420

Fourier Method of
Waveform Analysis
17.1 INTRODUCTION
In the circuits examined previously, the response was obtained for excitations having constant,
sinusoidal, or exponential form. In such cases a single expression described the forcing function for
all time; for instance, v = constant or v = V sin ot, as shown in Fig. 17-1(a) and (b).
Certain periodic waveforms, of which the sawtooth in Fig. 17-1(c) is an example, can be only locally
dened by single functions. Thus, the sawtooth is expressed by f (t) = (V,T)t in the interval 0 - t - T
and by f (t) = (V,T)(t T) in the interval T - t - 2T. While such piecemeal expressions describe the
waveform satisfactorily, they do not permit the determination of the circuit response. Now, if a periodic
function can be expressed as the sum of a nite or innite number of sinusoidal functions, the responses
of linear networks to nonsinusoidal excitations can be determined by applying the superposition
theorem. The Fourier method provides the means for solving this type of problem.
In this chapter we develop tools and conditions for such expansions. Periodic waveforms may be
expressed in the form of Fourier series. Nonperiodic waveforms may be expressed by their Fourier
transforms. However, a piece of a nonperiodic waveform specied over a nite time period may also be
expressed by a Fourier series valid within that time period. Because of this, the Fourier series analysis is
the main concern of this chapter.
Fig. 17-1
Copyright 2003, 1997, 1986, 1965 by The McGraw-Hill Companies, Inc. Click Here for Terms of Use.
17.2 TRIGONOMETRIC FOURIER SERIES
Any periodic waveformthat is, one for which f (t) = f (t T)can be expressed by a Fourier series
provided that
(1) If it is discontinuous, there are only a nite number of discontinuities in the period T;
(2) It has a nite average value over the period T;
(3) It has a nite number of positive and negative maxima in the period T.
When these Dirichlet conditions are satised, the Fourier series exists and can be written in trigonometric
form:
f (t) =
1
2
a
0
a
1
cos ot a
2
cos 2t a
3
cos 3ot
b
1
sin ot b
2
sin 2ot b
3
sin 3ot (1)
The Fourier coecients, as and bs, are determined for a given waveform by the evaluation
integrals. We obtain the cosine coecient evaluation integral by multiplying both sides of (1) by
cos not and integrating over a full period. The period of the fundamental, 2,o, is the period of the
series since each term in the series has a frequency which is an integral multiple of the fundamental
frequency.
_
2,o
0
f (t) cos not dt =
_
2,o
0
1
2
a
0
cos not dt
_
2,o
0
a
1
cos ot cos not dt

_
2,o
0
a
n
cos
2
not dt
_
2,o
0
b
1
sin ot cos not dt

_
2,o
0
b
2
sin 2ot cos nodt (2)
The denite integrals on the right side of (2) are all zero except that involving cos
2
not, which has the
value (,o)a
n
. Then
a
n
=
o

_
2,o
0
f (t) cos not dt =
2
T
_
T
0
f (t) cos
2nt
T
dt (3)
Multiplying (1) by sin not and integrating as above results in the sine coecient evaluation integral.
b
n
=
o

_
2,o
0
f (t) sin not dt =
2
T
_
T
0
f (t) sin
2nt
T
dt (4)
An alternate form of the evaluation integrals with the variable [ = ot and the corresponding period
2 radians is
a
n
=
1

_
2
0
F([) cos n[d[ (5)
b
n
=
1

_
2
0
F([) sin n[d[ (6)
where F([) = f ([,o). The integrations can be carried out from T,2 to T,2, to , or over any
other full period that might simplify the calculation. The constant a
0
is obtained from (3) or (5) with
n = 0; however, since
1
2
a
0
is the average value of the function, it can frequently be determined by
inspection of the waveform. The series with coecients obtained from the above evaluation integrals
converges uniformly to the function at all points of continuity and converges to the mean value at points
of discontinuity.
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 421
EXAMPLE 17.1 Find the Fourier series for the waveform shown in Fig. 17-2.
The waveform is periodic, of period 2,o in t or 2 in ot. It is continuous for 0 - ot - 2 and given therein
by f (t) = (10,2)ot, with discontinuities at ot = n2 where n = 0. 1. 2. . . . . The Dirichlet conditions are satised.
The average value of the function is 5, by inspection, and thus,
1
2
a
0
= 5. For n > 0, (5) gives
a
n
=
1

_
2
0
10
2
_ _
ot cos not d(ot) =
10
2
2
ot
n
sin not
1
n
2
cos not
_ _
2
0
=
10
2
2
n
2
(cos n2 cos 0) = 0
Thus, the series contains no cosine terms. Using (6), we obtain
b
n
=
1

_
2
0
10
2
_ _
ot sinnot d(ot) =
10
2
2

ot
n
cos not
1
n
2
sin not
_ _
2
0
=
10
n
Using these sine-term coecients and the average term, the series is
f (t) = 5
10

sinot
10
2
sin2ot
10
3
sin3ot = 5
10

o
n=1
sinnot
n
The sine and cosine terms of like frequency can be combined as a single sine or cosine term with a
phase angle. Two alternate forms of the trigonometric series result.
f (t) =
1
2
a
0

c
n
cos (not
n
) (7)
and f (t) =
1
2
a
0

c
n
sin (not
n
) (8)
where c
n
=

a
2
n
b
2
n
_
,
n
= tan
1
(b
n
,a
n
), and
n
= tan
1
(a
n
,b
n
). In (7) and (8), c
n
is the harmonic
amplitude, and the harmonic phase angles are
n
or
n
.
17.3 EXPONENTIAL FOURIER SERIES
A periodic waveform f (t) satisfying the Dirichlet conditions can also be written as an exponential
Fourier series, which is a variation of the trigonometric series. The exponential series is
f (t) =

o
n=o
A
n
e
jnot
(9)
To obtain the evaluation integral for the A
n
coecients, we multiply (9) on both sides by e
jnot
and
integrate over the full period:
422 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-2
_
2
0
f (t)e
jnot
d(ot) =
_
2
0
A
2
e
j2ot
e
jnot
d(ot)
_
2
0
A
1
e
jot
e
jnot
d(ot)

_
2
0
A
0
e
jnot
d(ot)
_
2
0
A
1
e
jot
e
jnot
d(ot)

_
2
0
A
n
e
jnot
e
jnot
d(ot) (10)
The denite integrals on the right side of (10) are all zero except
_
2
0
A
n
d(ot), which has the value 2A
n
.
Then
A
n
=
1
2
_
2
0
f (t)e
jnot
d(ot) or A
n
=
1
T
_
T
0
f (t)e
j2nt,T
dt (11)
Just as with the a
n
and b
n
evaluation integrals, the limits of integration in (11) may be the endpoints
of any convenient full period and not necessarily 0 to 2 or 0 to T. Note that, f (t) being real, A
n
= A
+
n
,
so that only positive n needed to be considered in (11). Furthermore, we have
a
n
= 2 Re A
n
b
n
= 2 ImA
n
(12)
EXAMPLE 17.2 Derive the exponential series (9) from the trigonometric series (1).
Replace the sine and cosine terms in (1) by their complex exponential equivalents.
sinnot =
e
jnot
e
jnot
2j
cos not =
e
jnot
e
jnot
2
Arranging the exponential terms in order of increasing n from o to o, we obtain the innite sum (9) where
A
0
= a
0
,2 and
A
n
=
1
2
(a
n
jb
n
) A
n
=
1
2
(a
n
jb
n
) for n = 1. 2. 3. . . .
EXAMPLE 17.3 Find the exponential Fourier series for the waveform shown in Fig. 17-2. Using the coecients
of this exponential series, obtain a
n
and b
n
of the trigonometric series and compare with Example 17.1.
In the interval 0 - ot - 2 the function is given by f (t) = (10,2)ot. By inspection, the average value of the
function is A
0
= 5. Substituting f (t) in (11), we obtain the coecients A
n
.
A
n
=
1
2
_
2
0
10
2
_ _
ote
jnot
d(ot) =
10
(2)
2
e
jnot
(jn)
2
(jnot 1)
_ _2
0
= j
10
2n
Inserting the coecients A
n
in (12), the exponential form of the Fourier series for the given waveform is
f (t) = j
10
4
e
j2ot
j
10
2
e
jot
5 j
10
2
e
jot
j
10
4
e
j2ot
(13)
The trigonometric series coecients are, by (12),
a
n
= 0 b
n
=
10
n
f (t) = 5
10

sinot
10
2
sin2ot
10
3
sin 3ot and so
which is the same as in Example 17.1.
17.4 WAVEFORM SYMMETRY
The series obtained in Example 17.1 contained only sine terms in addition to a constant term. Other
waveforms will have only cosine terms; and sometimes only odd harmonics are present in the series,
whether the series contains sine, cosine, or both types of terms. This is the result of certain types of
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 423
symmetry exhibited by the waveform. Knowledge of such symmetry results in reduced calculations in
determining the Fourier series. For this reason the following denitions are important.
1. A function f (x) is said to be even if f (x) = f (x).
The function f (x) = 2 x
2
x
4
is an example of even functions since the functional values for x and
x are equal. The cosine is an even function, since it can be expressed as the power series
cos x = 1
x
2
2!

x
4
4!

x
6
6!

x
8
8!

The sum or product of two or more even functions is an even function, and with the addition of a
constant the even nature of the function is still preserved.
In Fig. 17-3, the waveforms shown represent even functions of x. They are symmetrical with respect
to the vertical axis, as indicated by the construction in Fig. 17-3(a).
2. A function f (x) is said to be odd if f (x) = f (x).
The function f (x) = x x
3
x
5
is an example of odd functions since the values of the function for x
and x are of opposite sign. The sine is an odd function, since it can be expressed as the power series
sin x = x
x
3
3!

x
5
5!

x
7
7!

x
9
9!

The sum of two or more odd functions is an odd function, but the addition of a constant removes
the odd nature of the function. The product of two odd functions is an even function.
The waveforms shown in Fig. 17-4 represent odd functions of x. They are symmetrical with respect
to the origin, as indicated by the construction in Fig. 17-4(a).
424 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-3
Fig. 17-4 Fig. 17-5
3. A periodic function f (x) is said to have half-wave symmetry if f (x) = f (x T,2) where T is the
period. Two waveforms with half-wave symmetry are shown in Fig. 17-5.
When the type of symmetry of a waveform is established, the following conclusions are reached. If
the waveform is even, all terms of its Fourier series are cosine terms, including a constant if the wave-
form has a nonzero average value. Hence, there is no need of evaluating the integral for the coecients
b
n
, since no sine terms can be present. If the waveform is odd, the series contains only sine terms. The
wave may be odd only after its average value is subtracted, in which case its Fourier representation will
simply contain that constant and a series of sine terms. If the waveform has half-wave symmetry, only
odd harmonics are present in the series. This series will contain both sine and cosine terms unless the
function is also odd or even. In any case, a
n
and b
n
are equal to zero for n = 2. 4. 6. . . . for any
waveform with half-wave symmetry. Half-wave symmetry, too, may be present only after subtraction
of the average value.
Certain waveforms can be odd or even, depending upon the location of the vertical axis. The square
wave of Fig. 17-6(a) meets the condition of an even function: f (x) = f (x). A shift of the vertical axis to
the position shown in Fig. 17-6(b) produces an odd function f (x) = f (x). With the vertical axis placed
at any points other than those shown in Fig. 17-6, the square wave is neither even nor odd, and its series
contains both sine and cosine terms. Thus, in the analysis of periodic functions, the vertical axis should be
conveniently chosen to result in either an even or odd function, if the type of waveformmakes this possible.
The shifting of the horizontal axis may simplify the series representation of the function. As an
example, the waveform of Fig. 17-7(a) does not meet the requirements of an odd function until the
average value is removed as shown in Fig. 17-7(b). Thus, its series will contain a constant term and only
sine terms.
The preceding symmetry considerations can be used to check the coecients of the exponential
Fourier series. An even waveform contains only cosine terms in its trigonometric series, and therefore
the exponential Fourier coecients must be pure real numbers. Similarly, an odd function whose
trigonometric series consists of sine terms has pure imaginary coecients in its exponential series.
17.5 LINE SPECTRUM
A plot showing each of the harmonic amplitudes in the wave is called the line spectrum. The lines
decrease rapidly for waves with rapidly convergent series. Waves with discontinuities, such as the
sawtooth and square wave, have spectra with slowly decreasing amplitudes, since their series have strong
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 425
Fig. 17-6 Fig. 17-7
high harmonics. Their 10th harmonics will often have amplitudes of signicant value as compared to
the fundamental. In contrast, the series for waveforms without discontinuities and with a generally
smooth appearance will converge rapidly, and only a few terms are required to generate the wave. Such
rapid convergence will be evident from the line spectrum where the harmonic amplitudes decrease
rapidly, so that any above the 5th or 6th are insignicant.
The harmonic content and the line spectrum of a wave are part of the very nature of that wave and
never change, regardless of the method of analysis. Shifting the origin gives the trigonometric series a
completely dierent appearance, and the exponential series coecients also change greatly. However,
the same harmonics always appear in the series, and their amplitudes,
c
0
= [
1
2
a
0
[ and c
n
=

a
2
n
b
2
n
_
(n _ 1) (14)
or c
0
= [A
0
[ and c
n
= [A
n
[ [A
n
[ = 2[A
n
[ (n _ 1) (15)
remain the same. Note that when the exponential form is used, the amplitude of the nth harmonic
combines the contributions of frequencies no and no.
EXAMPLE 17.4 In Fig. 17-8, the sawtooth wave of Example 17.1 and its line spectrum are shown. Since there
were only sine terms in the trigonometric series, the harmonic amplitudes are given directly by
1
2
a
0
and [b
n
[. The
same line spectrum is obtained from the exponential Fourier series, (13).
17.6 WAVEFORM SYNTHESIS
Synthesis is a combination of parts so as to form a whole. Fourier synthesis is the recombination of
the terms of the trigonometric series, usually the rst four or ve, to produce the original wave. Often it
is only after synthesizing a wave that the student is convinced that the Fourier series does in fact
represent the periodic wave for which it was obtained.
The trigonometric series for the sawtooth wave of Fig. 17-8 is
f (t) = 5
10

sin ot
10
2
sin 2ot
10
3
sin 3ot
These four terms are plotted and added in Fig. 17-9. Although the result is not a perfect sawtooth wave,
it appears that with more terms included the sketch will more nearly resemble a sawtooth. Since this
wave has discontinuities, its series is not rapidly convergent, and consequently, the synthesis using only
four terms does not produce a very good result. The next term, at the frequency 4o, has amplitude 10/
4, which is certainly signicant compared to the fundamental amplitude, 10/. As each term is added
in the synthesis, the irregularities of the resultant are reduced and the approximation to the original wave
is improved. This is what was meant when we said earlier that the series converges to the function at all
points of continuity and to the mean value at points of discontinuity. In Fig. 17-9, at 0 and 2 it is clear
that a value of 5 will remain, since all sine terms are zero at these points. These are the points of
discontinuity; and the value of the function when they are approached from the left is 10, and from the
right 0, with the mean value 5.
426 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-8
17.7 EFFECTIVE VALUES AND POWER
The eective or rms value of the function
f (t) =
1
2
a
0
a
1
cos ot a
2
cos 2ot b
1
sin ot b
2
sin 2ot
is F
rms
=

(
1
2
a
0
)
2

1
2
a
2
1

1
2
a
2
2

1
2
b
2
1

1
2
b
2
2

_
=

c
2
0

1
2
c
2
1

1
2
c
2
2

1
2
c
3
3

_
(16)
where (14) has been used.
Considering a linear network with an applied voltage which is periodic, we would expect that the
resulting current would contain the same harmonic terms as the voltage, but with harmonic amplitudes
of dierent relative magnitude, since the impedance varies with no. It is possible that some harmonics
would not appear in the current; for example, in a pure LC parallel circuit, one of the harmonic
frequencies might coincide with the resonant frequency, making the impedance at that frequency
innite. In general, we may write
v = V
0

V
n
sin (not
n
) and i = I
0

I
n
sin (not [
n
) (17)
with corresponding eective values of
V
rms
=

V
2
0

1
2
V
2
1

1
2
V
2
2

_
and I
rms
=

I
2
0

1
2
I
2
1

1
2
I
2
2

_
(18)
The average power P follows from integration of the instantaneous power, which is given by the
product of v and i:
p = vi = V
0

V
n
sin (not
n
)
_ _
I
0

I
n
sin (not [
n
)
_ _
(19)
Since v and i both have period T, their product must have an integral number of its periods in T.
(Recall that for a single sine wave of applied voltage, the product vi has a period half that of the voltage
wave.) The average may therefore be calculated over one period of the voltage wave:
P =
1
T
_
T
0
V
0

V
n
sin (not
n
)
_ _
I
0

I
n
sin (not [
n
)
_ _
dt (20)
Examination of the possible terms in the product of the two innite series shows them to be of the
following types: the product of two constants, the product of a constant and a sine function, the product
of two sine functions of dierent frequencies, and sine functions squared. After integration, the product
of the two constants is still V
0
I
0
and the sine functions squared with the limits applied appear as
(V
n
I
n
,2) cos (
n
[
n
); all other products upon integration over the period T are zero. Then the average
power is
P = V
0
I
0

1
2
V
1
I
1
cos
1

1
2
V
2
I
2
cos
2

1
2
V
3
I
3
cos
3
(21)
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 427
Fig. 17-9
where
n
=
n
[
n
is the angle on the equivalent impedance of the network at the angular frequency no,
and V
n
and I
n
are the maximum values of the respective sine functions.
In the special case of a single-frequency sinusoidal voltage, V
0
= V
2
= V
3
= = 0, and (21)
reduces to the familiar
P =
1
2
V
1
I
1
cos
1
= V
eff
I
eff
cos
Compare Section 10.2. Also, for a dc voltage, V
1
= V
2
= V
3
= = 0, and (21) becomes
P = V
0
I
0
= VI
Thus, (21) is quite general. Note that on the right-hand side there is no term that involves voltage and
current of dierent frequencies. In regard to power, then, each harmonic acts independently, and
P = P
0
P
1
P
2

17.8 APPLICATIONS IN CIRCUIT ANALYSIS
It has already been suggested above that we could apply the terms of a voltage series to a linear
network and obtain the corresponding harmonic terms of the current series. This result is obtained by
superposition. Thus we consider each term of the Fourier series representing the voltage as a single
source, as shown in Fig. 17.10. Now the equivalent impedance of the network at each harmonic
frequency no is used to compute the current at that harmonic. The sum of these individual responses
is the total response i, in series form, to the applied voltage.
EXAMPLE 17.5 A series RL circuit in which R = 5 and L = 20 mH (Fig. 17-11) has an applied voltage
v = 100 50 sinot 25 sin3ot (V), with o = 500 rad/s. Find the current and the average power.
Compute the equivalent impedance of the circuit at each frequency found in the voltage function. Then obtain
the respective currents.
At o = 0, Z
0
= R = 5 and
I
0
=
V
0
R
=
100
5
= 20 A
At o = 500 rad/s, Z
1
= 5 j(500)(20 10
3
) = 5 j10 = 11.15 63.48 and
i
1
=
V
1.max
Z
1
sin(ot
1
) =
50
11.15
sin(ot 63.48) = 4.48 sin(ot 63.48) (A)
At 3o = 1500 rad/s, Z
3
= 5 j30 = 30.4 80.548 and
i
3
=
V
3.max
Z
3
sin(3ot
3
) =
25
30.4
sin (3ot 80.548) = 0.823 sin (3ot 80.548) (A)
The sum of the harmonic currents is the required total response; it is a Fourier series of the type (8).
i = 20 4.48 sin(ot 63.48) 0.823 sin (3ot 80.548) (A)
This current has the eective value
I
eff
=

20
2
(4.48
2
,2) (0.823
2
,2)
_
=

410.6
_
= 20.25 A
which results in a power in the 5- resistor of
P = I
2
eff
R = (410.6)5 = 2053 W
As a check, we compute the total average power by calculating rst the power contributed by each harmonic
and then adding the results.
At o = 0: P
0
= V
0
I
0
= 100(20) = 2000 W
At o = 500 rad/s: P
1
=
1
2
V
1
I
1
cos
1
=
1
2
(50)(4.48) cos 63.48 = 50.1 W
At 3o = 1500 rad/s: P
3
=
1
2
V
3
I
3
cos
3
=
1
2
(25)(0.823) cos 80.548 = 1.69 W
Then, P = 2000 50.1 1.69 = 2052 W
428 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Another Method
The Fourier series expression for the voltage across the resistor is
v
R
= Ri = 100 22.4 sin (ot 63.48) 4.11 sin(3ot 80.548) (V)
V
Reff
=

100
2

1
2
(22.4)
2

1
2
(4.11)
2
_
=

10 259
_
= 101.3 V and
Then the power delivered by the source is P = V
2
Reff
,R = (10 259),5 = 2052 W.
In Example 17.5 the driving voltage was given as a trigonometric Fourier series in t, and the
computations were in the time domain. (The complex impedance was used only as a shortcut; Z
n
and
n
could have been obtained directly from R, L, and no). If, instead, the voltage is represented by an
exponential Fourier series,
v(t) =

o
o
V
n
e
jnot
then we have to do with a superposition of phasors V
n
(rotating counterclockwise if n > 0, clockwise if
n - 0), and so frequency-domain methods are called for. This is illustrated in Example 17.6.
EXAMPLE 17.6 A voltage represented by the triangular wave shown in Fig. 17-12 is applied to a pure capacitor C.
Determine the resulting current.
In the interval - ot - 0 the voltage function is v = V
max
(27V
max
,)ot; and for 0 - ot - ,
v = V
max
(2V
max
,)ot. Then the coecients of the exponential series are determined by the evaluation integral
V
n
=
1
2
_
0

[V
max
(2V
max
,)ot]e
jnot
d(ot)
1
2
_

0
[V
max
(2V
max
,)ot]e
jnot
d(ot)
from which V
n
= 4V
max
,
2
n
2
for odd n, and V
n
= 0 for even n.
The phasor current produced by V
n
(n odd) is
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 429
Fig. 17-10 Fig. 17-11
Fig. 17-12
I
n
=
V
n
Z
n
=
4V
max
,
2
n
2
1,jnoC
= j
4V
max
oC

2
n
with an implicit time factor e
jnot
. The resultant current is therefore
i(t) =

o
o
I
n
e
jnot
= j
4V
max
oC

o
o
e
jnot
n
where the summation is over odd n only.
The series could be converted to the trigonometric form and then synthesized to show the current waveform.
However, this series is of the same form as the result in Problem 17.8, where the coecients are A
n
= j(2V,n) for
odd n only. The sign here is negative, indicating that our current wave is the negative of the square wave of Problem
17.8 and has a peak value 2V
max
oC,.
17.9 FOURIER TRANSFORM OF NONPERIODIC WAVEFORMS
A nonperiod waveform x(t) is said to satisfy the Dirichlet conditions if
(a) x(t) is absolutely integrable,
_
o
o
[x(t)[ dt - o, and
(b) the number of maxima and minima and the number of discontinuities of x(t) in every nite
interval is nite.
For such a waveform, we can dene the Fourier transform X( f ) by
X( f ) =
_
o
o
x(t)e
j2ft
dt (22a)
where f is the frequency. The above integral is called the Fourier integral. The time function x(t) is
called the inverse Fourier transform of X( f ) and is obtained from it by
x(t) =
_
o
o
X( f )e
j2ft
df (22b)
x(t) and X( f ) form a Fourier transform pair. Instead of f , the angular velocity o = 2f may also be
used, in which case, (22a) and (22b) become, respectively,
X(o) =
_
o
o
x(t)e
jot
dt (23a)
and x(t) =
1
2
_
o
o
X(o)e
jot
do (23b)
EXAMPLE 17.7 Find the Fourier transform of x(t) = e
at
u(t), a > 0. Plot X( f ) for o - f - o.
From (22a), the Fourier transform of x(t) is
X( f ) =
_
o
0
e
at
e
j2ft
dt =
1
a j2f
(24)
X( f ) is a complex function of a real variable. Its magnitude and phase angle, [X( f )[ and X( f ), respectively, shown
in Figs. 17-13(a) and (b), are given by
[X( f )[ =
1

a
2
4
2
f
2
_ (25a)
and X( f ) = tan
1
(2f ,a) (25b)
430 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Alternatively, X( f ) may be shown by its real and imaginary parts, Re [X( f )] and Im [X( f )], as in Figs. 17-14(a) and
(b).
Re [X( f )] =
a
a
2
4
2
f
2
(26a)
Im[X( f )] =
2f
a
2
4
2
f
2
(26b)
EXAMPLE 17.8 Find the Fourier transform of the square pulse
x(t) =
1 for T - t - T
0 otherwise
_
From (22a),
X( f ) =
_
T
T
e
j2ft
dt =
1
j2f
e
j2f
_ _
T
T
=
sin2fT
f
(27)
Because x(t) is even, X( f ) is real. The transform pairs are plotted in Figs. 17-15(a) and (b) for T =
1
2
s.
EXAMPLE 17.9 Find the Fourier transform of x(t) = e
at
u(t). a > 0.
X( f ) =
_
0
o
e
at
e
j2ft
dt =
1
a j2f
(28)
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 431
Fig. 17-13
Fig. 17-14
EXAMPLE 17.10 Find the inverse Fourier transform of X( f ) = 2a,(a
2
4
2
f
2
), a > 0.
By partial fraction expansion we have
X( f ) =
1
a j2f

1
a j2f
(29)
The inverse of each term in (29) may be derived from (24) and (28) so that
x(t) = e
at
u(t) e
at
u(t) = e
a[t[
for all t
See Fig. 17-16.
17.10 PROPERTIES OF THE FOURIER TRANSFORM
Some properties of the Fourier transform are listed in Table 17-1. Several commonly used trans-
form pairs are given in Table 17-2.
17.11 CONTINUOUS SPECTRUM
[X( f )[
2
, as dened in Section 17.9, is called the energy density or the spectrum of the waveform x(t).
Unlike the periodic functions, the energy content of a nonperiodic waveform x(t) at each frequency is
zero. However, the energy content within a frequency band from f
1
to f
2
is
432 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-15
Fig. 17-16
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 433
Table 17-1 Fourier Transform Properties
Time Domain x(t) =
_
o
o
X( f )e
j2ft
dt Frequency Domain X( f ) =
_
o
o
x(t)e
j2ft
dt
1. x(t) real X( f ) = X
+
(f )
2. x(t) even, x(t) = x(t) X( f ) = X(f )
3. x(t), odd, x(t) = x(t) X( f ) = X(f )
4. X(t) x(f )
5.
x(0) =
_
o
o
X( f ) df X(0) =
_
o
o
x(t) dt
6. y(t) = x(at)
Y( f ) =
1
[a[
X( f ,a)
7. y(t) = tx(t)
Y( f ) =
1
j2
dX( f )
df
8. y(t) = x(t) Y( f ) = X(f )
9. y(t) = x(t t
0
) Y( f ) = e
j2ft
0
X( f )
Table 17-2 Fourier Transform Pairs
x(t) X( f )
1. e
at
u(t). a > 0
1
a j2f
2. e
a[t[
. a > 0
2a
a
2
4
2
f
2
3. te
at
u(t). a > 0
1
(a j2f )
2
4. exp(t
2
,t
2
) t exp(f
2
t
2
)
5.
6.
7. 1 o( f )
8. o(t) 1
9. sin2f
0
t
o( f f
0
) o( f f
0
)
2j
10. cos 2f
0
t
o( f f
0
) o( f f
0
)
2
W = 2
_
f
2
f
1
[x( f )[
2
df (30)
EXAMPLE 17.11 Find the spectrum of x(t) = e
at
u(t) e
at
u(t), a > 0, shown in Fig. 17-17.
We have x(t) = x
1
(t) x
2
(t). Since x
1
(t) = e
at
u(t) and x
2
(t) = e
at
u(t),
X
1
( f ) =
1
a j2f
X
2
( f ) =
1
a j2f
X( f ) = X
1
( f ) X
2
( f ) =
j4f
a
2
4
2
f
2
Then
[X( f )[
2
=
16
2
f
2
(a
2
4
2
f
2
)
2
from which
EXAMPLE 17.12 Find and compare the energy contents W
1
and W
2
of y
1
(t) = e
[at[
and
y
2
(t) = e
at
u(t) e
at
u(t), a > 0, within the band 0 to 1 Hz. Let a = 200.
From Examples 17.10 and 17.11,
[Y
1
( f )[
2
=
4a
2
(a
2
4
2
f
2
)
2
and [Y
2
( f )[
2
=
16
2
f
2
(a
2
4
2
f
2
)
2
Within 0 - f - 1 Hz, the spectra and energies may be approximated by
[Y
1
( f )[
2
- 4,a
2
= 10
4
J,Hz and W
1
= 2(10
4
) J = 200 mJ
[Y
2
( f )
2
[ - 10
7
f
2
and W
2
- 0
The preceding results agree with the observation that most of the energy in y
1
(t) is near the low-frequency region in
contrast to y
2
(t).
Solved Problems
17.1 Find the trigonometric Fourier series for the square wave shown in Fig. 17-18 and plot the line
spectrum.
In the interval 0 - ot - , f (t) = V; and for - ot - 2, f (t) = V. The average value of the wave is
zero; hence, a
0
,2 = 0. The cosine coecients are obtained by writing the evaluation integral with the
functions inserted as follows:
434 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-17
a
n
=
1

0
V cos not d(ot)
_
2

(V) cos not d(ot)

1
n
sin not
_ _

1
n
sin not
_ _
2

_ _
= 0 for all n
Thus, the series contains no cosine terms. Proceeding with the evaluation integral for the sine terms,
b
n
=
1

0
V sinnot d(ot)
_
2

(V) sin not d(ot)

1
n
cos not
_ _

1
n
cos not
_ _
2

_ _
=
V
n
(cos n cos 0 cos n2 cos n) =
2V
n
(1 cos n)
Then b
n
= 4V,n for n = 1. 3. 5. . . . . and b
n
= 0 for n = 2. 4. 6. . . . . The series for the square wave is
f (t) =
4V

sinot
4V
3
sin3ot
4V
5
sin5ot
The line spectrum for this series is shown in Fig. 17-19. This series contains only odd-harmonic sine
terms, as could have been anticipated by examination of the waveform for symmetry. Since the wave in Fig.
17-18 is odd, its series contains only sine terms; and since it also has half-wave symmetry, only odd
harmonics are present.
17.2 Find the trigonometric Fourier series for the triangular wave shown in Fig. 17-20 and plot the line
spectrum.
The wave is an even function, since f (t) = f (t), and if its average value, V,2, is subtracted, it also has
half-wave symmetry, that is, f (t) = f (t ). For - ot - 0, f (t) = V (V,)ot; and for 0 - ot - ,
f (t) = V (V,)ot. Since even waveforms have only cosine terms, all b
n
= 0. For n _ 1,
a
n
=
1

_
0

[V (V,)ot] cos not d(ot)

0
[V (V,)ot] cos not d(ot)
=
V

cos not d(ot)

ot

cos not d(ot)

0
ot

cos not d(ot)

2
1
n
2
cos not
ot

sinnot
_ _
0

1
n
2
cos not
ot
n
sin not
_ _

0
_ _
=
V

2
n
2
[cos 0 cos(n) cos n cos 0] =
2V

2
n
2
(1 cos n)
As predicted from half-wave symmetry, the series contains only odd terms, since a
n
= 0 for n = 2. 4. 6. . . . .
For n = 1. 3. 5. . . . . a
n
= 4V,
2
n
2
. Then the required Fourier series is
f (t) =
V
2

4V

2
cos ot
4V
(3)
2
cos 3ot
4V
(5)
2
cos 5ot
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 435
Fig. 17-18 Fig. 17-19
The coecients decrease as 1,n
2
, and thus the series converges more rapidly than that of Problem 17.1. This
fact is evident from the line spectrum shown in Fig. 17-21.
17.3 Find the trigonometric Fourier series for the sawtooth wave shown in Fig. 17-22 and plot the line
spectrum.
By inspection, the waveform is odd (and therefore has average value zero). Consequently the series will
contain only sine terms. A single expression, f (t) = (V,)ot, describes the wave over the period from to
, and we will use these limits on our evaluation integral for b
n
.
b
n
=
1

(V,)ot sinnot d(ot) =

2
1
n
2
sin not
ot
n
cos not
_ _

=
2V
n
(cos n)
As cos n is 1 for even n and 1 for odd n, the signs of the coecients alternate. The required series is
f (t) =
2V

{sinot
1
2
sin 2ot
1
3
sin3ot
1
4
sin4ot ]
The coecients decrease as 1,n, and thus the series converges slowly, as shown by the spectrum in Fig. 17-23.
Except for the shift in the origin and the average term, this waveform is the same as in Fig. 17-8; compare the
two spectra.
17.4 Find the trigonometric Fourier series for the waveform shown in Fig. 17-24 and sketch the line
spectrum.
In the interval 0 - ot - , f (t) = (V,)ot; and for - ot - 2, f (t) = 0. By inspection, the average
value of the wave is V,4. Since the wave is neither even nor odd, the series will contain both sine and cosine
terms. For n > 0, we have
a
n
=
1

0
(V,)ot cos not d(ot) =
V

2
1
n
2
cos not
ot
n
sinnot
_ _

0
=
V

2
n
2
(cos n 1)
436 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-20 Fig. 17-21
Fig. 17-22 Fig. 17-23
When n is even, cos n 1 = 0 and a
n
= 0. When n is odd, a
n
= 2V,(
2
n
2
). The b
n
coecients are
b
n
=
1

0
(V,)ot sinnot d(ot) =
V

2
1
n
2
sinnot
ot
n
cos not
_ _

0
=
V
n
(cos n) = (1)
n1
V
n
Then the required Fourier series is
f (t) =
V
4

2V

2
cos ot
2V
(3)
2
cos 3ot
2V
(5)
2
cos 5ot

sin ot
V
2
sin 2ot
V
3
sin 3ot
The even-harmonic amplitudes are given directly by [b
n
[, since there are no even-harmonic cosine terms.
However, the odd-harmonic amplitudes must be computed using c
n
=

a
2
n
b
2
n
_
. Thus,
c
1
=

(2V,
2
)
2
(V,)
2
_
= V(0.377) c
3
= V(0.109) c
5
= V(0.064)
The line spectrum is shown in Fig. 17-25.
17.5 Find the trigonometric Fourier series for the half-wave-rectied sine wave shown in Fig. 17-26
and sketch the line spectrum.
The wave shows no symmetry, and we therefore expect the series to contain both sine and cosine terms.
Since the average value is not obtainable by inspection, we evaluate a
0
for use in the term a
0
,2.
a
0
=
1

0
V sinot d(ot) =
V

[cos ot]

0
=
2V

Next we determine a
n
:
a
n
=
1

0
V sin ot cos not d(ot)
=
V

n sin ot sinnot cos not cos ot

0
=
V
(1 n
2
)
(cos n 1)
With n even, a
n
= 2V,(1 n
2
); and with n odd, a
n
= 0. However, this expression is indeterminate for
n = 1, and therefore we must integrate separately for a
1
.
a
1
=
1

0
V sin ot cos ot d(ot) =
V

0
1
2
sin2ot d(ot) = 0
Now we evaluate b
n
:
b
n
=
1

0
V sin ot sinnot d(ot) =
V

n sinot cos not sinnot cos ot

0
= 0
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 437
Fig. 17-24 Fig. 17-25
Here again the expression is indeterminate for n = 1, and b
1
is evaluated separately.
b
1
=
1

0
V sin
2
ot d(ot) =
V

ot
2

sin 2ot
4
_ _

0
=
V
2
Then the required Fourier series is
f (t) =
V

1

2
sinot
2
3
cos 2ot
2
15
cos 4ot
2
35
cos 6ot
_ _
The spectrum, Fig. 17-27, shows the strong fundamental term in the series and the rapidly decreasing
amplitudes of the higher harmonics.
17.6 Find the trigonometric Fourier series for the half-wave-rectied sine wave shown in Fig. 17-28,
where the vertical axis is shifted from its position in Fig. 17-26.
The function is described in the interval - ot - 0 by f (t) = V sin ot. The average value is the
same as that in Problem 17.5, that is,
1
2
a
0
= V,. For the coecients a
n
, we have
a
n
=
1

_
0

(V sinot) cos not d(ot) =

_
0

(V sinot) cos ot d(ot) = 0

_
0

(V sin ot) sinnot d(ot) = 0

_
0

(V) sin
2
ot d(ot) =
V
2
Thus, the series is
f (t) =
V

1

2
sinot
2
3
cos 2ot
2
15
cos 4ot
2
35
cos 6ot
_ _
438 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-26 Fig. 17-27
Fig. 17-28
This series is identical to that of Problem 17.5, except for the fundamental term, which has a negative
coecient in this series. The spectrum would obviously be identical to that of Fig. 17-27.
Another Method
When the sine wave V sin ot is subtracted from the graph of Fig. 17.26, the graph of Fig. 17-28 results.
17.7 Obtain the trigonometric Fourier series for the repeating rectangular pulse shown in Fig. 17-29
and plot the line spectrum.
With the vertical axis positioned as shown, the wave is even and the series will contain only cosine terms
and a constant term. In the period from to used for the evaluation integrals, the function is zero
except from ,6 to ,6.
a
0
=
1

_
,6
,6
V d(ot) =
V
3
a
n
=
1

_
,6
,6
V cos not d(ot) =
2V
n
sin
n
6
Since sin n,6 = 1,2,

3
_
,2. 1.

3
_
,2. 1,2. 0. 1,2. . . . for n = 1. 2. 3. 4. 5. 6. 7. . . . , respectively, the series is
f (t) =
V
6

2V

_
1
2
cos ot

3
_
2
1
2
_ _
cos 2ot 1
1
3
_ _
cos 3ot

3
_
2
1
4
_ _
cos 4ot

1
2
1
5
_ _
cos 5ot
1
2
1
7
_ _
cos 7ot
_
f (t) =
V
6

2V

o
n=1
1
n
sin(n,6) cos not or
The line spectrum, shown in Fig. 17-30, decreases very slowly for this wave, since the series converges
very slowly to the function. Of particular interest is the fact that the 8th, 9th, and 10th harmonic
amplitudes exceed the 7th. With the simple waves considered previously, the higher-harmonic amplitudes
were progressively lower.
17.8 Find the exponential Fourier series for the square wave shown in Figs. 17-18 and 17-31, and
sketch the line spectrum. Obtain the trigonometric series coecients from those of the expo-
nential series and compare with Problem 17.1.
In the interval - ot - 0, f (t) = V; and for 0 - ot - , f (t) = V. The wave is odd; therefore,
A
0
= 0 and the A
n
will be pure imaginaries.
A
n
=
1
2
_
0

(V)e
jnot
d(ot)
_

0
Ve
jnot
d(ot)
_ _
=
V
2

1
(jn)
e
jnot
_ _
0

1
(jn)
e
jnot
_ _

0
_ _
=
V
j2n
(e
0
e
jn
e
jn
e
0
) = j
V
n
(e
jn
1)
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 439
Fig. 17-29 Fig. 17-30
For n even, e
jn
= 1 and A
n
= 0; for n odd, e
jn
= 1 and A
n
= j(2V,n) (half-wave symmetry). The
required Fourier series is
f (t) = j
2V
3
e
j3ot
j
2V

e
jot
j
2V

e
jot
j
2V
3
e
j3ot

The graph in Fig. 17-32 shows amplitudes for both positive and negative frequencies. Combining the
values at n and n yields the same line spectrum as plotted in Fig. 17-19.
The trigonometric-series cosine coecients are
a
n
= 2 Re A
n
= 0
b
n
= 2 ImA
n
=
4V
n
for odd n only and
These agree with the coecients obtained in Problem 17.1.
17.9 Find the exponential Fourier series for the triangular wave shown in Figs. 17-20 and 17-33 and
sketch the line spectrum.
In the interval - ot - 0, f (t) = V (V,)ot; and for 0 - ot - , f (t) = V (V,)ot. The wave
is even and therefore the A
n
coecients will be pure real. By inspection the average value is V,2.
A
n
=
1
2
_
0

[V (V,)ot]e
jnot
d(ot)
_

0
[V (V,)ot]e
jnot
d(ot)
_ _
=
V
2
2
_
0

ote
jnot
d(ot)
_

0
(ot)e
jnot
d(ot)
_

e
jnot
d(ot)
_ _
=
V
2
2
e
jnot
(jn)
2
(jnot 1)
_ _0

e
jnot
(jn)
2
(jnot 1)
_ _
0
_ _
=
V

2
n
2
(1 e
jn
)
For even n, e
jn
= 1 and A
n
= 0; for odd n, A
n
= 2V,
2
n
2
. Thus the series is
f (t) =
2V
(3)
2
e
j3ot

2V
()
2
e
jot

V
2

2V
()
2
e
jot

2V
(3)
2
e
j3ot

The harmonic amplitudes
c
0
=
V
2
c
n
= 2[A
n
[ =
0 (n = 2. 4. 6. . . .)
4V,
2
n
2
(n = 1. 3. 5. . . .)
_
are exactly as plotted in Fig. 17-21.
440 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-31 Fig. 71-32
17.10 Find the exponential Fourier series for the half-wave rectied sine wave shown in Figs. 17-26 and
17-34, and sketch the line spectrum.
In the interval 0 - ot - , f (t) = V sinot; and from to 2, f (t) = 0. Then
A
n
=
1
2
_

0
V sinot e
jnot
d(ot)
=
V
2
e
jnot
(1 n
2
)
(jn sin ot cos ot)
_ _
0
=
V(e
jn
1)
2(1 n
2
)
For even n, A
n
= V,(1 n
2
); for odd n, A
n
= 0. However, for n = 1, the expression for A
n
becomes
indeterminate. LHo pitals rule may be applied; in other words, the numerator and denominator are
separately dierentiated with respect to n, after which n is allowed to approach 1, with the result that
A
1
= j(V,4).
The average value is
A
0
=
1
2
_

0
V sin ot d(ot) =
V
2
_
cos ot
_

0
=
V

Then the exponential Fourier series is

V
3
e
j2ot
j
V
4
e
jot

j
V
4
e
jot

V
3
e
j2ot

V
15
e
j4ot

The harmonic amplitudes,
c
0
= A
0
=
V

c
n
= 2[A
n
[ =
2V,(n
2
1) (n = 2. 4. 6. . . .)
V,2 (n = 1)
0 (n = 3. 5. 7. . . .)
_
_
_
are exactly as plotted in Fig. 17-27.
17.11 Find the average power in a resistance R = 10 , if the current in Fourier series form is
i = 10 sin ot 5 sin 3ot 2 sin 5ot (A).
The current has an eective value I
eff
=

1
2
(10)
2

1
2
(5)
2

1
2
(2)
2
_
=

64.5
_
= 8.03 A. Then the average
power is P = I
2
eff
R = (64.5)10 = 645 W.
Another Method
The total power is the sum of the harmonic powers, which are given by
1
2
V
max
I
max
cos . But the
voltage across the resistor and the current are in phase for all harmonics, and
n
= 0. Then,
v
R
= Ri = 100 sin ot 50 sin 3ot 20 sin 5ot
and P =
1
2
(100)(10)
1
2
(50)(5)
1
2
(20)(2) = 645 W.
17.12 Find the average power supplied to a network if the applied voltage and resulting current are
v = 50 50 sin 5 10
3
t 30 sin 10
4
t 20 sin 2 10
4
t (V)
i = 11.2 sin (5 10
3
t 63.48) 10.6 sin (10
4
t 458) 8.97 sin (2 10
4
t 26.68) (A)
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 441
Fig. 17-33 Fig. 17-34
The total average power is the sum of the harmonic powers:
P = (50)(0)
1
2
(50)(11.2) cos 63.48
1
2
(30)(10.6) cos 458
1
2
(20)(8.97) cos 26.68 = 317.7 W
17.13 Obtain the constants of the two-element series circuit with the applied voltage and resultant
current given in Problem 17.12.
The voltage series contains a constant term 50, but there is no corresponding term in the current series,
thus indicating that one of the elements is a capacitor. Since power is delivered to the circuit, the other
element must be a resistor.
I
eff
=

1
2
(11.2)
2

1
2
(10.6)
2

1
2
(8.97)
2
_
= 12.6 A
The average power is P = I
2
eff
R, from which R = P,I
2
eff
= 317.7,159.2 = 2 .
At o = 10
4
rad/s, the current leads the voltage by 458. Hence,
1 = tan458 =
1
oCR
or C =
1
(10
4
)(2)
= 50 mF
Therefore, the two-element series circuit consists of a resistor of 2 and a capacitor of 50 mF.
17.14 The voltage wave shown in Fig. 17-35 is applied to a series circuit of R = 2 k and L = 10 H.
Use the trigonometric Fourier series to obtain the voltage across the resistor. Plot the line
spectra of the applied voltage and v
R
to show the eect of the inductance on the harmonics.
o = 377 rad/s.
The applied voltage has average value V
max
,, as in Problem 17.5. The wave function is even and
hence the series contains only cosine terms, with coecients obtained by the following evaluation integral:
a
n
=
1

_
,2
,2
300 cos ot cos not d(ot) =
600
(1 n
2
)
cos n,2 V
Here, cos n,2 has the value 1 for n = 2. 6. 10. . . . . and 1 for n = 4. 8. 12. . . . . For n odd, cos n,2 = 0.
However, for n = 1, the expression is indeterminate and must be evaluated separately.
a
1
=
1

_
,2
,2
300 cos
2
ot d(ot) =
300

ot
2

sin2ot
4
_ _
,2
,2
=
300
2
V
v =
300

1

2
cos ot
2
3
cos 2ot
2
15
cos 4ot
2
35
cos 6ot
_ _
(V) Thus,
In Table 17-3, the total impedance of the series circuit is computed for each harmonic in the voltage
expression. The Fourier coecients of the current series are the voltage series coecients divided by the Z
n
;
the current terms lag the voltage terms by the phase angles
n
.
442 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-35
I
0
=
300,
2
mA
i
1
=
300,2
4.26
cos (ot 628) (mA)
i
2
=
600,3
7.78
cos (2ot 75.18) (mA)
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Then the current series is
i =
300
2

300
(2)(4.26)
cos (ot 628)
600
3(7.78)
cos (2ot 75.18)

600
15(15.2)
cos (4ot 82.458)
600
35(22.6)
cos (6ot 84.928) (mA)
and the voltage across the resistor is
v
R
= Ri = 95.5 70.4 cos (ot 628) 16.4 cos (2ot 75.18)
1.67 cos (4ot 82.458) 0.483 cos (6ot 84.928) (V)
Figure 17-36 shows clearly how the harmonic amplitudes of the applied voltage have been reduced by
the 10-H series inductance.
17.15 The current in a 10-mH inductance has the waveform shown in Fig. 17-37. Obtain the trigono-
metric series for the voltage across the inductance, given that o = 500 rad/s.
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 443
Table 17-3
n no, rad/s R, k noL. k Z
n
. k
n
0 0 2 0 2 08
1 377 2 3.77 4.26 628
2 754 2 7.54 7.78 75.18
4 1508 2 15.08 15.2 82.458
6 2262 2 22.62 22.6 84.928
Fig. 17-36
Fig. 17-37
The derivative of the waveform of Fig. 17-37 is graphed in Fig. 17-38. This is just Fig. 17-18 with
V = 20,. Hence, from Problem 17.1,
di
d(ot)
=
80

2
(sin ot
1
3
sin3ot
1
5
sin 5ot ) (A)
v
L
= Lo
di
d(ot)
=
400

2
(sinot
1
3
sin 3ot
1
5
sin5ot ) (V) and so
Supplementary Problems
17.16 Synthesize the waveform for which the trigonometric Fourier series is
f (t) =
8V

2
{sinot
1
9
sin 3ot
1
25
sin5ot
1
49
sin7ot ]
17.17 Synthesize the waveform if its Fourier series is
f (t) = 5
40

2
(cos ot
1
9
cos 3ot
1
25
cos 5ot )

20

(sin ot
1
2
sin2ot
1
3
sin 3ot
1
4
sin4ot )
17.18 Synthesize the waveform for the given Fourier series.
f (t) = V
_
1
2

1

cos ot
1
3
cos 2ot
1
2
cos 3ot
1
15
cos 4ot
1
6
cos 6ot

1
4
sin ot
2
3
sin2ot
4
15
sin4ot
_
17.19 Find the trigonometric Fourier series for the sawtooth wave shown in Fig. 17-39 and plot the line spectrum.
Compare with Example 17.1.
Ans. f (t) =
V
2

V

(sin ot
1
2
sin2ot
1
3
sin 3ot )
444 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-38
17.20 Find the trigonometric Fourier series for the sawtooth wave shown in Fig. 17-40 and plot the spectrum.
Compare with the result of Problem 17.3.
Ans. f (t) =
2V

{sinot
1
2
sin 2ot
1
3
sin3ot
1
4
sin4ot ]
17.21 Find the trigonometric Fourier series for the waveform shown in Fig. 17-41 and plot the line spectrum.
Ans. f (t) =
4V

2
{cos ot
1
9
cos 3ot
1
25
cos 5ot ]
2V

{sin ot
1
3
sin3ot
1
5
sin 5ot ]
17.22 Find the trigonometric Fourier series of the square wave shown in Fig. 17-42 and plot the line spectrum.
Compare with the result of Problem 17.1.
Ans. f (t) =
4V

{cos ot
1
3
cos 3ot
1
5
cos 5ot
1
7
cos 7ot ]
17.23 Find the trigonometric Fourier series for the waveforms shown in Fig. 17-43. Plot the line spectrum of each
and compare.
Ans. (a) f (t) =
5
12

o
n=1
10
n
sin
n
12
_ _
cos not
10
n
1 cos
n
12
_ _
sinnot
_ _
(b) f (t) =
50
6

o
n=1
10
n
sin
n5
3
_ _
cos not
10
n
1 cos
n5
3
_ _
sinnot
_ _
17.24 Find the trigonometric Fourier series for the half-wave-rectied sine wave shown in Fig. 17-44 and plot the
line spectrum. Compare the answer with the results of Problems 17.5 and 17.6.
Ans. f (t) =
V

1

2
cos ot
2
3
cos 2ot
2
15
cos 4ot
2
35
cos 6ot
_ _
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 445
Fig. 17-39 Fig. 17-40
Fig. 17-41 Fig. 17-42
Fig. 17-43
17.25 Find the trigonometric Fourier series for the full-wave-rectied sine wave shown in Fig. 17-45 and plot the
spectrum.
Ans. f (t) =
2V

(1
2
3
cos 2ot
2
15
cos 4ot
2
35
cos 6ot )
17.26 The waveform in Fig. 17-46 is that of Fig. 17-45 with the origin shifted. Find the Fourier series and show
that the two spectra are identical.
Ans. f (t) =
2V

(1
2
3
cos 2ot
2
15
cos 4ot
2
35
cos 6ot )
17.27 Find the trigonometric Fourier series for the waveform shown in Fig. 17-47.
Ans. f (t) =
V
2

V
2
cos ot

o
n=2
V
(1 n
2
)
(cos n n sinn,2) cos not

V
4
sinot

o
n=2
nV cos n,2
(1 n
2
)
_ _
sin not
17.28 Find the trigonometric Fourier series for the waveform shown in Fig. 17-48. Add this series termwise to
that of Problem 17.27, and compare the sum with the series obtained in Problem 17.5.
Ans. f (t) =
V
2

V
2
cos ot

o
n=2
V(n sinn,2 1)
(n
2
1)
cos not
V
4
sinot

o
n=2
nV cos n,2
(1 n
2
)
sinnot
446 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-44 Fig. 17-45
Fig. 17-46 Fig. 17-47
Fig. 17-48 Fig. 17-49
17.29 Find the exponential Fourier series for the waveform shown in Fig. 17-49 and plot the line spectrum.
Convert the coecients obtained here into the trigonometric series coecients, write the trigonometric
series, and compare it with the result of Problem 17.4.
Ans. f (t) = V
_

1
9
2
j
1
6
_ _
e
j3ot
j
1
4
e
j2ot

2
j
1
2
_ _
e
jot

1
4

2
j
1
2
_ _
e
jot
j
1
4
e
j2ot

1
9
2
j
1
6
_ _
e
j3ot

_
17.30 Find the exponential Fourier series for the waveform shown in Fig. 17-50 and plot the line spectrum.
Ans. f (t) = V
_

1
9
2
j
1
6
_ _
e
j3ot
j
1
4
e
j2ot

2
j
1
2
_ _
e
jot

1
4

2
j
1
2
_ _
e
jot
j
1
4
e
j2ot

1
9
2
j
1
6
_ _
e
j3ot

_
17.31 Find the exponential Fourier series for the square wave shown in Fig. 17-51 and plot the line spectrum. Add
the exponential series of Problems 17.29 and 17.30 and compare the sum to the series obtained here.
Ans. f (t) = V j
1
3
e
j3ot
j
1

e
jot

1
2
j
1

e
jot
j
1
3
e
j3ot

_ _
17.32 Find the exponential Fourier series for the sawtooth waveform shown in Fig. 17-52 and plot the spectrum.
Convert the coecients obtained here into the trigonometric series coecients, write the trigonometric
series, and compare the results with the series obtained in Problem 17.19.
Ans. f (t) = V j
1
4
e
j2ot
j
1
2
e
jot

1
2
j
1
2
e
jot
j
1
4
e
j2ot

_ _
17.33 Find the exponential Fourier series for the waveform shown in Fig. 17-53 and plot the spectrum. Convert
the trigonometric series coecients found in Problem 17.20 into exponential series coecients and compare
them with the coecients of the series obtained here.
Ans. f (t) = V j
1
2
e
j2ot
j
1

e
jot
j
1

e
jot
j
1
2
e
j2ot

_ _
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 447
Fig. 17-50 Fig. 17-51
Fig. 17-52 Fig. 17-53
17.34 Find the exponential Fourier series for the waveform shown in Fig. 17-54 and plot the spectrum. Convert
the coecients to trigonometric series coecients, write the trigonometric series, and compare it with that
obtained in Problem 17.21.
Ans. f (t) = V
_

2
9
2
j
1
3
_ _
e
j3ot

2
j
1

_ _
e
jot

2
j
1

_ _
e
jot

2
9
2
j
1
3
_ _
e
j3ot

_
17.35 Find the exponential Fourier series for the square wave shown in Fig. 17-55 and plot the line spectrum.
Convert the trigonometric series coecients of Problem 17.22 into exponential series coecients and com-
pare with the coecients in the result obtained here.
Ans. f (t) =
2V

(
1
5
e
j5ot

1
3
e
j3ot
e
jot
e
jot

1
3
e
j3ot

1
5
e
j5ot
)
17.36 Find the exponential Fourier series for the waveform shown in Fig. 17-56 and plot the line spectrum.
Ans. f (t) =
V
2
sin
2
6
_ _
e
j2ot

sin

6
_ _
e
jot

V
6

V

sin

6
_ _
e
jot

V
2
sin
2
6
_ _
e
j2ot

17.37 Find the exponential Fourier series for the half-wave-rectied sine wave shown in Fig. 17-57. Convert these
coecients into the trigonometric series coecients, write the trigonometric series, and compare it with the
result of Problem 17.24.
Ans. f (t) =
V
15
e
j4ot

V
3
e
j2ot

V
4
e
jot

V
4
e
jot

V
3
e
j2ot

V
15
e
j4ot

17.38 Find the exponential Fourier series for the full-wave rectied sine wave shown in Fig. 17-58 and plot the line
spectrum.
Ans. f (t) =
2V
15
e
j4ot

2V
3
e
j2ot

2V

2V
3
e
j2ot

2V
15
e
j4ot

448 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-54 Fig. 17-55
Fig. 17-56 Fig. 17-57
17.39 Find the eective voltage, eective current, and average power supplied to a passive network if the applied
voltage is v = 200 100 cos (500t 308) 75 cos (1500t 608) (V) and the resulting current is
i = 3.53 cos (500t 758) 3.55 cos (1500t 78.458) (A). Ans. 218.5 V, 3.54 A, 250.8 W
17.40 A voltage v = 50 25 sin 500t 10 sin 1500t 5 sin 2500t (V) is applied to the terminals of a passive net-
work and the resulting current is
i = 5 2.23 sin(500t 26.68) 0.556 sin(1500t 56.38) 0.186 sin(2500t 68.28) (A)
Find the eective voltage, eective current, and the average power. Ans. 53.6 V, 5.25 A, 276.5 W
17.41 A three-element series circuit, with R = 5 , L = 5 mH, and C = 50 mF, has an applied voltage
v = 150 sin1000t 100 sin2000t 75 sin3000t (V). Find the eective current and the average power for
the circuit. Sketch the line spectrum of the voltage and the current, and note the eect of series resonance.
Ans. 16.58 A, 1374 W
17.42 A two-element series circuit, with R = 10 and L = 20 mH, has current
i = 5 sin 100t 3 sin 300t 2 sin 500t (A)
Find the eective applied voltage and the average power. Ans. 48 V, 190 W
17.43 A pure inductance, L = 10 mH, has the triangular current wave shown in Fig. 17-59, where o = 500 rad/s.
Obtain the exponential Fourier series for the voltage across the inductance. Compare the answer with the
result of Problem 17.8.
Ans. v
L
=
200

2
( j
1
3
e
j3ot
je
jot
je
jot
j
1
3
e
jot
) (V)
17.44 A pure inductance, L = 10 mH, has an applied voltage with the waveform shown in Fig. 17-60, where
o = 200 rad/s. Obtain the current series in trigonometric form and identify the current waveform.
Ans. i =
20

(sin ot
1
9
sin3ot
1
25
sin5ot
1
49
sin7ot ) (A); triangular
17.45 Figure 17-61 shows a full-wave-rectied sine wave representing the voltage applied to the terminals of an
LC series circuit. Use the trigonometric Fourier series to nd the voltages across the inductor and the
capacitor.
CHAP. 17] FOURIER METHOD OF WAVEFORM ANALYSIS 449
Fig. 17-58
Fig. 17-59 Fig. 17-60
Ans. v
L
=
4V
m

2oL
3 2oL
1
2oC
_ _ cos 2ot
4oL
15 4oL
1
4oC
_ _ cos 4ot
_
_
_
_
_

_
v
C
=
4V
m

1
2

1
3(2oC) 2oL
1
2oC
_ _ cos 2ot
1
15(4oC) 4oL
1
4oC
_ _ cos 4ot
_
_
_
_
_

_
17.46 A three-element circuit consists of R = 5 in series with a parallel combination of L and C. At
o = 500 rad/s, X
L
= 2 , X
C
= 8 . Find the total current if the applied voltage is given by
v = 50 20 sin 500t 10 sin 1000t (V). Ans. i = 10 3.53 sin(500t 28.18) (A)
450 FOURIER METHOD OF WAVEFORM ANALYSIS [CHAP. 17
Fig. 17-61