Networks and Linear Systems

Unit – II
Time response of circuits - complex frequency - poles and zeros -
frequency response from pole-zero configuration – Driving point
impedances - two-port networks.

Dr. Karthik Thirumala

Email: [email protected]

Reference textbooks:
1) D. Roy Choudhury, ‘Networks and Systems’, New Age International Publications, 1st Edition,
2013.
2) James W. Nilsson and Susan A. Riedel, ‘Electric Circuits’, Pearson Education Publications, 9th
Edition, 2011.
3) F.F.Kuo, ‘Network Analysis and Synthesis’, John Wiley Inc Publications, 2nd Edition, 2010.
4) M.E. Van Valkenburg, ‘Network Analysis’, PHI Learning Publications, 3rd Edition, 2014.
5) Hayt, W. H, Kemmerly J. E. & Durbin, ‘Engineering Circuit Analysis’, McGraw Hill Publications,
8th Edition, 2013.
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

Lecture 1
Time Response of Electric Circuits
In general, the complete time response of a circuit is sum of transient response (momentary) and steady
state response (exists a long time). The source-free response may be called the natural response or the
transient response or complementary solution. This happens when the inductor or capacitor is abruptly
disconnected from its dc source, the stored energy in an inductor or capacitor is suddenly released to a
resistive network. When we consider independent sources acting on a circuit, part of the response will
resemble the nature of the particular source (or forcing function) used; this part of the response, called the
particular solution or steady-state response or the forced response.
In order to obtain the complete response, we use KVL, KCL and ohms law to write the integro-
differential equations in terms of iL and vc. Solve the differential equations to obtain i(t) and v(t). We will
not attempt the solution of integro-differential equations here.
Natural Response
The currents and voltages that arise in a circuit with no independent sources are referred to as the natural
response of the circuit. It emphasizes the nature of the circuit itself, not external sources of excitation.
RL and RC circuits are also known as first-order circuits, because their voltages and currents are
described by first-order differential equations. We will consider several different methods of solving these
differential equations. The mathematical manipulation, however, is not circuit analysis. Our greatest
interest lies in the solutions themselves, and their interpretation, and we will try to become sufficiently
familiar with the form of the response that we are able to write down answers for new circuits.

1) First Order Series RL circuit

Finding the natural response requires finding the voltage and current at the terminals of the resistor after
the switch has been opened, that is, after the source has been disconnected and the inductor begins
releasing energy. If we let t = 0 denote the instant when the switch is opened, the problem becomes one of
finding v(t) and i(t) for f > 0.
The goal is an expression for i(t) which satisfies the below ordinary differential equation and also has the
value I0 at t = 0. The solution may be obtained by several different methods. The coefficients in the
equation, R and L, are constants; that is, they are not functions of either the dependent variable i or the
independent variable f. Thus the equation can also be described as an ordinary differential equation with
constant coefficients.

2
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Direct approach:
Separate the variables, integrate the equation and do necessary mathematical manipulations.

Hence
The coefficient of t (namely, R/L) determines the rate at which the current or voltage approaches zero.
The reciprocal of this ratio is the time constant of the circuit, denoted by τ.
Alternative approach:
The direct approach is easy but can be used only if variables are separable. We simply guess or assume a
form for the solution and then test our assumptions, first by substitution in the differential equation, and
then by applying the given initial conditions. Since we cannot be expected to guess the exact numerical
expression for the solution, we will assume a solution containing several unknown constants and select
the values for these constants in order to satisfy the differential equation and the initial conditions. Many
of the differential equations encountered in circuit analysis have a solution which may be represented by
the exponential function or by the sum of several exponential functions. Let us assume a solution in a
exponential form,

Where A and s1 are assumed constants. After substituting this equation in the original differential
equation, we have

The characteristic equation is LS+R = 0 → S = -R/L.

Obtain the solution and apply the initial condition i(0)=I0. Then the final solution is

2) First-order series RC circuit

When the switch is moved from position a to position b (at t = 0), the voltage on the capacitor is Vg.
Because there can be no instantaneous change in the voltage at the terminals of a capacitor, the problem
reduces to solving the circuit shown next with only R and C. The Vo denotes the initial voltage on the
capacitor.

3
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
3) Series and Parallel RLC circuits
We easily obtain the differential equation for the voltage by summing the currents away from the top
node, where each current is expressed as a function of the unknown voltage v:

We can't solve the equation above by separating the variables and integrating as we were able to do with
the first-order equations Thus, we assume that the solution is of exponential form

The differential equation now becomes

This equation is called the characteristic equation of the differential equation because the roots of this
quadratic equation determine the mathematical character of v(t). Solution of a quadratic equation are

The natural reponse is

The nature of the roots s1 and s2 depends on the values of α and ωo. There are three possible outcomes

We have been solving problems in which energy sources are

suddenly removed from the circuit; now we must consider that type
of response which results when energy sources are suddenly applied
to a circuit.

4
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Step response
The step response is the currents and voltages that arise when energy is being acquired by an inductor or
capacitor due to the sudden application of a dc voltage or current source. Stored energy may or may not
be present at the time the abrupt changes take place.
We will find v(t) or i(t) by writing the appropriate circuit equation and then solving it by separation of the
variables and integration. After we obtain the answer and investigate the two parts of which it is
composed, we will see that there is physical significance to each of these two terms. With a more intuitive
understanding of how each term originates, we will be able to produce more rapid and more meaningful
solutions to every problem involving the sudden application of any source.

1) First-order series RL circuit

Energy stored in the inductor at the time the switch is closed is given in terms of a nonzero initial current
i(0) = I0. The task is to find the expressions for the current in the circuit and for the voltage across the
inductor after the switch has been closed.

Now separate the variables to get

Integration on both sides gives

From which we get

If the initial current is zero, then the equation reduces to

This indicates that after the switch has been closed, the current increases exponentially from zero to a
final value of Vs/R.
The voltage across the inductor is

2) First-order series RC circuit

Similarly,

5
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
In general,

The natural response or complementary solution approaches zero as time increases without limit. The
natural response depends on the initial value of the complete response and thus on the initial value of the
forcing function also. The forced response is obtained by inspection of the final circuit. The f(0+) can be
found by equating the capacitor voltage or inductor current before switching and after switching i.e., or .

3) Series and Parallel RLC circuits

The general solution is obtained by the same procedure that was followed for RL and RC circuits. If there
are none acting in the circuit after the switching or discontinuity is completed, then the circuit is source-
free and the natural response accounts for the complete response. If independent sources are still present,
then the circuit is driven and a forced response must be determined. The forced response may be found by
inspection or, if necessary, by drawing the dc equivalent, with the addition of a source. The complete
response is then sum of natural and forced response.

Sinusoidal Response
It is worthwhile pointing out, however, that when the voltage forcing functions are sinusoidal functions of
time, it will be possible to define a voltage-current ratio (called impedance) or a current-voltage ratio
(called admittance) for each of the three passive elements.

Impulse response
The impulse response is defined as the zero-state (zero initial energy) response of a linear time-invariant
network due to a unit impulse sequence δ[n] at the system input.

1) Series RC network with an impulse voltage

The differential equation is

Current response is

Voltage response for a impulse current is

The same analysis can be done for the other series and parallel circuits with impulse forcing function.

6
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

Lecture 2
Zero-state response of a linear time-invariant circuit to an arbitrary input

C δ[n] → C h[n] and C δ[n±k] → C h[n±k]

The calculation of the response of any linear time-invariant circuit to an arbitrary input reduces to
• Determination of the impulse response h(t) or h[n]
• Convolution process
a. Folding: Fold h[k] about origin and obtain h[-k]
b. Time shifting: Shift h[k] by n unit to the right to obtain h[-k + n)] = h[n-k]
c. Multiplication: Multiply x[k] by h[n-k] to obtain f[k].
d. Summation: Sum all the values of the product f[k] to obtain value of the output at n.

7
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Summary
We wish to consider the transient behavior of circuits whose describing equations consist of node-voltage
or mesh current differential equation. We wish to determine the transient response of circuits whose
signal sources vary in ways more complicated than the simple dc level jumps. In the previous sections, we
discussed the classical methods for solving differential equations. The solutions are obtained directly in
time-domain and every step in solving is function of time. In this module, we use Laplace Transform to
transform the differential equation to the frequency domain, where the independent variable is complex
frequency, s.

Finally, we wish to relate, in a systematic fashion, the time-domain behavior of a circuit to its frequency-
domain behavior.
Our constant goal has been one of analysis: given some forcing function at one point in a linear circuit,
determine the response at some other point. In the step response lecture, we worked with dc forcing
functions and responses of the form V0e0. However, after the introduction of inductance and capacitance,
the sudden dc excitation of simple RL and RC circuits produced responses varying exponentially with
time: V0eσt. When we considered the RLC circuit, the responses took on the form of the exponentially
varying sinusoid, V0eσtcos(ωt+θ). All this work was accomplished in the time domain, and the dc forcing
function was the only one we considered. As we advanced to the use of the sinusoidal forcing function,
the tedium and complexity of solving the integro-differential equations caused us to begin casting about
for an easier way to work problems. The phasor transform was the result, and we might remember that we
were led to it through consideration of a complex forcing function of the form V0e j(ωt+θ). As soon as we
concluded that we did not need the factor containing t, we were left with the phasor V0e jθ we had arrived
at the frequency domain.
Now a little flexing of our brain has caused us to apply a forcing function of the form V0eσtej(ωt+θ), leading
to the invention of the complex frequency s, and thereby relegating all our previous functional forms to
special cases: dc (s = 0), exponential (s = σ), sinusoidal (s = jω), and exponential sinusoid (s = σ + jω).
Once we are able to find the forced response to this damped sinusoid, we will also have found the forced
response to a dc voltage, an exponential voltage, and a sinusoidal voltage.
By analogy to our previous experience with phasors, we saw that in these cases we may omit the factor
containing t, and once again obtain a solution by working in the frequency domain. For a function to have
a Fourier Transform, it must obey the condition
As a result, a ramp function or a step function does not possess Fourier Transform in strict sense, but
possess a generalized Fourier Transform containing impulses in frequency. But the same function will
have Laplace transform because of the added convergence factor e-σt (σ > 0).
Complex frequency
We introduce the notion of complex frequency by considering a (purely real) exponentially damped
sinusoidal function, such as the voltage

8
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
where σ (sigma) is a real quantity and is usually negative. Although we often refer to this function as
being “damped’’ it is conceivable that we might occasionally encounter a situation where σ > 0 and hence
the sinusoidal amplitude is increasing.
When s= σ + jω = 0,
When ω = 0 and s = σ,
When σ =0 and s = jω,
Thus, the damped sinusoid function includes as special cases the dc, exponential and sinusoidal functions.
Let us determine the complex frequencies associated with the exponentially damped sinusoidal function.

We find that a conjugate complex pair of frequencies, s1 = σ + jω and s2 = σ - jω, is also required to
describe the exponentially damped sinusoid. The complex frequency is .
The real part of s is associated with the exponential variation; if it is negative, the function decays as t
increases; if it is positive, the function increases; and if it is zero, the sinusoidal amplitude is constant. The
larger the magnitude of the real part of s, the greater is the rate of exponential increase or decrease. We
will also refer to σ as the neper frequency, the name arising from the dimensionless unit of the exponent
of e. Thus, given e7t, the dimensions of 7t are nepers (Np), and 7 is the neper frequency in nepers per
second. The imaginary part of s describes the sinusoidal variation; it is specifically the radian frequency.
A large magnitude for the imaginary part of s indicates a more rapidly changing function of time. The
complex frequency s is measured in units which are variously termed complex nepers per second or
complex radians per second.

The forcing function v(t) can be represented in terms of complex frequency s as

or
Either representation is suitable, and the two expressions should remind us that a pair of
conjugate complex frequencies is associated with a sinusoid or an exponentially damped
sinusoid.

When we apply the exponentially damped sinusoid to a circuit, the forced response has form of forcing
function as given below
or

Two-sided or Bilateral Laplace Transform

We know that sinusoidal forcing functions lead to sinusoidal responses, and also that exponential forcing
functions lead to exponential responses. However, as practicing engineers we will encounter many

9
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
waveforms that are neither sinusoidal nor exponential, such as square waves, sawtooth waveforms, and
pulses beginning at arbitrary instants of time. When such forcing functions are applied to a linear circuit,
we will see that the response is neither similar to the form of the excitation waveform nor exponential. As
a result, we are not able to eliminate the terms containing ‘t’ to form a frequency-domain response.
There is a solution, however, which makes use of a technique that allows us to expand any function into a
sum of exponential waveforms, each with its own complex frequency. Since we are considering linear
circuits, we know that the total response of our circuit can be obtained by simply adding the individual
response to each exponential waveform. And, in dealing with each exponential waveform, we may once
again neglect any terms containing t, and work instead in the frequency domain.
The Laplace transform of a general function f(t) is

The LT of f(t) becomes identical to Fourier transform of the f(t) multiplied by e-σt.
Two-sided or Bilateral Inverse Laplace Transform

where the real constant σ0 is included in the limits to ensure convergence of this improper integral.

Unilateral Laplace Transform

In many of our circuit analysis problems, the forcing and response functions do not exist forever in time,
but rather they are initiated at some specific instant that we usually select as t = 0. Thus, for time
functions that do not exist for t < 0, or for those time functions whose behaviour for t < 0 is of no interest,
the time-domain description can be thought of as v(t)u(t).
Another point regarding the lower limit concerns the situation when f(t) has a discontinuity at the origin.
Should we make the lower limit 0- and include the discontinuity, or should we exclude the discontinuity
by making the lower limit 0+? (We use the notation 0- and 0+ to denote values of t just to the left and right
of the origin, respectively.) Actually, we may choose either as long as we are consistent. If we are using 0-
as the lower limit, we note immediately that the integration from 0- to 0+ is zero. The only exception is
when the discontinuity at the origin is an impulse function. The defining integral for the Laplace
transform is taken with the lower limit at t = 0− in order to include the network condition just before
switching at t = 0.
The corresponding Laplace transform is then

The inverse transform expression remains unchanged, but when evaluated, it is understood to be
valid only for t > 0.

10
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Inference:
In order for a function to possess a Laplace transform, it must obey the condition

In the transient problems, the Laplace transform is preferred to the Fourier transform, not only because a
larger class of waveforms have Laplace transform but also because the LT takes directly into account
initial conditions at t= 0-.

Laplace Transform of some useful functions

In this section we will begin to build up a catalog of Laplace transforms for those time functions most
frequently encountered in circuit analysis. Because we are limiting to the unilateral or one-sided, Laplace
transform, we define all functions to be zero for t < 0-.
Unit Impulse function

Translated impulse function

Unit step function

Exponential decaying function

Ramp function

Sinusoidal function

11
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
f(t) = sin ωt

Powers of t

Few more functions

12
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Properties of Laplace Transform
Multiplication by constant

Addition or subtraction

Translation in Time domain

As we have seen in some of our earlier transient problems, not all forcing functions begin at t = 0. What
happens to the transform of a time function if that function is simply shifted in time by some known
amount? In particular, if the transform of f(t)u(t) is the known function F(s), then what is the transform of
f(t − a)u(t − a), the original time function delayed by a seconds (and not existing for t< a)? Working
directly from the definition of the Laplace transform, we get

Scale changing
The scale-change property gives the relationship between f(t) and F(s) when the time variable is
multiplied by a positive constant:

Multiplication by t or Frequency differentiation

Laplace Transform of is

Laplace Transform of is

Division by time or Frequency integration

Laplace Transform of is

Time differentiation

13
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Let us look at time differentiation first by considering a time function f(t) whose Laplace transform F(s) is
known to exist. We want the transform of the first derivative of f(t),

We find the Laplace transform of the nth derivative by successively applying the preceding process,
which leads to the general result

Time Integration
Integration in the time domain corresponds to dividing by s in the s domain. As before, we establish the
relationship by the defining integral:

Translation in Frequency domain

Translation in the frequency domain corresponds to multiplication by an
exponential in the time domain:

Convolution
Laplace Transform of is

Laplace Transform of Periodic functions

A periodic signal with time period T is

Initial and Final value theorem

The initial- and final-value theorems are useful because they enable us to determine from F(s) the
behavior of f(t) at 0+ and ∞. Hence, we can check the initial and final values of f(t) to see if they conform
with known circuit behavior, before actually finding the inverse transform of F(s).

The initial-value theorem states that

14
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
f(0) =

Example:

The final-value theorem states that

f(∞) =

Example:

Lecture 3
Relation between Fourier Transform and Laplace Transform
1. If σ < 0 or f(t) is absolute integrable and causal function (poles of F(s) lies in LHP)
|
F(jω) = FL(s) s=jw
This equation shows that the LT is related to the entire s-plane, whereas the FT is restricted to the
imaginary axis.
Example: e-atu(t), e-atcos(ωt), δ(t)
2. If σ > 0 or f(t) is causal and diverging function (poles of F(s) lies in RHP)
F(jω) doesn’t exist
Example: e2tu(t)

3. If σ =0 or f(t) is causal and neither converging nor diverging function

Example: u(t) and cosω0t

The LT is better suited for the analysis of transient analysis involving initial conditions, because it
permits inclusion of initial conditions, whereas the FT does not. The FT is especially useful for problems
in the steady state.

15
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Transformation of differential equations
We now illustrate how to use the Laplace transform to solve the ordinary integro-differential equations
that describe the behaviour of lumped parameter circuits. Consider the circuit shown in Figure below. We
assume that no initial energy is stored in the circuit at the instant when the switch,
which is shorting the dc current source, is opened. The problem is to find the time-domain expression for
v(t) when t ≥ 0.

We begin by writing the integro-differential equation that v(t) must satisfy. We need only a single node-
voltage equation to describe the circuit. Summing the currents away from the top node in the circuit
generates the equation:

After deriving the integro-differential equations, we transform the equations to the s domain. We will not
go through the steps of the transformation in detail, because in next lectures we will discover how to
bypass them and generate the s-domain equations directly. Use the necessary transforms to obtain an
algebraic equation in which V(s) is the unknown variable.

Next, we solve the algebraic equations for the unknowns.

To find v(t) we must inverse-transform the V(s) expression.

Note:
1) Depending on whether we wish to obtain the solution from t= 0- or 0+ we should substitute initial
conditions pertaining to t=0- or t=0+.
2) Differential equation is converted into an algebraic equation

Inverse Laplace Transform

The expression for F(s) is a rational function of s; that is, one that can be expressed in the form of a ratio
of two polynomials in s such that no non-integral powers of s appear in the polynomials. In fact, for
linear, lumped-parameter circuits whose component values are constant, the s-domain expressions for the
unknown voltages and currents are always rational functions of s. If we can inverse-transform rational
functions of s, we can solve for the time-domain expressions for the voltages and currents. The purpose of
this section is to present a straight-forward and systematic technique for finding the inverse transform of a
rational function.

16
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

The coefficients a and b are real constants, and the exponents m and n are positive integers. The ratio
N(s)/D(s) is called a proper rational function if m > n, and an improper rational function if m ≤ n. Only a
proper rational function can be expanded as a sum of partial fractions. This restriction poses no problem,
as we show at the end of this section.
1. Partial Fraction Expansion: Proper Rational Functions with distinct real roots of D(s)
A proper rational function is expanded into a sum of partial fractions by writing a term or a series
of terms for each root of D(s). Thus D(s) must be in factored form before we can make a partial
fraction expansion. For each distinct root of D(s), a single term appears in the sum of partial
fractions.
2. Partial Fraction Expansion: Proper Rational Functions with distinct complex roots of D(s)
The only difference between finding the coefficients associated with distinct complex roots and
finding those associated with distinct real roots is that the algebra in the former involves complex
numbers. We illustrate by expanding the rational function:

Again, we need to make some observations. First, in physically realizable circuits, complex roots
always appear in conjugate pairs. Second, the coefficients associated with these conjugate pairs
are themselves conjugates.

17
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

3. Partial Fraction Expansion: Proper Rational Functions with repeated real roots of D(s)

The function
Which we can be expanded as

To determine each constant, we first multiply the non-expanded version of V(s) by (s - p)n. The
constant an is found by evaluating the resulting expression at s = p. The remaining constants are
found by differentiating the expression (s-p)nV(s) the appropriate number of times prior to
evaluating at s = p, and dividing by a factorial term. The differentiation procedure removes the
constants previously found, and evaluating at s = p removes the remaining constants.

Example: Verify solutions for the below problems

1)

2)

4. Partial Fraction Expansion: Proper Rational Functions with repeated complex roots of D(s)
We handle repeated complex roots in the same way that we did repeated real roots; the only
difference is that the algebra involves complex numbers. Recall that complex roots always appear
in conjugate pairs and that the coefficients associated with a conjugate pair are also conjugates, so
that only half the Ks need to be evaluated. For example,

18
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

After factoring the denominator polynomial, we write

Now we need to evaluate only K1 and K2, because K1* and K2* are conjugate values. The value
of K1 is

The value of K2 is

We now group the partial fraction expansion by conjugate terms to obtain

5. Partial Fraction Expansion: Improper Rational Functions

An improper rational function can always be expanded into a polynomial plus a proper
rational function. The polynomial is then inverse-transformed into impulse functions and
derivatives of impulse functions. The proper rational function is inverse-transformed by
the techniques outlined in this section. To illustrate the procedure, we use the function

19
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

Lecture 4
Circuit analysis in S-domain
Use the Kirchhoff’s laws, Ohms law, Mesh analysis, nodal analysis and source transformation to solve
the networks.
1) Using the differential equations and laplace transform determine i(t) and v(t) for t > 0 in the series RC
circuit shown below. Find v(t) at t=0+, 0.1 s, 0.4 s and 1.2 s. Verify the initial and final value using
Laplace Transform.
Answer:

v(1.25) = 1.05,

2) Determine the step response of the parallel RLC circuit.

Solution: Summing the currents away from the top node generates the expression

Obtain V from the above equation

The response is

Expand the response in partial fractions

Verify the s-domain expression by finding the final value. The final inductor current must be 24 mA
as t→∞ or s→0.

20
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

Circuit elements in the S-domain

Instead of writing the differential equations and then apply Laplace transform, we can directly write the
algebraic equations in S-domain by constructing a circuit model that satisfies the relationship between the
s-domain current and voltage. We use the passive sign convention in all the derivations. Impedance in the
s-domain is measured in ohms, and admittance is measured in Siemens.

21
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Construct a transform diagram with initial conditions appropriately. From the transform diagram write the
algebraic equations then obtain Laplace transform of response quantity. Finally obtain response in time-
domain by applying inverse Laplace transform.

1) Draw the s-domain equivalent circuit for the below electric circuits with initial current of 1A.

Lecture 5
Transfer function
The transfer function of a linear-time invariant circuit is defined as the s-domain ratio of the Laplace
transform of the output (response) to the Laplace transform of the input (source). In computing the
transfer function, we restrict our attention to circuits where all initial conditions are zero. Firstly, once we
know the transfer function of a particular circuit, we can easily find the output that results from any input.
All we need to do is multiply H(s) by the input quantity, and take the inverse transform of the resulting
expression. Secondly, the form of the transfer function contains a great deal of information about the
behavior we might expect from a particular circuit (or system).
The transfer function is

| with zero initial conditions

Note that the transfer function depends on what is defined as the output signal and input signal. The H(s)
can be impedance function or Admittance function or ratio. If a circuit has multiple independent sources,
we can find the transfer function for each source and use superposition to find the response to all sources.
Thus, because circuits may have multiple sources and because the definition of the output signal of
interest can vary, a single circuit can generate many transfer functions. Remember that when multiple
sources are involved, no single transfer function can represent the total output—transfer functions
associated with each source must be combined using superposition to yield the total response.
Poles & Zeros
In order to evaluate the stability of a system it is necessary to determine the poles and zeros of the transfer
function H(s). For linear lumped-parameter circuits, H(s) is always a rational function of s. The rational
function of H(s) also may be expressed as the ratio of two factored polynomials, where a and b are the
coefficients of real positive value.

Let N(s) has n roots and D(s) has m roots. Then, we may write H(s) as

22
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

Where K is the constant an/bm also known as scaling factor.

The roots of the numerator polynomial, that is, - z1, - z2, - z3 , . . ., -zn, are called the zeros of H(s); they are
the values of s at which H(s) becomes zero. The roots of the denominator polynomial, that is, -p1, - p2, - p3
, . . ., - pm, are called the poles of H(s); they are the values of s at which H(s) becomes infinitely large.
If a circuit has three responses (y1, y2 and y3), then there would be three transfer functions corresponding
to Y1(s), Y2(s) and Y3(s). However, the denominator would remain same. So the characteristic equation is
D(s) = 0. The roots of the characteristic equation are the poles of the system, which are the natural
frequencies of the network.
Y(s) = H(s)X(s)
The expansion of Y(s) has terms containing poles of H(s) (corresponding to natural frequencies) and
terms poles of X(s) (corresponding to forced frequencies).

S- Plane or pole-zero plane

In what follows, you may find that being able to visualize the poles and zeros of F(s) as points on a
complex s plane is helpful. A complex plane is needed because the roots of the polynomials may be
complex. In the complex s plane, we use the horizontal axis to plot the real values of s and the vertical
axis to plot the imaginary values of s. The figure shows the poles and zeros plotted on the s plane,
where X represent poles and O represent zeros.

Complex poles, -6±j8 and zeros, -3±j4 always appear in

conjugate pairs. The zeros of H(s) may lie in either the right
half or the left half of the s plane. The poles of H(s) must lie
in the left half of the s plane if the response to a bounded
source (one whose values lie within some finite bounds) is to
be bounded.
• Poles on the x-axis (a real pole in the left half of s-plane defines an exponentially decaying
component) The rate of the decay is determined by the pole location; poles far from the origin in
the left-half plane correspond to components that decay rapidly, while poles near the origin
correspond to slowly decaying components.
𝐾
𝑆+𝑎
→ 𝐾𝑒 −𝑎𝑡

• Complex Pole Pair in the left half of the s-plane combine to generate a response component that
is a decaying sinusoid. The rate of decay is specified by σ; the frequency of oscillation is
determined by ω.

23
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
𝐾1 𝑆+𝐾2
(𝑆+𝛼)2 +𝜔02
→ 𝐴𝑒 −𝑎𝑡 sin⁡(𝜔𝑡 + 𝛽)

• Pole Pair on the imaginary axis generates an oscillatory component with a constant amplitude
determined by the initial conditions.
𝐾1 𝑆+𝐾2
𝑆 2 +𝜔02
→ 𝐴sin⁡(𝜔𝑡 + 𝛽)

• Poles on the RHP → A real pole in the right-half plane corresponds to an exponentially
increasing component in the homogeneous response; thus defining the system to be unstable.
𝐾 𝐾 𝑆+𝐾
→ 𝐾𝑒 𝑎𝑡 and (𝑆−𝛼)
1 2
2 +𝜔2
→ 𝐴𝑒 𝑎𝑡 sin⁡(𝜔𝑡 + 𝛽)
𝑆−𝑎 0

An illustration of the physical meaning of positive and negative values for σ and ω, as would be
represented on the complex-frequency plane. When ω = 0, a function will have no oscillatory component;
when σ = 0, the function is purely sinusoidal except when ω is also zero.

Lecture 6
3) Determine the impulse response of the circuit in figure below and use this to compute the forced
response vo(t) if the input vin(t) = 6e-tu(t) V.
Solution:

24
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

4) Calculate the voltage vx in the circuit below using nodal analysis technique.

Solution: The first step is to draw the corresponding s-domain circuit. We see that the 1/2 F capacitor
has an initial voltage of 2 V across it at t = 0-. Draw the S-domain circuit and then write the nodal
equation as below

So that

The nodal voltage vx is found by taking the inverse Laplace transform, and we find that

5) Simplify the circuit and determine an expression for the voltage v(t).

25
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

6) For a step response of a multiple mesh circuit, find out the branch currents using mesh analysis.

Solution: Here we want to find the branch currents i1 and i2 that arise when the 336 V dc voltage
source is applied suddenly to the circuit. The initial energy stored in the circuit is zero. The two mesh-
current equations are

and
Using Cramer’s method to solve for I1 and I2, we obtain

26
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

Therefore, the final values of i1 and i2 are

and

7) The switch in the circuit shown has been in position ‘a’ for a long time. At t = 0, the switch
is thrown to position ‘b’.
a. Find V0 as rational functions of s.
b. Find the time-domain expressions for v0. For t≥0.
Answer:

Lecture 7
Frequency response from pole-zero configuration
The transfer function can be

The frequency response function is a complex function of the frequency ω=2πf that describes the
response of a system to input of different frequencies. The frequency response may be written in terms of
the system poles and zeros by substituting jω for s directly into the transfer function

A Bode plot is a standard format for plotting frequency response of LTI systems. The format is a log
frequency scale on the horizontal axis and, on the vertical axis, phase in degrees and magnitude in

27
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
decibels. The magnitude of the complete frequency response may then be written in terms of the
magnitudes of these component vectors as

However, on a log scale, the product turns into a sum. The log magnitude of |𝐻(𝑗𝜔)| with unit decibel
(dB) is

Using polar form, we can say that the angle of the product (quotient) is the sum of the angles of each term
(except for division we subtract, so it’s the sum of the angles for the top terms, minus the sum of the
angles for the terms in the denominator).

For example, consider the sinusoidal response of a first-order system with a pole on the real axis at s =
−1, its Bode plot is shown below.

Bode plot of a first order system with single pole at s = -a (a=1)

The amplitude is -3 dB at ω = 1, and then the slope is −20 dB/decade for ω > 1. The angle plot is 0o for ω
< 0.1, −900 for ω > 10, and −450 at ω = 1, and it has a slope of−450/decade when 0.1 < ω < 10.

28
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Network Functions
A pair of terminals through which a signal is accessed to or from the network is called a port. A network
having only one such pair of terminals is called a one-port network (Example: two-terminal R, L or C).
Figure (a) is a representation of a one-port network. The pair of terminals is customarily connected to an
energy source, which is the driving force for the network. Figure (b) is a two-port network or two terminal
pair network. The port 1-1’ is assumed to be connected to the driving force and is called the input port
and 2-2’ is connected to a load and is called the output port. The three-terminal device such as transistors
can be configured into two-port networks. There may be n-port network as shown in figure (c).

Immittance is the word used generically for impedances and admittances. In general, an immittance can
be a driving point immittance or a transfer immittance.

Driving point immittance function

For an LTI one-port network, the ratio of current to voltage or voltage to current of same branch is driving
point function.
Driving point Impedance
𝑉1 (𝑆)
Z(s) = 𝐼1 (𝑆)
Poles of Z(s) are open circuit natural frequencies i.e., I(s) = 0.

Driving point admittance

𝐼 (𝑆)
Y(s) = 𝑉1 (𝑆)
1

Poles of Y(s) are short circuit natural frequencies i.e., V(s) = 0.

Calculation of driving point immittance (Z or Y)

• Series – parallel reduction
When s=0, Z(0) is DC equivalent resistance because inductor is short circuited and capacitor is
open circuited. When s=∞, Z(∞)

Example 1: Determine the driving point impedance of the circuit

Answer:

29
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
• Loop circuit method
First write a set of loop equations for a network

then the coefficient of each current will be an impedance Zij(s), and the circuit determinant, or
determinant of the coefficients, is

𝐼 (𝑆) ∆11 𝑉1 (𝑆) ∆

Y11(s) = 𝑉1 (𝑆) = Z11(s) = = ∆𝑍
1 ∆𝑍 𝐼1 (𝑆) 11

Example 2: Calculate the input impedance for the one-port resistive network shown below
Solution:

Example 3: For the T-network shown below, find the driving point admittance Y11 and transfer
admittance Y12.
Answer:

30
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
• Node voltage method
With similar procedure using nodal equations, the immittance can be obtained as
𝑉(𝑆)
Z(s) = 𝐼(𝑆)

The driving point immittance function of a ladder network shown in figure below is

Lecture 8
Two-port network
We will assume in all that follows that the network is composed of linear elements and contains no
independent sources. We will consider the two-port as it is shown in figure below; the voltage and current
at the input terminals are V1 and I1, and V2 and I2 are specified at the output port. The directions of I1 and
I2 are both customarily selected as into the network at the upper conductors (and out at the lower
conductors). Transfer function is defined as ratio of response to excitation, where response and excitation
are not associated in same branch. In general, the transfer function may have the following forms:
𝑉 (𝑆)
1) Voltage transfer function 𝐺21 (𝑆) = 𝑉2 (𝑆)
1
𝐼 (𝑆)
2) Current transfer function 𝛼21 (𝑆) = 𝐼2 (𝑆)
1
𝑉 (𝑆)
3) Transfer impedance function 𝑍21 (𝑆) = 𝐼 2(𝑆)
1
𝐼2 (𝑆)
4) Transfer admittance function 𝑌21 (𝑆) = 𝑉 (𝑆)
1
1
Note that for a two-port network, 𝑍21 (𝑆) ≠ 𝑌 .
21 (𝑆)

Two-port network parameters

Four terminal quantities: V1, I1, V2 and I2
In viewing a circuit as a two-port network, we are interested in relating the current and voltage at one port
to the current and voltage at the other port. However, the choice of two independent variables and two
dependent variables is a matter of convenience in a given application. For an n-port network, the number
of ways in which these 2n variables can be arranged in a two groups of n each, is equal to (2n)!/(n!)2. For
a two-port network, there are 4 variables and 6 possibilities.

31
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

The usefulness of the different methods of description comes clearly into evidence when the problem is
synthesizing networks such as filters, impedance-matching networks, wave-shaping networks, etc.

Open Circuit Impedance Parameters

Independent: I1 and I2 and Dependent: V1 and V2
Since the network is linear and contains no independent sources within it, V1 may be considered to be the
superposition of two components, one caused by I1 and the other by I2. When the same argument is
applied to V2. Thus, the terminal voltages can be related to the terminal currents as

Or in matrix form as

The most informative description of z parameters can be obtained by setting either input current or output
current to zero.

Since zero current results from an open-circuit termination, the z parameters are known as the open-
circuit impedance parameters. The z11 is the open-circuit input impedance or driving point impedance, z22
is open-circuit output impedance, z12 is open-circuited forward transfer impedance and z21 is open-
circuited reverse transfer impedance.

32
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
According to the above equations, we obtain z11 and z21 by connecting V1 to port 1 with port 2 open-
circuited. Then finding the I1 and V2 will give z11 and z21. Similarly, we can obtain z12 and z22 by
connecting a voltage V2 to port 2 with port 1 open-circuit. Then finding the I2 and V1 will give z12 and z22.

Symmetrical or reciprocal networks

When z11 = z22, the two-port network is said to be symmetrical network. This implies that the network has
mirror like symmetry about some arbitrary center line.
When the two-port network is linear and has no dependent sources, the transfer impedances are equal, and
such a network is reciprocal. In other ways, a two-port circuit is reciprocal if the points of excitation and
response are interchanged (interchange of an ideal voltage source at one port with an ideal ammeter at the
other port produces the same ammeter reading).

(a) (b)
The reciprocal network yields V = z12I according to V-I relations when connected as in Fig (a), but yields
V = z21I when connected as in Fig. (b). This is possible only if z12 = z21. Any two-port that is made
entirely of resistors, capacitors, and inductors must be reciprocal. For a reciprocal network, the T-
equivalent circuit in Fig. 1(a) (below) can be used. If the network is not reciprocal, a more general
equivalent network is shown in Fig. 1(b) (below); notice that this figure follows directly from the basic V-
I relations.

Example 1: Find the z parameters for the resistive two-port shown below.

33
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

Example 2: Find the four impedance parameters for the resistive two-port shown below.
Answer: z11= 30/7 Ω, z21= 20/7 Ω
z12 = 20/7 Ω and z22 = 60/7 Ω

Example 3: Find the currents I1 and I2 in the circuit shown below.

Solution: Directly by substituting in the V-I relationship

34
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Short Circuit Admittance Parameters
Dependant: I1 and I2 and Independent: V1 and V2
Some of the two-port networks do not have z-parameters, as they can’t be described by the V=ZI relation.
For example, an ideal transformer has no z parameters. These are other set of parameters, here the
terminal currents are expressed in terms of terminal voltages. Since the network is linear and contains no
independent sources within it, I1 may be considered to be the superposition of two components, one
caused by V1 and the other by V2. When the same argument is applied to I2, we may begin with the set of
equations.

Thus, we may write the matrix equation I = yV as

Each of the y parameters may be described as a current-voltage ratio with either V1 =0 (the input
terminals short circuited) or V2 =0 (the output terminals short circuited). Because each parameter is an
admittance which is obtained by short circuiting either the output or the input port, the y parameters are
known as the short-circuit admittance parameters.

The specific name of y11 is the short circuit input admittance or driving point admittance, y22 is the short-
circuit output admittance or driving point admittance, and y12 and y21 are the short-circuit transfer
admittances.
According to the above equations, we obtain y11 and y21 by connecting I1 to port 1 with port 2 short-
circuited. Then finding the I2 and V1 will give y11 and y21. Similarly, we can obtain y12 and y22 by
connecting a current source I2 to port 2 with port 1 short-circuit. Then finding the I1 and V2 will give y12
and y22.
For a two-port network that is linear and has no dependent sources, the transfer admittances are equal (y12
= y21). This can be proved in the same way as for the z parameters. A reciprocal network (y12 = y21) can be
modeled by the π-equivalent circuit in Fig. (a). If the network is not reciprocal, a more general equivalent
network is shown in Fig. (b).

35
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Example 4: For the circuit, determine the Y parameters.

Example 5: For the circuit, determine the Y parameters.

Solution: To determine y11, we short-circuit the output and find the ratio
of I1 to V1. Similarly, to find y12, we short-circuit the input terminals and
apply V2 at the output terminals.

They are easily related to the short-circuit admittance parameters as

Y = inv(Z) but z11 ≠ 1/y11

Lecture 9
Hybrid Parameters
The hybrid parameters would find wide usage in electronic circuits, especially in constructing models for
transistors. The parameters of a transistor can be easily measured with h-parameters than to measure
short-circuit admittance or open-circuit impedance parameters. Some kind of parameter representation is
required by which some parameters are measured by open-circuiting the input port, while the rest of the
parameters can be measured by short-circuiting the output port. This is the so called hybrid parameter
representation. Set of equations result when we express V1 and I2 in terms of I1 and V2.

36
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

Since the parameters represent impedance, admittance, voltage gain, and current gain, they are called the
“hybrid’’ parameters. The procedure to calculate these parameters is similar to that used for the z aor y
parameters.

For reciprocal networks,

The hybrid parameters can be related to the y parameters as

= =

Example 6: Find the hybrid parameters for the two-port network shown below.

Example 7: Find the hybrid parameters for the two-port network

shown.
Answer: h11 = 1.2Ω , h12 = 0.4, h21 = -0.4, h22 = 0.4 S.

37
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Example 8: Find the hybrid parameters for the resistive network shown below

Transmission Parameters or ABCD Parameters

The t parameters, ABCD parameters, chain parameters or simply the transmission parameters equation
of the two-port network can be written as
&
In matrix form,

These two-port parameters provide a measure of how a circuit transmits voltage and current from a source
to a load. Transmission parameters are used in the analysis of power transmission line, where they are
also known as general circuit parameters. Note that the negative sign for the I2 arises because I2 in the 2-
port network is opposite to the conventional current direction. This is done merely for conventional
reasons; when you cascade two-ports (output to input), it is most logical to think of I2 as leaving the two-
port. It is also customary in the power industry to consider I2 as leaving the two-port.
Other widely used nomenclature for this set of parameters is

1/A is open circuit voltage gain, -1/B is short-circuit transfer admittance, 1/C is the open-circuit transfer
impedance and -1/D is a short-circuit current gain.

If a two-port circuit is reciprocal, the following relationships exist among the port parameters:

38
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
The transformation between y, z, h and t parameters is as follows

Lecture 10
Example

39
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022

Interconnected Networks
A large, complex network may be divided into sub networks for the purposes of analysis and design. The
sub networks are modeled as two-port networks, interconnected to form the original network. The two-
port networks may therefore be regarded as building blocks that can be interconnected to form a complex
network. Two-port circuits may be interconnected five ways: (a) in cascade, (b) in series, (c) in parallel,
(d) in series-parallel, and (e) in parallel-series. Figure below depicts these five basic interconnections.

Although the interconnected network can be described by any of the six parameter sets, a certain set of
parameters may have a definite advantage. For example, when the networks are in series, their individual
z parameters add up to give the z parameters of the larger network. When they are in parallel, their
individual y parameters add up to give the y parameters the larger network. When they are cascaded, their

40
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
individual transmission parameters can be multiplied together to get the transmission parameters of the
larger network.

Parallel Networks: The y parameters are useful in describing the parallel connection of two-ports.
For network A →
For network B →

Thus, for a two parallel network

and the corresponding Y parameters are

Series Networks: The y parameters are useful when two-ports are interconnected in parallel, and, in a
dual manner, the z parameters simplify the problem of a series connection of networks.
It follows that

Cascaded Networks: When the two networks are cascaded as shown below, the t parameters are helpful
in finding out the parameters of the combined network as explained below

For network A For network B

Combining these results, we have

Therefore, the ABCD parameters for the cascaded networks are found by the matrix product.

41
NLS, UNIT-II, KARTHIK THIRUMALA JULY 2022
Example: The three-element network shown in figure has an input
impedance ZA(s) that has a zero at s = −10 + j0. If a 20 Ω resistor is
placed in series with the network, the zero of the new impedance
shifts to s = −3.6 + j0. Calculate R and C.
Answer: R = 20 Ω and C = 25 mF

Example: Find the y parameters for the circuit shown below

Solution:

42