matrix
A matrix is simply an ordered array of elements, and has no numerical

value. On the other hand, the determinant of a square matrix has a
numerical value, which is given the symbol ~, or det A or lAI. This value
can be used in the computation of the value of unknown variables in the
equations represented by the matrix equations.
The value of a determinant of order 2 is calculated as folIows.
For example
I-~ 41
-3
= 2(-3) - 4(-5) = 14
For a determinant of order 3
all a 12 a \3
a 21 a 22 a 23 = alla2Za33 + alZaZ2a31 + a\3a 2.a 32 - a\3a 22a 31 - a12a21a33
a 31 a 32 a 33 -allaZ3a3Z
For example
2 -3 4 2.6.7 + (-3).( -2).( -3) + 4.5.(-4)

5 6 -2 = -4.6.(-3) - (-3).5.7 - 2.(-2).(-4) = -63
-3 -4 7
15.13 Minors and cofactors
The minor of the element a ij (row i, columnj) of a determinant is obtained

by deleting row i and column j of the determinant; the minor of this
element is given the symbol M ij' The value of the minor is multiplied by
( -1) i + j to give the cofactor <;>f a ij; the cofactor is given the symbol ~ ij'
For example, in the determinant of order 3 in section 15.12
all
= I a 31 aa 33 1
13
M 22
and
a l1 a 13 a l1 a 13
~22 = (_1)2+21 a 31 a 33 1=+ I a 31 a 33 1=
+ a Ha 33 - a 13a 31
15.14 Evaluating a determinant
The value of a determinant of order N is the sum of the N products of each

element in a selected row (or column) and its cofactor (great care should be
taken in ensuring that the cofactor has the correct mathematical sign, see
section 15.13).
Consider the following determinant of order 3, which can be evaluated
by (for example) selecting the elements in the first row as follows
a 11 a 12 a 13
a 21 a 22 a 23 = al1~l1 + aI2~12 + a13~13
a 31 a 32 a 33
= a 11 I a 22 a 23 1_ a 12 1 a 21 a 23 1+ a 13 1
a 32 a 33 a 31 a 33
Alternatively, if the elements in the second column are selected, then

the value of the determinant is calculated from
a 12~ 12 + a22~22 + a32~32
Both of the above calculations produce the result given earlier in section
15.12.
15.15 The rule of Sarrus
A determinant of order 3 can be evaluated using the rule of Sarrus as

follows
The determinant is written down, and the first two columns are repeated
to the right of the determinant. Diagonal lines are drawn joining sets of
three elements together; the product of the diagonally downwards terms
are given a positive sign, and the product of the diagonally upwards terms
are given a negative sign. The value of the determinant is the sum of these
products. For example, by the rule of Sarrus
5 6 7
2 -3 4 = 5.(-3).3 + 6.4.1. + 7.2.(-2) -7.(-3).1- 6.2.3 - 5.4.(-2)
1 -2 3 = -45 + 24 - 28 + 21 - 36 + .40 = -24
15.16 Cramer's rule
Linear simultaneous equations can be solved by Cramer's rule as follows.

Consider the following matrix form of equation
a iN
a ZN
aMI a MZ a MN
The value of x K in the Kth row is obtained from the computations
a I(K - 1) Y 1 a I(K + 1) a iN
aZ(K-I) Yz aZ(K+I) a ZN
aMI aM(K-I) YM aZ(K+ 1) a MN

For example, solve for 1 2 in the following

5 = 101 1 - 31 2 - 51 3
10 = -31 1 + 71 2 - 41 3
-15 = -51 1 - 41 2 + 91 3
The matrix form of the equation is
5
10 = [~~ -~ ~
-15 -5 -4 9
From Cramer's rule
10 5 -5 10 -3 -5
12 =
and using the rule of Sarrus

-3 10 -4
-5 -15 9 I -3 7 -4
-5 -4 9
10.10.9 + 5.(-4).(-5) + (-5).(-3).(-15) - (-5).10.(-5) - 5.(-3).9 - 10.(-4).(-15)

12 =----------------------------------------------------
10.7.9 + (-3).(-4).(-5) + (-5).(-3).(-4) - (-5).7.(-5) - (-3).(-3).9 -10.(-4).(-4)
60
= - = 0.638
94
15.17 Matrices and determinants containing complex numbers
Matrices and determinants containing complex numbers can be handled by

the methods described above, but the reader must use the methods·
described in sections 15.5 to 15.8 when dealing with complex numbers.
15.18 Partial fractions
Functions of s can be transformed directly using integral calculus but, in

practice, it is easier to arrange the functions so that they fit one or more of
the terms in a table of Laplace transforms. One method of doing this is by
the use of partial fractions.
Analysis of electrical problems by the Laplace transform method gener-
ally requires the derivation of the inverse transform from a result which,
usually, is in the form of the ratio of two polynomials in s.
If
N(s)
F(s) = - -
D(s)
where N(s) and D(s) are polynomials in s, and the degree of N(s) is less
than the degree of D(s), then:
1. For every linear factor (As + B) in D(s) there is a corresponding partial
fraction
1
+B As
2. For every quadratic factor (As 2 + Bs + C)2 in D(s) which has real roots,
there is a corresponding partial fraction
Ps +Q
As 2 + Bs + C
3. For every repeated factor (As + B)2 in D(s) there is a corresponding
partial fraction
P Q
---+----
As + B (As + B)2
4. For every repeated quadratic factor (As 2 + Bs + C)2 in D(s) there is a
corresponding partial fraction
Ps + Q Rs + T
------+ ------
As 2 + Bs +C (As 2 + Bs + C)2
5. For every thrice-repeated character (As + B)3 in D(s) there is a
corresponding partial fraction
P Q R
---+ +----
As + B (As + B) 2 (As + B) 3
6. For every cubic factor As 3 + Bs 2 + Cs + D in D(s) there is a corre-
sponding partial fraction
Ps 2 + Qs + R
Worked example 15.1
Determine the partial fraction expansion of

200
ses + 1)(s + 2)
Solution
Applying the mIes laid down above

200 ABC
------=-+-'-+--
s(s + l)(s + 2) s s +1 s +2
A, Band C in the above expression are known as the residues.
The residues are quickly evaluated in this case by the cover-up rule as
folIows.
(i) Determine the value of s which makes the denominator of that particu-
lar term zero.
(ii) Substitute this value into the full expression (both numerator and
denominator) and, ignoring or 'covering up' the factor in question, the
residue is the result of the calculation.
The residue A is calculated by letting s = O. Substituting this value in the
original equation whilst 'covering up' the factor s gives
200 200
A=------ - - - - - - = 100
(s + l)(s + 2) s = 0
«0) + 1)«0) + 2)
The residue B is calculated by letting s = - 1, as follows
200 200
B = -200
s(s + 2) s = -1
(-1)«-1) + 2)
and the residue C is calculated by letting s = - 2 as follows
200 200
C=--- - - - - - = 100
s(s + 1) s = -2
(-2)« -2) + 1)
That is
200 100 200 100
------= -- - - - +--
s(s + l)(s + 2) s s +1 s+2
Worked example 15.2
Determine the partial fraction expansion of

54
Solution
This can be re-written in the form

54 54
(4s 2 - 5s + 1)2 (1 - s)2(1 - 4S)2
Since there are repeated factors in the denominator , the partial fraction
expansion is written as follows
ABC D
--+ +---+----
1- s (1 - S)2 1 - 4s (1 - 4S)2
Since (1 - s)2 and (1 - 4S)2 appear in the original equation, we can
evaluate the residues Band D by the cover-up rule as follows
54 54
B=--- ----=6
(1 - 4S)2 (1 - 4(1»2
s = 1
54 54
D=--- ----=96
(1-S)2s~O.25 (1 - 0.25)2
The residues A and C are calculated by substituting the known residues,

and multiplying both sides of the equation by the denominator of the
polynomial as follows
54 = A(1 - s)(1 - 4s)2 + 6(1 - 4S)2 + C(1 - s)2(1 - 4s) + 96(1 - S)2
Equating coefficients of S3 gives -16A - 4C = 0, and equating co-
efficients of S2 gives 8A + 3C = -64. Solving for A and C yields A = 16,
C = -64. That is
54 16 6 64 96
------ = -- + - --- +----
(4s 2 -5s+1)2 1-s (1-S)2 1-4s (1-4s)2
Bibliography
There are many excellent texts in the field of circuit analysis, the following
being a representative selection. In particular, the attention ofthe reader is
directed towards those marked with a *; these texts contain introductory
material on the computer program SPICE.
1. W. Banzhaf, Computer-Aided Circuit Analysis using SPICE, Pren-
tice-Hall, 1989*
2. L. S. Bobrow, Elementary Linear Circuit Analysis, 2nd edn, Holt,
Rinehart and Winston, 1987*
3. S. A. Boctor, Electric Circuit Analysis, Prentice-Hall, 1987*
4. R. L. Boylestad, Introductory Circuit Analysis, 5th edn, Merrill Pub-
lishing Co., 1987
5. M. D. Ciletti, Introduction to Circuit Analysis and Design, Holt,
Rinehart and Winston, 1988. Supplement available on SPICE
6. T. L. Floyd, Principles 01 Electric Circuits, 2nd edn, Merrill Publishing
Co., 1985
7. J. D. Irwin, Basic Engineering Circuit Analysis, 2nd edn, Macmillan,
1987*
8. M. C. Kelly and B. Nichols, Introductory Linear Electrical Circuits and
Electronics, John Wiley, 1988
9. N. M. Morris and F. W. Senior, Electric Circuits, Macmillan WORK
OUT series, 1991 *
10. J. W. Nilsson, Electric Circuits, 3rd edn, Addison-Wesley, 1990*
11. P. W. Tuinenga, SPICE - A Guide to Circuit Simulation & Analysis
Using PSpice, Prentice-Hall, 1988*
PSpice software suitable for personal computers both in full and edu-
cational versions are available from ARS Microsystems, Herriard Business
Centre, Alton Road, Herriard, Basingstoke, Hampshire. The educational
version is available at low cost from many Shareware suppliers.
401
Index
ABCD- parameters 186-9 Conduetor 5
Active element 8 Conjugate complex pair 388
Admittanee 111 Conjugate of complex number 388
Admittance parameters 176-80 Controlled source 8
Alternating eurrent (a.e.) 100 Corner frequeney 284
Angular frequeney 101, 102 Cotree 50
a-parameters 186-9 link in 50
Apparent power 116 Cramer's rule 396, 397
Average value 100 Critical damping 225, 252
Current 1
Balanced polyphase load 147 Current division in a parallel
Balanced polyphase supply 147, 148 cireuit 22
Band-pass filter 318 Current gain 175
Bandwidth 305,311 Current source 8
BASIC programs 36, 129, 130 Cut-off frequency 273
Bilateral element 30
Bilateral network 30 Damping, eritieal 225,252
Bode diagram 280-95 Damping faetor 225
Braneh 29 Decibel 281
Delta eonneetion 150, 155--60
Capacitive reactance 110 Delta funetion 229
Capacitor 79 Delta-star transformation 74, 136,
Capaeitors in parallel 82 137
Capacitors in series 82, 83 Dependent souree 8, 138-40
voltage division between 83,84 Determinant 394
Cartesian· form of complex Distributed-eonstant element 29
number 387 Dot notation 92, 93
Cireuit 28 Double-energy cireuit 218
CIVIL, mnemonie 110 Double-suffix eonvention 14
Coefficient of magnetie coupling 93 Dual 54
Complex eonjugate 112 Duality 54, 90
Complex frequeney 266, 267 Dynamie impedance 309
Complex number 106, 267, 387 Dynamie resistanee 309
conjugate 388
manipulation of 388-90
Complex power 121 Effeetive value 102
Complex value 106 e.m.f. 3
Complex wave 325 Energy 7
circuit response to 338-40 in eapacitor 81, 82
power factor of 341 in induetor 67
power supplied by 340,341 Excitation eurrent 85
Conduetance 6 Exponential form of complex
Conduetance matrix 43 number 387
402
Index 403
Filter Laplace transform 226, 227

active 277, 278, 320 final value theorem 241
band-pass 318 initial value theorem 241
band-stop 319 of periodic function 239
Final value theorem 241 properties of 237-41
First-order circuit 218 Laplace transform pairs, table of 233
Fourier series 326 Line spectrum 335-7
Frequency 102 Line voltage 147
angular 101, 102 Linear circuit element 4
Frequency response 272 Linearity 61
Frequency scaling 316, 317 Link, in cotree 50
Fundamental frequency 325 Loop, in circuit 29
Loop analysis 49, 52-4, 130
Generalised impedance convertor Lumped-constant element 29
(GIe) 279, 280
Magnetic coupling coefficient 93
Half-power frequency 273, 305 Magnetically coupled circuit, input
Half-power point 273, 305 impedance of 212
Harmonic 325 Magnitude scaling 315-17
Harmonic analysis 343-5 Matrix 33, 391-4
h-parameters 183-6 conductance 43
Hybrid parameters 183-6 determinant of 394-7
cofactor of 395
minor of 395
Ideal element 5 manipulation of 392-4
Ideal source 8 resistance 33
Ideal transformer 198-200, 207 square 33
Imaginary number 386 Maximum power transfer
Imaginary operator 386 theorem 69,70,137,138
Impedance 107-10 Mean value 100
Impedance (z) parameters 180-3 Mesh 30
Impulse function 229 Mesh analysis 31-4, 37-41, 130, 132
Independentsource 8 Mesh connection 150
Inductance 84 Mesh current 31
mutual 91 Millman's theorem 70-2, 135, 136
self 85 Minor, of deterrninant 395
Inductive reactance 109 Mutual inductance 91
Inductors
parallel 88, 89 Negative phase sequence 146, 164
series 87, 88 Network 28
Initial ~ondition generator 242, 243 Network topology 49--54
Initial value theorem 241 Neutral conductor 152
Input adrnittance 172 Neutral point 146
Input impedlUlce 172 Nodal analysis 41-9, 130, 131
Instantaneous value 1, 2 Node 29
Insulator 5 Non-linear element 5
Internal conductance 17 Non-planar network 30
Internal resistance 16 Norton's theorem 65-7, 133-5
lunction 29 Ohm's law 5

Open-circuit 28
Kirchhoff's laws 11-14 Operational amplifier 9, 277-80
404 Index
Output conductance 17 Q (quality) factor 302, 310

Output impedance 173
Output resistance 16 Ramp function 230
Overdamped response 225,252 Reactance
capacitive 110
Parallel circuit 21, 113 inductive 109
current division in 22 Reactive power 116
Parallel generator theorem 70-2, 135 Real part 387
Parallel resonance 308 Reciprocity theorem 67, 68
Parameters Rectangular form of complex
ABCD 186-9 number 387
h 183-6 Rectangular pulse train, Laplace
t 186-9 transform of 239
transformation between 189-91 Reference direction 104
y 176-80 Reference node 14
z 180-3 Reftected impedance 212
Partial fractions 397-400 Resistance 5
Pass-band 319 Resistance matrix 33
Passive element 4 Resonance 299-322
Path, in circuit 29 magnetically coupled circuit 321
Periodic time 102 parallel 308-15
Periodic waveform 325 selective 320
Phase angle 103-5 series 300-8
Phase sequence 146 Resonant frequency 274, 276, 290,
Phase voltage 147 300, 308, 313
Phasor 105 Root-mean-square (Lm.s.)
Phasor diagram 105 value 102, 103
Pi-tee transformation 74,75 Rosen's theorem 72, 73, 136
Planar network 30
non- 30 Sarrus, rule of 396
Polar form of complex number 387 Scaling 315
Pole 274 frequency 316, 317
Polyphase circuit 144 magnitude 315-17
Port 171 Second-order circuit 218
Positive phase sequence 146, 164 Secondary winding 91
Potential difference 3 Selective resonance 320
Power 4, 6, 115, 118-21 Se1ectivity 306, 311
apparent 116 Self-inductance 85
average 4, 102, 115 Semiconductor 5
complex 121 Series circuit 19, 112
instantaneous 102 voltage division in 20
reactive 116 Series resonance 300
symmetrical component 166, 167 Short-circuit 29
three-phase 160-4 Simultaneous equations, so~ution
Power factor 116 of 34-6, 128-30
Power factor of complex wave 341 Single-energy circuit 218
Power gain 175 Source
Power supplied by complex wave 340 ideal 8
Power triangle 116 practical 16
Practical current source 8, 17 transformation of 17
Practical element 5 Source conductance 17
Practical voltage source 8, 16 Source resistance 16
Primary winding 91 SPIeE 280, 349-85
Index 405
circuit description using 351 Transient analysis 218-62
element description using 351-7 classical solution
programming example using first-order 219-23
357-85 second-order 224-6
s-plane 268-9 Laplace transform solution
Star connection 145, 147-55, 159, first-order 244-51
160 second-order 251-7
Star point 146 magnetically coupled circuit 258-62
Star-delta transformation 73, 74, 136 Transmission parameters 186-9
Star-mesh transformation, general Tree, in network 50
72-3 Trigonometric form of complex
Steady-state value 2 number 387
Stop band 319 Triple-n harmonics 343
Superconductor 5 Tuned coupled circuits 321
Supermesh 38 Turns ratio 200
Supernode 46 Two-port network 67, 171-94
Superposition principle 61-3, 132, interconnection between 191-4
133 transformer as 207
Susceptance 111 Two-wattmeter method of power
Symmetrical components 164-7 measurement 163, 164
Symmetry, in waveform 332
Underdamped natural frequency 225
Tank circuit 309 Unit-impulse function (delta
Tee-pi transformation 73 function) 229
Tee-wye transformation 73 Unit-ramp function 230, 231
Thevenin's theorem 63-5,67, 133, Unit-step function 227-9
134 Universal resonant circuit 317
Three-phase system 144-67
Time constant 223
Time domain 106 VA (volt-amperes) 117
V Ar (volt-amperes reactive) 117
t-parameters 186-9
Transfer function 257, 258, 276-80 Voltage 3
Transformer 198 Voltage division in aseries circuit 20
Voltage gain 174
core construction of 203
equivalent network of 203 Voltage source 8
ideal 198-200, 207 dependent 8
impedance matching with 204 ideal 8
independent 8
input impedance of 212
leakage flux of 203 practical 63
leakage reactance of 204
linear 211-15 Waveform symmetry 332-5
phasor diagram of 201-12 Wye-delta transformation 74, 75
n-equivalent circuit of 214
reflected impedance of 212 Y-connection 145
shell construction of 203 y-parameters 176-80
T-equivalent circuit of 214
transient response of 258 Zero, on s-plane 274
turns ratio of 200 Zero phase sequence component 164
voltage regulation of 208, 209 z-parameters 180-3

Electrical Circuit Analysis and Design by Noel M. Morris

Uploaded by

Copyright:

Available Formats

Electrical Circuit Analysis and Design by Noel M. Morris

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Electrical Circuit Analysis and Design by Noel M. Morris

Uploaded by

Copyright:

Available Formats

Electrical Circuit Analysis and Design

All rights reserved. No reproduction, copy or transmission of

First published 1993 by

ISBN 978-0-333-55483-8 ISBN 978-1-349-22560-6 (eBook)

A catalogue record for this book is available

1 Elements and Laws 1

2.4 General mIes for writing mesh equations 34

4 Energy Storage Elements 79

4.13 Mutual inductance 91

5 The Sinewave, Phasors and Power 100

6 Sinusoidal Steady-state Analysis 126

7 Polyphase Circuits 144

8 Two-port Networks 171

8.8 Interconnection between two-port networks 191

9 The Transformer 198

10 Transient Solution of Electrical Circuits 218

11 Complex Frequency, the s-Plane and Bode Diagrams 266

11.6 Frequency response as a function of w 271

13 Harmonics and Fourier Analysis 325

In order to understand circuit analysis fully, the reader should have a

impedance into its equivalent s-domain impedance is covered. Frequency

1.2 Electric current

The diagrams in figure 1.1 illustrate the concept of movement of charge

Figure 1.2 Representation 01 current on a circuit diagram.

1.3 Voltage, e.m.f. and p.d.

The voltage 'between' or 'across' a pair of terminals is a measure of the

(a) (b) (c) (d)

When an arrow is used, the terminal which the arrowhead points to

1.4 Power in d.c. circuits

1.5 Linear passive circuit elements

1.6 Resistance, conductance and Ohm's law

Figure 1.5 Ohm's law.

1. 7 Power in a resistive circuit

If v = 10 V and i = -15 mA, calculate the power consumed.

p = 10 X (-15 X 10- 3 ) = -0.15 W

Worked example 1.7.2

If P = 200 mWand I = 10 mA, calculate the resistance of the circuit.

R = P/J2 = 200 x 10- 3/(10 X 10-3 )2 = 2000 Q

A current of 0.1 A produces 5 W of power in a circuit. Calculate the

G = P/P = 0.1 2/5 = 0.002 S

1.8 Energy consumed in a resistive circuit

Power is the rate of energy transfer, that is

where W is the energy transferred in the interval t 1 to t 2 • If the power

1.9 Independent and dependent supply sources

A source is an active element, that is, it can deliver energy to an external

Figure 1.6 An ideal or independent voltage souree (a) delivering apower of 9 W,

There is another group of voltage and current sources, namely depen-

(a) (b) (c) (d)

1.9.1 The operational amplijier

The operational amplifier (often abbreviated to op-amp) is of fundamental

Figure 1.8 (a) Idealised operational amplijier, (b) cireuit representation.

Worked example 1.9.1

Deduce an expression for the voltage gain of the operational amplifier

Figure 1.9 Circuit Jor worked example 1.9.1.

and the overall voltage gain is

1.10 Kirchhotf's laws

A simple way of expressing this is ~i = 0 or, more precisely

where N is the number of wires meeting at the node. Alternatively, we can

Figure 1.10 Kirchhoff's current law (KeL).

or, if we take the path ACBA we get