Networks and Linear Systems

Unit – I
Classification of signals – representation in terms of
elementary signals - impulse functions Frequency response -
Fourier series - Harmonic analysis of simple circuits – Fourier
integral - Fourier transforms – application to simple circuits

Dr. Karthik Thirumala

Email: [email protected]

Reference:
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-I, KARTHIK THIRUMALA - JULY 2021
Lecture - 1
Signal
In general, signal is a function of time. In the absence of signals, the network or system is completely
lifeless. And therefore, an analysis of signals and knowledge of the various kinds of signals that one
comes across is important for us. The signals can also be described in terms of frequency information.
Between time and frequency, the translation is effected by the Fourier series, the Fourier integral and the
Laplace transform.

Classification of Signals
- based on nature of independent variable, number of independent variables, nature of
dependant variable, based on periodicity,
1. Continuous-time signal
In case of continuous-time signals, the independent variable is continuous and thus these signals are
defined for continuum of values of the independent variable. The continuous-time signals are also
called analog signals. We will use the symbol ‘t’ to denote continuous-time independent variable.
They are denoted by x(t). They are continuous in amplitude as well as in time. Most of the signals
available are continuous-time signals.
2. Discrete-time signal
On the other hand, discrete-time signals are defined only at discrete instants of times. The discrete-
time signals are continuous in amplitude, but discrete in time. Use ‘n’ to denote discrete-time
independent variable. Let t=nT, n is integer (0, ±1, ±2…) and T is the sampling time. Since they are
defined only at discrete instants of time, they are denoted by a sequence x(nT) or simply by x(n)
where n is an integer. The discrete-time signals may be inherently discrete (count of events or things)
or may be discrete versions of the continuous-time signals.

3. Digital signal
Discrete-valued signal is takes on values from a finite set of possible values, which are equidistant.
Quantization is a process of converting a continuous-valued signal into a discrete-valued signal (eg.
rounding or truncating). Digital signals are discrete-time signals having a set of discrete values.
4. Periodic signal
A periodic continuous-time signal is a function that repeats itself every finite time internal of T
seconds. It is described by the equation
Why the interest in periodic functions? One reason is that many electrical sources of practical value
generate periodic waveforms.

2
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
5. Even signal
A continuous signal s(t) is said to be as an even signal if it is identical to its time-reversed counterpart
i.e., with its reflection about the origin. It obeys the relation
Example: Cosine wave is even function
6. Odd signal
An odd function obeys the relation
This may be noted that an odd continuous time signal will be zero at origin i.e., f(0)=0 at t=0.
Example: Sine wave is odd signal.

3
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

Elementary Signals
Since we deal with signals, a reference frame of interest has to be made of signals with well know
properties, which reflect the kind of information we want to extract. For example, if we hear a musical
note, and we want to determine which note it is in the scale, we need to compare it with sinusoidal signals
(i.e. "tones") at all frequencies and determine which frequency is the closest.
We may encounter functions that have a discontinuity, or jump, at the origin. For example, we know from
earlier discussions of transient behavior that switching operations create abrupt changes in currents and
voltages. We accommodate these discontinuities mathematically by introducing the step and impulse
functions. We have been speaking of the “sudden application’’ of an energy source, and by this phrase we
imply its application in zero time. The operation of a switch in series with a battery is thus equivalent to a
forcing function which is zero up to the instant that the switch is closed and is equal to the battery voltage
thereafter. The forcing function has a break, or discontinuity, at the instant the switch is closed. Certain
special forcing functions which are discontinuous or have discontinuous derivatives are called singularity
functions, the two most important of these singularity functions being the unit-step function and the unit-
impulse function. Here present some of the elementary discrete-time and continuous signals which will be
at the basis of the analysis developed in the next lectures.

• Unit step function

The physical analogy of a unit step excitation corresponds to a switch S, which closes at t=0 and
connects a battery of 1 V to a circuit.

The unit step can be represented in terms of impulse function as

4
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
The unit step function is not continuous and differentiable at t=0. Assuming it is linear in the interval
of 0- to 0+, u(0) = 0.5 (average value).

FIGURE (a) A voltage-step forcing function is shown as the source driving a general network. (b) A simple circuit which,
although not the exact equivalent of part (a), may be used as its equivalent in many cases. (c) An exact equivalent of part (a).

• Unit Impulse function

The unit impulse function or delta function is a mathematical anomaly. P. A. M. Dirac first used it in
his writings on quantum mechanics. An impulse is a signal of infinite amplitude and zero duration.
Such signals don't exist in nature, but some circuit signals come very close to approximating this
definition, so we find a mathematical model of an impulse useful. Impulsive voltages and currents
occur in circuit analysis either because of a switching operation or because the circuit is excited by
an impulsive source.

Continuous time unit impulse function

The integral equation states that the area under the impulse function is constant and is equal to one. If
the area under the impulse function curve is other than unity, the impulse function is denoted Kδ(t),
where K is the area. K is often referred to as the strength of the impulse function. An impulse that
occurs at t = a is denoted Kδ(t - a).
Sampling property

• Continuous sinusoidal signals

A continuous time sinusoidal signal is defined as . As we have seen
in the Fourier Analysis section, the sinusoidal signals are one of the building blocks of more general
signals. A discrete time sinusoid is obtained
by sampling a continuous time sinusoid with sampling interval Ts, as

5
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Sampling a sinusoid

By defining the digital frequency, we can write the sampled sinusoid as

• Complex exponential signal

The complex exponential signal or sequence is defined by X[n] = 𝑒 𝑗𝑤𝑛

• Rectangular or Gate function

• Unit Ramp function

r(-t), -r(t), -r(-t), -4r(t-1)?

Representation of signals in terms of elementary functions

The significance of this representation is the fact that any signal, no matter how complicated, can be
decomposed into elementary function properly scaled in amplitude and shifted in time.
In terms of Impulse function
Any discrete time signal can be expanded into the superposition of elementary shifted impulses, each one
representing each of the samples. The representation of discrete time signals in terms of impulses. The
key idea is to express an arbitrary discrete time signal as weighted sum of time shifted impulses. Given an
arbitrary sequence we can write it as a linear combination of shifted unit impulses,
where the weights of their combination are x[k]. For any given n, in the summation is expressed as

The sequence shown in figure below can be expanded as

6
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

X[n] = x[-2] δ[n+2] + x[-1] δ[n+1] + x[0] δ[n] + x[1] δ[n-1] + x[2] δ[n-2] + x[3] δ[n-3]

In terms of step function

With the use of the change of amplitude and the shifting properties of step function, we can construct a
family of pulse waveforms. The square pulse and stair-case waveforms below can be constructed by the
sum of step functions.

The same gate function can be expressed as the product of two step functions as s(t) = 4u(t-1) * u(-t+2).

Problem 1: The staircase function shown below can be mathematically expressed as

Problem 2:

𝑥(𝑡) = −4𝑢(𝑡 + 5) + 𝑢(𝑡 + 4) + 5𝑢(𝑡 + 3) + 𝑢(𝑡 − 4) − 3𝑢(𝑡 − 6)

Problem 3: Express the waveform x(t) in terms of unit step and unit ramp functions

→
x(t) = r(t+2) - 2r(t+1) + r(t-1) + 2u(t-3) - u(t-4)

7
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Problem 4

After simplification
f(t) = 2tu(t)-(4t-4)u(t-1)+(4t-12)u(t-3)-(2t-8)u(t-4)
Using ramp functions:
f(t) = 2r(t)-4r(t-1)+4r(t-3)-2r(t-4)

Lecture 2
System: is a collection of components or objects united together to serve a particular purpose. These
components or system components as they are called are unified through some kind of interdependence
among them. They react with each other. So, the whole system functions as a whole to serve a particular
purpose. You may think of a power system, the national educational system, hospital systems, railway
systems, transportation systems and so on. For example, if you are taking about an electro mechanical
system you may have the electrical variables voltages and currents associated with electrical components.
Mechanical variables like force and velocity which describe the status of a mechanical component at a
particular point in the system.
Network is a structure which resembles a kind of system. In network, the components or variables are of
all same kind.

8
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Network analysis is concerned with determining the response given the excitation and the network. In
network synthesis, the problem is to design the network given the excitation and the desires response.

Linear and Non-linear System

A system (network) is said to be linear if it obeys the principle of superposition and the principle of
homogeneity. A system which does not obey the principle of superposition and homogeneity is called a
non-linear system.
Superposition principle: If an input is the sum of several signals, then the output is the sum of the
responses of the system of each of those signals.
Homogeneity property: a system which produces an output y(t) for an input x(t) must produce an
output ay(t) for an input ax(t)

Example of linear systems: tx(t), t2x(t), sin(6t)x(t), etc

Example of non-linear system: mx(t)+c, mx2(t)

Passive Network
A linear network is passive if the energy delivered to the network is non-negative for any arbitrary
excitation, and (b) if no voltages or currents appear between any two terminals before an excitation is
applied.

Causal and Non-Causal System

A system is causal if the output depends only on values of the input at present as well as of past but not on
future inputs. This kind of system is non-anticipative as the system output does not anticipate the future
values of the input. The series RC circuit is causal, since the capacitor voltage responds only to the
present and past values of the source voltage. The motion of a car is causal, since it does not anticipate
future actions of the driver.

Causal example: y(n) = 2 x(t) + 3 x(t-3)

Non-causal example: y(n) = 2 x(t) + 3 x(t-3) + 6x(t + 3)

Continuous-time and discrete-time system

In a continuous-time system the input output relations are defined for every instant of time in a
continuous basis. In discrete time system you would have an input output relation defined at
particular discrete points along the time axis not necessarily at every point of time.

9
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Static and Dynamic System
In static or memoryless system, the outputs at present instant depends only on present inputs. These
systems are also called as memory less systems as the system output at given time is dependent only on
the inputs at that same time. Example: y(t) = 2 x(t)
Dynamic systems are those in which the output at present instant depends on past inputs and past outputs.
These are also called as systems with memory as the system output needs to store information regarding
the past inputs or outputs. Example: y(t) = 2 x(t) + 3 x(t-3)

Stable and Unstable System

Most of the control system theory involves estimation of stability of systems. Stability is an important
parameter which determines its applicability. Stability of a system is formulated in bounded input
bounded output sense i.e. a system is stable if its response is bounded for a bounded input (bounded
means finite). Example: y (t) = x2(t)
An unstable system is one in which the output of the system is unbounded for a bounded input. The
response of an unstable system diverges to infinity. Example: y (t) = ∫x(t)dt

Time-variant and time-invariant System

A system is said to be time variant system if its response varies with time. If the system response to an
input signal does not change with time such system is termed as time invariant system. The behaviour and
characteristics of time variant system are fixed over time.

10
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Lecture 3
The steady state response of a linear system to a dc and sinusoidal excitations can be found
easily by using the impedance concept. However, in practise, the input signals are of more
complex nature of non-sinusoidal periodic and non-periodic waveforms. Our aim is to generalize
the process of determining the forced response of linear network to such non-sinusoidal
functions.
Such complex functions may be represented as the sum of an infinite number of sine and cosine
functions which are harmonically related. Therefore, since the forced response to each sinusoidal
component can be determined easily by sinusoidal steady-state analysis, the response of the
linear network to the general periodic forcing function may be obtained by superposing the
partial responses.
Moreover, non-sinusoidal periodic functions are important in the analysis of nonelectrical
systems. Problems involving mechanical vibration, fluid flow, and heat flow all make use of
periodic functions. In fact, the study and analysis of heat flow in a metal rod led the French
mathematician Jean Baptiste Joseph Fourier (1768-1830) to the trigonometric series
representation of a periodic function. This series bears his name and is the starting point for
finding the steady-state response to periodic excitations of electric circuits. Fourier series can be
represented either in the form of infinite trigonometric series or infinite exponential series.
Trigonometric Form of the Fourier series
What Fourier discovered in investigating heat-flow problems is that a periodic function can be
represented by an infinite sum of sine or cosine functions that are harmonically related. In other
words, the period of any trigonometric term in the infinite series is an integral multiple, or
harmonic, of the fundamental period T of the periodic function.
The Dirichlet conditions on a periodic function f(t) are:

These are sufficient conditions, not necessary conditions. Thus if f(t) meets these requirements,
we know that we can express it as a Fourier series. However, if f(t) does not meet these
requirements, we still may be able to express it as a Fourier series. The necessary conditions on
f(t) are not known.

11
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Signal shown below is an example of signal with infinite maxima and minima

Fourier showed that this periodic function f(t) satisfying the Dirichlet conditions can be
expressed as

and where a0, an, and bn are known as Fourier coefficients that depend upon n and f(t). Equation
[1] is the trigonometric form of the Fourier series for f (t), and the process of determining the
values of the constants a0, an, and bn is called Fourier analysis.

Some useful Trigonometric Integrals

12
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

Evaluation of the Fourier coefficients

The evaluation of the unknown constants in the Fourier series may now be accomplished readily.
If we integrate each side of Eq. [1] over a full period, we obtain

But every term in the summation is of the form of Eq. [2] or [3], and thus

or

This constant a0 is simply the average value of f (t) over a period, and we therefore describe it as
the dc component of f (t).
To evaluate one of the cosine coefficients, for example, ak, the coefficient of cos kω0t, we first
multiply each side of Eq. [1] by cos kω0t and then integrate both sides of the equation over a full
period:

13
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
From Eqs. [3], [4], and [6] we note that every term on the right-hand side of this equation is zero
except for the single an term where k = n. We evaluate that term using Eq. [8], and in so doing
we find an

In a similar way, we obtain bk by multiplying by sin kω0t, integrating over a period, noting that
all but one of the terms on the right-hand side are zero, and performing that single integration by
Eq. [7]. The result is

Note that a0 is the average value of f(t), an is twice the average value of f(t) cos nω0t, and bn is
twice the average value of f(t) sin nω0t.

Problems
1) Find the Fourier Series for the periodic voltage shown below.

Solution:
The expression for v(t) between 0 and T0 is

14
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

The Fourier series for v(t) is

Our solution can be checked by truncating the Fourier series after a specific number of terms and
plugging values into. Another approach is to plot the functions and verify.

2) Find the Fourier series representation of the voltage response of a half-wave rectifier
circuit to which a sinusoidal input is applied.

Solution:
From the graph,

15
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

3) Write the Fourier series for the three voltage waveforms shown in below figure.

Answer:

16
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
4) A periodic waveform f (t) is described as follows: f (t) = −4, 0 < t < 0.3; f (t) = 6, 0.3 < t <
0.4; f (t) = 0, 0.4 < t < 0.5; T = 0.5. Evaluate a0, a3 and b1.
Answer: −1.200, 1.383, −4.44

5) Derive the expressions for a0, ak, and bk for the periodic voltage function shown if Vm =
9π V. If T = 125.66 ms, what is the frequency of the third harmonic in hertz? Also, write
the Fourier series up to and including the fifth harmonic.

Answer:

Practice:
6) Determine the Fourier series of a triangular waveform having maximum amplitude of 1 and time
period of 2π.
8 1
Answer: f(t) = ∑∝𝑛=0 𝑐𝑜𝑠(2𝑛 + 1)𝑡
𝜋2 (2𝑛+1)2

7) Determine the Fourier series of a saw tooth waveform having maximum amplitude of 1 and time
period of 2π.
Answer: a0 =0, an = 0, bn = -2/πncos 𝑛𝜋
2 1 2 1
f(t) = 𝜋 𝑠𝑖𝑛 𝑡 − 𝜋 𝑠𝑖𝑛 2𝑡 + 3𝜋 𝑠𝑖𝑛 3𝑡 − 2𝜋 𝑠𝑖𝑛 4𝑡

8) Determine the Fourier series of a trapezoidal waveform having maximum amplitude of 1 and
time period of T = 2π.
𝑛𝜋
12 𝑠𝑖𝑛( )
Answer: f(t) = ∑∝ 3
𝑠𝑖𝑛(𝑛𝑡)
𝜋 2 𝑛=1,3,5 𝑛 2

Lecture 4
Finding the Fourier coefficients, in general, is tedious. Therefore, anything that simplifies the task is
beneficial. Fortunately, a periodic function that possesses certain types of symmetry greatly reduces the
amount of work involved in finding the coefficients. It is possible to anticipate the absence of certain
terms in a Fourier series, before any integrations are performed, by an inspection of the symmetry of the
given time function.

17
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Four types of symmetry may be used to simplify the task of evaluating the Fourier coefficients:
• even-function symmetry: exhibits mirror symmetry about the vertical axis
• odd-function symmetry:
• half-wave symmetry:

The evenness, or oddness, of a periodic function can be destroyed by shifting the function along the time
axis. In other words, the judicious choice of where t = 0 may give a periodic function even or odd
symmetry.

In waveform synthesis we can see that sum of even functions is an even function. Sum of odd functions is
odd function. However, sum of even and odd function is neither even nor odd. Also, product of two even
functions is even function, product of two odd functions is also even function. However, product of an
even and an odd function is an odd function.

Even function symmetry

Now let us investigate the effect that even symmetry produces in a Fourier series. If we think of the
expression which equates an even function f (t) and the sum of an infinite number of sine and cosine
functions, then it is apparent that the sum must also be an even function. A sine wave, however, is an odd
function, and no sum of sine waves can produce any even function other than zero (which is both even
and odd). It is thus plausible that the Fourier series of any even function is composed of only a constant
and cosine functions.

18
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

In each derivation, we select t0 = -T/2 and then break the interval of integration into the range from -T/2
to 0 and 0 to T/2,

which shows that the integration from –T/2 to 0 is identical to that from 0 to T/2; therefore, a0 is

A similar examination of the expression for an leads to an integral over the half period extending from t =
0 to t = T/2:

But

The fact is that an may be obtained for an even function by taking “twice the integral over half
the range” should seem logical.
19
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Odd function symmetry
A function is defined as odd if

Addition of a constant removes the odd nature of the function f(t). For every point, such as f(t1) at t1, there
is a corresponding point –f(t1) at –t1. Therefore, a function having odd symmetry can contain no constant
term or cosine terms in its Fourier expansion.

The values of bn may again be obtained by integrating over half the range:

Half wave symmetry

What are odd harmonics and even harmonics?
Which harmonics are present in a square wave (even or odd function)?
• Will not contain even harmonics, which means only odd harmonics are present
• In sense, an and bn are zero for even values of n.
• This result is caused by another type of symmetry, called half-wave symmetry

20
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
A periodic function is said to have half-wave symmetry if

Or
Equation states that a periodic function has half-wave symmetry if, after it is shifted one-half period and
inverted, it is identical to the original function.

It may be shown that the Fourier series of any function which has half-wave symmetry contains only odd
harmonics. Let us consider the coefficients an. We have again

21
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

The dc coefficient,
It should be noted that half-wave symmetry may be present in a waveform which also shows odd
symmetry or even symmetry. When a waveform possesses half-wave symmetry and either even or odd
symmetry, then it is possible to reconstruct the waveform if the function is known over any quarter-period
interval.

If a signal has none of the symmetry discussed so far, we can obtain the even and odd functions of the
original waveforms. Fourier expansion of such non-symmetric signal will be addition of Fourier series of
odd and even functions.

22
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Problems
1) Identify which terms are absent in the Fourier series of the given waveform (T=2 s).

Answer: Half-wave and odd, so it has only odd harmonic sine terms
Half-wave and even, so it has only odd harmonic cosine terms

2) Complete the waveform for at least one full cycle. The given function has only odd-ordered sine
terms.

Solution: Since the function has only sine terms, the

function is odd. Also, the function contains only odd-ordered sine terms which means it is half-wave odd
function.

3) Complete the waveform for at least one full cycle. The given function has only odd-ordered cosine
terms.

Solution: Since
the function has only cosine terms, the function is even. Also, the function contains only odd-ordered
cosine terms which means it is half-wave even function.

4) Derive the Fourier series for the periodic voltage shown below

Answer:

23
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Lecture 5

5) Find the amplitude of significant harmonics in the input and output voltages of the given
circuit. R = 1kΩ, C = 15.9 uF, f0 = 50 Hz.

The first step in finding the steady-state response is to represent the periodic excitation source with its
Fourier series.

24
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
6) The input to circuit of figure shown below is a rectified sine wave as shown in fig. 4.26(b).

25
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
7) A linear RC circuit is excited by a periodic voltage as shown below. Determine the periodic
response vo(t).

The Fourier series of the periodic function is

Tlie voltage source as expressed above is the equivalent of infinitely many series-connected
sinusoidal sources, each source having its own amplitude and frequency. To find the contribution
of each source to the output voltage, we use the principle of superposition. For any one of the
sinusoidal sources, the phasor-domain expression for the output voltage is

All the voltage sources are expressed as sine functions, so we interpret a phasor in terms of the
sine instead of the cosine. The phasor output voltage owing to the fundamental frequency of the
sinusoidal source is

26
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

The Fourier series representation of the output voltage is

But, although we have an analytic expression for the steady-state output, what v0(t) looks like is
not immediately apparent from Eq. 16.58. As we mentioned earlier, this shortcoming is a
problem with the Fourier series approach.

8) The periodic triangular-wave voltage is applied to the circuit shown below. Derive the first
three nonzero terms in the Fourier series that represents the steady-state voltage V0 if Vm =
281.25𝜋 2 mV and the period of the input voltage is 200 𝜋 ms.

Answer:

27
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Lecture 6

Alternative Trigonometric form of Fourier Series

We have seen that the amplitude of each frequency component depends on both an and bn; that is,
the sine term and the cosine term both contribute to the amplitude. In circuit applications of the
Fourier series, we combine the cosine and sine terms in the series into a single term for
convenience. Doing so allows the representation of each harmonic of v(t) or i(t) as a single

phasor quantity. The exact expression for this amplitude is √𝑎𝑛 2 + 𝑏𝑛 2 . It is also possible to
obtain the amplitude directly by using alternative trigonometric form of Fourier series in which
each term is a cosine function with a phase angle; the amplitude and phase angle are functions of
f(t) and n. The cosine and sine terms may be merged in either a cosine expression or a sine
expression. Because we chose the cosine format in the phasor method of analysis, we choose the
cosine expression here for the alternative form of the series.

Phasor representation of cosine and sine terms is as follows

Where
When we inverse the phasor to time domain

Thus we write the alternative Fourier series as

Exponential form of Fourier series

An even more convenient and concise form of the Fourier series is obtained if the sines and
cosines are expressed as exponential functions with complex multiplying constants.

28
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

We now define a complex constant

n = 1, 2, 3,…
We may therefore express f (t) as

Evaluation of complex coefficients is obtained by substituting the Fourier coefficients an and bn

in cn

Merits
• Only Single coefficient has to be evaluated
• Simpler integration

The effect of waveform symmetry is also present in the exponential form of representation.

29
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

Problem
1) Determine the complex coefficient for the square wave

This square wave possesses both even and half-wave symmetry. If we ignore the symmetry and
use our general equation, with T = 2 and ω0 = 2π/2 = π, we have

The results are the same as those obtained when the symmetry of the waveform is not taken into account.

30
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
2) A certain function f(t) is a train of rectangular pulses of amplitude V0 and duration τ,
recurring periodically every T seconds, as shown in figure below. Find the exponential
Fourier series for f(t).

Lecture 7
Frequency Spectra
We can present graphically the description of a periodic function in terms of the amplitude and
phase angle of each term in the Fourier series of f(t). The plot of the amplitude of each term
versus the frequency is called the amplitude spectrum of f(t), and the plot of the phase angle
versus the frequency is called the phase spectrum of f(t). Because the amplitude and phase angle
data occur at discrete values of the frequency (that is, at w0, 2w0, 3w0,...), these plots also are
referred to as line spectra. The An is the general amplitude quantity which we must be shown in
a single sided line spectrum. The sample amplitude and phase spectrum are shown if Fig below.
The x-axis can be order of harmonics ‘n’ or frequency ‘nω’.
Example: Plot an amplitude and phase spectrum of a periodic function shown below with Vm
as 40 V and n = 0,1,2,3,…

Alternative trigonometric form

31
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

Single sided spectrum Two-sided spectrum

Exponential form

Fourier Spectrum from the Exponential Form

The phase angle in complex form ∅𝑛 is negative of the phase angle 𝜃𝑛 (∅𝑛 = −𝜃𝑛) . The 𝐶−𝑛 is
the conjugate of 𝐶𝑛 , which means angle associated with ‘-n’ is negative of angle associated with
‘n’ i.e., ∅−𝑛 = 𝜃𝑛 .

Average Power and Effective RMS Calculation

If we have the Fourier series representation of the voltage and current at a pair of terminals in a
linear lumped-parameter circuit, we can easily express the average power at the terminals as a
function of the harmonic voltages and currents. Using the alternative trigonometric form of the
Fourier series, we write the periodic voltage and current at the terminals of a network as

32
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

We assume that the current reference is in the direction of the reference voltage drop across the
terminals (using the passive sign convention), so that the instantaneous power at the terminals is
vi. The average power is

To find the expression for the average power, we substitute Eqs. 16.73 and 16.74 into Eq. 16.75
and integrate. At first glance, this appears to be a formidable task, because the product vi requires
multiplying two infinite series. If we expand, the integral of product of a constant and a cosine
term is zero. The integral of product of two cosine functions of different frequencies over a
period T will also yield zero. The only terms to survive integration are the products of voltage
and current at the same frequency and DC terms. Therefore Eq. 16.75 reduces to

33
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Equation 16.78 is particularly important because it states that in the case of an interaction
between a periodic voltage and the corresponding periodic current, the total average power is the
sum of the average powers obtained from the interaction of currents and voltages of the same
frequency. Currents and voltages of different frequencies do not interact to produce average
power. Therefore, in average-power calculations involving periodic functions, the total average
power is the superposition of the average powers associated with each harmonic voltage and
current.

Total Harmonic Active Power?

RMS Value

The integral of the squared time function simplifies because the only terms to survive integration
over a period are the product of the dc term and the harmonic products of the same frequency.
All other products integrate to zero. Therefore aforementioned equation reduces to

The above equation states that the RMS value of a periodic function is the square root of the sum
obtained by adding the square of the RMS value of each harmonic to the square of the dc value.
This will yield an estimate of the near true value.

Frms in terms of Cn ?

34
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Example
1. Let's assume that a periodic voltage is represented by the finite series

The RMS value of voltage is

2. Assume that the periodic square-wave voltage is applied across the terminals of a 15-ohm
resistor. The value of Vm is 60 V, and that of T is 5 ms. a) Estimate the RMS value of the voltage
for the five terms b) Calculate the average power
associated with first five non-zero terms. c) Calculate
the total average power delivered to the 15Ω resistor
d) What percentage of the total power is delivered by
the first five terms of the Fourier series?

a)

Thus, using the first five nonzero terms of the Fourier series,

35
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

b) The voltage is applied to the terminals of a resistor, so we can find the power associated with each term
as

The total power delivered by the first five nonzero terms is P = 55.15 W
c) To obtain the total average power delivered to the 15-ohm resistor, we first calculate the RMS value of
v(t) and the average power delivered to the resistor.

d) The percentage of power delivered is 55.15/60*100 = 91.92%.

3. The triangular-wave voltage source vg(t) is applied to the circuit in figure below. Estimate
the average power delivered to the resistor when the circuit is in steady-state operation.

Solution:

36
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

2
P = 𝑉𝑅𝑀𝑆 /R = 1.85 W

4. A circuit shown below has an applied voltage v(t). Determine the effective RMS voltage,
effective RMS current, average power and power factor.
Answer: 138.07 V, 16.58 A, 1374 W and 0.6

37
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

P = 1374 W
S = 𝑉𝑅𝑀𝑆 𝐼𝑅𝑀𝑆 = 2289.15 VA
PF = P/S = 0.6

5.

Now by solving the above two equations

R = 2 Ω and C = 50 µF
The power of the circuit is P = 238 W, IRMS = 10.9041 A, VRMS = 64.8 V and S = 706.668
VA.

38
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Lecture 8
Additional Properties
The examples showed that the size of the various harmonic components and the powers
associated with the harmonics go down as the other harmonic increases you should like to see,
how fast these various co-efficient go down and that is; related to what is meant by, convergence
of the Fourier series you should look at the convergence like to look at the convergent properties
of the Fourier series.
1. Convergence: The Mean square error decreases monotonically with increase in ‘N’
number of terms.

2. Time shifting of functions, amplitude of the spectrum of f(t) and f(t-T) are equal

3. Multiplication of function by sinusoid

The Fourier series representation enables us to describe the periodic function in terms of the
frequency-domain attributes of amplitude and phase. The Fourier transform extends this
frequency-domain description to functions that are not periodic (aperiodic).

Case 1: T is small (periodic)

39
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

The line spectrum is then obtained by simply erecting a vertical line at each harmonic frequency, as
shown in the sketch. The amplitudes shown are those of the cn. The particular case sketched applies to the
case where τ/T = 1/(1.5π) = 0.212. In Fig. b, the amplitude of the sinusoidal component is plotted as a
function of frequency.
Frequency resolution? Δω

The remarks on the discrete spectrum

• The spectrum was discrete in the sense that it was not a smooth or continuous function of
frequency; instead, it had nonzero values only at specific frequencies.
• “width” of the envelope depends upon τ, and not upon T.
• A shift in the time origin does not change the amplitude and phase spectrum; that is, |cn| is not a
function of t0. The relative phases of the frequency components do change with the choice of t0.
• If the pulse period T is increased (or the pulse repetition frequency f0 is decreased), the bandwidth
1/τ does not change, but the number of spectral lines between zero frequency and 1/τ Hz increases,
albeit discontinuously; the amplitude of each line is inversely proportional to T.

40
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Fourier Transform and Fourier Integral
There are many important forcing functions, that are aperiodic, such as a single rectangular pulse, a step
function, a ramp function, or the impulse function. Frequency spectra may be obtained for such aperiodic
functions, but they will be continuous spectra in which some energy, in general, may be found in any
nonzero frequency interval, no matter how small. We will develop this concept by beginning with a
periodic function and then letting the period become infinite.

Case 2: T is large with the same τ

The envelope will decrease in amplitude without otherwise changing shape, and that more and
more frequency components will be found in any given frequency interval. In the limit-time
period, we should expect an envelope of vanishingly small amplitude, filled with an infinite
number of frequency components separated by vanishingly small frequency intervals.

Case 3: T is infinite
As period, T approaches infinity, it implies a non-periodic waveform having only one pulse of
duration τ, which never repeats. The discrete-line spectrum approaches the continuous spectrum.
The number of frequency components between 0 and 100 Hz, for example, becomes infinite, but
the amplitude of each one approaches zero. At first thought, a spectrum of zero amplitude is a
puzzling concept. We know that the line spectrum of a periodic forcing function
shows the amplitude of each frequency component. But what does the zero amplitude continuous
spectrum of an aperiodic forcing function signify?

We now modify the Fourier series expansion for periodic functions, such that it could represent
non-periodic transient functions. We begin the derivation of the Fourier transform, viewed as a
limiting case of a Fourier series, with the exponential form of the series:

As , the incremental separation Δω0 becomes smaller

and smaller, it approaches a differential separation dω0.
Thus,

As the period increases, the frequency moves from being a discrete variable to becoming a
continuous variable, or

As the period increases, the Fourier coefficients Cω get smaller. In the limit, Cω → 0 as T →∞.
This result makes sense, because we expect the Fourier coefficients to vanish as the function
loses its periodicity.
41
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
and

If we multiply each side of the equation by the period T and then undertake the limiting process,
a nontrivial result is obtained that is the coefficient density

This expression is a function of ω (and not of t), and we represent it by F(ω) or F(jω):

Now let us apply the limiting process to f(t). We begin by multiplying and dividing the
summation by T

As T →∞, the summation approaches integration and CwT → F(ω), the equation becomes

Fourier coefficient density spectrum (magnitude density: volt-sec, amp-sec)

Lecture 9
Properties of Fourier Transform
The first mathematical property we call to your attention is that F(w) is a complex quantity and
can be expressed in either rectangular or polar form. Thus from the defining integral,

Where

Note:
1. The A(ω) is an even function of ω; in other words, A(ω) = A(-ω). The imaginary part of
F(ω), B(ω) is an odd function of ω; in other words, B(ω) = -B(-ω).
2. Magnitude is even function and phase angle
is an odd function.

42
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
3. If f(t) is an even function, its Fourier transform is an even function, and if f(t) is an odd
function, its Fourier transform is an odd function.
4. If f(t) is an even function, F(ω) is real, i.e., B(ω) =0 and

5. If f(t) is an odd function, F(ω) is imaginary i.e., A(ω)=0 and

Multiplication by a Constant
From the defining integral, if

Thus, multiplication of f(t) by a constant corresponds to multiplying F(ω) by that same constant.

Addition or subtraction

43
NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

44
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

Reversal in Time doamin

Linearity

Summary:

45
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Fourier Transform of Basic functions

46
NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

let

47
NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

let

48
NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

49
NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

50
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

Problem: Find Fourier Transform of the given function x(t) using Translation and differentiation
properties

51
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Lecture 10
System function
In practice we are often faced with circuits to which arbitrary sources can be connected, and
require an efficient means of determining the new output each time. This is easily accomplished
if we can characterize the basic circuit by a transfer function called the system function. The
analysis can proceed in either the time domain or the frequency domain, although it is generally
more useful to work in the frequency domain.
As we have noted several times before, the output vo(t) at some point in a linear circuit can be
obtained by convolving the input vi(t) with the unit impulse response h(t). However, we must
remember that the impulse response results from the application of a unit impulse at t = 0 with
all initial conditions zero. Subsequently, we found that it was often more convenient to perform
such operations in the frequency domain, as the Fourier transform of the convolution of two
functions is simply the product of each function in the frequency domain.

By these means some relatively complicated integral expressions will be reduced to simple
functions of jw, and the mathematical operations of integration and differentiation will be
replaced by the simpler operations of algebraic multiplication and division.
In such situations, we have a simple four-step process in the frequency domain:
1. Determine the circuit system function H(jw) (if not already known)
• obtain the system function H(jω) by assuming that the input and output voltages are
both sinusoids described by their corresponding phasors
2. Obtain the Fourier transform of the forcing function to be applied i.e., Fi(jw)
3. Multiply this transform and the system function; and finally Fo(jw) = H(jw) Fi(jw)
4. Perform an inverse transform operation on the product to find the output. i.e, Fo(t)

Circuit application problems

1. Find the voltage across the inductor of the circuit shown in Fig. when the input voltage is a
simple exponentially decaying pulse, as indicated.
Solution:

52
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

2. Use Fourier transform techniques on the circuit of Fig below to find i1(t) at t = 1.5ms if is(t)
equals (a) δ(t) A; (b) u(t) A; (c) cos 500t A.

Answer:
3.

Answers:

3. When the input voltage to the system shown in figure below is 15u(t) V, the output voltage is
What is the output voltage if Vj = 15sgn(t) V?

53
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

4. Use the Fourier transform method to find v0 in the circuit when

Also, find , and do the voltage values obtained make sense in terms of
known circuit behavior? Explain.

54
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

5. Find the Fourier transform of given function

Answer:

Lecture 11

THE FOURIER TRANSFORM OF A GENERAL PERIODIC TIME FUNCTION

Consider a periodic time function f (t) with period T and Fourier series expansion is

Keeping in mind that the Fourier transform of a sum is just the sum of the FT of the terms in the
sum, and that cn is not a function of time, we can write
Using the properties of Fourier Transform

55
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021

After obtaining the transform of ejnω0t from expression, we have

This shows that F(ω) is a discrete spectrum consisting of impulses located at points on the ω axis
given by ω = nωo, n =...,-2,-1,0,1,.... The strength of each impulse is 2π times the value of the
corresponding Fourier coefficient appearing in the complex form of the Fourier series expansion
for f(t).

To demonstrate this, consider f (t) = cosω0t. First we evaluate the Fourier coefficients cn.

Then

This expression has values that are nonzero only when n = ±1, and it follows, therefore, that the
entire summation reduces to

which is precisely the expression that we obtained before

Finding Fourier series from the Fourier Transform

56
 NLS, UNIT-I, KARTHIK THIRUMALA - JULY 2021
Find the Fourier series and Fourier Transform of rectangular Wave

For periodic signals,

 
2  n0 d 
F ( j ) = 2  c  ( − n )
n =−
n 0 F ( j ) =  n sin 
n =− 2 
 ( − n0 )

57