Circuit Analysis II
Steven T. Karris is the president and founder of Orchard Publications. He earned a bachelors
degree in electrical engineering at Christian Brothers University, Memphis, Tennessee, a masters
degree in electrical engineering at Florida Institute of Technology, Melbourne, Florida, and has
done postmaster work at the latter. He is a registered professional engineer in California and
Florida. He has over 30 years of professional engineering experience in industry. In addition, he
has over 25 years of teaching experience that he acquired at several educational institutions as
an adjunct professor. He is currently with UC Berkeley Extension.
ISBN 0970951191
$39.95 U.S.A.
Circuit Analysis II
with MATLAB Applications
Steven T. Karris
Orchard Publications
www.orchardpublications.com
ISBN 0970951191
Disclaimer
The publisher has used his best effort to prepare this text. However, the publisher and author makes no warranty of any
kind, expressed or implied with regard to the accuracy, completeness, and computer codes contained in this book, and
shall not be liable in any event for incidental or consequential damages in connection with, or arising out of, the
performance or use of these programs.
Preface
This text is written for use in a second course in circuit analysis. The reader of this book should have
the traditional undergraduate knowledge of an introductory circuit analysis material such as Circuit
Analysis I with MATLAB Applications by this author. Another prerequisite would be knowledge of
differential equations, and in most cases, engineering students at this level have taken all required
mathematics courses. It encompasses a spectrum of subjects ranging from the most abstract to the
most practical, and the material can be covered in one semester or two quarters. Appendix B serves as
a review of differential equations with emphasis on engineering related topics and it is recommended
for readers who may need a review of this subject.
There are several textbooks on the subject that have been used for years. The material of this book is
not new, and this author claims no originality of its content. This book was written to fit the needs of
the average student. Moreover, it is not restricted to computer oriented circuit analysis. While it is true
that there is a great demand for electrical and computer engineers, especially in the internet field, the
demand also exists for power engineers to work in electric utility companies, and facility engineers to
work in the industrial areas.
Chapter 1 is an introduction to second order circuits and it is essentially a sequel to first order circuits
that were discussed in the last chapter of as Circuit Analysis I with MATLAB Applications. Chapter 2
is devoted to resonance, and Chapter 3 presents practical methods of expressing signals in terms of
the elementary functions, i.e., unit step, unit ramp, and unit impulse functions. Thus, any signal can be
represented in the compex frequency domain using the Laplace transformation.
Chapters 4 and 5 are introductions to the unilateral Laplace transform and Inverse Laplace transform
respectively, while Chapter 6 presents several examples of analyzing electric circuits using Laplace
transformation methods. Chapter 7 begins with the frequency response concept and Bode magnitude
and frequency plots. Chapter 8 is devoted to transformers with an introduction to self and mutual
inductances. Chapter 9 is an introduction to one and twoterminal devices and presents several
practical examples. Chapter 10 is an introduction to threephase circuits.
It is not necessary that the reader has previous knowledge of MATLAB. The material of this text
can be learned without MATLAB. However, this author highly recommends that the reader studies
this material in conjunction with the inexpensive MATLAB Student Version package that is available
at most college and university bookstores. Appendix A of this text provides a practical introduction
to MATLAB. As shown on the front cover of this text the magnitude and phase plots can be easily
obtained with a one line MATLAB code. Moreover, MATLAB will be invaluable in later studies such
as the design of analog and digital filters.
Contents
Chapter 1
Second Order Circuits
Chapter 2
Resonance
Chapter 3
Elementary Signals
Chapter 4
The Laplace Transformation
Chapter 5
The Inverse Laplace Transformation
Chapter 6
Circuit Analysis with Laplace Transforms
ii
Chapter 7
Frequency Response and Bode Plots
Chapter 8
Self and Mutual Inductances  Transformers
SelfInductance ............................................................................................................................81
The Nature of Inductance ............................................................................................................81
Lenzs Law ..................................................................................................................................83
Mutually Coupled Coils...............................................................................................................83
Establishing Polarity Markings..................................................................................................811
Energy Stored in a Pair of Mutually Coupled Inductors ...........................................................814
Circuits with Linear Transformers.............................................................................................820
Reflected Impedance in Transformers .......................................................................................825
The Ideal Transformer ...............................................................................................................828
Impedance Matching..................................................................................................................832
A Simplified Transformer Equivalent Circuit ...........................................................................833
Thevenin Equivalent Circuit ......................................................................................................834
Summary ....................................................................................................................................838
Exercises ....................................................................................................................................842
Solutions to Exercises ................................................................................................................844
Chapter 9
One and Twoport Networks
iii
The y Parameters..........................................................................................................................................912
The z parameters..........................................................................................................................................919
The h Parameters .........................................................................................................................................924
The g Parameters .........................................................................................................................................929
Reciprocal TwoPort Networks.................................................................................................................934
Summary........................................................................................................................................................938
Exercises........................................................................................................................................................943
Solutions to Exercises .................................................................................................................................945
Chapter 10
ThreePhase Systems
Appendix A
Introduction to MATLAB
iv
Appendix B
Differential Equations
Appendix C
Matrices and Determinants
Appendix D
Constructing Semilog Plots with Microsoft Excel
Appendix E
Scaling
NOTES
vi
Chapter 1
Second Order Circuits
his chapter discusses the natural, forced and total responses in circuits containing resistors,
inductors and capacitors. These circuits are characterized by linear secondorder differential
equations whose solutions consist of the natural and the forced responses. We will consider
both DC (constant) and AC (sinusoidal) excitations.
* State variables and state equations are discussed in Signals and Systems with MATLAB Applications, ISBN 0970951132 by this author.
11
for t ! 0 .
R
vS u0 t
+
`L
i t
C
i dt + V 0 = v S
t!0
(1.1)
and by differentiation
2
d i i dv
di
R  + L 2 +  = S t ! 0
dt
C
dt
dt
To find the forced response, we must first specify the nature of the excitation v S , that is, DC or AC.
If v S is DC ( v S =constant), the right side of (1.1) will be zero and thus the forced response component i f = 0 . If v S is AC ( v S = V cos Zt + T , the right side of (1.1) will be another sinusoid and
therefore i f = I cos Zt + M . Since in this section we are concerned with DC excitations, the right
side will be zero and thus the total response will be just the natural response.
The natural response is found from the homogeneous equation of (1.1), that is,
2
d i i
di
R  + L 2 +  = 0
dt
C
dt
(1.2)
* The unit step function is discussed in detail in Chapter 3. For our present discussion it will suffice to state that
u 0 t = 0 for t 0 and u 0 t = 1 for t ! 0 .
12
from which
R
R
1
s 1 s 2 =  r 2 2L
LC
4L
(1.3)
Zn S =
Z0 DS
D or Damping
Resonant
Beta
Damped Natural
Coefficient
Frequency
Coefficient
Frequency
DS Z0
ES =
R
D S = 2L
(1.4)
where the subscript s stands for series circuit. Then, we can express (1.3) as
2
s 1 s 2 = D S r D S Z 0 = D S r E S if D S ! Z 0
(1.5)
or
2
s 1 s 2 = D S r Z 0 D S = D S r Z n S if Z 0 ! D S
(1.6)
Case I: If D 2S ! Z 20 , the roots s 1 and s 2 are real, negative, and unequal. This results in the overdamped natural response and has the form
in t = k1 e
s1 t
+ k2 e
s2 t
(1.7)
Case II: If D 2S = Z 20 , the roots s 1 and s 2 are real, negative, and equal. This results in the critically
damped natural response and has the form
in t = e
DS t
k1 + k2 t
(1.8)
Case III: If Z 20 ! D 2S , the roots s 1 and s 2 are complex conjugates. This is known as the underdamped
or oscillatory natural response and has the form
in t = e
DS t
k 1 cos Z n S t + k 2 sin Z n S t = k 3 e
DS t
cos Z n S t + M
(1.9)
A typical overdamped response is shown in Figure 1.2 where it is assumed that i n 0 = 0 . This plot
was created with the following MATLAB code:
13
A typical critically damped response is shown in Figure 1.3 where it is assumed that i n 0 = 0 . This
plot was created with the following MATLAB code:
t=0: 0.01: 6; ft=420.*t.*(exp(2.45.*t)); plot(t,ft); grid; xlabel('t');...
ylabel('f(t)'); title('Critically Damped Response for 420.*t.*(exp(2.45.*t))')
A typical underdamped response is shown in Figure 1.4 where it is assumed that i n 0 = 0 . This
plot was created with the following MATLAB code:
14
+
15u 0 t V
it
` 1 mH
100 e 6 mF
Figure 1.5. Circuit for Example 1.1
Solution:
This circuit can be represented by the integrodifferential equation
15
i dt + v C 0 = 15 t ! 0
(1.10)
Differentiating and noting that the derivatives of the constants v C 0 and 15 are zero, we obtain the
homogeneous differential equation
2
d i i
di
R  + L 2 +  = 0
dt
dt C
or
2
di
i
d i R
2 +   +  = 0
L
dt
LC
dt
(1.11)
The roots of the characteristic equation of (1.11) are s 1 = 200 and s 2 = 300 . The total response
is just the natural response and for this example it is overdamped. Therefore, from (1.7),
i t = in t = k1 e
s1 t
+ k2 e
s2 t
= k1 e
200 t
+ k2 e
300 t
(1.12)
The constants k 1 and k 2 can be evaluated from the initial conditions. Thus from the first initial condition i L 0 = i 0 = 5 A and (1.12) we get
0
i 0 = k1 e + k2 e = 5
k1 + k2 = 5
(1.13)
We need another equation in order to compute the values of k 1 and k 2 . With this equation we will
dv
dt
make use of the second initial condition, that is, v C 0 = 2.5 V . Since i C t = i t = C C , we differentiate (1.12), we evaluate it at t = 0 + , and we equate it with this initial condition. Then,
di
di
 = 200k 1 e 200 t 300k 2 e 300 t and dt
dt
= 200k 1 300 k 2
t=0
(1.14)
Also, at t = 0 + ,
di
+
Ri 0 + L dt
16
+ v c 0 = 15
t=0
we get
t=0
di
dt
t=0
15 0.5 u 5 2.5
= = 10000
3
10
(1.15)
(1.16)
Simultaneous solution of (1.13) and (1.16) yields k 1 = 115 and k 2 = 110 . By substitution into (1.12)
we find the total response as
i t = i n t = 115e
200 t
110 e
300 t
(1.17)
17
In the above example, differentiation eliminated (set equal to zero) the right side of the differential
equation and thus the total response was just the natural response. A different approach however,
may not set the right side equal to zero, and therefore the total response will contain both the natural
and forced components. To illustrate, we will use the following approach.
t
1
The capacitor voltage, for all time t , may be expressed as v C t =  i dt and as before, the circuit
C
18
i dt = 15 u 0 t
(1.18)
we rewrite (1.18) as
2
dv C
dv
RC C + LC + v C = 15 u 0 t
2
dt
dt
(1.19)
We observe that this is a nonhomogeneous differential equation whose solution will have both the
natural and the forced response components. Of course, the solution of (1.19) will give us the capacitor voltage v C t . This presents no problem since we can obtain the current by differentiation of the
expression for v C t .
Substitution of the given values into (1.19) yields
2
dv
dv
50
 10 3 2C + v C = 15 u 0 t
 u 10 3 C + 1 u 10 3 u 100
6
6
dt
dt
or
2
dv
dv C
5
2 + 500 C + 60000v C = 9 u 10 u 0 t
dt
dt
(1.20)
The characteristic equation of (1.20) is the same as of that of (1.11) and thus the natural response is
vC n t = k1 e
s1 t
+ k2 e
s2 t
= k1 e
200 t
+ k2 e
300 t
(1.21)
Since the right side of (1.20) is a constant, the forced response will also be a constant and we denote it
as v C f = k 3 . By substitution into (1.20) we get
0 + 0 + 60000k 3 = 900000
or
v C f = k 3 = 15
(1.22)
The total solution then is the summation of (1.21) and (1.22), that is,
v C t = v C n t + v Cf = k 1 e
200 t
+ k2 e
300 t
+ 15
(1.23)
As before, the constants k 1 and k 2 will be evaluated from the initial conditions. First, using
v C 0 = 2.5 V and evaluating (1.23) at t = 0 , we get
0
v C 0 = k 1 e + k 2 e + 15 = 2.5
or
19
(1.24)
Also,
dv
i
dv
dv
i L = i C = C C C = L and C
dt
dt
C
dt
t=0
iL 0
5
 = =  = 300
100
C
3
 u 10
6
(1.25)
= 200k 1 300k 2
(1.26)
t=0
or
k 1 1.5k 2 = 1.5
(1.27)
From (1.24) and (1.27), we get k 1 = 34.5 and k 2 = 22 . By substitution into (1.23), we obtain the
total solution as
v C t = 22e
300 t
34.5 e
200 t
+ 15 u 0 t
(1.28)
y1 =
6600*exp(300*t)+6900*exp(200*t)
y2=diff(y0,2)
y2 =
1980000*exp(300*t)1380000*exp(200*t)
y=y2+500*y1+60000*y0
y =
900000
Using the expression for v C t we can find the current as
dv
100
3
200t
300t
200t
300t
i = i L = i C = C C =  u 10 6900e
6600 e
= 115e
110 e
A
dt
6
110
(1.29)
111
t!0
(1.30)
+
i t
` 1 mH
100 e 6 mF
Figure 1.8. Circuit for Example 1.2
whose solution consists of the summation of the natural and forced responses. We know its natural
response from the previous example. We start with
i t = in t + if t = k1 e
200 t
+ k2 e
300 t
+ if t
(1.31)
where the constants k 1 and k 2 will be evaluated from the initial conditions after i f t has been
found. The steady state (or forced) response will have the form i f t = k 3 cos 10 000t + T in the
time domain ( t domain) and has the form k 3 T in the frequency domain ( jZ domain).
To find i f t we will use the phasor analysis relation I = V e Z where I is the phasor current, V is
the phasor voltage, and Z is the impedance of the phasor circuit which, as we know, is
Z = R + j ZL 1 e ZC =
R + ZL 1 e ZC tan ZL 1 e ZC e R
(1.32)
X L = ZL = 10 u 10
= 10 :
and
3
1
1
X C =  = = 6 u 10 :
4
3
ZC
10 u 100 e 6 10
Then,
R
3 2
Also,
1
3
1
1
10 6 u 10  = tan 9.994

tan ZL 1 e ZC e R = tan  0.5
0.5
and this yields T = 1.52 rads = 87.15q . Then, by substitution into (1.32),
Z =
0.25 + 99.88 T
= 10 87.15
and thus
o
o
o
V
200 0
I =  = o = 20 87.15 20 cos 10000t 87.15 = i f t
Z 10 87.15
112
200 t
+ k2 e
300 t
(1.33)
The constants k 1 and k 2 are evaluated from the initial conditions. From (1.33) and the first initial
condition i L 0 = 5 A we get
0
i 0 = k 1 e + k 2 e + 20 cos 87.15 = 5
i 0 = k 1 + k 2 + 20 u 0.05 = 5
k1 + k2 = 4
(1.34)
We need another equation in order to compute the values of k 1 and k 2 . This equation will make use
dv
dt
of the second initial condition, that is, v C 0 = 2.5 V . Since i C t = i t = C C , we differentiate
(1.33), we evaluate it at t = 0 , and we equate it with this initial condition. Then,
di
 = 200k 1 e 200 t 300k 2 e 300 t 2 u 105 sin 10000t 87.15 o
dt
(1.35)
and at t = 0 ,
di
dt
(1.36)
t=0
Also, at t = 0 +
di
+
Ri 0 + L dt
and solving for di
dt
we get
t=0
di
dt
t=0
(1.37)
or
k 1 + 1.5k 2 = 25
113
(1.38)
200 t
+ 42e
300 t
(1.39)
The plot is shown in Figure 1.9 and was created with the following MATLAB code:
t=0: 0.005: 0.20; i1=38.*(exp(200.*t)); i2=42.*(exp(300.*t));...
i3=20.*cos(10000.*t87.15.*pi./180); iT=i1+i2+i3; plot(t,i1,t,i2,t,i3,t,iT); grid; xlabel('t');...
ylabel('i1, i2, i3, iT'); title('Response iT for Example 1.2')
iG
iS u0 t
vt
iC
iL
G
`L
114
or
1
Gv + L
dv
v dt + I 0 + C  = i S
dt
t!0
By differentiation,
2
dv v di
dv
C 2 + G  +  = S
dt L dt
dt
t!0
(1.40)
To find the forced response, we must first specify the nature of the excitation i S , that is DC or AC.
If i S is DC ( v S =constant), the right side of (1.40) will be zero and thus the forced response component v f = 0 . If i S is AC ( i S = I cos Zt + T , the right side of (1.40) will be another sinusoid and
therefore v f = V cos Zt + M . Since in this section we are concerned with DC excitations, the right
side will be zero and thus the total response will be just the natural response.
The natural response is found from the homogeneous equation of (1.40), that is,
2
dv v
dv
C 2 + G  +  = 0
dt L
dt
(1.41)
or
G
2
i = 0
s +  s + C
LC
from which
2
(1.42)
Z nP =
Z0 DP
DP Z0
EP =
1
Z 0 = LC
G
D P = 2C
D or Damping
Resonant
Beta
Damped Natural
Coefficient
Frequency
Coefficient
Frequency
(1.43)
where the subscript p stands for parallel circuit, we can express (1.42) as
115
(1.44)
(1.45)
s 1 s 2 = D P r D P Z 0 = D P r E P if D P ! Z 0
or
2
s 1 s 2 = D P r Z 0 D P = D P r Z nP if Z 0 ! D P
s1 t
+ k2 e
s2 t
(1.46)
Case II: If D 2P = Z 20 , the roots s 1 and s 2 are real, negative, and equal. This results in the critically
damped natural response and has the form
vn t = e
DP t
k1 + k2 t
(1.47)
Case III: If Z 20 ! D 2P , the roots s 1 and s 2 are complex conjugates. This results in the underdamped
or oscillatory natural response and has the form
vn t = e
DP t
k 1 cos Z nP t + k 2 sin Z nP t = k 3 e
DP t
cos Z nP t + M
(1.48)
116
iR
iC
iL
vt
32 :
10u 0 t A
10 H
1 e 640 F
Solution:
We could write the integrodifferential equation that describes the given circuit, differentiate, and find
the roots of the characteristic equation from the homogeneous differential equation as we did in the
previous section. However, we will skip these steps and start with
v t = vf t + vn t
(1.49)
or
2
D P = 100
and
2
1
1
Z 0 =  =  = 64
LC
10 u 1 e 640
Then
2
s 1 s 2 = D P r D P Z 0 = 10 r 6
or s 1 = 4 and s 2 = 16 . Therefore, the natural response is overdamped and from (1.46) we get
v t = vn t = k1 e
s1 t
+ k2 e
s2 t
= k1 e
4 t
+ k2 e
16 t
(1.50)
and the constants k 1 and k 2 will be evaluated from the initial conditions.
From the initial condition v C 0 = v 0 = 5 V and (1.50) we get
117
v 0 = k1 e + k2 e = 5
or
k1 + k2 = 5
(1.51)
The second equation that is needed for the computation of the values of k 1 and k 2 is found from
dv
dt
dv
dt
other initial condition, that is, i L 0 = 2 A . Since i C t = C C = C  , we differentiate (1.50),
evaluate it at t = 0 + , and we equate it with this initial condition.Then,
dv
 = 4k 1 e 4 t 16k 2 e 16 t and dv
dt
dt
= 4k 1 16 k 2
t=0
(1.52)
Also, at t = 0 +
dv
1
 v 0 + + i L 0 + + C dt
R
and solving for dv
dt
= 10
t=0
we get
t=0
dv
dt
t=0
10 5 e 32 2 = 502
= 1 e 640
(1.53)
or
2k 1 8 k 2 = 251
(1.54)
Simultaneous solution of (1.51) and (1.54) yields k 1 = 291 e 6 , k 2 = 261 e 6 , and by substitution
into (1.50) we get the total response as
291 4 t 261 16 t
1
4 t
16 t
v t = v n t =  e  e
=  291e 261 e V
6
6
6
(1.55)
y1 =
194*exp(4*t)+696*exp(16*t)
118
y2 =
776*exp(4*t)11136*exp(16*t)
y=y2/640+y1/32+y0/10
y =
0
The plot is shown in Figure 1.12 where we have used the following MATLAB code:
t=0: 0.01: 1; v1=(291./6).*(exp(4.*t)); v2=(261./6).*(exp(16.*t));...
vT=v1+v2; plot(t,v1,t,v2,t,vT); grid; xlabel('t');...
ylabel('v1, v2, vT'); title('Response vT for Example 1.3')
From the plot of Figure 1.12, we observe that v t attains its maximum value somewhere in the interval 0.10 and 0.12 sec., and the maximum voltage is approximately 24 V . If we desire to compute precisely the maximum voltage and the exact time it occurs, we can find the derivative of (1.55), set it
equal to zero, and solve for t . Thus,
dv
dt
= 1164e
4 t
+ 4176e
16 t
12t
+ 4176 = 0
= 0
(1.56)
t=0
or
119
12t
348
= 97
or
348
12t = ln  = 1.2775
97
and
1.2775
t = t max =  = 0.106 s
12
(1.57)
A useful quantity, especially in electronic circuit analysis, is the settling time, denoted as t S , and it is
defined as the time required for the voltage to drop to 1% of its maximum value. Therefore, t S is an
indication of the time it takes for v t to dampout, meaning to decrease the amplitude of v t to
approximately zero. For this example, 0.01 u 23.76 = 0.2376 V , and we can find t S by substitution
into (1.55). Then,
1
4t
16t
0.01v max = 0.2376 =  291e 261e
6
(1.58)
and we need to solve for the time t . To simplify the computation, we neglect the second term inside
the parentheses of (1.58) since this component of the voltage damps out much faster than the other
component. This expression then simplifies to
4 ts
1
0.2376 =  291e
6
or
4 t S = ln 0.005 = 5.32
or
t S = 1.33 s
(1.59)
Example 1.4
For the circuit of Figure 1.13, i L 0 = 2 A and v C 0 = 5 V , and the resistor is to be adjusted so
that the natural response will be critically damped. Compute and sketch v t for t ! 0 .
120
iR
iC
iL
vt
10u 0 t A
10 H
1 e 640 F
Solution:
Since the natural response is to be critically damped, we must have Z 20 = 64 because the L and C values are the same as in the previous example. Please refer to (1.43). We must also have
G
1
D P =  =  = Z 0 =
2C 2RC
1
 = 8
LC
or
2
11 = 8 u  = 640
40
R
DP t
8t
k 1 + k 2 t or v t = v n t = e k 1 + k 2 t
(1.60)
v 0 = e k1 + k2 0 = 5
or
k1 = 5
(1.61)
v t = e 5 + k2 t
(1.62)
As before, we need to compute the derivative dv e dt in order to apply the second initial condition and
find the value of the constant k 2 .
We obtain the derivative using MATLAB as follows:
syms t k2; v0=exp(8*t)*(5+k2*t); v1=diff(v0);
% v1 is 1st derivative of v0
v1 =
8*exp(8*t)*(5+k2*t)+exp(8*t)*k2
121
and
dv
dt
= 40 + k 2
(1.63)
t=0
i
dv
 = C and
Also, i C = C  or dv
dt
dt
dv
dt
t=0
iC 0 IS iR 0 iL 0
=  = C
C
(1.64)
or
dv
dt
t=0
IS vC 0 e R iL 0
10 5 e 40 2
7.875
=  =  =  = 5040
C
1 e 640
1 e 640
(1.65)
or
k 2 = 5080
(1.66)
v t = e 5 + 5080t V
(1.67)
y1 =
8*exp(8*t)*(5+5080*t)+5080*exp(8*t)
y2=diff(y0,2)
y2 =
64*exp(8*t)*(5+5080*t)81280*exp(8*t)
y=y2/640+y1/40+y0/10
y =
0
The plot is shown in Figure 1.14 where we have used the following MATLAB code:
122
By inspection of (1.67), we see that at t = 0 , v t = 5 V and thus the second initial condition is satisfied. We can verify that the first initial condition is also satisfied by differentiation of (1.67). We can
also show that v t approaches zero as t approaches infinity with LHpitals rule as follows:
8t
5 + 5080t = lim d 5 + 5080t e dt = lim 5080 = 0
lim v t = lim e 5 + 5080t = lim 8t
8t
tof
tof
tof
t o f 8e 8t
e
d e e dt
tof
(1.68)
Example 1.5
For the circuit of Figure 1.15, i L 0 = 2 A and v C 0 = 5 V . Compute and sketch v t for t ! 0 .
iR
vt
10u 0 t A
iC
iL
50 :
10 H
1 e 640 F
Solution:
This is the same circuit as the that of the two previous examples except that the resistance has been
increased to 50 : . For this example,
123
or
2
D P = 40.96
and as before,
2
1
1
Z 0 =  =  = 64
LC
10 u 1 e 640
Z0 DP =
64 40.96 =
23.04 = 4.8
Since v f = 0 , the total response is just the natural response. Then, from (1.48),
v t = v n t = ke
DP t
cos Z nP t + M = ke
6.4t
cos 4.8t + M
(1.69)
and the constants k and M will be evaluated from the initial conditions.
From the initial condition v C 0 = v 0 = 5 V and (1.69) we get
0
v 0 = ke cos 0 + M = 5
or
k cos M = 5
(1.70)
To evaluate the constants k and M we differentiate (1.69), we evaluate it at t = 0 , we write the equation which describes the circuit at t = 0 + , and we equate these two expressions. Using MATLAB we
get:
syms t k phi; y0=k*exp(6.4*t)*cos(4.8*t+phi); y1=diff(y0)
y1 =
32/5*k*exp(32/5*t)*cos(24/5*t+phi)24/5*k*exp(32/5*t)*sin(24/
5*t+phi)
pretty(y1)
(1.71)
and
124
= 32 4.8k sin M
(1.72)
t=0
i
dv
 = C and
Also, i C = C  or dv
dt
dt
dv
dt
t=0
iC 0 IS iR 0 iL 0
=  = C
C
or
dv
dt
t=0
IS vC 0 e R iL 0
5 e 50 2 = 7.9 u 640 = 5056
=  = 10
1 e 640
C
(1.73)
or
k sin M = 1060
(1.74)
or
1
or
5
 = 1042
k = cos 1.566
6.4t
(1.75)
The plot is shown in Figure 1.16 where we have used the following MATLAB code:
125
We can also use a spreadsheet to plot (1.75). From the columns of that spreadsheet we can read the
following maximum and minimum values and the times these occur.
t (sec)
v (V)
Maximum
0.13
266.71
Minimum
0.79
4.05
Alternately, we can find the maxima and minima by differentiating the response of (1.75) and setting
it equal to zero.
126
iR
vt
iS
iC
iL
50 :
10 H
1 e 640 F
Solution:
This is the same circuit as the previous example where the DC source has been replaced by an AC
source. The total response will consist of the natural response v n t which we already know from the
previous example, and the forced response v f t which is the AC steadystate response, will be found
by phasor analysis.
The t domain to jZ domain transformation yields
i s t = 20 sin 6400t + 90q = 20 cos 6400t I = 20 0q
The admittance Y is
1 =
Y = G + j ZC 
ZL
2
1 2 tan 1 ZC 1  e G
G + ZC 
ZL
ZL
where
1
1
1
1
1
1
G =  =  , ZC = 6400 u  = 10 and  =  = ZL 6400 u 10 64000
640
R 50
and thus
Y =
1
1 2
1
1 2
1
 + 10  tan 10  e  = 10 89.72q
50
50
64000
64000
127
6.4t
(1.76)
or
k cos M  5
(1.77)
To make use of the second initial condition, we differentiate (1.76) using MATLAB as follows, and
then we evaluate it at t = 0 .
syms t k phi; y0=k*exp(6.4*t)*cos(4.8*t+phi)+2*cos(6400*t1.5688);
y1=diff(y0);
% Differentiate v(t) of (1.76)
y1 =
32/5*k*exp(32/5*t)*cos(24/5*t+phi)24/5*k*exp(32/5*t)*sin(24/
5*t+phi)12800*sin(6400*t1961/1250)
or
6.4t
6.4t
dv
 = 6.4ke
cos 4.8t + M 4.8ke
sin 4.8t + M 12800 sin 6400t 1.5688
dt
and
dv
dt
= 6.4k cos M 4.8k sin M 12800 sin 1.5688 = 6.4k cos M 4.8k sin M + 12800
(1.78)
t=0
(1.79)
t=0
i
dv
 = C and
Also, i C = C  or dv
dt
dt
dv
dt
t=0
iC 0
iS 0 iR 0 iL 0
=  = C
C
or
dv
dt
t=0
iS 0 vC 0 e R iL 0
5 e 50 2 = 11456
= = 20
1 e 640
C
(1.80)
128
(1.81)
or
M = 1.53 rad = 89q
6.4t
(1.82)
129
40 K:
R2
R1
50 K:
+ 200 K:
vin
C2 10 nF
C1
v1 R
3
v2
vout
25 nF
Solution:
At node v 1 :
v 1 v in
dv v 1 v out v 1 v 2
 + C 1 1 +  +  = 0 t ! 0
dt
R3
R1
R2
(1.83)
dv out
v2 v1
 = C 2 dt
R3
(1.84)
At node v 2 :
(1.85)
and
dv out
v 1 = R 3 C 2 dt
(1.86)
2
dv 1
dv out
 = R 3 C 2 2
dt
dt
(1.87)
130
dt
2 u 10
4 u 10
4 u 10
5 u 10
or
3
0.05 u 10 v 1 + 25 u 10
dv
1 0.25 u 10 4 v out = 0.5 u 10 5 v in
dt
dv out
4
v 1 = 5 u 10 dt
(1.88)
(1.89)
dv
4 d v out
1 = 5 u 10 2
dt
dt
(1.90)
0.05 u 10 5 u 10
+ 25 u 10 5 u 10
2
dt
dt
3
(1.91)
or
125 u 10
13
2
dv out
d v out
7
 0.25 u 10 4 v out = 10 4 v in

0.25
u
10
2
dt
dt
dv out
d v out
3
6
5
 + 2 u 10  + 2 u 10 v out = 1.6 u 10 v in
2
dt
dt
or
2
dv out
d v out
3
 + 2 u 10 6 v out = 10 6 cos 6280t
+
2
10
u
2
dt
dt
(1.92)
y0 =
[ 1000+1000*i]
[ 10001000*i]
that is,
131
We cannot classify the given circuit as series or parallel and therefore, we should not use the damping
ratio D S or D P . Instead, for the natural response v n t we will use the general expression
v n t = Ae
s1 t
+ Be
s2 t
= e
Dt
k 1 cos Et + k 2 sin Et
(1.93)
where
s 1 ,s 2 = D r jE = 1000 r j1000
1000t
(1.94)
Since the right side of (1.92) is a sinusoid, the forced response has the form
v f t = k 3 cos 6280t + k 4 sin 6280t
(1.95)
Of course, for the derivation of the forced response we could use phasor analysis but we must first
derive an expression for the impedance or admittance because the expressions weve used earlier are
valid for series and parallel circuits only.
The coefficients k 3 and k 4 will be found by substitution of (1.95) into (1.92) and then by equating
like terms. Using MATLAB we get:
syms t k3 k4; y0=k3*cos(6280*t)+k4*sin(6280*t); y1=diff(y0)
y1 =
6280*k3*sin(6280*t)+6280*k4*cos(6280*t)
y2=diff(y0,2)
y2 =
39438400*k3*cos(6280*t)39438400*k4*sin(6280*t)
y=y2+2*10^3*y1+2*10^6*y0
y =
37438400*k3*cos(6280*t)37438400*k4*sin(6280*t)12560000*k3*sin(6280*t)+12560000*k4*cos(6280*t)
Equating like terms with (1.92) we get
6
(1.96)
132
y =
k3: [1x1 sym]
k4: [1x1 sym]
y.k3
ans =
0.0240
y.k4
ans =
0.0081
that is, k 3 = 0.024 and k 4 = 0.008 . Then, by substitution into (1.95)
v f t = 0.024 cos 6280t 0.008 sin 6280t
(1.97)
v out t = v n t + v f t = e
k 1 cos 1000t + k 2 sin 1000t
+ 0.024 cos 6280t 0.008 sin 6280t
(1.98)
We will use the initial conditions v C1 = v C2 = 0 to evaluate k 1 and k 2 . We observe that v C2 = v out
and at t = 0 relation (1.98) becomes
0
1000t
(1.99)
To evaluate the constant k 2 , we make use of the initial condition v C1 0 = 0 . We observe that
v C1 = v 1 and by KCL at node v 1 we have:
dv out
v1 v2
 + C 2  = 0
dt
R3
or
dv out
v1 0
8
4 = 10 dt
5 u 10
133
dv out
dt
= 0
(1.100)
t=0
The last step in finding the constant k 2 is to differentiate (1.99), evaluate it at t = 0 , and equate it
with (1.100). This is done with MATLAB as follows:
y0=exp(1000*t)*(0.024*cos(1000*t)+k2*sin(1000*t))...
+0.024*cos(6280*t)0.008*sin(6280*t);
y1=diff(y0)
y1 =
1000*exp(1000*t)*(3/125*cos(1000*t)+k2*sin(1000*t))+exp(1000*t)*(24*sin(1000*t)+1000*k2*cos(1000*t))3768/
25*sin(6280*t)1256/25*cos(6280*t)
or
dv out
1000t 3
 cos 1000t + k 2 sin 1000t
 = 1000e
125
dt
+e
1000t
1256
3768
 sin 6280t  cos 6280t
25
25
and
dv out
dt
t=0
3
= 1000  + 1000k 2 1256
 125
25
(1.101)
or
k 2 = 0.026
1000t
(1.102)
134
Voltage (V)
0.02
0.01
0.00
0.01
0.02
0.03
0.000
0.002
0.004
0.006
0.008
Time (s)
135
s 1 s 2 = D S r D S Z 0 = D S r E S if D S ! Z 0
or
2
s 1 s 2 = D S r Z 0 D S = D S r Z n S if Z 0 ! D S
where
R
D S = 2L
1
Z 0 = LC
ES =
DS Z0
Zn S =
Z0 DS
If D 2S ! Z 20 , the roots s 1 and s 2 are real, negative, and unequal. This results in the overdamped natural response and has the form
in t = k1 e
s1 t
+ k2 e
s2 t
If D 2S = Z 20 , the roots s 1 and s 2 are real, negative, and equal. This results in the critically damped
natural response and has the form
in t = e
136
DS t
k1 + k2 t
Summary
If Z 20 ! D 2S , the roots s 1 and s 2 are complex conjugates. This is known as the underdamped or
oscillatory natural response and has the form
in t = e
DS t
k 1 cos Z n S t + k 2 sin Z n S t = k 3 e
DS t
cos Z n S t + M
x For a parallel GLC circuit, the roots s 1 and s 2 are found from
2
s 1 s 2 = D P r D P Z 0 = D P r E P if D P ! Z 0
or
2
s 1 s 2 = D P r Z 0 D P = D P r Z nP if Z 0 ! D P
where
GD P = 2C
1
Z 0 = LC
EP =
DP Z0
Z nP =
Z0 DP
If D 2P ! Z 20 , the roots s 1 and s 2 are real, negative, and unequal. This results in the overdamped natural response and has the form
vn t = k1 e
s1 t
+ k2 e
s2 t
If D 2P = Z 20 , the roots s 1 and s 2 are real, negative, and equal. This results in the critically damped
natural response and has the form
vn t = e
DP t
k1 + k2 t
If Z 20 ! D 2P , the roots s 1 and s 2 are complex conjugates. This results in the underdamped or oscillatory natural response and has the form
vn t = e
DP t
k 1 cos Z nP t + k 2 sin Z nP t = k 3 e
DP t
cos Z nP t + M
x If a second order circuit is neither series nor parallel, the natural response if found from
yn = k1 e
s1 t
+ k2 e
s2 t
or
yn = k1 + k2 t e
or
yn= e
Dt
s1 t
k 3 cos Et + k 4 sin E t = e
Dt
k 5 cos Et + M
depending on the roots of the characteristic equation being real and unequal, real and equal, or
complex conjugates respectively.
137
1. For the circuit of Figure 1.21, it is known that v C 0 = 0 and i L 0 = 0 . Compute and sketch
v C t and i L t for t ! 0 .
iL t
10 :
0.2 H
+
vC t
8 mF
+
100u 0 t V
2. For the circuit of Figure 1.22, it is known that v C 0 = 0 and i L 0 = 0 . Compute and sketch
v C t and i L t for t ! 0 .
iL t
4:
5H
+
100u 0 t V
21.83 mF
+
vC t
3. In the circuit of Figure 1.23, the switch S has been closed for a very long time and opens at
t = 0 . Compute v C t for t ! 0 .
100 :
+
20 H
t = 0
+
v t
C
400 :
100 V
1 e 120 F
138
Exercises
4. In the circuit of Figure 1.24, the switch S has been closed for a very long time and opens at t = 0 .
Compute v C t for t ! 0 .
100 :
20 H
vS
+
t = 0
+
v t
C
400 :
1 e 120 F
v S = 100 cos t u 0 t V
5. In the circuit of Figure 1.25, the switch S has been in position A for closed for a very long time
and it is placed in position B at t = 0 . Find the value of R that will cause the circuit to become
critically damped and then compute v C t and i L t for t ! 0
t = 0
3:
6:
S
B
+
2:
12 V
iL t
3H
+
vC t
1 e 12 F
6. In the circuit of Figure 1.26, the switch S has been closed for a very long time and opens at t = 0 .
Compute v AB t for t ! 0 .
t = 0
+
4:
B
A
2H
12 V
2:
1e4 F
139
140
Solutions to Exercises
1.
iL t
10 :
+
0.2 H
+
vC t
8 mF
100u 0 t V
di
Ri + L  + v C = 100
dt
t!0
dv
dt
dv
d vC
RC C + LC + v C = 100
2
dt
dt
2
d v C R dv C
1
100
+   +  v C = 2
LC
L
dt
LC
dt
2
d v C 10 dv C
1
100
3 v C = +   + 
2
3
0.2 dt 0.2 u 8 u 10
0.2 u 8 u 10
dt
2
dv
d vC
+ 50 C + 625 v C = 62500
2
dt
dt
s + 50s + 625 = 0
DS t
k 1 + k 2 t = 100 + e
25 t
k 1 + k 2 t (1)
With the first initial condition v C 0 = 0 the above expression becomes 0 = 100 + e 0 k 1 + 0 or
k 1 = 100 and by substitution into (1) we get
v C t = 100 + e
25 t
k 2 t 100 (2)
To evaluate k 2 we make use of the second initial condition i L 0 = 0 and since i L = i C , and
i = i C = C dv C e dt , we differentiate (2) using the following MATLAB code:
141
v1 =
25*exp(25*t)*(k2*t100)+exp(25*t)*k2
Thus,
dv C
25t
25t
 = k 2 e 25e k 2 t 100
dt
and
dv
Cdt
dv
dt
i
C
= k 2 + 2500 (3)
t=0
i
C
t=0
iL 0
=  = 0 (4)
C
From (3) and (4) k 2 + 2500 = 0 or k 2 = 2500 and by substitution into (2)
v C t = 100 e
25 t
iL =
1/5*exp(25*t)*(100+2500*t)20*exp(25*t)
Thus,
i L t = i C t = 0.2e
25t
25t
142
Solutions to Exercises
iL(t)
0
0.005 0.7191
2.206
0.010 2.6499
3.894
100
0.015 5.4977
5.155
80
0.020 9.0204
6.065
60
0.025
13.02
6.691
0.030 17.336
7.085
20
0.035 21.838
7.295
0.040 26.424
7.358
0.045 31.011
7.305
0.050 35.536
7.163
0.055 39.951
6.953
0.060 44.217
6.694
0.065 48.311
6.4
0.070 52.212
6.082
0.075
55.91
5.751
0.080 59.399
5.413
0.085 62.677
5.076
0.090 65.745
4.743
0.095 68.608
4.418
0.100
71.27
4.104
0.105 73.741
3.803
0.110 76.027
3.516
0.115 78.139
3.244
vC(t)
Volts
0.000
40
0.0
0.1
0.2
0.3
0.4
0.5
0.4
0.5
Time
iL(t)
8
6
Amps
vC(t)
4
2
0
0.0
0.1
0.2
0.3
Time
2.
iL t
4:
5H
+
100u 0 t V
21.83 mF
+
vC t
The general form of the differential equation that describes this circuit is same as in Exercise 1,
that is,
2
d v C R dv C 1
100
+   +  v C = 2
L
dt
LC
LC
dt
t!0
2
dv
d vC
C + 9.16v C = 916
+
0.8
2
dt
dt
From the characteristic equation s 2 + 0.8s + 9.16 = 0 and the MATLAB code below
143
ans =
0.4000 + 3.0000i
0.4000  3.0000i
we find that s 1 = 0.4 + j3 and s 2 = 0.4 j3 . Therefore, the total solution is
v C t = v C f + v C n = 100 + ke
DS t
cos Z n S t + M
where
D S = R e 2L = 0.4
and
Zn S =
Z0 DS =
1 e LC R e 4L =
9.16 0.16 = 3
Thus,
v C t = 100 + ke
0.4t
cos 3t + M (1)
v1 =
2/5*k*exp(2/5*t)*cos(3*t+phi)3*k*exp(2/5*t)*sin(3*t+phi)
Thus,
dv C
0.4t
0.4t
 = 0.4k e
cos 3t + M 3ke
sin 3 t + M
dt
dv C
dt
i
C
= 40 3k sin M (3)
t=0
i
C
144
t=0
iL 0
=  = 0 (4)
C
Solutions to Exercises
From (3) and (4)
3k sin M = 40 (5)
The value of k can be found from either (2) or (5). From (2)
k cos 0.1236 = 100
100
k =  = 100.8
cos 0.1236
0.4t
iL =
137529/156250*exp(2/5*t)*cos(3*t663/5000)+412587/62500*exp(2/5*t)*sin(3*t663/5000)
137529/156250, 412587/62500
ans =
0.8802
ans =
6.6014
i L t = 0.88e
0.4t
0.4t
sin 3t 7.6q
145
vC(t)
iL(t)
vC(t)
0.198
0.020 0.1677
0.395
0.030
0.394
0.591
0.040 0.7094
0.784
0.050 1.1129
0.975
0.060 1.6034
1.164
0.070 2.1798
1.35
0.080 2.8407
1.534
0.090 3.5851
1.714
0.100 4.4115
1.892
0.110 5.3185
2.066
0.120 6.3046
2.238
0.130 7.3684
2.405
0.140 8.5082
2.57
0.150 9.7224
2.73
0.160 11.009
2.887
150
50
50 0
12
12
iL(t)
10
10
3.
100 :
+
20 H
iL 0
+
v 0
C
400 :
100 V
1 e 120 F
At this time the inductor behaves as a short and the capacitor as an open. Then,
i L 0 = 100 e 100 + 400 = I 0 = 0.2 A
146
Solutions to Exercises
20 H
100 :
+
v t
C
+
100 V
1 e 120 F
The general form of the differential equation that describes this circuit is same as in Exercise 1,
that is,
2
d v C R dv C 1
100
+   +  v C = 2
L
dt
LC
LC
dt
t!0
2
dv C
d vC
 + 6v C = 600
+
5
2
dt
dt
From the characteristic equation s 2 + 5s + 6 = 0 we find that s 1 = 2 and s 2 = 3 and the total
response for the capacitor voltage is
v C t = v C f + v C n = 100 + k 1 e
s1 t
+ k2 e
s2 t
= 100 + k 1 e
2t
+ k2 e
3t
(1)
or
k 1 + k 2 = 20 (2)
i
C
= 2k 1 3k 2 (3)
t=0
i
C
t=0
iL 0
0.2  = 24 (4)
=  = 1 e 120
C
147
2t
+ 16e
3t
4.
100 :
20 H
t = 0
vS
+
+
v t
C
400 :
1 e 120 F
v S = 100 cos t u 0 t V
This is the same circuit as in Exercise 3 where the DC voltage source has been replaced by an AC
source that is being applied at t = 0 + . No initial conditions were given so we will assume that
i L 0 = 0 and v C 0 = 0 . Also, the circuit constants are the same and thus the natural
We will find the forced (steadystate) response using phasor circuit analysis where Z = 1 ,
jZL = j20 , j e ZC = j120 , and 100 cos t 100 0q . The phasor circuit is shown below.
100 :
j20 :
VS
+
j 120 :
V S = 100 0q V
+
V
C
2t
+ k2 e
3t
(1)
148
Solutions to Exercises
v C 0 = 0 = 60 2 cos 135q + k 1 + k 2
and
dv C
dt
= 60 2 sin 45q 2k 1 3k 2
t=0
dv
Cdt
dv
dt
i
C
= 60 2k 1 3k 2 (3)
t=0
i
C
t=0
iL 0
=  = 0 (4)
C
Simultaneous solution of (2) and (5) yields k 1 = 120 and k 2 = 60 . Then, by substitution into (1)
2t
3t
5.
3:
t = 0
A
6:
S
B
+
12 V
2:
iL t
3H
+
vC t
1 e 12 F
`
We must first find the value of R before we can establish initial conditions for i L 0 = 0 and
v C 0 = 0 . The condition for critical damping is
149
D P Z 0 = 0 where D P = G e 2C = 1 e 2R'C
+ v6 :
1:
3:
+
vC 0
+
12 V
iL 0
and
v6 :
7.2
i L 0 =  =  = 1.2 A
6
6
1:
iR t
2:
iC t
iL t
3H
+
vC t
1 e 12 F
Since the circuit is critically damped, the solution has the form
vC t = e
DP t
k1 + k2 t
150
2 t
k 1 + k 2 t (1)
Circuit Analysis II with MATLAB Applications
Orchard Publications
Solutions to Exercises
With the initial condition v C 0 = 7.2 V relation (1) becomes 7.2 = e0 k 1 + 0 or k 1 = 7.2 V
and (1) simplifies to
vC t = e
2 t
7.2 + k 2 t (2)
and
dv
Cdt
i
C
dv
dt
t=0
iC 0
0
 =  = 0 (4)
= C
C
2 t
2t
6.
4:
2:
12 V
iL 0
151
+
2H
1e4 F
+
vC 0
R1
R2
4:
B
+
1e4 F
vC t
C
2H
d vC
dv
R 1 + R 2 C C + LC + vC = 0
2
dt
dt
2
d v C R 1 + R 2 dv C 1
+   +  v C = 0
2
dt LC
L
dt
2
dv
d vC
+ 3 C + 2v C = 0
2
dt
dt
The characteristic equation of the last expression above yields s 1 = 1 and s 2 = 2 and thus
t
vC t = k1 e + k2 e
2t
(1)
and
dv
Cdt
dv
dt
i
C
= k 1 2k 2 (3)
t=0
i
C
152
Solutions to Exercises
dv C
dt
t=0
iL 0
6
=  =  = 24 (4)
C
1e4
and from (2) and (5) k 1 = 48 and k 2 = 36 . By substitution into (1) we get
t
v C t = 48e 36 e
2t
Then,
2
di
d iC
 vC t
v AB = v L t v C t = L L v C t = LC 2
dt
dt
2
d
t
2t
t
2t
= 0.5 2 48e 36 e 48e 36 e
dt
2t
= 24 e 108 e
153
2t
2t
2t
= 24 e + 4.5e
154
Chapter 2
Resonance
his chapter defines series and parallel resonance. The quality factor Q is then defined in terms
of the series and parallel resonant frequencies. The halfpower frequencies and bandwidth are
also defined in terms of the resonant frequency.
jZL
1 e jZC
The impedance Z is
V
1
1
Phasor Voltage
Impedance = Z =  = S = R + j Z L +  = R + j Z L 
I
ZC
jZC
Phasor Current
(2.1)
or
Z =
R + Z L 1 e Z C tan Z L 1 e Z C e R
(2.2)
R + ZL 1 e ZC
(2.3)
and
1
T Z = tan Z L 1 e Z C e R
(2.4)
21
Chapter 2 Resonance
Magnitude of Impedance
Z
ZL
R
Z0
1 / ZC
ZL1 / ZC
Radian Frequency
The frequency at which the capacitive reactance X C = 1 e Z C and the inductive reactance X L = Z L
are equal is called the resonant frequency. The resonant frequency is denoted as Z 0 or f 0 and these
can be expressed in terms of the inductance L and capacitance C by equating the reactances, that is,
1
Z 0 L = Z0C
2
1
LC
Z 0 = 
1
LC
Z 0 = 
(2.5)
1 f 0 = 2 S LC
(2.6)
and
We observe that at resonance Z 0 = R where Z 0 denotes the impedance value at resonance, and
T Z = 0 . In our subsequent discussion the subscript zero will be used to indicate that the circuit vari
22
Series Resonance
jX L = j10 :
R = 1.2 :
L=0.2 mH
VS
I
120 0q V
j XC
Solution:
At resonance,
jX L = jX C
and thus
Z 0 = R = 1.2 :
Then,
120 V = 100 A
I 0 = 1.2 :
Since
X L0 = Z 0 L = 10 :
it follows that
10
L
10
0.2 u 10
Therefore,
1
X C0 = X L0 = 10 = Z0 C
or
1
C =  = 2 PF
10 u 50000
Now,
V R0 = RI 0 = 1.2 u 100 = 120 V
V L0 = Z 0 LI 0 = 50000 u 0.2 u 10
u 100 = 1000
and
1
1
 u 100 = 1000 V
V C0 =  I 0 = 6
Z0 C
50000 u 2 u 10
23
Chapter 2 Resonance
VL0 = 1000 V
VR0 = 120 V
VC0 = 1000 V
Figure 2.4. Phasor diagram for Example 2.1
Figure 2.4 reveals that V L0 = V C0 = 1000 V and these voltages are much higher than the applied
voltage of 120 V . This illustrates the useful property of resonant circuits to develop high voltages
across capacitors and inductors.
Z 0 L = Z0C
and
VS
I 0 = R
Then
Z0 L
VS
V L0 = Z 0 LI 0 = Z 0 L  =  V S
R
R
(2.7)
1
1 VS
1
V C0 =  I 0 =   =  V S
Z0C
Z0 C R
Z 0 RC
(2.8)
and
At series resonance the left sides of (2.7) and (2.8) are equal and therefore,
Z0 L
1
= R
Z 0 RC
*
We denote the quality factor for series resonant circuits as Q 0S , and the quality factor for parallel resonant circuits as Q 0P .
24
(2.9)
In a practical circuit, the resistance R in the definition of Q 0S above, represents the resistance of the
inductor and thus the quality factor Q 0S is a measure of the energy storage property of the inductance
L in relation to the energy dissipation property of the resistance R of that inductance.
In terms of Q 0S , the magnitude of the voltages across the inductor and capacitor are
V L0 = V C0 = Q 0S V S
(2.10)
and therefore, we say that there is a resonant rise in the voltage across the reactive devices and it is
equal to the Q 0S times the applied voltage. Thus in Example 2.1,
V C0
V L0
1000
25
Q 0S =  =  =  = VS
VS
120
3
The quality factor Q is also a measure of frequency selectivity. Thus, we say that a circuit with a high
Q has a high selectivity, whereas a low Q circuit has low selectivity. The high frequency selectivity is
more desirable in parallel circuits as we will see in the next section.
Figure 2.5 shows the relative response versus Z for Q = 25 50 , and 100 where we observe that
highest Q provides the best frequency selectivity, i.e., higher rejection of signal components outside
the bandwidth BW = Z 2 Z 1 which is the difference in the 3 dB frequencies.
1.2
1.0
0.8
Q=50
0.6
Q=100
0.4
0.2
0.0
Z1 Z0 Z2
Zr/s
25
Chapter 2 Resonance
We will see later that
Z0
Frequency = Resonant
Q = Bandwidth
Z2 Z1
(2.11)
We also observe from (2.9) that selectivity depends on R and this dependence is shown on the plot
of Figure 2.6.
High and Low Q Curves
Dependence on R
Relative Response
2R
If we keep one reactive device, say L , constant while varying C , the relative response shifts as
shown in Figure 2.7, but the general shape does not change.
Relative Response
0.5C
26
Parallel Resonance
V G
IG L
IS
IL
IC
Figure 2.8. Parallel GLC circuit for defining parallel resonance
Phasor Voltage
jZL
V
ZL
or
Y =
G + Z C 1 e Z L tan 1 Z C 1 e Z L e G
2
(2.12)
G + ZC 1 e ZL
(2.13)
and
1 Z C 1 e Z L
T Y = tan G
(2.14)
The frequency at which the inductive susceptance B L = 1 e Z L and the capacitive susceptance
B C = Z C are equal is, again, called the resonant frequency and it is also denoted as Z 0 We can find
Z0
Since
1
Z 0 C Z0 L
then,
1
Z 0 C = Z0 L
and
1
LC
Z 0 = 
(2.15)
27
Chapter 2 Resonance
Magnitude of Admittance
Y
ZC
G
Z0
1 / ZL
ZC1/ZL
Radian Frequency
(2.16)
TY = 0
(2.17)
and
Example 2.2
For the circuit of Figure 2.10, i S t = 10 cos 5000t mA . Compute i G t , i L t , and i C t .
iS t
v t
iG t
0.01:
L
1
iL t C
10 mH
iC t
4 PF
Solution:
The capacitive and inductive susceptances are
B C = Z C = 5000 u 4 u 10
= 0.02 :
and
1
1
1
B L =  =  = 0.02 :
3
ZL
5000 u 10 u 10
and since B L = B C , the given circuit operates at parallel resonance with Z 0 = 5000 rad e s . Then,
28
and
i G t = i S t = 10 cos 5000t mA
In phasor form,
v 0 t = cos 5000t V V 0 = 1 0q
Now,
I L0 = jB L V 0 = 1 90 q 0.02 1 0q = 0.02 90 q A
or
i L0 t = 20 sin 5000t mA
Similarly,
I C0 = jB C V 0 = 1 90q 0.02 1 0q = 0.02 90q A
or
i C0 t = 20 sin 5000t mA
Z 0 C = Z0 L
and
I
V 0 = SG
Then,
Z0 C
IS
I C0 = Z 0 CV 0 = Z 0 C  =  I S
G
G
(2.18)
Also,
Circuit Analysis II with MATLAB Applications
Orchard Publications
29
Chapter 2 Resonance
1
1 VS
1
I L0 =  V 0 =   =  I S
Z0 L
Z0 L G
Z 0 GL
(2.19)
At parallel resonance the left sides of (2.18) and (2.19) are equal and therefore,
Z0C
1  = G
Z 0 GL
Now, by definition
Z0C
1
 = Q 0P = G
Z 0 GL
(2.20)
The above expressions indicate that at parallel resonance, it is possible to develop high currents
through the capacitors and inductors. This was found to be true in Example 2.2.
(2.21)
Essentially, the resonant frequency is the frequency at which the inductor gives up energy just as fast
as the capacitor requires it during one quarter cycle, and absorbs energy just as fast as it is released by
the capacitor during the next quarter cycle. This can be seen from Figure 2.11 where at the instant of
maximum current the energy is all stored in the inductance, and at the instant of zero current all the
energy is stored in the capacitor.
WL & WC in Series RLC Circuit
WC
Energy (J)
WL
Z
vC
iL
210
Then,
Ip
v C =  sin Z t
ZC
Also,
1 2
1 2
2
W L =  Li =  LI p cos Z t
2
2
(2.22)
and
2
WC
1 2
1 Ip
2
 sin Z t
=  Cv =  2
2
2Z C
(2.23)
(2.24)
and this expression is true for any series circuit, that is, the circuit need not be at resonance. However, at resonance,
1
Z 0 L = Z0 C
or
1L = Z 20 C
(2.25)
and (2.25) shows that the total energy W T is dependent only on the circuit constants L , C and resonant frequency, but it is independent of time.
Next, using the general definition of Q we get:
2
Q 0S
1 e 2 I p L
f0 L
Maximum Energy Stored
 = 2 S = 2 S  = 2 S 2
R
Energy Dissipated per Cycle
1 e 2 I p R e f 0
211
Chapter 2 Resonance
or
Z0 L
Q 0S = R
(2.26)
Q 0S
1 e 2 I p 1 e Z 20 C
f0
Maximum Energy Stored
= 2 S  = 2 S = 2 S 2
2
Energy Dissipated per Cycle
Z 0 RC
1 e 2 I p R e f 0
or
Z0
1 = Q 0S = 2
Z 0 RC
Z 0 RC
(2.27)
(2.28)
Vp
0.707Vp
Z1 Z0
Z2
Z
212
impedance of a parallel resonant circuit, is less than the magnitude at resonance by a factor of 2 as
shown above. We observe also, that Z 1 and Z 2 are not exactly equidistant from Z 0 . However, it is
convenient to assume that they are equidistant, and unless otherwise stated, this assumption will be
followed in the subsequent discussion.
We call Z 1 the lower halfpower point, and Z 2 the upper halfpower point. The difference Z 2 Z 1 is
the halfpower bandwidth BW , that is,
Bandwidth = BW = Z 2 Z 1
(2.29)
The names halfpower frequencies and halfpower bandwidth arise from the fact that the power at
2
ZL
Z0
, we get
Multiplying the j term by G Z0 G
ZZ 0 C
Z0
 Y = G + jG  Z G ZZ LG
0
0
by substitution we get
Z
Z
 0
Y = G 1 + jQ 0P  Z0 Z
(2.30)
and if Z = Z 0 , then
Y = G
213
Chapter 2 Resonance
Next, we want to find the bandwidth Z 2 Z 1 in terms of the quality factor Q 0P . At the halfpower
points, the magnitude of the admittance is 2 e 2 Y p and, if we use the halfpower points as reference, then to obtain the admittance value of
Y max =
2G
we must set
Z
Q 0P 2 0 = 1
Z0 Z2
for Z = Z 2 .
We must also set
Z
Z
Q 0P 1 0 = 1
Z0 Z1
for Z = Z 1 .
Recalling that 1 r j1 =
1 2
1
1 +  +  2Q 0P
2Q 0P
(2.31)
Z1 =
1 2
1 1 +   2Q 0P
2Q 0P
(2.32)
and
0
BW = Z 2 Z 1 = Q 0P
(2.33)
or
f0
BW = f 2 f 1 = Q 0P
(2.34)
As mentioned earlier, Z 1 and Z 2 are not equidistant from Z 0 In fact, the resonant frequency Z 0 is
the geometric mean* of Z 1 and Z 2 , that is,
Z0 =
(2.35)
Z1 Z2
* The geometric mean of n positive numbers a 1 , a 2 ,..., a n is the nth root of the product. a 1 a 2 } a n
214
0.001:
1 mH
0.4 PF
Solution:
a.
2
1
LC
Z 0 =  =  = 25 u 10
1 u 10
u 0.4 u 10
or
Z 0 = 50000 r e s
b.
4
6
Z0C
5 u 10 u 0.4 u 10  = 20
Q 0P =  = 3
G
10
c.
Z0
50000
BW =  =  = 2500 = rad e s
Q 0P
20
d.
BW
e.
BW
215
Chapter 2 Resonance
2.8 A Practical Parallel Resonant Circuit
In our previous discussion, we assumed that the inductors are ideal, but a real inductor has some
resistance. The circuit shown in Figure 2.14 is a practical parallel resonant circuit.
Y
R
IC
IL
C
R
Figure 2.15. Simplified network for derivation of the resonant frequency
and
V
I C =  = j Z C V
1 e jZC
where
R
V
Re ^ I L ` = 2
2
R + ZL
and
Z L V
Im ^ I L ` = 2
2
R + ZL
Also,
Re ^ I C ` = 0
and
Im ^ I C ` = Z C V
216
(2.36)
= > Re ^ I T ` + Im ^ I T ` @V
Z0L
Im ^ I T ` = Im ^ I L ` + Im ^ I C ` = Z 0 C  V = 0
2
2
R + Z0 L
1 R
LC L 2
(2.37)
or
1 1 R2
f 0 =   2S LC L 2
(2.38)
1
We observe that for R = 0 , (2.37) reduces to Z 0 = as before.
LC
217
Chapter 2 Resonance
Antenna
Speaker
Local
Oscillator
Radio
Frequency
Amplifier
Mixer
Intermediate
Frequency
Amplifier
Detector
Audio
Frequency
Amplifier
The antenna picks up signals from several stations and these are fed into the Radio Frequency ( RF )
Amplifier which improves the SignaltoNoise ( S e N ) ratio. The RF amplifier also serves as a preselector. This preselection suppresses the imagefrequency interference as explained below.
When we tune to a station of, say 740 KHz , we are setting the RF circuit to 740 KHz and at the
same time the local oscillator is set at 740 KHz + 456 KHz = 1196 KHz . This is accomplished by
the capacitor in the RF amplifier which is also ganged to the local oscillator. These two signals, one
of 740 KHz and the other of 1196 KHz , are fed into the mixer whose output into the Intermediate
Frequency ( IF ) amplifier is 456 KHz ; this is the difference between these two frequencies
( 1196 KHz 740 KHz = 456 KHz ).
The IF amplifier is always set at 456 KHz and therefore if the antenna picks another signal from
another station, say 850 KHz , it would be mixed with the local oscillator to produce a frequency of
1196 KHz 850 KHz = 346 KHz but since the IF amplifier is set at 456 KHz , the unwanted
850 KHz signal will not be amplified. Of course, in order to hear the signal at 850 KHz the radio
receiver must be retuned to that frequency and the local oscillator frequency will be changed to
850 KHz + 456 KHz = 1306 KHz so that the difference of these frequencies will be again
456 KHz .
Now let us assume that we select a station at 600 KHz . Then, the local oscillator will be set to
600 KHz + 456 KHz = 1056 KHz so that the IF signal will again be 456 KHz . Now, let us suppose
that a powerful nearby station broadcasts at 1512 KHz and this signal is picked up by the mixer circuit. The difference between this signal and the local oscillator will also be 456 KHz
1512 KHz 1056 KHz = 456 KHz . The IF amplifier will then amplify both signals and the result
will be a strong interference so that the radio speaker will produce unintelligent sounds. This interference is called imagefrequency interference and it is reduced by the RF amplifier before entering
the mixer circuit and for this reason the RF amplifier is said to act as a preselector.
The function of the detector circuit is to convert the IF signal which contains both the carrier and
the desired signal to an audio signal and this signal is amplified by the Audio Frequency ( AF ) Amplifier whose output appears at the radio speaker.
218
Solution:
a.
2
1
LC
Z 0 = 
or
2
1
f 0 = 2
4 S LC
Then,
1
C =  = 77.2 pF
3
3 2
2
4 S 0.5 u 10 u 810 u 10
b.
Z0C
Q 0P = G
or
5
12
2 S f0 C
1
2 S u 8.1 u 10 u 77.2 u 10
G =  = = 5.4 P:
75
Q 0P
c.
I
I
I
I
V 810 KHz =  =  =  = 6
Y 810 KHz
Y0
G
5.24 u 10
(2.39)
Also,
I V 740 KHz = Y 740 KHz
where
Y 740 KHz =
2
1 2
G + Z C 
ZL
219
Chapter 2 Resonance
or
Y 740 KHz =
6 2
12
2
1

3
3
2S u 740 u 10 u 0.5 u 10
or
Y 740 KHz = 71.2 P:
and
I
V 740 KHz = 6
71.2 u 10
(2.40)
6
6
V 810 KHz
71.2 u 10
I e 5.24 u 10
 =  =  = 13.6
6
6
V 740 KHz
I e 71.2 u 10
5.24 u 10
(2.41)
that is, the voltage developed across the parallel circuit when it is tuned at f = 810 KHz is 13.6
times larger than the voltage developed at f = 740 KHz .
2.10 Summary
x In a series RLC circuit, the frequency at which the capacitive reactance X C = 1 e Z C and the
Z 0 = 
and
1
f 0 = 2 S LC
x The quality factor Q 0S at series resonance is defined as
Z0 L
1 = Q 0S = R
Z 0 RC
x In a parallel GLC circuit, the frequency at which the inductive susceptance B L = 1 e Z L and the
capacitive susceptance B C = Z C are equal is, again, called the resonant frequency and it is also
denoted as Z 0 As in a series RLC circuit, the resonant frequency is
220
Summary
1
LC
Z 0 = 
magnitude of the input admittance of a parallel resonant circuit, is greater than the magnitude at
resonance by a factor of 2 , or equivalently, the frequencies at which the magnitude of the input
impedance of a parallel resonant circuit, is less than the magnitude at resonance by a factor of 2 .
x We call Z 1 the lower halfpower point, and Z 2 the upper halfpower point. The difference
Z2 Z1
0
BW = Z 2 Z 1 = Q 0P
or
f0
BW = f 2 f 1 = Q 0P
221
Chapter 2 Resonance
2.11 Exercises
1. A series RLC circuit is resonant at f 0 = 1 MHz with Z 0 = 100 : and its halfpower bandwidth
is BW = 20 KHz . Find R , L , and C for this circuit.
2. For the network of Figure 2.17, the impedance Z 1 is variable, Z 2 = 3 + j4 and Z 3 = 4 j3 . To
what value should Z 1 be adjusted so that the network will operate at resonant frequency?
Z1
Z IN
Z3
Z2
3. For the circuit of Figure 2.18 with the capacitance C adjusted to 1 PF , the halfpower frequencies are f 1 = 925 KHz and f 2 = 1075 KHz .
a. Compute the approximate resonant frequency.
b. Compute the exact resonant frequency.
c. Using the approximate value of the resonant frequency, compute the values of Q op , G , and L .
4. The GLC circuit of Figure 2.19, is resonant at f 0 = 500 KHz with V 0 = 20 V and its halfpower
bandwidth is BW = 20 KHz .
a. Compute L , C , and I 0 for this circuit.
b. Compute the magnitude of the admittances Y 1 and Y 2 corresponding to the halfpower frequencies f 1 and f 2 . Use MATLAB to plot Y in the 100 KHz d f d 1000 KHz range.
+
V
222
Exercises
5. For the circuit of Figure 2.20, v s = 170 cos Z t and Q 0 = 50 . Find:
a. Z 0
b. BW
c. Z 1 and Z 2
d. V C0
L
R1
1 mH
1:
C
10 :
R2
1 PF
vs
6. The seriesparallel circuit of Figure 2.21, will behave as a filter if the parallel part is made resonant
to the frequency we want to suppress, and the series part is made resonant to the frequency we
wish to pass. Accordingly, we can adjust capacitor C 2 to achieve parallel resonance which will
reject the unwanted frequency by limiting the current through the resistive load to its minimum
value. Afterwards, we can adjust C 1 to make the entire circuit series resonant at the desired frequency thus making the total impedance minimum so that maximum current will flow into the
load.
For this circuit, we want to set the values of capacitors so that v LOAD will be maximum at
f 1 = 10 KHz and minimum at f 2 = 43 KHz . Compute the values of C 1 and C 2 that will
achieve these values. It is suggested that you use MATLAB to plot v LOAD versus frequency f in
the interval 1 KHz d f d 100 KH to verify your answers.
C1
100 :
v = 170 cos Z t
S
C2
R1
2 mH
+ RL
v LOAD
1:
223
Chapter 2 Resonance
2.12 Solutions to Exercises
1. At series resonance Z 0 = R = 100 and thus R = 100 : . We find L from Q 0S = Z 0 L e R where
Z 0 = 2Sf 0 . Also,
6
Z0
Z0
2S u 10  = 50
 =  = Q 0S = 3
Z2 Z1
BW
2S u 20 u 10
Then,
R Q 0S
100 u 50 = 0.796 mH
L =  = 6
Z0
2S u 10
and from Z 20 = 1 e LC
1
1 =  = 31.8 pF
C = 2
2
6
4
Z0 L
2S u 10 u 7.96 u 10
R = 100.00 Ohms
L = 7.96e004 H
C = 3.18e011 F
2.
Z1
Z IN
Z3
Z2
Z IN = Z 1 + Z 2 __ Z 3
where
j
j9 + j16 + 12 7
3 + j4 4 j3  = 12
 Z 2 __ Z 3 = 7j
7+j
3 + j4 + 4 j3
168 + j49 j24 + 7 175 + j25
=  =  = 3.5 + j0.5
2
2
50
7 +1
224
Solutions to Exercises
Equating real and imaginary parts we get
R IN = R 1 + 3.5
0 = jX 1 + j0.5
and while R 1 can be any real number, we must have jX 1 = j0.5 and thus
Z 1 = R 1 j0.5 :
3.
a. BW = f 2 f 1 = 1075 925 = 150 KHz . Then,
f 0 = f 1 + BW e 2 = 925 + 150 e 2 = 1000 KHz
f1 f2 =
Z0 C
f0
 = 20 e 3 . Also, Q 0P =  = 1000
 . Then
c. Q 0P = 
f2 f1
150
6
6
Z0 C
2Sf 0 C
1
3S
2S u 10 u 10
G =  =  =  =  = 0.94 :
Q 0P
Q 0P
20 e 3
10
and
1
1
1
L =  =  =  = 0.025 PH
2 2
2
12
6
Z0 C
4S u 10 u 10
4S f 0 C
4.
Z0 C
f0
 = 25 . Also, Q 0P =  or
 = 500
a. Q 0P = 
BW
20
3
Q 0P G
25 u 10  = 7.96 u 10 9 F = 7.96 nF
C =  = 5
Z0
2S u 5 u 10
6
1
1
1
L =  =  =  = 12.73 u 10 H = 12.73 PH
2 2
2
10
9
Z0 C
4S f 0 C
4S u 25 u 10 u 7.96 u 10
I 0 = V 0 Y 0 = V 0 G = 20 u 10
A = 20 mA
225
Chapter 2 Resonance
Y
f = f1
1
= G + j Z 1 C 
Z L
1
= 10
+ j 2S u 490 u 10 u 7.96 u 10
1

3
6
2S u 490 u 10 u 12.73 u 10
1

3
6
2S u 510 u 10 u 12.73 u 10
Likewise,
Y
f = f2
1
= G + j Z 1 C 
Z L
1
= 10
+ j 2S u 510 u 10 u 7.96 u 10
226
Solutions to Exercises
5.
L
1 mH
R1
jZL
1:
C
Z IN
1 PF
1
jZC
10 :
R2
a. It is important to remember that the relation Z 0 = 1 e LC applies only to series RLC and
parallel GLC circuits. For any other circuit we must find the input impedance Z IN , set the
imaginary part of Z IN equal to zero, and solve for Z 0 . Thus, for the given circuit
1
1 e j Z C 10 + j Z L
Z IN = R 1 +  __ R 2 + j Z L = 1 + jZC
10 + j Z L 1 e Z C
10 + j Z L 1 e Z C + 10 e j Z C + L e C 10 j Z L 1 e Z C
=  10 + j Z L 1 e Z C
10 j Z L 1 e Z C
100 + j10 Z L 1 e Z C + 100 e j Z C + 10L e C j10 Z L 1 e Z C
= 2
100 + Z L 1 e Z C
2
Z L 1 e Z C 10 e Z C Z L 1 e Z C jL e C Z L 1 e Z C
+ 2
100 + Z L 1 e Z C
2
100 + 10L e C + Z L 1 e Z C 10 e Z C Z L 1 e Z C
= 2
100 + Z L 1 e Z C
100 e j Z C jL e C Z L 1 e Z C
+ 2
100 + Z L 1 e Z C
= 0
2
100
L = 0
 + Z 0 L Z0
Z0C
2
L C Z 0 + 100C L = 0
227
Chapter 2 Resonance
2
1
LC
100
100
10
u 10
10
and thus
Z0 =
9 u 10 = 30 000 r e s
b.
BW = Z 0 e Q = 30 000 e 50 = 600 r e s
c.
Z 1 = Z 0 BW e 2 = 30 000 300 = 29 700 r e s
Z 2 = Z 0 + BW e 2 = 30 000 + 300 = 30 300 r e s
4
VS
1:
V C0
j30 :
10 :
170 0q V
j100 e 3 :
Vc0 = 168.32
6. First, we will find the appropriate value of C 2 . We recall that at parallel resonance the voltage is
maximum and the current is minimum. For this circuit the parallel resonance was found as in
(2.37), that is,
228
Solutions to Exercises
1 R
LC L 2
Z0 =
or
2S u 43 000 =
10 1
 3
6
2 u 10 C 2 4 u 10
4 2
10  + 2S u 4.3 u 10 4 2 = 10
10  = + 2S u 4.3 u 10 u 4 u 10 
6
6
2C 2
4 u 10
4 u 10
6
4 u 10
 = 6.62 u 10 9 F = 6.62 nF
C 2 = 500 4
4 2
6
10 + 2S u 4.3 u 10 u 4 u 10
Next, we must find the value of C 1 that will make the entire circuit series resonant (minimum
impedance, maximum current) at f = 10 KHz . In the circuit below we let z 1 = jX C1 ,
z 2 = jX C2 , z 3 = R 1 + jX L , and z LOAD = 1 .
C2
jX C1
Z IN
R1
100 :
jX C2
jX L
C1
2 mH
V S = 170 0q V
+ RL
v LOAD
1:
Then,
Z IN = z 1 + z 2 __ z 3 + z LOAD
and
Z IN f = 10 KHz = z 1 + z 2 __ z 3
where z 2 __ z 3
f = 10 KHz
f = 10 KHz
+ z LOAD = z 1 + z 2 __ z 3
f = 10 KHz
+ 1 (1)
Zp =
111.12 + 127.72i
and by substitution into (1)
Circuit Analysis II with MATLAB Applications
Orchard Publications
229
Chapter 2 Resonance
Z IN f = 10 KHz = z 1 + 111.12 + j127.72 + 1 = z 1 + 113.12 + j127.72 : (2)
The expression of (2) will be minimum if we let z 1 = j127.72 : at f = 10 KHz . Then, the
capacitor C 1 value must be such that 1 e Z C = 127.72 or
7
1
C 1 =  = 1.25 u 10 F = 0.125 PF
4
2S u 10 u 127.72
Shown below is the plot of V LOAD versus frequency and the MATLAB code that produces this
plot.
f=1000: 100: 60000; w=2*pi*f; Vs=170; C1=1.25*10^(7); C2=6.62*10^(9);...
L=2.*10.^(3);...
R1=100; Rload=1; z1=j./(w.*C1); z2=j./(w.*C2); z3=R1+j.*w.*L; Zload=Rload;...
Zin=z1+z2.*z3./(z2+z3); Vload=Zload.*Vs./(Zin+Zload); magVload=abs(Vload);...
plot(f,magVload); axis([1000 60000 0 2]);...
xlabel('Frequency f'); ylabel('Vload'); grid
This circuit is considered to be a special type of filter that allows a specific frequency (not a band
of frequencies) to pass, and attenuates another specific frequency.
230
Chapter 3
Elementary Signals
his chapter begins with a discussion of elementary signals that may be applied to electric networks. The unit step, unit ramp, and delta functions are then introduced. The sampling and
sifting properties of the delta function are defined and derived. Several examples for expressing
a variety of waveforms in terms of these elementary signals are provided.
vS
We wish to describe v out in a math form for the time interval f t +f . To do this, it is convenient to divide the time interval into two parts, f t 0 , and 0 t f .
For the time interval f t 0 the switch is open and therefore, the output voltage v out is zero. In
other words,
v out = 0 for f t 0
(3.1)
For the time interval 0 t f the switch is closed. Then, the input voltage v S appears at the output,
i.e.,
v out = v S for 0 t f
(3.2)
(3.3)
31
vS
0
The waveform of Figure 3.2 is an example of a discontinuous function. A function is said to be discontinuous if it exhibits points of discontinuity, that is, the function jumps from one value to another
without taking on any intermediate values.
0
u0 t =
1
(3.4)
t!0
In the waveform of Figure 3.3, the unit step function u 0 t changes abruptly from 0 to 1 at t = 0 .
But if it changes at t = t 0 instead, it is denoted as u 0 t t 0 . Its waveform and definition are as
shown in Figure 3.4 and relation (3.5).
1
u0 t t0
t
t0
* In some books, the unit step function is denoted as u t , that is, without the subscript 0. In this text, however, we
will reserve the u t designation for any input.
32
0
u0 t t0 =
1
(3.5)
t ! t0
u0 t + t0
t
t0 0
0
u0 t + t0 =
1
t ! t0
(3.6)
Example 3.1
Consider the network of Figure 3.6, where the switch is closed at time t = T .
R
t = T
vS
Express the output voltage v out as a function of the unit step function, and sketch the appropriate
waveform.
Solution:
For this example, the output voltage v out = 0 for t T , and v out = v S for t ! T . Therefore,
v out = v S u 0 t T
(3.7)
33
v out
Other forms of the unit step function are shown in Figure 3.8.
7
(a)
Au 0 t
Au 0 t + T
(d)
A
A u0 t
(e)
7
7
A u0 t + T
A
7 0
(f)
7
(h)
(c)
A u0 t + T
Au 0 t T
A
0
(g)
A
A u0 t T
A u0 t
(b)
A
A
7
A u0 t T
(i)
A
Unit step functions can be used to represent other timevarying functions such as the rectangular
pulse shown in Figure 3.9.
u0 t
1
0
1
a
0
b
1
0
t
c
u0 t 1
Figure 3.9. A rectangular pulse expressed as the sum of two unit step functions
Thus, the pulse of Figure 3.9(a) is the sum of the unit step functions of Figures 3.9(b) and 3.9(c) is
represented as u 0 t u 0 t 1 .
34
angular, or triangular, or sawtooth, or any other recurring pulse, it can be represented as a sum (difference) of unit step functions.
Example 3.2
Express the square waveform of Figure 3.10 as a sum of unit step functions. The vertical dotted lines
indicate the discontinuities at T 2T 3T and so on.
vt
A
}
T
2T
3T
0
A
Solution:
Line segment { has height A , starts at t = 0 , and terminates at t = T . Then, as in Example 3.1, this
segment is expressed as
v1 t = A > u0 t u0 t T @
(3.8)
Line segment  has height A , starts at t = T and terminates at t = 2T . This segment is expressed
as
v 2 t = A > u 0 t T u 0 t 2T @
(3.9)
Line segment } has height A , starts at t = 2T and terminates at t = 3T . This segment is expressed as
v 3 t = A > u 0 t 2T u 0 t 3T @
(3.10)
(3.11)
Thus, the square waveform of Figure 3.10 can be expressed as the summation of (3.8) through (3.11),
that is,
35
(3.12)
= A > u 0 t u 0 t T @ A > u 0 t T u 0 t 2T @
+A > u 0 t 2T u 0 t 3T @ A > u 0 t 3T u 0 t 4T @
v t = A > u 0 t 2u 0 t T + 2u 0 t 2T 2u 0 t 3T + } @
Example 3.3
Express the symmetric rectangular pulse of Figure 3.11 as a sum of unit step functions.
A
T e 2
it
Te2
Solution:
This pulse has height A , starts at t = T e 2 , and terminates at t = T e 2 . Therefore, with reference to
Figures 3.5 and 3.8 (b), we get
T
T
T
T
i t = Au 0 t +  Au 0 t  = A u 0 t +  u 0 t 
2
2
2
(3.14)
Example 3.4
Express the symmetric triangular waveform of Figure 3.12 as a sum of unit step functions.
1
T e 2
vt
Te2
Solution:
We first derive the equations for the linear segments { and  shown in Figure 3.13.
36
vt
2
 t + 1
T
{
T e 2
Te2
(3.15)
2
T
v 2 t =  t + 1 u 0 t u 0 t 
T
(3.16)
2
T
(3.17)
Example 3.5
Express the waveform of Figure 3.14 as a sum of unit step functions.
v t
3
2
1
t
0
Solution:
As in the previous example, we first find the equations of the linear segments { and  shown in Figure 3.15.
37
2t + 1
t+3

t
or
v t = 2t + 1 u 0 t + > 2t + 1 + 3 @u 0 t 1
+ > 3 + t + 3 @u 0 t 2 t + 3 u 0 t 3
(3.18)
Two other functions of interest are the unit ramp function, and the unit impulse or delta function. We
will introduce them with the examples that follow.
Example 3.6
In the network of Figure 3.16, where i S is a constant source, the switch is closed at time t = 0 .
R
iS
t = 0
C
+
vC t
38
f i
C W dW
(3.19)
iC t = iS u0 t
and assuming that v C t = 0 for t 0 , we can write (3.19) as
i S u 0 W dW =
i
SC
f u0 W dW
1
v C t = C
iS
+ C
0 u 0 W dW
(3.21)
or
iS
v C t =  tu 0 t
C
(3.22)
Therefore, we see that when a capacitor is charged with a constant current, the voltage across it is a
linear function and forms a ramp with slope i S e C as shown in Figure 3.17.
vC t
slope = i S e C
0
Figure 3.17. Voltage across a capacitor when charged with a constant current source.
* Since the initial condition for the capacitor voltage was not specified, we express this integral with f at the
lower limit of integration so that any nonzero value prior to t 0 would be included in the integration.
39
(3.23)
f u0 W dW
1
t
Figure 3.18. Area under the unit step function from fto t
Therefore, we define u 1 t as
0
u1 t =
t
t0
tt0
(3.24)
(3.25)
Higher order functions of t can be generated by repeated integration of the unit step function. For
example, integrating u 0 t twice and multiplying by 2, we define u 2 t as
0
u2 t = 2
t
t0
0
u3 t = 3
t
t0
or
u2 t = 2
or
u3 t = 3
tt0
f u1 W dW
(3.26)
Similarly,
tt0
f u2 W dW
(3.27)
and in general,
310
0
un t = n
t
or
tt0
un t = 3
f un 1 W dW
(3.28)
Also,
1d
u n 1 t =   u n t
n dt
(3.29)
Example 3.7
In the network of Figure 3.19, the switch is closed at time t = 0 and i L t = 0 for t 0 .
R
t = 0
iS
v t
`
L
iL t
(3.30)
iL t = iS u0 t
(3.31)
d
v L t = Li S  u 0 t
dt
(3.32)
311
f G W dW
(3.33)
= u0 t
and
(3.34)
G t = 0 for all t z 0
To better understand the delta function G t , let us represent the unit step u 0 t as shown in Figure
3.20 (a).
Figure (a)
H
1
2H
Area =1
Figure (b)
0
H
The function of Figure 3.20 (a) becomes the unit step as H o 0 . Figure 3.20 (b) is the derivative of
Figure 3.20 (a), where we see that as H o 0 , 1 e 2 H becomes unbounded, but the area of the rectangle
remains 1 . Therefore, in the limit, we can think of G t as approaching a very large spike or impulse
at the origin, with unbounded amplitude, zero width, and area equal to 1 .
Two useful properties of the delta function are the sampling property and the sifting property.
(3.35)
f t G t = f 0 G t
(3.36)
or, when a = 0 ,
312
(3.37)
ft = f0 + >ft f0@
(3.38)
We rewrite f t as
f W G W dW =
f 0 G W dW +
f > f W f 0 @G W dW
(3.39)
The first integral on the right side of (3.39) contains the constant term f 0 ; this can be written outside the integral, that is,
t
f 0 G W dW = f 0
f G W dW
(3.40)
The second integral of the right side of (3.39) is always zero because
G t = 0 for t 0 and t ! 0
and
>fW f 0 @
W=0
= f 0 f0 = 0
f W G W dW = f 0
f G W dW
(3.41)
(3.42)
313
f f t G t D dt
= fD
(3.43)
that is, if we multiply any function f t by G t D and integrate from f to +f , we will obtain the
value of f t evaluated at t = D .
Proof:
Let us consider the integral
b
a f t G t D dt
where a D b
(3.44)
We will use integration by parts to evaluate this integral. We recall from the derivative of products
that
d xy = xdy + ydx or xdy = d xy ydx
(3.45)
x dy
(3.46)
= xy y dx
Now, we let x = f t ; then, dx = f c t . We also let dy = G t D ; then, y = u 0 t D . By substitution into (3.46), we get
b
f t G t D dt = f t u 0 t D
a
a u0 t D f c t dt
(3.47)
We have assumed that a D b ; therefore, u 0 t D = 0 for D a , and thus the first term of the
right side of (3.47) reduces to f b . Also, the integral on the right side is zero for D a , and therefore, we can replace the lower limit of integration a by D . We can now rewrite (3.47) as
b
a
and letting
f t G t D dt = f b
D f c t d t
= fb fb + fD
, we get
f f t G t D dt = f D
(3.48)
Sifting Property of G t
314
G
n
G t =  > u 0 t @
dt
(3.49)
The function G' t is called doublet, G'' t is called triplet, and so on. By a procedure similar to the
derivation of the sampling property of the delta function, we can show that
f t G' t a = f a G' t a f ' a G t a
(3.50)
Also, the derivation of the sifting property of the delta function can be extended to show that
n
n
nd
f t G t D dt = 1 n > f t @
f
dt
(3.51)
t=D
Example 3.8
Evaluate the following expressions:
4
a. 3t G t 1
b.
f tG t 2 dt
2
c. t G' t 3
Solution:
4
3t G t 1 = ^ 3t
4
t=1
`G t 1 = 3G t 1
f f t G t D dt
Then,
f
f tG t 2 dt = f 2 = t t = 2 = 2
c. The given expression contains the doublet; therefore, we use the relation
315
t G' t 3 = t
t=3
Example 3.9
a. Express the voltage waveform v t shown in Figure 3.21 as a sum of unit step functions for the
time interval 1 t 7 s .
b. Using the result of part (a), compute the derivative of v t and sketch its waveform.
vt V
3
2
1
1
t s
1
2
Solution:
a. We first derive the equations for the linear segments of the given waveform. These are shown in
Figure 3.22.
Next, we express v t in terms of the unit step function u 0 t , and we get
v t = 2t > u 0 t + 1 u 0 t 1 @ + 2 > u 0 t 1 u 0 t 2 @
+ t + 5 > u0 t 2 u0 t 4 @ + > u0 t 4 u0 t 5 @
(3.52)
+ t + 6 > u0 t 5 u0 t 7 @
316
vt V
t+5
3
2
t+6
1
t s
1
2
2t
or
v t = 2tu 0 t + 1 + 2t + 2 u 0 t 1 + t + 3 u 0 t 2
+ t 4 u 0 t 4 + t + 5 u 0 t 5 + t 6 u 0 t 7
b. The derivative of v t is
dv
 = 2u 0 t + 1 + 2tG t + 1 2u 0 t 1 + 2t + 2 G t 1
dt
u 0 t 2 + t + 3 G t 2 + u 0 t 4 + t 4 G t 4
(3.53)
u 0 t 5 + t + 5 G t 5 + u 0 t 7 + t 6 G t 7
From the given waveform, we observe that discontinuities occur only at t = 1 , t = 2 , and
t = 7 . Therefore, G t 1 = 0 , G t 4 = 0 , and G t 5 = 0 , and the terms that contain these
delta functions vanish. Also, by application of the sampling property,
2tG t + 1 = ^ 2t
t = 1
`G t + 1 = 2G t + 1
t + 3 G t 2 = ^ t + 3
t 6 G t 7 = ^ t 6
t=2
t=7
`G t 2 = G t 2
`G t 7 = G t 7
317
(3.54)
+ G t 2 + u0 t 4 u0 t 5 + u0 t 7 + G t 7
V e s
1
Gt 7
Gt 2
1
1
t s
1
2G t + 1
Figure 3.23. Plot of the derivative of the waveform of Figure 3.23.
We observe that a negative spike of magnitude 2 occurs at t = 1 , and two positive spikes of
magnitude 1 occur at t = 2 , and t = 7 . These spikes occur because of the discontinuities at
these points.
MATLAB* has builtin functions for the unit step, and the delta functions. These are denoted by the
names of the mathematicians who used them in their work. The unit step function u 0 t is referred
to as Heaviside(t), and the delta function G t is referred to as Dirac(t). Their use is illustrated with
the examples below.
syms k a t;
u=k*sym('Heaviside(ta)')
u =
k*Heaviside(ta)
d=diff(u)
d =
k*Dirac(ta)
318
Summary
int(d)
ans =
Heaviside(ta)*k
3.8 Summary
x The unit step function u 0 t that is defined as
t0
0
u0 t =
1
t!0
x The unit step function offers a convenient method of describing the sudden application of a volt
f u0 W dW
x The unit impulse or delta function, denoted as G t , is the derivative of the unit step u 0 t . It is also
defined as
t
f G W dW
= u0 t
and
G t = 0 for all t z 0
x The sampling property of the delta function states that
f t G t a = f a G t
or, when a = 0 ,
f t G t = f 0 G t
x The sifting property of the delta function states that
f
f f t G t D dt
= fD
319

b. cos 2tG t S

c. cos t G t S
2
2

d. tan 2tG t S
2 t
e. t e G t 2 dt
f

f. sin t G 1 t S
2
2
2.
a. Express the voltage waveform v t shown in Figure 3.24, as a sum of unit step functions for
the time interval 0 t 7 s .
b. Using the result of part (a), compute the derivative of v t , and sketch its waveform.
vt V
vt
20
2t
10
0
1
t s
10
20
320
Solutions to Exercises
3.10 Solutions to Exercises
1. We apply the sampling property of the G t function for all expressions except (e) where we apply
the sifting property. For part (f) we apply the sampling property of the doublet.
We recall that the sampling property states that f t G t a = f a G t a . Thus,
 = sin t
a. sin tG t S
6
t = Se6
b. cos 2tG t S
 = cos 2t
S
S
S
S
G t  = sin  G t  = 0.5G t 
6
6
6 6
t = Se4
S
S
S
G t  = cos  G t  = 0
4
2 4
c. cos t G t S
 =  1 + cos 2t
2
2
2
d. tan 2tG t S
 = tan 2t
t = Se8
1
1
S
S
S
G t  =  1 + cos S G t  =  1 1 G t  = 0
2
2
2
2
2
t = Se2
S
S
S
S
G t  = tan  G t  = G t 
8
8
4 8
2 t
f t e
2 t
G t 2 dt = t e
t=2
= 4e
f f t G t D dt
= f D . Thus,
= 0.54
We recall that the sampling property for the doublet states that
f t G' t a = f a G' t a f ' a G t a
Thus,
2 1
2
S
sin t G t  = sin t
2
t = Se2
d
2
1
S
G t   sin t
2 dt
1
=  1 cos 2t
2
f.
t = Se2
t = Se2
1
S
G t  sin 2t
2
S
G t 
2
t = Se2
S
G t 
2
1
1
1
S
S
S
=  1 + 1 G t  sin SG t  = G t 
2
2
2
2
2.
a.
v t = e
2t
or
321
2t
u0 t e
2t
2t
u0 t + e
2t
+ 10t + 70 u 0 t 7
b.
2t
2t
2t
2t
dv
 = 2e u 0 t + e G t + 2e + 10 u 0 t 2 + e + 10t 30 G t 2
dt
(1)
10u 0 t 7 + 10t + 70 G t 7
2t
+ 10t 30 G t 2 = e
2t
+ 10t 30
20t + 80 G t 3 = 20t + 80
20t 120 G t 5 = 20t 120
t=3
t=5
t=2
G t 2  10G t 2
G t 3 = 20G t 3
G t 5 = 20 G t 5
2t
u 0 t + 2e
2t
u 0 t 2 + 10u 0 t 2 10G t 2
2t
20G t 3
10
10
t s
10 G t 2
20
2e
322
1
2t
20G t 5
Chapter 4
The Laplace Transformation
his chapter begins with an introduction to the Laplace transformation, definitions, and properties of the Laplace transformation. The initial value and final value theorems are also discussed
and proved. It concludes with the derivation of the Laplace transform of common functions
of time, and the Laplace transforms of common waveforms.
f f t e
1
^ F s ` = f t = 2Sj
st
V + jZ
V jZ
(4.1)
dt
st
(4.2)
F s e ds
1
where L ^ f t ` denotes the Laplace transform of the time function f t , L ^ F s ` denotes the
Inverse Laplace transform, and s is a complex variable whose real part is V , and imaginary part Z ,
that is, s = V + jZ .
In most problems, we are concerned with values of time t greater than some reference time, say
t = t 0 = 0 , and since the initial conditions are generally known, the twosided Laplace transform
pair of (4.1) and (4.2) simplifies to the unilateral or onesided Laplace transform defined as
L ^ft`= Fs =
f te
st
dt =
1
^ F s ` = f t = 2Sj
V + jZ
V jZ
0 f t e
st
st
F s e ds
dt
(4.3)
(4.4)
The Laplace Transform of (4.3) has meaning only if the integral converges (reaches a limit), that is, if
f
0 f t e
st
dt f
(4.5)
To determine the conditions that will ensure us that the integral of (4.3) converges, we rewrite (4.5)
Circuit Analysis II with MATLAB Applications
Orchard Publications
41
0 f t e
Vt jZt
(4.6)
dt f
jZt
The term e
in the integral of (4.6) has magnitude of unity, i.e., e
for convergence becomes
f
0 f t e
Vt
jZt
(4.7)
dt f
Fortunately, in most engineering applications the functions f t are of exponential order*. Then, we
can express (4.7) as,
f
f t e
Vt
dt
0 ke
V 0 t Vt
dt
(4.8)
and we see that the integral on the right side of the inequality sign in (4.8), converges if V ! V 0 .
Therefore, we conclude that if f t is of exponential order, L ^ f t ` exists if
(4.9)
Re ^ s ` = V ! V 0
ft Fs
42
V0 t
for all t t 0 .
respectively, and
c 1 c 2 } c n
c1 f1 t + c2 f2 t + } + cn fn t c1 F1 s + c2 F2 s + } + cn Fn s
Proof:
L ^ c1 f1 t + c2 f2 t + } + cn fn t ` =
> c 1 f 1 t + c 2 f 2 t + } + c n f n t @ dt
= c1
f1 t e
st
dt + c 2
f2 t e
st
dt + } + c n
fn t e
st
dt
= c1 F1 s + c2 F2 s + } + cn Fn s
Note 1:
It is desirable to multiply f t by u 0 t to eliminate any unwanted nonzero values of f t for t 0 .
2. Time Shifting Property
The time shifting property states that a right shift in the time domain by a units, corresponds to multiplication by e
as
as
(4.12)
Fs
Proof:
L ^ f t a u 0 t a ` =
0e
st
dt +
a ft ae
st
dt
(4.13)
Now, we let t a = W ; then, t = W + a and dt = dW . With these substitutions, the second integral
on the right side of (4.13) becomes
f
fWe
s W + a
dW = e
as
0 f W e
sW
dW = e
as
Fs
exponential function e where a is an arbitrary positive constant, this multiplication will produce a
shift of the s variable in the complex frequency domain by a units. Thus,
Circuit Analysis II with MATLAB Applications
Orchard Publications
43
at
(4.14)
ft Fs + a
Proof:
L ^e
at
ft` =
at
fte
st
dt =
0 f t e
s + a t
dt = F s + a
Note 2:
A change of scale is represented by multiplication of the time variable t by a positive scaling factor
a . Thus, the function f t after scaling the time axis, becomes f at .
4. Scaling Property
Let a be an arbitrary positive constant; then, the scaling property states that
1 s
f at  F 
a a
(4.15)
Proof:
L ^ f at ` =
0 f at e
st
dt
fW e
s W e a
W
1
d  =  a
a
0 f W e
s e a W
1 s
d W =  F 
a a
Note 3:
Generally, the initial value of f t is taken at t = 0 to include any discontinuity that may be present
at t = 0 . If it is known that no such discontinuity exists at t = 0 , we simply interpret f 0 as f 0 .
5. Differentiation in Time Domain
The differentiation in time domain property states that differentiation in the time domain corresponds
to multiplication by s in the complex frequency domain, minus the initial value of f t at t = 0 .
Thus,
d
f ' t =  f t sF s f 0
dt
(4.16)
Proof:
L ^f 't` =
44
0 f ' t e
st
dt
v du
we let du = f ' t and v = e
st
(4.17)
= uv u dv
. Then, u = f t , dv = se
L ^ f ' t ` = f t e
st f
= lim > e
aof
sa
+s
fte
st
st
, and thus
dt = lim
f t e
aof
st a
+ sF s
f a f 0 @ + sF s = 0 f 0 + sF s
(4.18)
d3
 f t s 3 F s s 2 f 0 sf ' 0 f '' 0
3
dt
(4.19)
and in general
n
d n
n1
n2
ft s Fs s
f0 s
f '0 } f
n
dt
n1
(4.20)
L ^ g ' t ` = sL ^ g t ` g 0
Then,
= s 2 F s sf 0 f ' 0
45
(4.21)
Proof:
L ^ f t` = Fs =
0 f t e
st
dt
Differentiating with respect to s, and applying Leibnitzs rule* for differentiation under the integral, we
get
d
dF s = ds
ds
=
f te
st
0 > tf t @ e
dt =
st
w e st f t dt =
ws
0 t e
st
f t dt
dt = L > tf t @
In general,
n
n
nd
t f t 1 n F s
ds
(4.22)
The proof for n t 2 follows by taking the second and higherorder derivatives of F s with respect
to s .
7. Integration in Time Domain
This property states that integration in time domain corresponds to F s divided by s plus the initial
value of f t at t = 0 , also divided by s . That is,
t
F s  + f0
f W dW s
s
f
(4.23)
a f x D dx
where f is some
known function of integration x and the parameter D , a and b are constants independent of x and D , and the pardFtial derivative wf e wD exists and it is continuous, then =
dD
46
w x D
 dx .
a w D
f W dW =
f W dW +
(4.24)
0 f W dW
The first integral on the right side of (4.24), represents a constant value since neither the upper, nor
the lower limits of integration are functions of time, and this constant is an initial condition denoted
as f 0 . We will find the Laplace transform of this constant, the transform of the second integral
on the right side of (4.24), and will prove (4.23) by the linearity property. Thus,
L ^f 0 ` =
0 f 0 e
st
dt = f 0
0 e
st
st f
e
dt = f 0 s
(4.25)
f0
f0
= f 0 u 0  =  s
s
This is the value of the first integral in (4.24). Next, we will show that
t
Fs
0 f W dW s
We let
t
gt =
0 f W dW
then,
g' t = f W
and
g 0 =
0 f W dW
= 0
Now,
L ^ g' t ` = G s = sL ^ g t ` g 0 = G s 0
sL ^ g t ` = G s
Gs
L ^ g t ` = s
0 f W dW =
F
s s
(4.26)
47
f t

t
(4.27)
s F s ds
Proof:
Fs =
0 f t e
st
dt
F s ds =
s 0 f t e
st
dt ds
F s ds =
0 s
st
ds f t dt
and performing the inner integration on the right side integral with respect to s , we get
f
F s ds =
1
 e
t
st f
s
f t dt =
ft
0 t  e
st
f t
dt = L 
t
9. Time Periodicity
The time periodicity property states that a periodic function of time with period T corresponds to
the integral
0 f t e
st
dt divided by 1 e
sT
be a periodic function with period T , that is, f t = f t + nT , for n = 1 2 3 } we get the transform pair
T
0 f t e
st
dt
f t + nT  sT
1e
48
(4.28)
f te
st
dt =
f te
st
dt +
2T
f te
st
dt +
3T
2T f t e
st
dt + }
In the first integral of the right side, we let t = W , in the second t = W + T , in the third t = W + 2T ,
and so on. The areas under each period of f t are equal, and thus the upper and lower limits of
integration are the same for each integral. Then,
L ^ft` =
fWe
sW
dW +
f W + T e
s W + T
dW +
0 f W + 2T e
s W + 2T
dW + }
(4.29)
sT
+e
2sT
+ }
0 f W e
sW
dW
(4.30)
(4.31)
sW
f W e dW
L ^ f W ` =  sT
We
lim f t = lim sF s = f 0
to0
(4.32)
sof
Proof:
From the time domain differentiation property,
d
 f t sF s f 0
dt
or
49
L  f t = sF s f 0 =
dt
d
 f t e st dt
dt
sof
sof
 f t e
T o f H dt
lim
st
dt
Ho0
 f t
T o f H dt
sof
lim e
sof
st
dt
Ho0
and since
lim e
sof
st
= 0
lim > sF s f 0 @ = 0
sof
or
lim sF s = f 0
sof
tof
so0
(4.33)
Proof:
From the time domain differentiation property,
d
 f t sF s f 0
dt
or
d
L  f t = sF s f 0 =
dt
d
 f t e st dt
dt
410
so0
 f t e
T o f H dt
lim
so0
st
dt
Ho0
 ft
dt
Tof H
so0
lim e
st
so0
dt
Ho0
Also, since
lim e
st
= 1
so0
so0
Tof H
Ho0
d
 f t dt = lim
dt
Tof
Ho0
H f t
= lim > f T f H @ = f f f 0
Tof
Ho0
and therefore,
lim sF s = f f
so0
f 1 t *f 2 t F 1 s F 2 s
Proof:
L ^ f 1 t *f 2 t ` = L
=
f 1 W f 2 t W dW =
f
0 f1 W 0 f2 t W e
st
0 0 f1 W f2 t W dW
st
dt
(4.35)
dt dW
Convolution is the process of overlapping two signals. The convolution of two time functions f 1 t and f 2 t is
denoted as f 1 t *f 2 t , and by definition, f 1 t *f 2 t =
f f1 W f2 t W dW
discussed in detail Signals and Systems with MATLAB Applications by this author.
411
f1 W
f2 O e
s O + W
dO dW =
f 1 W e
sW
dW
0 f2 O e
sO
dO
= F 1 s F 2 s
(4.36)
Proof:
L ^ f 1 t f 2 t ` =
0 f1 t f2 t e
st
(4.37)
dt
V + jZ
V jZ
Pt
F 1 P e dP
1
= 2Sj
1
2Sj
V + jZ
V jZ
Pt
F 1 P e dP f 2 t e
V + jZ
V jZ
F1 P
0 f2 t e
s P t
st
dt
dt dP
V + jZ
V jZ F1 P F2 s P dP
1
=  F 1 s *F 2 s
2Sj
For easy reference, we have summarized the Laplace transform pairs and theorems in Table 4.1.
412
Linearity
Time Domain
c1 f1 t + c2 f2 t
c1 F1 s + c2 F2 s
+ } + cn fn t
+ } + cn Fn s
Time Shifting
Frequency Shifting
Time Scaling
Time Differentiation
d
 f t
See also (4.18) through (4.20) dt
sF s f 0
Frequency Differentiation
See also (4.22)
d
 F s
ds
Time Integration
Frequency Integration
f t
t
Time Periodicity
f t + nT
as
f t a u 0 t a
as
Fs + a
ft
f at
tf t
t
f f W dW
Fs
1 s
 F a a
Fs f 0
 + s
s
f
s F s ds
T
0 f t e
st
dt
 sT
1e
10
lim f t
to0
lim sF s = f 0
sof
11
lim f t
tof
lim sF s = f f
so0
12
Time Convolution
f 1 t *f 2 t
F 1 s F 2 s
13
Frequency Convolution
f 1 t f 2 t
1
 F 1 s *F 2 s
2Sj
413
0 f t e
st
dt
1e
st
st f
e
dt = s
1
1
= 0  =  s
s
(4.38)
for Re ^ s ` = V ! 0 .*
Example 4.2
Find L ^ u 1 t `
Solution:
We apply the definition
L ^ f t` = F s =
0 f t e
st
dt
0 t e
st
dt
u dv
(4.39)
= uv v du
We let
u = t and dv = e
st
then,
st
e
du = 1 and v = s
414
t e
L ^ t ` = 
s 0
st
st
st
t e  e
e  dt = 2
s
s
s
(4.40)
0
Since the upper limit of integration in (4.40) produces an indeterminate form, we apply L Hpitals
rule*, that is,
lim te
tof
st
d
t
t
d
t
1 = 0
= lim  = lim  = lim st
st
tof e
tof d
t o f se st
e
dt
(4.41)
for V ! 0 .
Example 4.3
n
Find L ^ t u 0 t `
Solution:
To find the Laplace transform of this function, we must first review the gamma or generalized factorial function * n defined as
*n =
0 x
n 1 x
(4.42)
e dx
f x
Often, the ratio of two functions, such as  , for some value of x, say a, results in an indeterminate form. To
gx
fx
work around this problem, we consider the limit lim , and we wish to find this limit, if it exists. To find this
xoa
gx
d
d
limit, we use LHpitals rule which states that if f a = g a = 0 , and if the limit  f x e  g x as x
dx
dx
dd
f x = lim f x e  g x
approaches a exists, then, lim dx
x o a gx
x o a dx
415
The integral of (4.42) can be evaluated by performing integration by parts. Thus, in (4.39) we let
u = e
and dv = x
n1
Then,
n
x
xdu = e dx and v = n
x e
* n = n
1
+ n
x=0
n x
0 x e
dx
(4.43)
With the condition that n ! 0 , the first term on the right side of (4.43) vanishes at the lower limit
x = 0 . It also vanishes at the upper limit as x o f . This can be proved with L Hpitals rule by differentiating both numerator and denominator m times, where m t n . Then,
d
n x
m1
m1
nx
n1
x e  = lim dx
x  = lim x
lim  = lim d
 = }
m1
x
n
x o f ne x
xof d m
xof
d
x
ne
ne
m
m1
dx
dx
xof
nm
n 1 n 2 } n m + 1
n n 1 n 2 } n m + 1 x
= lim  = lim  = 0
x
mn x
xof
xof
ne
x
e
n x
0 x e
dx
a.
a f x dx
b.
a f x dx
416
where f(x) becomes infinite at a value x between the lower and upper limits of integration inclusive.
1
x
e dx = n
n x
0 x e
(4.44)
dx
(4.45)
n* n = * n + 1
(4.46)
or
It is convenient to use (4.45) for n 0 , and (4.46) for n ! 0 . From (4.45), we see that * n becomes
infinite as n o 0 .
For n = 1 , (4.42) yields
*1 =
e dx = e
x f
0
(4.47)
= 1
*1 = 1
(4.49)
* 3 = 2 * 2 = 2 1 = 2!
* 4 = 3 * 3 = 3 2 = 3!
and in general
(4.50)
* n + 1 = n!
for n = 1 2 3 }
The formula of (4.50) is a noteworthy relation; it establishes the relationship between the * n
function and the factorial n!
n
We now return to the problem of finding the Laplace transform pair for t u 0 t , that is,
n
L ^ t u0 t ` =
0 t
n st
dt
(4.51)
To make this integral resemble the integral of the gamma function, we let st = y , or t = y e s , and
417
L ^ t u0 t ` =
n
1
y e y d y = n+1
s
s
s
n y
0 y e
* n + 1  = n! dy = n+1
n+1
s
s
(4.52)
0 G t e
st
dt
0 G t e
st
dt = e
s 0
= 1
Gt 1
for all V .
Example 4.5
Find L ^ G t a `
Solution:
L ^Gt a` =
0 G t a e
st
dt
and again, using the sifting property of the delta function, we get
L ^Gt a` =
418
0 G t a e
st
dt = e
as
as
(4.54)
for V ! 0 .
Example 4.6
Find L ^ e
at
u0 t `
Solution:
L ^e
at
u0 t ` =
at st
dt =
0 e
1  e s + a t
=  s + a
s + a t
dt
1= s+a
at
1
u 0 t s+a
(4.55)
for V ! a .
Example 4.7
n at
Find L t e
u0 t
Solution:
For this example, we will use the transform pair of (4.52), i.e.,
n! n
t u 0 t n+1
s
(4.56)
at
f t F s + a
(4.57)
419
t e
n!
u 0 t n+1
s + a
(4.58)
where n is a positive integer, and V ! a Thus, for n = 1 , we get the transform pair
te
at
1
u 0 t 2
s + a
(4.59)
for V ! a .
For n = 2 , we get the transform
2 at
t e
2!
u 0 t 3
s + a
(4.60)
n!
u 0 t n+1
s + a
(4.61)
and in general,
n at
t e
for V ! a
Example 4.8
Find L ^ sin Zt u 0 t `
Solution:
L ^ sin Zt u 0 t ` =
sin Zt e
st
dt = lim
aof 0
sin Zt e
st
dt
ax
a sin bx b cos bx 
e sin bx dx = e2
2
a +b
Then,
1 jZt jZt
1
at
This can also be derived from sin Zt =  e e
, and the use of (4.55) where e u 0 t  . By the linj2
s+a
earity property, the sum of these terms corresponds to the sum of their Laplace transforms. Therefore,
1
L > sin Ztu 0 t @ = j2
420
1  1
Z
=  s jZ s + jZ
2
s + Z2
e s sin Zt Z cos Zt
L ^ sin Zt u 0 t ` = lim 2
2
aof
s +Z
= lim
aof
as
(4.62)
for V ! 0
Example 4.9
Find L ^ cos Z t u 0 t `
Solution:
L ^ cos Z t u 0 t ` =
cos Zt e
st
dt = lim
cos Zt e
a o f 0
st
dt
ax
e acos bx + b sin bx
e cos bx dx = 2
2
a +b
Then,
st
e s cos Zt + Z sin Zt
L ^ cos Z t u 0 t ` = lim 2
2
aof
s +Z
= lim
aof
as
1 jZt
jZt
We can use the relation cos Zt =  e + e
and the linearity property, as in the derivation of the transform
2
d
of sin Z t on the footnote of the previous page. We can also use the transform pair  f t sF s f 0 ; this
dt
is the time differentiation property of (4.16). Applying this transform pair for this derivation, we get
1d
1
d
1
Z
s
L > cos Z tu 0 t @ = L   sin Z tu 0 t =  L  sin Z tu 0 t =  s = 2
Z dt
Z dt
Z s2 + Z2
s + Z2
421
(4.63)
for V ! 0
Example 4.10
Find L ^ e
at
sin Zt u 0 t `
Solution:
Since
Z
sin Ztu 0 t 2
2
s +Z
ft Fs + a
(4.64)
Z
sin Zt u 0 t 2
2
s + a + Z
(4.65)
at
for V ! 0 and a ! 0 .
Example 4.11
Find L ^ e
at
cos Z t u 0 t `
Solution:
Since
s
cos Z t u 0 t 2
2
s +Z
using the frequency shifting property of (4.14), we replace s with s + a , and we get
e
at
s+a
cos Z t u 0 t 2
2
s + a + Z
(4.66)
for V ! 0 and a ! 0 .
For easy reference, we have summarized the above derivations in Table 4.2.
422
Fs
u0 t
1es
t u0 t
1es
n!
t u0 t
n+1
s
Gt
Gt a
6
7
at
u0 t
n at
t e
u0 t
as
1
s+a
n!
n+1
s + a
sin Zt u 0 t
Z
2
2
s +Z
cos Z t u 0 t
s 2
2
s +Z
10
11
at
at
sin Zt u 0 t
Z
2
2
s + a + Z
cos Z t u 0 t
s+a
2
2
s + a + Z
423
t
a
Solution:
We first express the given waveform as a sum of unit step functions. Then,
(4.67)
fP t = A > u0 t u0 t a @
as
Fs
and
Au 0 t a e
as A
s
Then, by the linearity property, the Laplace transform of the pulse of Figure 4.1 is
A as A A
as
A > u 0 t u 0 t a @  e  =  1 e
s
s s
Example 4.13
Find the Laplace transform for the waveform f L t of Figure 4.4. The subscript L stands for line.
fL t
1
t
0
424
fL t
1
t
1
Figure 4.3. Waveform for Example 4.13 with the equation of the linear segment
as
Fs
(4.68)
Example 4.14
Find the Laplace transform for the triangular waveform f T t of Figure 4.4.
Solution:
We must first derive the equations of the linear segments. These are shown in Figure 4.5. Then, we
express the given waveform in terms of the unit step function.
fT t
1
0
425
fT t
t+2
1 t
0
Figure 4.5. Waveform for Example 4.13 with the equations of the linear segments
as
Fs
Then,
1 2s 1
1 2e s + e 2tu 0 t 2 t 1 u 0 t 1 + t 2 u 0 t 2 2
2
s
s
s
or
1
s
2s
tu 0 t 2 t 1 u 0 t 1 + t 2 u 0 t 2 2 1 2e + e
s
(4.69)
Example 4.15
Find the Laplace transform for the rectangular periodic waveform f R t of Figure 4.6.
426
fR t
A
t
a
2a
3a
A
Figure 4.6. Waveform for Example 4.15
Solution:
This is a periodic waveform with period T = 2a , and thus we can apply the time periodicity property
T
0 f W e
sW
dW
L ^ f W ` =  sT
1e
2a
fR t e
a
st
1 dt =  2as
1e
st
st
A
e
e

= +
2as
s 0
s
1e
Ae
st
dt +
2a
A e
st
dt
2a
or
A
 e as + 1 + e 2as e as
L ^ f R t ` =  2as
s1 e
as
A
A1 e
 1 2e as + e 2as = =  2as
as
as
s1 e
s1 + e 1 e
2
A 1 e as A e as e 2 e as e 2 e as e 2 e as e 2
 =  
=   s 1 + e as s e as e 2 e as e 2 + e as e 2 e as e 2
as e 2 as e 2
as e 2
e
e
A sinh as e 2
Ae
  =  =  s e as e 2 e as e 2 + e as e 2 s cosh as e 2
or
A
as
f R t  tanh 
s
2
(4.70)
427
f HW t
sint
3S
2S
4S
Solution:
This is a periodic waveform with period T = 2S . We will apply the time periodicity property
T
0 f W e
sW
dW
L ^ f W ` =  sT
1e
2S
f t e
st
1
dt =  2Ss
1e
0 sin t e
st
s sin t cos t 
1  e=  2Ss
2
s +1
1e
st
dt
Ss
1  1 + e = 2
2Ss
s + 1 1 e
Ss
1 + e
1
L ^ f HW t ` = 2
Ss
Ss
s + 1 1 + e 1 e
or
1
f HW t 2
Ss
s + 11 e
428
(4.71)
Summary
4.5 Summary
x The twosided or bilateral Laplace Transform pair is defined as
L ^ft `= Fs =
f f t e
1
^ F s ` = f t = 2Sj
st
dt
V + jZ
V jZ
st
F s e ds
1
fte
st
dt =
0 f t e
st
dt
x We denote transformation from the time domain to the complex frequency domain, and vice
versa, as
ft Fs
x The linearity property states that
c1 f1 t + c2 f2 t + } + cn fn t c1 F1 s + c2 F2 s + } + cn Fn s
as
Fs
at
ft Fs + a
429
and in general
n
d
 f t s n F s s n 1 f 0 s n 2 f ' 0 } f
n
dt
n1
where the terms f 0 f ' 0 f '' 0 , and so on, represent the initial conditions.
x The differentiation in complex frequency domain property states that
d
tf t  F s
ds
and in general,
n
d
t f t 1 n F s
ds
n
Fs f 0
f W dW  + s
s
f
s F s ds
t exists.
provided that the limit lim fto0
0 f t e
st
dt
f t + nT  sT
1e
x The initial value theorem states that
lim f t = lim sF s = f 0
to0
430
sof
Summary
x The final value theorem states that
lim f t = lim sF s = f f
tof
so0
x Convolution in the time domain corresponds to multiplication in the complex frequency domain,
that is,
f 1 t *f 2 t F 1 s F 2 s
x Convolution in the complex frequency domain divided by 1 e 2Sj , corresponds to multiplication
The Laplace transforms of some common functions of time are shown below.
u0 t 1 e s
t 1 e s2
n!
n
t u 0 t n+1
s
Gt 1
Gt a e
e
te
at
at
n at
t e
1
u 0 t s+a
1
u 0 t 2
s + a
2 at
t e
as
2!
u 0 t 3
s + a
n!
u 0 t n+1
s + a
Z
sin Zt u 0 t 2
2
s +Z
s
cos Z t u 0 t 2
2
s +Z
431
Z
sin Zt u 0 t 2
2
s + a + Z
at
s+a
cos Z t u 0 t 2
2
s + a + Z
t
a
A as A A
as
A > u 0 t u 0 t a @  e  =  1 e
s
s s
fL t
1
t
0
s 1
t 1 u 0 t 1 e 2s
fT t
1
0
1
s 2
f T t 2 1 e
s
fR t
t
2a
3a
A
A
as
f R t  tanh 
2
s
432
Summary
f HW t
sint
2S
3S
4S
1
f HW t 2
Ss
s + 1 1 e
433
e. 4t u 0 t
2. Find the Laplace transform of the following time domain functions:
a. j8
b. j5 90q
c. 5e
5t
u0 t
7 5t
d. 8t e
u0 t
e. 15G t 4
3. Find the Laplace transform of the following time domain functions:
3
a. t + 3t + 4t + 3 u 0 t
b. 3 2t 3 G t 3
c. 3 sin 5t u 0 t
d. 5 cos 3t u 0 t
e. 2 tan 4t u 0 t Be careful with this! Comment and skip derivation.
4. Find the Laplace transform of the following time domain functions:
a. 3t sin 5t u 0 t
2
b. 2t cos 3t u 0 t
c. 2e
434
5t
sin 5t
Exercises
d. 8e
3t
cos 4t
e. cos t G t S e 4
5. Find the Laplace transform of the following time domain functions:
a. 5tu 0 t 3
2
b. 2t 5t + 4 u 0 t 3
c. t 3 e
2t
d. 2t 4 e
e. 4te
3t
u0 t 2
2t 2
u0 t 3
cos 2t u 0 t
4t
b. d 3e
dt
2
c. d t cos 2t
dt
2t
d. d e sin 2t
dt
2 2t
e. d t e
dt
b.
sin W
 dW
0 W
sin at
c. t
d.
cos W
 dW
W
435
f W
edW
W
8. Find the Laplace transform for the sawtooth waveform f ST t of Figure 4.8.
f ST t
A
3a
2a
9. Find the Laplace transform for the full rectification waveform f FR t of Figure 4.9.
f FR t
1
2a
3a
4a
436
Solutions to Exercises
4.7 Solutions to Exercises
1. From the definition of the Laplace transform or from Table 4.2 we get:
a. 12 e s b. 6 e s c. e
12s
5!
24
2
 d. 5 e s e. 4 6s
s
2. From the definition of the Laplace transform or from Table 4.2 we get:
7!
5
4s
a. j8 e s b. 5 e s c.  d. 8 8 e. 15e
s+5
s + 5
3.
3 u 2! 4 3
a. From Table 4.2 and the linearity property 3!
 +  +  + 4
3
2
s
b. 3 2t 3 G t 3 = 3 2t 3
t=3
G t 3 = 9G t 3 and 9G t 3 9e
2
3s
4 e s + 2
s
sin 4t
5
 d. 5  e. 2 tan 4t = 2  2  = 8
 . This answer looks susc. 3 2
2
2
2
2
2
cos 4t
s +3
s e s + 2 s
picious because 8 e s 8u 0 t and the Laplace transform is unilateral, that is, there is onetos +5
one correspondence between the time domain and the complex frequency domain. The fallacy
with this procedure is that if we assume that f 1 t F 1 s and f 2 t F 2 s , we cannot conf t
f2 t
F s
F2 s
1
1
1
 . For this exercise f 1 t f 2 t = sin 4t  and as weve learned
clude that cos 4t
0 2 tan 4t e
st
dt and this
requires integration by parts. We skip this analytical derivation. The interested reader may try
to find the answer with the MATLAB code syms s t; 2*laplace(sin(4*t)/cos(4*t))
4. From (4.22)
n
n
nd
t f t 1 n F s
ds
Then,
a.
5 
5 2s
1 d
30s
= 3 2 = 3 1  2
2
ds s 2 + 5 2
2
s + 25
s + 25
437
s + 9 2s 2 s + 9 2s s + 9 
= 2 4
2
s + 9
2
s + 9 2s 4s s + 9 
2s 18s + 4s 36s= 2 = 2 3
3
2
2
s + 9
s + 9
3
2s 54s
2s s 27 4s s 2 27
 = = 2 3 = 2 3
3
2
2
2
s + 9
s + 9
s + 9
c.
10
2u5
= 2
2
2
s + 5 + 25
s + 5 + 5
d.
8s + 3
8s + 3
 = 2
2
2
s + 3 + 4
s + 3 + 16
e.
cos t
Se4
G t S e 4 = 2 e 2 G t S e 4 and 2 e 2 G t S e 4 2 e 2 e
S e 4 s
5.
a.
5tu 0 t 3 = > 5 t 3 + 15 @u 0 t 3 e
3s
5 3s
5 + 15
 + 3
 =  e 1
s
2
s
s
s
b.
2
2t 5t + 4 u 0 t 3 = > 2 t 3 + 12t 18 5t + 4 @u 0 t 3
2
= > 2 t 3 + 7t 14 @u 0 t 3
2
= > 2 t 3 + 7 t 3 + 21 14 @u 0 t 3
2
= > 2 t 3 + 7 t 3 + 7 @u 0 t 3 e
u 2! 7 7
 +  + 
3
2
s
s
s
3s 2
c.
t 3 e
2t
u 0 t 2 = > t 2 1 @e
e
438
2s
2 t 2
e u0 t 2
4
2s s + 1
1
1
  = e e
2
2 s + 2
s + 2
s + 2
Solutions to Exercises
d.
2t 4 e
2t 2
u 0 t 3 = > 2 t 3 + 6 4 @e
e
3s
2 t 3
e u0 t 3
s+4
2
3s
2
2
 +  = 2e e
2
2
s
+
3
s + 3
s + 3
e.
4te
3t
d
s+3
s+3
1 d
 = 4   cos 2t u 0 t 4 1  2
2
2
ds s + 6s + 9 + 4
ds s + 3 + 2
2
d
s+3
s + 6s + 13 s + 3 2s + 6
 = 4  4  2
2
2
ds s + 6s + 13
s + 6s + 13
2
2
s + 6s + 13 2s 6s 6s 18
4 s + 6s + 5
 =  4 2
2
2
2
s + 6s + 13
s + 6s + 13
6.
a.
d
 f t sF s f 0
dt
3
sin 3t 2
2
s +3
f 0 = sin 3t
= 0
t=0
3
3s
d
 0 =  sin 3t s 2
2
2
dt
s +3
s +9
b.
3e
4t
d
 f t sF s f 0
dt
3
s+4
f 0 = 3e
4t
t=0
= 3
3
4t
3s
3 s + 4 12
d
3e s  3 =   = s+4
s+4
s+4
s+4
dt
c.
2
s
cos 2t 2
2
s +2
s
2
2 d
t cos 2t 1 2 2
ds s + 4
2
2
2
s + 4s + 4 s 2s 
d d  s + 4 2s s + 4 s + 4 2 2s 
=
=
2
2
4
2
ds 2
ds
2
s + 4
s + 4
s + 4
2
s + 4 2s s + 4 4s 2s 8s + 4s 16s 2s s 12
=  =  = 3
3
3
2
2
2
s + 4
s + 4
s + 4
Thus,
2
2s s 12
2
t cos 2t 3
2
s + 4
Circuit Analysis II with MATLAB Applications
Orchard Publications
439
2 2
2s s 12 
2s s 12
0
s =
3
3
2
2
s + 4
s + 4
d.
2
sin 2 t 2
2
s +2
2t
d
 f t sF s f 0
dt
2
sin 2t 2
s + 2 + 4
2
2s
d e 2t sin 2t s  0 = 2
2
dt
s + 2 + 4
s + 2 + 4
e.
2!
2
t 3s
2 2t
t e
2!
3
s + 2
d
f t sF s f 0
dt
2!
d 2 2t
2s
t e s 3 0 = 3
dt
s + 2
s + 2
7.
a.
sin t
1 sin t exists. Since
sin t but to find L  we must show that the limit lim 2
t
t
t
o
0
s +1
sin t
sin x = 1 this condition is satisfied and thus 
lim t
xo0 x
1
dx
x2
2
+a
1
1
=  tan x e a + C . Then,
a
 ds = tan
s2
+1
1
 ds . From tables of integrals
2
s +1
b.
1
sin t
From (a) above  tan 1 e s and since
sin W
 dW  tan
0 s
W
440
f 0 , it follows that
s  + f W dW F
s
s
f
1 e s
Solutions to Exercises
c.
1 s
sin t
1
From (a) above  tan 1 e s and since f at  F  , it follows that
a
1
sin at
1
1es
1
sin at
  tan  or  tan a e s
a
at
a
t
d.
f
s cos t
s  ds , and from tables of integrals
,
cos t 2
2
t
s s +1
s +1
s
1
1
x
 dx =  ln x 2 + a 2 + C . Then,  ds =  ln s 2 + 1 + C and the constant of inte2
2
2
2
2
s +1
x +a
get
F s f 0
f W dW  +  we
s
s
f
cos W
1
 dW  ln s 2 + 1
W
2s
e.
t
e
1
t
e  , 
s+1
t
1
 dx
ax + b
 ds
s s+1
1
1
=  ln ax + b . Then,  ds = ln s + 1 + C and the constant of integration C
s+1
2
tof
get
so0
so0
and using
Fs f 0
f W dW  +  we
s
s
f
f W
e
1
 dW  ln s + 1
s
W
8.
f ST t
A
A
 t
a
2a
3a
441
A
A
 te st dt =  as
a
0
a1 e
0 te
st
dt (1)
a
0
0 te
st
st
st
te
e
dt =  2
s
s
st
st
te
e
=  + 2
s
s
as
as
1
e  = 1 ae
 > 1 1 + as e as @
 = 2
2
2
s
s
s
s
a
0
1
1
as
as
= 2 > 1 + as 1 + as e as @ = 2 > 1 + as 1 e as @
s
s
9.
This is a periodic waveform with period T = a = S and its Laplace transform is
1
F s =  sT
1e
f t e
st
1
dt =  Ss
1 e
0 sin te
st
dt
ax
e asin bx b cos bx
sin bxe dx = 2
2
a +b
Then,
st
1
e s sin t cos t
 F s =  Ss
2
1e
s +1
Ss
1
1+e
 =  Ss
2
1e
s +1
Ss
1
1+e
1
Ss
  =  coth 
= 2
Ss
2
2
s +1 1e
s +1
442
Chapter 5
The Inverse Laplace Transformation
his chapter is a continuation to the Laplace transformation topic of the previous chapter and
presents several methods of finding the Inverse Laplace Transformation. The partial fraction
expansion method is explained thoroughly and it is illustrated with several examples.
1
^ F s ` = f t = 2Sj
V + jZ
V jZ
st
F s e ds
(5.1)
This integral is difficult to evaluate because it requires contour integration using complex variables
theory. Fortunately, for most engineering problems we can refer to Tables of Properties, and Common Laplace transform pairs to lookup the Inverse Laplace transform.
(5.2)
m1
m2
+ bm 2 s
+ } + b1 s + b0
bm s + bm 1 s
Ns
F s =  = n
n
1
n
2
Ds
an s + an 1 s
+ an 2 s
+ } + a1 s + a0
(5.3)
The coefficients a k and b k are real numbers for k = 1 2 } n , and if the highest power m of N s
is less than the highest power n of D s , i.e., m n , F s is said to be expressed as a proper rational
function. If m t n , F s is an improper rational function.
In a proper rational function, the roots of N s in (5.3) are found by setting N s = 0 ; these are
called the zeros of F s . The roots of D s , found by setting D s = 0 , are called the poles of F s .
We assume that F s in (5.3) is a proper rational function. Then, it is customary and very convenient
Circuit Analysis II with MATLAB Applications
Orchard Publications
51
(5.4)
The zeros and poles of (5.4) can be real and distinct, or repeated, or complex conjugates, or combinations of real and complex conjugates. However, we are mostly interested in the nature of the poles, so
we will consider each case separately.
Case I: Distinct Poles
If all the poles p 1 p 2 p 3 } p n of F s are distinct (different from each another), we can factor the
denominator of F s in the form
Ns
F s =  s p1 s p2 s p3 } s pn
(5.5)
where p k is distinct from all other poles. Next, using the partial fraction expansion method,*we can
express (5.5) as
rn
r2
r3
r1
 +  +  + } + F s =  s p1 s p2 s p3
s pn
(5.6)
s = pk
(5.7)
Example 5.1
Use the partial fraction expansion method to simplify F 1 s of (5.8) below, and find the time domain
function f 1 t corresponding to F 1 s .
3s + 2
F 1 s = 2
s + 3s + 2
(5.8)
* The partial fraction expansion method applies only to proper rational functions. It is used extensively in integration, and in finding the inverses of the Laplace transform, the Fourier transform, and the ztransform. This
method allows us to decompose a rational polynomial into smaller rational polynomials with simpler denominators from which we can easily recognize their integrals and inverse transformations. This method is also being
taught in intermediate algebra and introductory calculus courses.
52
(5.9)
= 1
(5.10)
= 4
(5.11)
s = 1
and
3s + 2
r 2 = lim s + 2 F s = s + 1
s o 2
s = 2
(5.12)
at
1
u 0 t s+a
(5.13)
Then,
4  e t + 4e 2t u t = f t
1  + F 1 s = 0
1
s + 1 s + 2
(5.14)
The residues and poles of a rational function of polynomials such as (5.8), can be found easily using
the MATLAB residue(a,b) function. For this example, we use the code
Ns = [3, 2]; Ds = [1, 3, 2]; [r, p, k] = residue(Ns, Ds)
53
3s + 2s + 5 F 2 s = 2
3
s + 12s + 44s + 48
(5.15)
Solution:
First, we use the MATLAB factor(s) symbolic function to express the denominator polynomial of
F 2 s in factored form. For this example,
syms s; factor(s^3 + 12*s^2 + 44*s + 48)
ans =
(s+2)*(s+4)*(s+6)
Then,
2
2
r1
r2
r3
3s + 2s + 5  = 3s + 2s + 5  =  +  + F 2 s = 2
3
s + 2 s + 4 s + 6
s + 2 s + 4 s + 6
s + 12s + 44s + 48
(5.16)
54
3s + 2s + 5
r 1 = s + 4s + 6
s = 4
3s + 2s + 5
r 3 = s + 2s + 4
(5.17)
37
= 4
(5.18)
89
= 8
(5.19)
s = 2
3s + 2s + 5
r 2 = s + 2s + 6
9
= 8
s = 6
9e8
37 e 4 89 e 8
3s + 2s + 5
F 2 s =  =  +  + 2
3
s + 2 s + 4 s + 6
s + 12s + 44s + 48
(5.20)
at
1
u 0 t s+a
(5.21)
Then,
4t 89 6t
89 e 8 9
37 e 4 + 9 e 8  +  e 2t 37
F 2 s =  e +  e u 0 t = f 2 t
8
8
s + 2 s + 4 s + 6
4
(5.22)
ft =
37/4*exp(4*t)+9/8*exp(2*t)+89/8*exp(6*t)
Case II: Complex Poles
Quite often, the poles of F s are complex*, and since complex poles occur in complex conjugate
pairs, the number of complex poles is even. Thus, if p k is a complex root of D s , then, its complex
conjugate pole, denoted as p k
, is also a root of D s . The partial fraction expansion method can
also be used in this case, but it may be necessary to manipulate the terms of the expansion in order to
express them in a recognizable form. The procedure is illustrated with the following example.
Example 5.3
Use the partial fraction expansion method to simplify F 3 s of (5.23) below, and find the time
* A review of complex numbers is presented in Appendix B of Circuit Analysis I with MATLAB Applications.
55
(5.23)
Solution:
Let us first express the denominator in factored form to identify the poles of F 3 s using the MATLAB factor(s) function. Then,
syms s; factor(s^3 + 5*s^2 + 12*s + 8)
ans =
(s+1)*(s^2+4*s+8)
The factor(s) function did not factor the quadratic term. We will use the roots(p) function.
p=[1 4 8]; roots_p=roots(p)
roots_p =
2.0000 + 2.0000i
2.0000  2.0000i
Then,
s+3
s+3
F 3 s =  = 2
3
s + 1 s + 2 + j2 s + 2 j2
s + 5s + 12s + 8
or
r 2
r1
r2
s+3
 + + F 3 s =  = 2
3
s + 1 s + 2 + j2 s + 2 j 2
s + 5s + 12s + 8
(5.24)
s = 2 j2
s = 1
2
= 5
1 j2
1 j2
= =  8 + j4
1 j2 j4
1 j2 8 j4
j316 + j12 = 1
=   =  +  8 + j4 8 j4
5 20
80
1 j3
1 j3
r 2
=  +  =   5 20
5 20
(5.25)
(5.26)
(5.27)
56
(5.28)
The last two terms on the right side of (5.28), do not resemble any Laplace transform pair that we
derived in Chapter 2. Therefore, we will express them in a different form. We combine them into a
single term*, and now (5.28) is written as
2e5
1
2s + 1
F 3 s =    s + 1 5 s 2 + 4s + 8
(5.29)
For convenience, we denote the first term on the right side of (5.29) as F 31 s , and the second as
F 32 s . Then,
2 t
2e5
F 31 s =   e = f 31 t
5
s + 1
(5.30)
2s + 1
1
F 32 s =  2
5 s + 4s + 8
(5.31)
Next, for F 32 s
Z
sin Ztu 0 t 2
2
s + a + Z
at
s+a
cos Ztu 0 t 2
2
s + a + Z
e
e
(5.32)
we express F 32 s as
1 3 3
s +  +  
2 2 2
2 3 e 2
s+2
 = 2
F 32 s =   + 
2
2
2
2
5 s + 2 + 2
5 s + 2 + 2 s + 2 2 + 22
2
s + 2  6 e 10 
 = 2
+
2 s + 2 2 + 22
5 s + 2 2 + 22
(5.33)
s + 2  2
3
2

=  +  2
2
10 s + 2 2 + 2 2
5 s + 2 + 2
* Here, we used MATLAB with simple((1/5 +3j/20)/(s+2+2j)+(1/5 3j/20)/(s+22j)). The simple function, after several simplification tools that were displayed on the screen, returned (2*s1)/
(5*s^2+20*s+40)
57
s + 1 5 s + 2 + 2
10 s + 2 2 + 2 2
3 2t
2 t 2 2t
 e  e cos 2t +  e sin 2t = f 3 t
10
5
5
ft =
2/5*exp(t)2/5*exp(2*t)*cos(2*t)
+3/10*exp(2*t)*sin(2*t)
Case III: Multiple (Repeated) Poles
In this case, F s has simple poles, but one of the poles, say p 1 , has a multiplicity m . For this condition, we express it as
Ns
F s = m
s p 1 s p 2 } s p n 1 s p n
(5.34)
Denoting the m residues corresponding to multiple pole p 1 as r 11 r 12 r 13 } r 1m , the partial fraction expansion of (5.34) is written as
r 11
r 12
r 13
r 1m
F s =  +  +  + } + m
m1
m2
s p1
s p1
s p1
s p1
(5.35)
rn
r2
r3
 +  + } + +  s p2 s p3
s pn
For the simple poles p 1 p 2 } p n , we proceed as before, that is, we find the residues as
r k = lim s p k F s = s p k F s
s o pk
(5.36)
s = pk
The residues r 11 r 12 r 13 } r 1m corresponding to the repeated poles, are found by multiplication of
m
s p 1 F s = r 11 + s p 1 r 12 + s p 1 r 13 + } + s p 1
m1
rn
r3
r2
m
+ s p 1  +  + } +  s p2 s p3
s p n
58
r 1m
(5.37)
s o p1
m1
s o p1
+ lim
s o p1
r 1m @
rn
r3
r2
m
 +  + } +  s p 1  s p2 s p3
s p n
or
m
(5.38)
r 11 = lim s p 1 F s
s o p1
and thus (5.38) yields the residue of the first repeated pole.
The residue r 12 for the second repeated pole p 1 , is found by differentiating (5.37) with respect to s
and again, we let s o p 1 , that is,
d > s p m F s @
r 12 = lim 1
s o p 1 ds
(5.39)
s p 1 F s = r 11 + r 12 s p 1 + r 13 s p 1 + }
(5.40)
k 1 !r 1k
d
m
1 = lim  k 1 > s p 1 F s @
s o p 1 k 1 ! ds
(5.41)
or
k1
d m
1  r 1k = lim > s p1 F s @
k
1
s o p 1 k 1 ! ds
(5.42)
Example 5.4
Use the partial fraction expansion method to simplify F 4 s of (5.43) below, and find the time
domain function f 4 t corresponding to F 4 s .
s+3
F 4 s = 2
s + 2s + 1
(5.43)
Solution:
We observe that there is a pole of multiplicity 2 at s = 1 , and thus in partial fraction expansion
form, F 4 s is written as
Circuit Analysis II with MATLAB Applications
Orchard Publications
59
(5.44)
s = 1
= 1
s = 2
= 2
s = 1
s + 2 s + 3
= 2
s + 2
= 1
s = 1
The value of the residue r 22 can also be found without differentiation as follows:
Substitution of the already known values of r 1 and r 21 into (5.44), and letting s = 0 *, we get
s+3
2
s + 1 s + 2
s=0
1
= s + 2
2
+ 2
s = 0 s + 1
s=0
r 22
+ s + 1
s=0
or
3
1
 =  + 2 + r 22
2
2
(5.45)
We can use the following code to check the partial fraction expansion.
syms s
Ns = [1 3];
expand((s + 1)^2);
d1 = [1 2 1];
d2 = [0 1 2];
510
[r,p,k]=residue(Ns,Ds)
r =
1.0000
1.0000
2.0000
p =
2.0000
1.0000
1.0000
k =
[]
Example 5.5
Use the partial fraction expansion method to simplify F 5 s of (5.46) below, and find the time
domain function f 5 t corresponding to the given F 5 s .
2
s + 3s + 1
F 5 s = 3
2
s + 1 s + 2
(5.46)
Solution:
We observe that there is a pole of multiplicity 3 at s = 1 , and a pole of multiplicity 2 at s = 2 .
Then, in partial fraction expansion form, F 5 s is written as
r 21
r 11
r 12
r 13
r 22
+ +  + + F 5 s = 3
2
2
s + 1 s + 2
s + 2
s + 1
s + 1
(5.47)
s + 3s + 1
r 11 = 2
s + 2
2
d s + 3 s + 1
r 12 =  
ds s + 2 2
= 1
s = 1
s = 1
s + 2 2s + 3 2 s + 2 s + 3 s + 1 
= 4
s + 2
s = 1
s+4
= 3
s + 2
=3
s = 1
511
s = 1
2
1 d d s + 3 s + 1

=    2 ds ds s + 2 2
3
1d s+4
=   3
2 ds s + 2
s = 1
1 s + 2 3s 12
=  4
2
s + 2
s = 1
1 s + 2 3 s + 2 s + 4 
=  6
2
s + 2
s = 1
s5
= 4
s + 2
s = 1
= 4
s = 1
s + 3s + 1
r 21 = 3
s + 1
= 1
s = 2
and
2
d s + 3 s + 1

r 22 =  ds s + 1 3
s = 2
s + 1 2s + 3 3 s + 1 s + 3 s + 1
= 6
s + 1
s + 1 2s + 3 3 s + 3 s + 1
= 4
s + 1
s = 2
s 4s
= 4
s + 1
s = 2
=4
s = 2
(5.48)
We will check the values of these residues with the MATLAB code below.
syms s;
Ds =
s^5+7*s^4+19*s^3+25*s^2+16*s+4
Ns=[1 3 1]; Ds=[1 7 19 25 16 4]; [r,p,k]=residue(Ns,Ds)
r =
4.0000
1.0000
4.0000
3.0000
1.0000
512
Case for m t n
p =
2.0000
2.0000
1.0000
1.0000
1.0000
k =
[]
From Table 2.2 of Chapter 2
e
at
1
s+a
te
at
1
2
s + a
n 1 at
n 1 !
n
s + a
(5.49)
ft = 1/2*t^2*exp(t)+3*t*exp(t)4*exp(t)
+t*exp(2*t)+4*exp(2*t)
F s = k0 + k1 s + k2 s + } + km n s
mn
Ns
+ Ds
(5.50)
s + 2s + 2
F 6 s = s+1
(5.51)
513
1 +1+s
s + 2s + 2 = F 6 s = s+1
s+1
and
1 Gt
but
s?
and
u 0'' t = G' t
where G' t is the doublet of the delta function. Also, by the time differentiation property
2
2
2 1
u 0'' t = G' t s F s sf 0 f ' 0 = s F s = s  = s
s
(5.52)
1  + 1 + s e t + G t + G' t = f t
s + 2s + 2 = F 6 s = 6
s+1
s+1
(5.53)
and thus,
2
In general,
n
d
 G t s n
n
dt
(5.54)
r =
1
514
Here, the direct terms k= [1 1] are the coefficients of G t and G' t respectively.
(5.55)
n
2. We repeat step 1 for each of the distinct linear and quadratic factors of D s
3. We set the given F s equal to the sum of these partial fractions
4. We clear the resulting expression of fractions and arrange the terms in decreasing powers of s
5. We equate the coefficients of corresponding powers of s
6. We solve the resulting equations for the residues
Example 5.7
Express F 7 s of (5.56) below as a sum of partial fractions using the method of clearing the fractions.
2s + 4
F 7 s = 2
2
s + 1s 1
(5.56)
515
(5.57)
With Step 4,
2
(5.58)
2s + 4 = r 1 s + A s 1 + r 21 s + 1 + r 22 s 1 s + 1
2s + 4 = r 1 + r 22 s + 2r 1 + A r 22 + r 21 s
(5.59)
+ r 1 2A + r 22 s + A r 22 + r 21
Relation (5.59) will be an identity is s if each power of s is the same on both sides of this relation.
Therefore, we equate like powers of s and we get
0 = r 1 + r 22
0 = 2r 1 + A r 22 + r 21
(5.60)
2 = r 1 2A + r 22
4 = A r 22 + r 21
or
(5.61)
r1 = 2
or
(5.62)
r 22 = 2
Next, substitution of (5.61) and (5.62) into the third equation of (5.60) yields
2 = 2 2A 2
or
(5.63)
A = 1
Finally by substitution of (5.61), (5.62), and (5.63) into the fourth equation of (5.60), we get
516
or
(5.64)
r 21 = 1
(5.65)
Example 5.8
Use partial fraction expansion to simplify F 8 s of (5.66) below, and find the time domain function
f 8 t corresponding to F 8 s .
s+3
F 8 s = 2
3
s + 5s + 12s + 8
(5.66)
Solution:
This is the same transform as in Example 5.3, where we found that the denominator D s can be
expressed in factored form of a linear term and a quadratic. Thus, we write F 8 s as
s+3
F 8 s = 2
s + 1 s + 4s + 8
(5.67)
(5.68)
As in Example 5.3,
s+3 r 1 = 2
s + 4s + 8
s = 1
= 2
5
(5.69)
Next, to compute r 2 and r 3 , we follow the procedure of this section and we get
2
s + 3 = r 1 s + 4s + 8 + r 2 s + r 3 s + 1
(5.70)
Since r 1 is already known, we only need two equations in r 2 and r 3 . Equating the coefficient of s 2
on the left side, which is zero, with the coefficients of s 2 on the right side of (5.70), we get
0 = r1 + r2
(5.71)
517
(5.72)
as before.
The remaining steps are the same as in Example 5.3, and thus f 8 t is the same as f 3 t , that is,
2 t 2 2t
3 2t
f 8 t = f 3 t =  e  e cos 2t +  e sin 2t u 0 t
5
10
5
5.5 Summary
x The Inverse Laplace Transform Integral defined as
L
1
^ F s ` = f t = 2Sj
V + jZ
V jZ
st
F s e ds
is difficult to evaluate because it requires contour integration using complex variables theory.
x For most engineering problems we can refer to Tables of Properties, and Common Laplace trans
forms in a recognizable form from which we can obtain the equivalent timedomain functions.
x If the highest power m of the numerator N s is less than the highest power n of the denominator D s , i.e., m n , F s is said to be expressed as a proper rational function. If m t n , F s is
F s = k0 + k1 s + k2 s + } + km n s
mn
Ns
+ Ds
x In a proper rational function, the roots of numerator N s are called the zeros of F s and the
roots of the denominator D s are called the poles of F s .
518
Summary
x The partial fraction expansion method can be applied whether the poles of F s are distinct, com
and in general,
n
d
 G t s n
n
dt
x The method of clearing the fractions is an alternate method of partial fraction expansion.
519
s+3
4
b. 2
s + 3
4
c. 4
s + 3
3s + 4
d. 5
s + 3
2
s + 6s + 3e. 5
s + 3
4s + 5
b. 2
s + 5s + 18.5
2
s + 3s + 2
c. 3
2
s + 5s + 10.5s + 9
2
s 16
d. 3
2
s + 8s + 24s + 32
s+1
e. 3
2
s + 6s + 11s + 6
5s + 3
b. 2 (See hint on next page)
2
s + 4
520
Exercises
2
1
s
 sin Dt + Dt cos Dt 
2
2
2
2D
s + D
Hint:
1
1

3 sin Dt Dt cos Dt 2
2 2
s + D
2D
2s + 3 c. 2
s + 4.25s + 1
3
s + 8s + 24s + 32
d. 2
s + 6s + 8
e. e
2s
3 3
2s + 3
4. Use the Initial Value Theorem to find f 0 given that the Laplace transform of f t is
2s + 3
2
s + 4.25s + 1
521
d.
1
1
s + 3 5 e 3
3s + 4 e 3 + 5 e 3 5 e 3
3s + 4
 = 3  5  =  = 3 5
4
5
5
5
s + 3
s + 3
s + 3
s + 3
s + 3
3 3 3t 5 4 3t 1 3 3t 5 4 3t
 t e  t e =  t e  t e
4!
12
3!
2
e.
2
2
2
1
s + 3 6
1
s+ 6s + 3 = s+ 6s + 9 6 = =  6 5
5
5
5
5
3
s + 3
s + 3
s + 3
s + 3
s + 3 s + 3
1 2 3t 6 4 3t 1 2 3t 1 4 3t
 t e  t e =  t e  t e
2
4!
2
2!
2.
a.
s + 2
1
2u9
s + 2 2 e 3
3s + 4
3s + 4 e 3 + 2 e 3 2 e 3
 = 3    =  = 3 2
2
2
2 9
2
2
2
2
s + 2 + 9
s + 2 + 9
s + 2 + 9
s + 4s + 85
s + 2 + 81
s + 2
2
9
2
   3e 2t cos 9t  e 2t sin 9t
= 3 2
2 9
2
2
9
s + 2 + 9
s + 2 + 9
b.
s+5e4
4s + 5
4s + 5
4s + 5  = = 4  = 2
2
2
2
2
2
s + 2.5 + 3.5
s + 5s + 6.25 + 12.25
s + 2.5 + 3.5
s + 5s + 18.5
s + 10 e 4 10 e 4 + 5 e 4
s + 2.5
1
5 u 3.5
 = 4   = 4 2
2
2
2 3.5
2
2
s + 2.5 + 3.5
s + 2.5 + 3.5
s + 2.5 + 3.5
10
3.5
s + 2.5
10 2.5t
2.5t
 = 4  4e
cos 3.5t  e
sin 3.5t
2
2
2
2
7
7
s + 2.5 + 3.5
s + 2.5 + 3.5
ans = (s+2)*(s+1)
ans = 1/2*(s+2)*(2*s^2+6*s+9)
522
Solutions to Exercises
Then,
2
s + 3s + 2
s + 1s + 2
s + 1
s+1
 =  =  = 3
2
2
2
2
s + 5s + 10.5s + 9
s + 2 s + 3s + 4.5
s + 3s + 4.5
s + 3s + 2.25 2.25 + 4.5
1
0.5 u 1.5
s + 1.5 1.5 + 1
s + 1.5
=  =   2
2
2
2
2
2 1.5
s
+
1.5
+
1.5
s + 1.5 + 1.5
s + 1.5 + 1.5
1
1.5
1 1.5t
1.5t
s + 1.5
=   e
cos 1.5t  e
sin 1.5t
2
2
2
2 3
3
s + 2.5 + 3.5
s + 1.5 + 1.5
d.
2
s 16
s + 4 s 4
s 4
s+224
 =  =  = 3
2
2
2
2
2
2
s + 8s + 24s + 32
s + 4 s + 4s + 8
s + 2 + 2
s + 2 + 2
6u2
1
s+2
=   2
2
2
2 2
s + 2 + 2
s + 2 + 2
2
s+2
 e 2t cos 2t 3e 2t sin 2t
3 = 2
2
2
2
s + 2 + 2
s + 2 + 2
e.
1
s + 1
s+1
 =  = 3
2
s
+
2
s + 3
s
+
1
s
+
2
s
+
3
s + 6s + 11s + 6
r2
r1
1
 =  + = s + 2 s + 3 s + 2 s + 3
1r 1 = s+3
=1
s = 2
1r 2 = s+2
= 1
s = 3
2t
3t
1
1
1
=  =   e e
s + 2 s + 3
s+2 s+3
3.
a.
s
2
5
2
1 2u5
3s + 2
3s
 = 3  +   3 cos 5t +  sin 5t
 =  +  2
2
2
2
2
2
2
2
2
5 s +5
5
5 s +5
s +5
s + 25
s +5
b.
2
2
1
1
5s + 3
5s
3
 =  +  5  sin 2t + 2t cos 2t + 3  sin 2t 2t cos 2t
2
2u8
2u2
2
2
2 2
2
2 2
s + 4
s + 2
s + 2
17
23
5 3
5 3
 +  sin 2t +   2t cos 2t =  sin 2t +  t cos 2t
4 16
4 16
8
16
523
5  = 4
= 3
15 e 4
s = 4
+ 3r 2 = 2s
s+4
s = 1 e 4
5 e 2 = 2
= 15 e 4 3
2
4t
t e 4
4e3
2e3
 +   2e + e
3
s+4 s+1e4
d.
3
s + 8s + 24s + 32 s + 4 s + 4s + 8 s + 4s + 8
 =  =  and by long division
2
s + 2 s + 4
s + 2
s + 6s + 8
2
2t
4
s + 4s + 8
 = s + 2 +  G' t + 2G t + 4e
s+2
s+2
e.
e
2s
3
3
2s + 3
2s
F s f t 2 u 0 t 2
3
3e8
3 2 3 e 2 t
3 1 2 3 e 2 t
3
3e2
3e8
F s =  =  =  = 3   t e
=  t e
3
3
3
3
16
8 2!
s + 3 e 2
2s + 3 e 2
2s + 3
> 2s + 3 e 2 @
2s
Fs= e
2s
3
3
  t 2 2 e 3 e 2 t 2 u 0 t 2
3
16
2s + 3
sof
2s + 3
2s + 3s
 = lim f 0 = lim s s o f s 2 + 4.25s + 1
s o f s 2 + 4.25s + 1
2
2+3es
2s e s + 3s e s
= lim  = lim =2
2
2
2
2
s o f s e s + 4.25s e s + 1 e s
s o f 1 + 4.25 e s + 1 e s 2
The value f 0 = 2 is the same as in the time domain expression that we found in Exercise 3(c).
As 1
5. We are given that F s =  and lim f t = lim sF s = 10 . Then,
ss + 1
tof
so0
As 1
s 1  = A = 10 . Therefore,
lim s  = A lim s o 0 ss + 1
s o 0 s + 1
r2
r1
20  10 20e t u t , that is,
10 s 1  =   +  = 10
F s = 0
s s+1
ss + 1
s s+1
t
524
Chapter 6
Circuit Analysis with Laplace Transforms
his chapter presents applications of the Laplace transform. Several examples are given to illustrate how the Laplace transformation is applied to circuit analysis. Complex impedance, complex admittance, and transfer functions are also defined.
iR t
vR t
v R t = Ri R t
vR t
i R t = R
V R s = RI R s
IR s
VR s
VR s
I R s = R
Figure 6.1. Resistive circuit in time domain and complex frequency domain
b. Inductor
The time and complex frequency domains for purely inductive circuits is shown in Figure 6.2.
Time Domain
di L
v L t = L dt
+
vL t
iL t
1
i L t = L
f vL dt
V L s = sLI L s Li L 0
+
sL
VL s
`
IL s
VL s iL 0
+ I L s = Ls
s
Li L 0
Figure 6.2. Inductive circuit in time domain and complex frequency domain
c. Capacitor
The time and complex frequency domains for purely capacitive circuits is shown in Figure 6.3.
61
Time Domain
+
vC t
C + iC t
dv C
i C t = C dt
1
v C t = C
1sC
IC s
VC s
f iC dt
I C s = sCV C s Cv C 0
+
+
IC s vC 0
 + V C s = sC
s
vC 0
s
Figure 6.3. Capacitive circuit in time domain and complex frequency domain
Note:
In the complex frequency domain, the terms sL and 1 e sC are called complex inductive impedance,
and complex capacitive impedance respectively. Likewise, the terms and sC and 1 e sL are called complex capacitive admittance and complex inductive admittance respectively.
Example 6.1
Use the Laplace transform method to find the voltage v C t across the capacitor for the circuit of
Figure 6.4, given that v C 0 = 6 V .
R
vS
:
+
1F
12u 0 t V
vC t
Solution:
We apply KCL at node A as shown in Figure 6.5.
R
vS
iR
:
A
C
+ iC
1F
12u 0 t V
vC t
Then,
iR + iC = 0
or
62
or
dv
C + v C t = 12u 0 t
dt
(6.1)
or
12
s + 1 V C s =  + 6
s
or
6s + 12V C s = ss + 1
= 12
s=0
= 6
s = 1
Therefore,
t
t
12
6
V C s =   12 6e = 12 6e u 0 t = v C t
s s+1
Example 6.2
Use the Laplace transform method to find the current i C t through the capacitor for the circuit of
Figure 6.6, given that v C 0 = 6 V .
vS
+
12u 0 t V
:
+
1F
iC t
vC t
63
:
vS
iC t
+
1F
12u 0 t V
vC t
= 12u 0 t
f iC t dt
f iC t dt
(6.2)
= 12u 0 t
IC s vC 0
12
 +  = I C s + s
s
s
12 6
6
1 + 1
 I C s =   = 
s
s s
s
s+1
 I s = 6
 s C
s
or
t
6
I C s =  i C t = 6e u 0 t
s+1
v C t = 12 6e u 0 t
Then,
dv C
dv
t
t
d
i C t = C  = C = 12 6e u 0 t = 6e u 0 t + 6G t
dt
dt
dt
(6.3)
The presence of the delta function in (6.3) is a result of the unit step that is applied at t = 0 .
64
R1
S1
2:
2A
L 1 0.5 H
1F
S2
t = 0
is t
i L1 t
0.5 H
R2
v 0 = 3 V
C
1:
v
`
L
out t
Solution:
Since the circuit contains a capacitor and an inductor, we must consider two initial conditions One
is given as v C 0 = 3 V . The other initial condition is obtained by observing that there is an initial
current of 2 A in inductor L 1 ; this is provided by the 2 A current source just before switch S 2
opens. Therefore, our second initial condition is i L1 0 = 2 A .
1/s
0.5s
1
1V
1
For t ! 0 , we transform the circuit of Figure 6.8 into its sdomain* equivalent shown in Figure 6.9.
3/s
0.5s
`V
out s
Figure 6.9. Transformed circuit of Example 6.3
In Figure 6.9 the current in L 1 has been replaced by a voltage source of 1 V . This is found from the
relation
1
L 1 i L1 0 =  u 2 = 1 V
2
(6.4)
* Henceforth, for convenience, we will refer the time domain as tdomain and the complex frequency domain as sdomain
65
(6.5)
2s s + 3
V out s = 3
2
s + 8s + 10s + 4
(6.6)
We will use MATLAB to factor the denominator D s of (6.6) into a linear and a quadratic factor.
p=[1 8 10 4]; r=roots(p)
r =
6.5708
0.7146 + 0.3132i
0.7146  0.3132i
y=expand((s + 0.7146 0.3132j)*(s + 0.7146 + 0.3132j))% Find quadratic form
y =
s^2+3573/2500*s+3043737/5000000
3573/2500
% Find coefficient of s
ans =
1.4292
3043737/5000000
ans =
0.6087
Therefore,
2s s + 3
2s s + 3  = V out s = 3
2
2
s + 6.57 s + 1.43s + 0.61
s + 8s + 10s + 4
(6.7)
66
(6.8)
= 1.36
(6.9)
s = 6.57
(6.10)
(6.11)
or
s + 0.715 0.91 1.36
V out s =  + 0.64 2
2
s + 6.57
s + 0.715 + 0.316
1.36
0.64 s + 0.715
0.58
=  +  s + 6.57 s + 0.715 2 + 0.316 2 s + 0.715 2 + 0.316 2
(6.12)
1.36
0.64 s + 0.715
1.84 u 0.316
=  +  2
2
s + 6.57 s + 0.715 + 0.316 s + 0.715 2 + 0.316 2
0.64s 0.12 * We perform these steps to express the term in a form that resembles the transform pairs
2
s + 1.43s + 0.61
s+a Z
at
at
 . The remaining steps are carried out in
e cos Ztu 0 t and e sin Ztu 0 t 2
2
2
2
s + a + Z
s + a + Z
(6.12).
67
6.57t
+ 0.64e
0.715t
0.715t
sin 0.316t u 0 t
(6.13)
sL
Is
VS s
1
sC
+
V out s
VS s
I s = R + sL + 1 e sC
(6.14)
(6.15)
(6.16)
1
Z s = R + sL + sC
(6.17)
where
68
VS s
1es
Figure 6.11. Circuit for Example 6.4
Solution:
First Method:
We will first find I s , and we will compute Z s from (6.15). We assign the voltage V A s at node
A as shown in Figure 6.12.
1 VA s
A
I s
VS s
1es
Figure 6.12. Circuit for finding I s in Example 6.4
By nodal analysis,
VA s VS s VA s VA s
 +  +  = 0
1
s
s+1es
1
1 + 1
 +  V A s = V S s
s s+1es
3
s +1
 VS s
V A s = 3
2
s + 2s + s + 1
VS s VA s
2s + 1
s +1
 VS s
 = 1  V S s = I s = 3
2
3
2
1
s + 2s + s + 1
s + 2s + s + 1
and thus,
3
2
VS s
s + 2s + s + 1
 = Z s = 2
Is
2s + 1
(6.18)
69
Z1
Zs
1es
s
Z2
Z3
`Z
Figure 6.13. Computation of the impedance of Example 6.4 by series parallel combinations
To find the equivalent impedance Z s , looking to the right of terminals a and b , we start on the
right side of the network and we proceed to the left combining impedances as we combine resistances. Then,
Z s = > Z 3 + Z 4  Z 2 @ + Z 1
2
ss + 1 e s
s +s
s + 2s + s + 1
s +1
Z s =  + 1 =  + 1 =  + 1 = 2
2
2
s+s+1es
2s + 1
2s + 1
2s + 1 e s
(6.19)
1
sL
sC
sL
610
(6.20)
(6.21)
1
Y s = G +  + sC
sL
(6.22)
where
Compute Z s and Y s for the circuit of Figure 6.15. All values are in : (ohms). Verify your
answers with MATLAB.
13s
10
8es
20
Zs
Ys
` 5s
16 e s
Solution:
It is convenient to represent the given circuit as shown in Figure 6.16.
Z1
Z s Y s
Z2
Z3
611
8
13s + 8
Z 1 = 13s +  = s
s
Z 2 = 10 + 5s
16
4 5s + 4
Z 3 = 20 +  = s
s
Then,
4 5s + 4
10 + 5s 
Z2 Z3
s
13s + 8
Z s = Z1 + =  + s
5s + 4 Z2 + Z3
10 + 5s + 4
s
2
4 5s + 4
10 + 5s 
13s 2 + 8 20 5s 2 + 14s + 8
s
13s + 8
=  +  =  + 2
2
s
s
5s + 10s + 4 5s + 4
5s + 30s + 16
s
2
z =
13*s+8/s+(5*s+10)*(20+16/s)/(5*s+30+16/s)
z10 = simplify(z)
z10 =
(65*s^4+490*s^3+528*s^2+400*s+128)/s/(5*s^2+30*s+16)
pretty(z10)
4
3
2
65 s + 490 s + 528 s + 400 s + 128
2
s (5 s + 30 s + 16)
The complex input admittance Y s is found by taking the reciprocal of Z s , that is,
2
1
s 5s + 30s + 16
Y s =  = 4
3
2
Zs
65s + 490s + 528s + 400s + 128
612
(6.23)
Transfer Functions
6.4 Transfer Functions
In an s domain circuit, the ratio of the output voltage V out s to the input voltage V in s under
zero state conditions, is of great interest in network analysis. This ratio is referred to as the voltage
transfer function and it is denoted as G v s , that is,
V out s
G v s { V in s
(6.24)
Similarly, the ratio of the output current I out s to the input current I in s under zero state conditions, is called the current transfer function denoted as G i s , that is,
I out s
G i s { I in s
(6.25)
The current transfer function of (6.25) is rarely used; therefore, from now on, the transfer function
will have the meaning of the voltage transfer function, i.e.,
V out s
G s { V in s
(6.26)
Example 6.6
Derive an expression for the transfer function G s for the circuit of Figure 6.17, where R g represents the internal resistance of the applied (source) voltage V S , and R L represents the resistance of
the load that consists of R L , L , and C .
+
RL
Rg
L
vg
v out
C
Solution:
No initial conditions are given, and even if they were, we would disregard them since the transfer
function was defined as the ratio of the output voltage V out s to the input voltage V in s under
Circuit Analysis II with MATLAB Applications
Orchard Publications
613
`V
1
sC
out s
The transfer function G s is readily found by application of the voltage division expression of the
s domain circuit of Figure 6.18, i.e.,
R L + sL + 1 e sC
V out s =  V in s
R g + R L + sL + 1 e sC
Then,
R L + Ls + 1 e sC
V out s
 = G s = V in s
R g + R L + Ls + 1 e sC
(6.27)
Example 6.7
Compute the transfer function G s for the circuit of Figure 6.19 in terms of the circuit constants
R 1 R 2 R 3 C 1 and C 2 Then, replace the complex variable s with jZ , and the circuit constants with
their numerical values and plot the magnitude G s = V out s e V in s versus radian frequency Z .
R2
vin
40 K
C2 10 nF
R1
R3
200 K
50K
C1
vout
25 nF
Solution:
The s domain equivalent circuit is shown in Figure 6.20.
614
Transfer Functions
R1
Vin (s)
1/sC2
R2
R3
V1 s
2
V2 s
1/sC1
Vout (s)
(6.28)
V out s
V2 s V1 s
 = 1 e sC 2
R3
(6.29)
At node 2,
(6.30)
and by substitution of (6.30) into (6.28), rearranging, and collecting like terms, we get:
1
1
1
1
1
 +  +  + sC 1 sR 3 C 2  V s =  V in s
R1 R2 R3
R1
R 2 out
or
V out s
1
G s =  = V in s
1
1
1
1
R1
 +  +  + sC 1 sR 3 C 2 +  R1 R2 R3
R2
(6.31)
By substitution of s with jZ and the given values for resistors and capacitors, we get
1
G j Z = 8
5
4
8
1
1
 + j 2.5 u 10 Z j 5 u 10 u 10 Z + 4
2 u 10
20 u 10 3
4 u 10
or
615
(6.32)
We use MATLAB to plot the magnitude of (6.32) on a semilog scale with the following code:
w=1:10:10000; Gs=1./(2.5.*10.^(6).*w.^25.*j.*10.^(3).*w+5);
semilogx(w,abs(Gs)); grid; hold on
xlabel('Radian Frequency w'); ylabel('Vout/Vin');
title('Magnitude Vout/Vin vs. Radian Frequency')
The plot is shown in Figure 6.21. We observe that the given op amp circuit is a second order lowpass filter whose cutoff frequency ( 3 dB ) occurs at about 700 r e s .
616
Summary
6.5 Summary
x The Laplace transformation provides a convenient method of analyzing electric circuits since
integrodifferential equations in the t domain are transformed to algebraic equations in the s 
domain.
x In the s domain the terms sL and 1 e sC are called complex inductive impedance, and complex
capacitive impedance respectively. Likewise, the terms and sC and 1 e sL are called complex
is a complex quantity, and it is referred to as the complex input impedance of an s domain RLC
series circuit.
x In the s domain the current I s can be found from
VS s
I s = Zs
x The expression
1 + sC
Y s = G + sL
is a complex quantity, and it is referred to as the complex input admittance of an s domain GLC
parallel circuit.
x In the s domain the voltage V s can be found from
IS s
V s = Ys
x In an s domain circuit, the ratio of the output voltage V out s to the input voltage V in s under
zero state conditions is referred to as the voltage transfer function and it is denoted as G s , that
is,
V out s
G s { V in s
617
R1
10 :
L
20 :
1 mH
i t `
R2
32 V
2. In the circuit of Figure 6.23, switch S has been closed for a long time, and opens at t = 0 . Use
the Laplace transform method to compute v c t for t ! 0 .
t = 0
R1
6 K:
R2
R3
R4
30 K:
20 K:
C
60 K:
v t
C
40
 PF
10 K:
9
72 V
R5
3. Use mesh analysis and the Laplace transform method, to compute i 1 t and i 2 t for the circuit
of Figure 6.24, given that i L (0 = 0 and v C (0 = 0 .
L1
R2
2H
3:
R1
+
v1 t = u0 t
i1 t
1:
1F
L2
i2 t
`1 H
+
v 2 t = 2u 0 t
618
Exercises
4. For the s domain circuit of Figure 6.25,
a. compute the admittance Y s = I 1 s e V 1 s
b. compute the t domain value of i 1 t when v 1 t = u 0 t , and all initial conditions are zero.
VC s
3:
1es
R2
1:
R3
1:
V1 s
R1
I1 s
R4
V 2 s = 2V C s
2:
5. Derive the transfer functions for the networks (a) and (b) of Figure 6.26.
+
C
V in s
(a)
V out s
V in s
+
R
(b)
V out s
6. Derive the transfer functions for the networks (a) and (b) of Figure 6.27.
R
+
R
V in s
(a)
V out s
V in s
+
L
(b)
`V
out s
7. Derive the transfer functions for the networks (a) and (b) of Figure 6.28.
619
V in s
V out s
(a)
V in s
(b)
`V
out s
8. Derive the transfer function for the networks (a) and (b) of Figure 6.29.
R2
V in s
R1
R2
R1
V in s
V out s
V out s
(a)
(b)
Figure 6.29. Networks for Exercise 8
9. Derive the transfer function for the network of Figure 6.30. Using MATLAB, plot G s versus
frequency in Hertz, on a semilog scale.
R4
R1 = 11.3 k:
R2 = 22.6 k:
R1
V in s
R3
R3=R4 = 68.1 k:
R2
C1=C2 = 0.01 PF
V out s
C1
C2
620
Solutions to Exercises
6.7 Solutions to Exercises
1. At t = 0
inductor.
the t domain circuit is as shown below and the 20 : resistor is shorted out by the
S
10 :
L
20 :
1 mH
i t `
32 V
Then,
iL t
t=0
32
=  = 3.2 A
10
i L 0 = 3.2 A
20 :
10 s
3
Li L 0 = 3.2 u 10 V
20000t
3.2 u 10
3.2
u0 t = iL t
I L s =  =  3.2e
3
20 + 10 s s + 20000
2. At t = 0
6 K:
20 K:
i2 t
60 K:
30 K:
vC t
10 K:
72 V
Then,
Circuit Analysis II with MATLAB Applications
Orchard Publications
621
and
1
i 2 0 =  i T 0 = 1 mA
2
20 K:
1
6 40 e 9 u 10 s
60 K:
VR = VC s
9
u 10
40s
VC s
10 K:
30 e s
22.5 K:
VR
30 e s
60 K: + 30 K: __ 20 K: + 10 K: = 22.5 K:
3
3
22.5 u 10
30
30 u 22.5 u 10
  = V C s = V R = 6
3
6
3
9 u 10 e 40s + 22.5 u 10 s
9 u 10 e 40 + 22.5 u 10 s
3
30 u 22.5 u 10 e 22.5 u 10
30
30
=  =  = 6
3
6
4
10
+s
9 u 10 e 40 u 22.5 u 10 + s
9 u 10 e 90 u 10 + s
Then,
10t
30
V C s =  30e u 0 t V = v C t
s + 10
z1
2s
`s
1
+
1es
622
I1 s
1es +
z2
I2 s
2es
Solutions to Exercises
Then,
z 1 + z 2 I 1 s z 2 I 2 s = 1 e s
z 2 I 1 s + z 2 + z 3 I 2 s = 2 e s
z2
z2
z2 + z3
I1 s
I2 s
1es
2 e s
Is1 =
2
2 s  1 + s
2
3
(6 s + 3 + 9 s + 2 s ) conj(s)
Is2 =
2
4 s + s + 1
 2
3
(6 s + 3 + 9 s + 2 s ) conj(s)
Therefore,
2
s + 2s 1  (1)
I 1 s = 3
2
2s + 9s + 6s + 3
2
4s + s + 1
I 2 s =  (2)
3
2
2s + 9s + 6s + 3
We express the denominator of (1) as a product of a linear and quadratic term using MATLAB.
p=[2 9 6 3]; r=roots(p); fprintf(' \n'); disp('root1 ='); disp(r(1));...
disp('root2 ='); disp(r(2)); disp('root3 ='); disp(r(3)); disp('root2+root3 ='); disp(r(2)+r(3));...
disp('root2 * root3 ='); disp(r(2)*r(3))
root1 =
3.8170
root2 =
Circuit Analysis II with MATLAB Applications
Orchard Publications
623
r1 = 0.48
r2 = 0.52
r3 = 0.31
By inspection, the Inverse Laplace of first term on the right side of (4) is
0.48 3.82t
 0.48e
(5)
s + 3.82
The second term on the right side of (4) requires some manipulation. Therefore, we will use the
MATLAB ilaplace(s) function to find the Inverse Laplace as shown below.
syms s t
IL=ilaplace((0.52*s0.31)/(s^2+0.68*s+0.39));
624
Solutions to Exercises
pretty(IL)
1217
17
1/2
1/2
  exp(  t) 14
sin(7/50 14
t)
4900
50
13
17
1/2
+  exp(  t) cos(7/50 14
t)
25
50
Thus,
i 1 t = 0.48e
3.82t
0.93e
0.34t
0.34t
cos 0.53t
4s + s + 1 I 2 s = 3
2
2s + 9s + 6s + 3
r1 = 4.49
r2 = 0.49
r3 = 0.20
By inspection, the Inverse Laplace of first term on the right side of (7) is
Circuit Analysis II with MATLAB Applications
Orchard Publications
625
The second term on the right side of (7) requires some manipulation. Therefore, we will use the
MATLAB ilaplace(s) function to find the Inverse Laplace as shown below.
syms s t
IL=ilaplace((0.49*s0.20)/(s^2+0.68*s+0.39)); pretty(IL)
167
17
1/2
 exp(  t) 14
9800
50
1/2
sin(7/50 14
t)
49
17
1/2
+  exp(  t) cos(7/50 14
t)
100
50
Thus,
i 2 t = 4.47 e
3.82t
+ 0.06e
0.34t
0.34t
cos 0.53t
4.
I1 s
I2 s
V1 s
1es
VC s
V 2 s = 2V C s
a. Mesh 1:
2 + 1 e s I1 s I2 s = V1 s
or
6 2 + 1 e s I 1 s 6I 2 s = 6V 1 s (1)
Mesh 2:
I 1 s + 6I 2 s = V 2 s = 2 e s I 1 s (2)
or
11 + 8 e s I 1 s = 6V 1 s
626
Solutions to Exercises
and thus
I1 s
6s 6  =  = Y s = V 1 s 11 + 8 e s 11s + 8
b. With V 1 s = 1 e s we get
6 8 e 11 t
1
6s
6 e 11  6  = e
= i1 t
I 1 s = Y s V 1 s =   = 11
11s + 8 s 11s + 8 s + 8 e 11
5.
Circuit (a):
R
+
1 e Cs
V in s
V out s
1 e Cs
V out s =  V in s
R + 1 e Cs
and
V out s
1 e RC 1  = 1 e Cs
1 e Cs  =  =  = G s = s + 1 e RC
RCs + 1
RCs + 1 e Cs
R + 1 e Cs
V in s
Circuit (b):
+
V in s
+
R
V out s
R
V out s =  V in s
Ls + R
and
V out s
ReL R  = G s =  = s+ReL
Ls + R
V in s
627
+
R
V in s
V out s
R
V out s =  V in s
1 e Cs + R
and
V out s
R
RCs
s
 =  =  = G s = V in s
1 e Cs + R
RCs + 1
s + 1 e RC
Circuit (b):
R
+
L
V in s
` V
out s
Ls
V out s =  V in s
R + Ls
and
V out s
Ls
s
G s =  =  = R + Ls
s+ReL
V in s
Circuit (a):
+
V in s
V out s
R
V out s =  V in s
Ls + 1 e Cs + R
628
Solutions to Exercises
and
V out s
R
RCs
R e L s
 =  =  = G s = 2
2
V in s
Ls + 1 e Cs + R
LCs + 1 + RCs
s + R e L s + 1 e LC
V in s
`V
out s
Ls + 1 e Cs
V out s =  V in s
R + Ls + 1 e Cs
and
2
2
V out s
Ls + 1 e Cs
LCs + 1
s + 1 e LC
G s =  =  =  = 2
2
V in s
R + Ls + 1 e Cs
LCs + RCs + 1
s + R e L s + 1 e LC
R u 1 e Cs
R 2 + 1 e Cs
R2
V out s
V s
V in s
z
z1
2
out
 and since for inverting opamp  = 2 , for this circuit
Let z 1 = R 1 and z 2 = 
> R 2 u 1 e Cs e R 2 + 1 e Cs @
R 2 u 1 e Cs
R1 C
V out s
G s =  =  =  = V in s
R1
R 1 R 2 + 1 e Cs
s + 1 e R2 C
629
R1
V in s
V out s
V s
V in s
z
z1
out
 = 2 , for this circuit
Let z 1 = R 1 + 1 e Cs and z 2 = R 2 and since for inverting opamp 
R2
R 2 e R 1 s
V out s
 =  = G s = V in s
R 1 + 1 e Cs
s + 1 e R1C
R1 = 11.3 K:
R2 = 22.6 K:
R3
R1
R3=R4 = 68.1 K:
R2
V1
C1=C2 = 0.01 PF
V3
V2
V in s
V out s
C1
C2
At Node V 1 :
V 1 s V 1 s V out s
 +  = 0
R3
R4
or
1
1 V s = 1 +   V s (1)
1
R
R 4 out
R4
3
At Node V 3 :
V3 s V2 s V3 s
 +  = 0
R2
1 e C1 s
630
Solutions to Exercises
and since V 3 s = V 1 s we express the last relation above as
V1 s V2 s
 + C 1 sV 1 s = 0
R2
or
1
1
 + C 1 s V 1 s =  V 2 s (2)
R
R
2
2
At Node V 2 :
V 2 s V in s V 2 s V 1 s V 2 s V out s
 +  +  = 0
R1
R2
1 e C2 s
or
V in s V 1 s
1
1
 +  + C 2 s V 2 s =  +  + C 2 sV out s (3)
R
R1
R2
R2
1
From (1)
R3
1 e R4
V 1 s =  V out s =  V out s (4)
R3 + R4 e R3 R4
R3 + R4
From (2)
1
V 2 s = R 2  + C 1 s V 1 s = 1 + R 2 C 1 s V 1 s
R
or
R3
3 1 + R2 C1 s
1
1
1 + C s R
1 +     C 2 s V out s =  V in s
2
R
R3 + R4
R
R
R
+
R
R
2
3
4
1
1
2
and thus
V out s
1
 = G s = V in s
R3 1 + R2 C1 s 1
R3
1
1
R 1  +  + C 2 s    C 2 s
R
R3 + R4
R2 R3 + R4
R2
1
631
7.83 u 10
G s = 2
4
7
s + 1.77 u 10 s + 5.87 u 10
w=1:10:10000; s=j.*w; Gs=7.83.*10.^7./(s.^2+1.77.*10.^4.*s+5.87.*10.^7);...
semilogx(w,abs(Gs)); grid; hold on
xlabel('Radian Frequency w'); ylabel('Vout/Vin');
title('Magnitude Vout/Vin vs. Radian Frequency')
The plot above indicates that this circuit is a secondorder lowpass filter.
632
Chapter 7
Frequency Response and Bode Plots
his chapter discusses frequency response in terms of both amplitude and phase. This topic will
enable us to determine which frequencies are dominant and which frequencies are virtually
suppressed. The design of electric filters is based on the study of the frequency response. We
will also discuss the Bode method of linear system analysis using two separate plots; one for the magnitude of the transfer function, and the other for the phase, both versus frequency. These plots reveal
valuable information about the frequency response behavior.
Note: Throughout this text, the common (base 10) logarithm of a number x will be denoted as
log x while its natural (base e) logarithm will be denoted as ln x . However, we should remember
that in MATLAB the log x function displays the natural logarithm, and the common (base 10) logarithm is defined as log 10 x .
7.1 Decibels
The ratio of any two values of the same quantity (power, voltage or current) can be expressed in
decibels ( dB ). For instance, we say that an amplifier has 10 dB power gain or a transmission line
has a power loss of 7 dB (or gain 7 dB ). If the gain (or loss) is 0 dB , the output is equal to the
input. We should remember that a negative voltage or current gain A V or A I indicates that there is a
180q phase difference between the input and the output waveforms. For instance, if an amplifier has
a gain of 100 (dimensionless number), it means that the output is 180q outofphase with the
input. For this reason we use absolute values of power, voltage and current when these are expressed
in dB terms to avoid misinterpretation of gain or loss.
By definition,
P out
dB = 10 log P in
(7.1)
Therefore,
10 dB represents a power ratio of 10
10n dB represents a power ratio of 10
71
From these, we can estimate other values. For instance, 4 dB = 3 dB + 1 dB which is equivalent to a
power ratio of approximately 2 u 1.25 = 2.5 Likewise, 27 dB = 20 dB + 7 dB and this is equivalent
to a power ratio of approximately 100 u 5 = 500 .
2
Since y = log x = 2 log x and P = V e R = I R , if we let R = 1 the dB values for the voltage and
current ratios become:
V out
V out 2
dB v = 10 log  = 20 log V in
V in
(7.2)
I out
I out 2
dB i = 10 log  = 20 log I in
I in
(7.3)
and
Example 7.1
P out
1w
10 w
Solution:
P out
10
dB W = 10 log  = 10 log  = 10 log 10 = 10 u 1 = 10 dB W
1
P in
Example 7.2
Compute the gain in dB V for the amplifier shown in Figure 7.2 given that log 2 = 0.3 .
V in
V out
1v
2v
72
V out
0.707
Bandwith
Z2
Z1
As shown in Figure 7.3, the bandwidth is BW = Z 2 Z 1 where Z 1 and Z 2 are the lower and upper
cutoff frequencies respectively. At these frequencies, V out =
are known as the 3 dB down or halfpower points. They derive their name from the fact that since
2
power p = v e R = i R , for R = 1 and for v = 0.707 V out or i = 0.707 I out the power is 1 e 2 ,
that is, it is halved. Alternately, we can define the bandwidth as the frequency band between halfpower points.
Most amplifiers are used with a feedback path which returns (feeds) some or all its output to the
input as shown in Figure 7.4.
INPUT
GAIN AMPLIFIER
OUTPUT
+
FEEDBACK CIRCUIT
Figure 7.5 shows an amplifier where the entire output is fed back to the input.
73
GAIN AMPLIFIER
+
OUTPUT
FEEDBACK PATH
The symbol 6 (Greek capital letter sigma) inside the circle indicates the summing point where the
output signal, or portion of it, is combined with the input signal. This summing point may be also
indicated with a large plus (+) symbol inside the circle. The positive (+) sign below the summing
point implies positive feedback which means that the output, or portion of it, is added to the input.
On the other hand, the negative () sign implies negative feedback which means that the output, or
portion of it, is subtracted from the input. Practically, all amplifiers use used with negative feedback
since positive feedback causes circuit instability.
(7.4)
If these frequencies are such that Z 2 = 2Z 1 , we say that these frequencies are separated by one
octave and if Z 2 = 10Z 1 , they are separated by one decade.
Let us now consider a transfer function G s whose magnitude is evaluated at s = jZ , that is,
CG s = k
s
s = jZ
C= G Z = k
Z
(7.5)
Taking the log of both sides of (7.5) and multiplying by 20, we get
k
or
(7.6)
G Z dB = 20klog 10 Z + 20log 10 C
Relation (7.6) is an equation of a straight line in a semilog plot with abscissa log 10 Z where
dB
slope = 20k decade
74
40
GZ
dB
G Z axis intercept
30
C
20 dB e decade = 6 dB e octave
20
10
0
1
log10 Z
10
100
1000
With these concepts in mind, we can now proceed to discuss Bode Plots and Asymptotic Approximations.
10) scale, and the radian frequency Z is equally spaced between powers of 10 such as 10 , 10 ,
1
10 , 10 and so on.
The ordinate ( dB axis) of the magnitude plot has a scale in dB units, and the ordinate of the phase
plot has a scale in degrees as shown in Figure 7.7.
90q
10
0
10
10
100
Frequency Z r/s
Magnitude (dB)
20
45q
0
45q
10
100
Frequency Z r/s
90q
20
Bode Magnitude Plot
75
100
1 jZ dB + dB
 = 20 dB + 1 + 
1 + jZ
100
1 + jZ
and the Bode plots then can be approximated by straight lines called asymptotes.
7.5 Construction of Bode Plots when the Zeros and Poles are Real
Let us consider the transfer function
A s + z1 s + z2 } s + zm
G s = s s + p 1 s + p2 s + p3 s + pn
(7.7)
where A is a real constant, and the zeros z i and poles p i are real numbers. We will consider complex
zeros and poles in the next section. Letting s = jZ in (7.7) we get
A jZ + z 1 jZ + z 2 } jZ + z m
G jZ = jZ jZ + p 1 jZ + p 2 jZ + p 3 jZ + p n
(7.8)
Next, we multiply and divide each numerator factor jZ + z i by z i and each denominator factor
jZ + p i by p i and we get:
A z 1 jZ
 + 1 z 2 jZ
 + 1 } z m jZ
 + 1
z
1
2
m
G jZ = jZ
jZ
jZ
jZ p 1  + 1 p 2  + 1 } p n  + 1
p
(7.9)
Letting
m
zi
A z1 z2 } zm
i=1
K = = A n
p1 p2 } pn
pi
(7.10)
i=1
76
Construction of Bode Plots when the Zeros and Poles are Real
jZ
jZ
jZ
G Z = 20 log K + 20 log  + 1 + 20 log  + 1 + } + 20 log  + 1
z
1
2
m
jZ
jZ
jZ
20 log jZ 20 log  + 1 20 log  + 1 } 20 log  + 1
p
1
2
n
(7.11)
jZ
jZ
jZ
G Z = K +  + 1 +  + 1 + } +  + 1
z
(7.12)
jZ + 1 } jZ
jZ + 1  + 1
jZ p
Magnitude (dB)
The constant K can be positive or negative. Its magnitude is K and its phase angle is 0q if K ! 0 ,
and 180 q if K 0 . The magnitude and phase plots for the constant K are shown in Figure 7.8.
20logK
0
Frequency Z r/s
K!0
0q
Frequency Z r/s
K0
180q
For a zero of order n , that is, jZ at the origin, the Bode plots for the magnitude and phase are as
shown in Figures 7.9 and 7.10 respectively.
n
2 n
2 ne2
a + Z = a + Z
(7.13)
(7.14)
(7.15)
77
120
100
80
Magnitude in dB
60
40
20
0
20
40
60
80
n=1
n=2
n=3
100
120
0.01
0.10
1.00
10.00
100.00
Z (r/s)
360
n=3
270
n=2
180
n=1
90
0
0.01
0.10
1.00
10.00
100.00
Z(rad/s)
(7.16)
78
Construction of Bode Plots when the Zeros and Poles are Real
120
100
n=3
80
Magnitude in dB
60
n=2
40
20
n=1
0
20
40
60
80
100
120
0.01
0.10
1.00
10.00
100.00
Z (r/s)
0
90
180
270
360
0.01
n=1
n=2
n=3
0.10
1.00
10.00
100.00
Z(rad/s)
For u 1 the first term of (7.16) becomes 10n log 1 = 0 dB . For u 1 , this term becomes approxi2
mately 10n log u = 20n log u and this has the same form as G jZ = jZ which is shown in Figure 7.9 for n = 1 , n = 2 , and n = 3 .
The frequency at which two aymptotes intersect each other forming a corner is referred to as the
corner frequency. Thus, the two lines defined by the first term of (7.16), one for u 1 and the other
for u 1 intersect at the corner frequency u = 1 .
The second term of (7.16) represents the ordinate axis intercept defined by this straight line.
79
u{Zea
and
1
(7.18)
I u = tan u
Then,
n
2 n e 2 jn I u
a + jZ = a 1 + ju = a 1 + u tan u = a 1 + u
(7.19)
Magnitude in dB
100
Asymptotes
80
n=3
60
n=2
40
n=1
20
Corner Frequencies
0
0.01
0.10
1.00
10.00
100.00
Frequency u (r/s)
As shown in Figure 7.13, a quick sketch can be obtained by drawing the straight line asymptotes given
2
nI u = T u = n tan u
710
(7.20)
Construction of Bode Plots when the Zeros and Poles are Real
we get
1
(7.21)
1
nS
lim T u = lim n tan u = 2
uof
(7.22)
uo0
uo0
and
uof
(7.23)
360
n=3
270
n=2
180
n=1
90
0
0.01
0.10
1.00
10.00
100.00
u (rad/s)
(7.24)
T u = n tan u
(7.25)
The plots for (7.24) and (7.25) are shown in Figures 7.15 and 7.16 respectively.
711
n=1
Magnitude in dB
20
40
60
n=2
n=3
80
100
120
0.01
Asymptotes
0.10
1.00
10.00
100.00
Frequency u (r/s)
0
n=1
90
n=2
180
n=3
270
360
0.01
0.10
1.00
10.00
100.00
u (rad/s)
7.6 Construction of Bode Plots when the Zeros and Poles are Complex
The final type of terms appearing in the transfer function G s are quadratic term of the form
2
as + bs + c whose roots are complex conjugates. In this case, we express the complex conjugate
roots as
712
Construction of Bode Plots when the Zeros and Poles are Complex
2
s + D jE s + D + jE = s + D + E
= s + 2Ds + D + E
(7.26)
and letting
(7.27)
D = ]Z n
and
2
(7.28)
D + E = Zn
(7.29)
s + 2Ds + D + E = s + 2]Z n s + Z n
Next, we let
2
(7.30)
G s = s + 2]Z n s + Z n
Then,
2
G jZ = jZ + j2Z n Z + Z n
=
2
Zn
(7.31)
Z + j2Z n Z
2 2
Z n Z + 4] Z n Z
(7.32)
(7.33)
u { Z e Zn
2 2
or
713
2 2
2
2
Zn Z
4 Zn Z
 + 4] 2 Z 4n Z
 = 10 log Z 4n  + 4] 2 Z 4n Z
20 log G ju = 10 log Z n 2
2
Z 2n
Z 2n
Zn
Zn
2 2
2 2
2 2
(7.35)
2 2
The first term in (7.35) is a constant which represents the ordinate axis intercept defined by this
2
straight line. For the second term, if u 1 , this term reduces to approximately 10 log 1 = 0 dB and
2
if u 1 , this term reduces to approximately 10 log u and this can be plotted as a straight line
increasing at 40 dB e decade . Using these two straight lines as asymptotes for the magnitude curve
we see that the asymptotes intersect at the corner frequency u = 1 . The exact shape of the curve
depends on the value of ] which is called the damping coefficient.
A plot of (7.35) for ] = 0.2 , ] = 0.4 , and ] = 0.707 is shown in Figure 7.17.
2
The phase shift associated with Z n Z + j2Z n Z is also simplified by the substitution u { Z e Z n
and thus
1 2]u
T u = tan 2
1u
(7.36)
uo0
uo0
1u
(7.37)
1 2]u
lim T u = lim tan 2 = S
uof
uof
1u
(7.38)
and
2
uo1
1u
(7.39)
A plot of the phase for ] = 0.2 , ] = 0.4 , and ] = 0.707 is shown in Figure 7.18.
714
Construction of Bode Plots when the Zeros and Poles are Complex
Zeros of (Zn2Z2)+j2]ZnZ
u = Z/Zn, Zn = 1
10logZn4+10log{(1u2)2+4]2u2}
40
Magnitude in dB
30
20
]=0.707
10
0
10
20
0.01
]=0.2
]=0.4
0.10
1.00
10.00
100.00
Frequency u (r/s)
2 2
2 2
Zeros of (Zn2Z2)+j2]ZnZ
u = Z/Zn, Zn = 1
2
T(u) = (arctan(2]u/(1u ))*180/S
180
]=0.707
]=0.4
90
]=0.2
0
0.01
0.10
1.00
10.00
100.00
u (rad/s)
4
2 2
2 2
715
G jZ = Z n Z + j2Z n Z
2 2
2 2
(7.40)
1 2]u
T u = tan 2
1u
(7.41)
A plot of (7.40) for ] = 0.2 , ] = 0.4 , and ] = 0.707 is shown in Figure 7.19.
20
]=0.2
Magnitude in dB
10
]=0.4
0
10
]=0.707
20
30
40
0.01
0.10
1.00
10.00
100.00
Frequency u (r/s)
2 2
2 2
A plot of the phase for ] = 0.2 , ] = 0.4 , and ] = 0.707 is shown in Figure 7.20.
716
Construction of Bode Plots when the Zeros and Poles are Complex
2
T(u) = (arctan(2]u/(1u ))*180/S
0
]=0.4
]=0.707
90
]=0.2
180
0.01
0.10
1.00
10.00
100.00
u (rad/s)
2 2
2 2
100 PF
+
100 mH
R
110 :
vs u0 t
+
v out t
Solution:
a. We transform the given circuit to its equivalent in the s domain shown in Figure 7.22.
717
C
4
10 e s
0.1s
+
R
110
V in s
V out s
(7.42)
b. Letting s = jZ we get
1100jZ
G jZ =  jZ + 100 jZ + 1000
or in standard form
0.011jZ
G jZ =  1 + jZ e 100 1 + jZ e 1000
(7.43)
10
100
(7.44)
We observe that the first term on the right side of (7.44) is a constant whose value is
20 log 0.011 = 39.17 . The second term is a straight line with slope equal to 20 dB e decade . For
Z 100 r e s the third term is approximately zero and for Z ! 100 it decreases with slope equal to
20 dB e decade Likewise, for Z 1000 r e s the fourth term is approximately zero and for
Z ! 1000 it also decreases with slope equal to 20 dB e decade
For Bode plots we use semilog paper. Instructions to construct semilog paper with Microsoft
Excel are provided in Appendix D.
718
Construction of Bode Plots when the Zeros and Poles are Complex
In the Bode plot of Figure 7.23 the individual terms are shown with dotted lines and the sum of
these with a solid line.
80
20 log 10 jZ
60
20 log 10 1 + jZ
40
20
0
20
20 log 10 1 + jZ e 1000
40
20 log 10 1 + jZ e 100
60
20 log 10 0.011
80
1.E+00
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
c. The plot of Figure 7.23 shows that the magnitude of (7.43) at Z = 30 r e s is approximately
9 dB and at Z = 4000 r e s is approximately 10 dB . The actual values are found as follows:
At Z = 30 r e s , (7.43) becomes
0.011 u j30
G j30 =  1 + j0.3 1 + j0.03
mag = 0.32
magdB = 10.01 dB
Therefore,
G j30 = 0.32
and
20 log G j30 = 20 log 0.32  10 dB
Circuit Analysis II with MATLAB Applications
Orchard Publications
719
mag = 0.27
magdB = 11.48 dB
Therefore,
G j4000 = 0.27
and
20 log G j4000 = 20 log 0.27 = 11.48 dB
d. From the Bode plot of Figure7.23, we see that the value of A dB at Z = 5000 r e s is approxiy
12

20
= 0.25
and therefore
V out max = A V S = 0.25 u 10 = 2.5 V
If we wish to obtain a more accurate value, we substitute Z = 5000 into (7.43) and we get
g5000=0.011*5000j/((1+50j)*(1+5j));...
fprintf(' \n'); fprintf('mag = %6.2f \t',abs(g5000));...
fprintf('phase = %6.2f deg.',angle(g5000)*180/pi); fprintf(' \n'); fprintf(' \n')
mag = 0.22
Then,
V out max = A u 10 = 0.22 u 10 = 2.2 V
720
Construction of Bode Plots when the Zeros and Poles are Complex
We can use the MATLAB function bode(sys) to draw the Bode plot of a Linear Time Invariant
(LTI) System where sys = tf(num,den) creates a continuoustime transfer function sys with
numerator num and denominator den, and tf creates a transfer function. With this function, the freq u e n c y r a n g e a n d n u m b e r o f p o i n t s a r e c h o s e n a u t o m a t i c a l l y. T h e f u n c t i o n
bode(sys,{wmin,wmax}) draws the Bode plot for frequencies between wmin and wmax (in radians/second) and the function bode(sys,w) uses the usersupplied vector w of frequencies, in radians/second, at which the Bode response is to be evaluated. To generate logarithmically spaced frequency vectors, we use the command l o g s p a c e ( f i r s t _ e x p o n e n t , l a s t _ e x p o n e n t ,
number_of_values). For example, to generate plots for 100 logarithmically evenly spaced points
1
and the MATLAB code to generate the magnitude and phase plots is
num=[0 1100 0]; den=[1 1100 10^5]; w=logspace(0,5,100); bode(num,den,w)
However, since for this example we are interested in the magnitude only, we will use the code
num=[0 1100 0]; den=[1 1100 10^5]; sys=tf(num,den);...
w=logspace(0,5,100); bodemag(sys,w); grid
and upon execution, MATLAB displays the plot shown in Figure 7.24.
Example 7.4
721
Solution:
a. From (7.43) of Example 7.3
0.011jZ
G jZ =  1 + jZ e 100 1 + jZ e 1000
(7.45)
where
D = 90q
E = tan Z e 100
J = tan Z e 1000
For Z = 100
1
E = tan 1 = 45q
For Z = 1000
1
J = tan 1 = 45q
722
Construction of Bode Plots when the Zeros and Poles are Complex
180
TZ = D E J
135
90
D = 90q
45
45
J = tan Z e 1000
1
E = tan Z e 100
90
135
180
10
10
10
10
10
10
Figure 7.26 shows the magnitude and phase plots generated with the following MATLAB code:
num=[0 1100 0]; den=[1 1100 10^5]; w=logspace(0,5,100); bode(num,den,w)
b. From the Bode plot of Figure 7.25 we find that the phase is zero degrees at approximately
Z = 310 r e s
c. From (7.45)
0.011jZ
G jZ =  1 + jZ e 100 1 + jZ e 1000
723
Figure 7.26. Bode plots for Example 7.4 generated with MATLAB
This is a trigonometric equation and we will solve it for Z with the solve(equ) MATLAB function as follows:
syms w; x=solve(atan(w/100)+atan(w/1000)pi/2); combine(x)
ans =
316.2278
Therefore, Z = 316.23 r e s
d. Evaluating (7.45) at Z = 316.23 r e s we get:
0.011 j316.23
G j316.23 =  1 + j316.23 e 100 1 + j316.23 e 100 0
(7.46)
magGj316 = 1.00
724
The phase angle of the input voltage is given as T in = 60q and with T j316.23 = 0q we find that
the phase angle of the output voltage is
T out = T in + T j316.23 = 60q + 0q = 60q
and thus
V out = 10 60q
or
v out t = 10 cos 316.23t + 60q
Z = Zn
= r 20 log 1 + j = r 20 log 2 = r 3 dB
(7.47)
where the plus (+) sign applies to a first order zero, and the minus () to a first order pole.
Similarly,
A dB
Z = Zn e 2
= r 20 log 1 + j e 2 = r 20 log 5
 = r 0.97 dB  r 1 dB
4
(7.48)
(7.49)
and
A dB
Z = 2 Zn
As we can seen from Figure 7.27, the straight line approximations, shown by dotted lines, yield 0 dB
at half the corner frequency and at the corner frequency. At twice the corner frequency, the straight
line approximations yield r 6 dB because Z n and 2Z n are separated by one octave which is equivalent to r 3 dB per decade. Therefore, the corrections to be made are r 1 dB at half the corner frequency Z n e 2 , r 3 dB at the corner frequency Z n , and r 1 dB at twice the corner frequency 2Z n .
The corrected amplitude plots for a first order zero and first order pole are shown by solid lines in
Figure 7.27.
725
Magnitude in dB
20
15
10
7 dB
6 dB
3 dB
1 dB
0
1 dB
5
3 dB
6 dB
7 dB
10
15
20
Zn e 2
Zn
Z in r/s
Zn e 2
We observe that as the damping coefficient ] becomes smaller and smaller, larger and larger peaks in
the amplitude occur in the vicinity of the corner frequency Z n . We also observe that when
] t 0.707 , the amplitude at the corner frequency Z n lies below the straight line approximation.
We can obtain a fairly accurate amplitude plot by computing the amplitude at four points near the
corner frequency Z n as shown in Figure 7.28.
726
20
]=0.2
Magnitude in dB
10
]=0.4
0
10
]=0.707
20
30
40
0.01
0.10
1.00
10.00
100.00
Frequency u (r/s)
G s = s + 2]Z n s + Z n
which was derived earlier for complex zeros, the transfer function for complex poles is
C
G s = 2
2
s + 2]Z n s + Z n
(7.50)
where C is a constant.
Dividing each term of the denominator of (7.50) by Z n we get
C
1
G s = 2 2
Z n s e Z n + 2] s e Z n + 1
2
(7.51)
727
(7.52)
K
G ju = 2
1 u + j2]u T
(7.53)
2 2
(7.54)
4
(7.55)
In (7.54) the term 20 log K is constant and thus the amplitude A dB , as a function of frequency, is
dependent only the second term on the right side. Also, from this expression, we observe that as
u o 0,
4
(7.56)
10 log > u + 2u 2] 1 + 1 @ o 0
and as u o f ,
4
(7.57)
We are now ready to compute the values of A dB at points 1 , 2 , 3 , and 4 of the plot of Figure 7.29.
At point 1, the corner frequency Z n corresponds to u = 1 . Then, from (7.54)
u
4
2
2
A dB Z n e 2 = A dB  = 10 log > u + 2u 2] 1 + 1 @
2
u = 1e2
1
2
2 1
1
1
= 10 log  + 2  2] 1 + 1 = 10 log  + ]  + 1
4
2
16
16
(7.58)
= 10 log ] + 0.5625
728
point 1
6
5
4
Point 2 at Z = Z max
Point 3 at Z = Z n
Point 1 at Z = Z n e 2
2
Point 4 at Z = Z 0 dB
1
0
1
2
3
4
5
6
Zn e 2
Z
Z max n Z 0 dB
Figure 7.29. Corrections for magnitude Bode plots with complex poles when ] = 0.4
To find the amplitude at point 2, in (7.54) we let K = 1 and we form the magnitude in dB . Then,
A dB
point 2
1
= 20 log 2
1 Z e Z n + j2]Z e Z n
(7.59)
We now recall that the logarithmic function is a monotonically increasing function and therefore
(7.59 has a maximum when the absolute magnitude of this expression is maximum. Also, the square
of the absolute magnitude is maximum when the absolute magnitude is maximum.
The square of the absolute magnitude is
1
2 2
2
> 1 Z e Z n @ + 4 ]Z e Z n
(7.60)
729
(7.61)
To find the maximum, we take the derivative with respect to Z and we set it equal to zero, that is,
2
4Z e Z n 4Z e Z n 8] Z e Z n
 = 0
2
2
2
2
> 1 Z e Z n @ + 4 ]Z e Z n
(7.62)
The expression of (7.62) will be zero when the numerator is set to zero, that is,
2
(7.63)
Z e Z n 4 4Z e Z n 8] = 0
4 4Z e Z n 8] = 0
or
2
4Z e Z n = 4 8]
from which
Z max = Z = Z n 1 2]
(7.64)
u=
1 2]
2
= 10 log > 1 2] + 2 1 2] 2] 1 + 1 @
2
(7.65)
= 10 log > 4] 1 ] @
730
A dB Z n = 10 log > u + 2u 2] 1 + 1 @
u=1
= 10 log > 1 + 2 2] 1 + 1 @
(7.66)
Finally, at point 4 , the dB value of the amplitude crosses the 0 dB axis. Therefore, at this point we
are interested not in A dB Z 0 dB but in the location of Z 0 dB in relation to the corner frequency Z n .
at this point we must have from (7.57)
4
0 dB = 10 log > u + 2u 2] 1 + 1 @
u + 2u 2] 1 + 1 = 1
4
u + 2u 2] 1 = 0
2
u u + 2 2] 1 = 0
or
2
u + 2 2] 1 = 0
Z 0 dB = Z n 2 1 2]
From (7.67),
Z max = Z n 1 2]
therefore, if we already know the frequency at which the dB amplitude is maximum, we can compute
the frequency at point 4 from
Z 0 dB =
2Z max
(7.67)
Example 7.5
731
R
0.2 :
10 mH
C
40 mF
v in u 0 t
v out t
Solution:
a. We transform the given circuit to its equivalent in the s domain shown in Figure 7.31 where
R = 1 , Ls = 0.05s , and 1 e Cs = 125 e s .
L
R
0.2
+
V in s
0.01s
C
25 e s
V out s
732
(7.68)
K
G s = 2
2
s + 2]Z n s + Z n
(7.69)
b. From (7.50)
Z n = 50 rad e s
(7.71)
d. For Z Z n , the straight line approximation lies along the 0 dB axis, whereas for Z ! Z n , the
straight line approximation has a slope of 40 dB . The corner frequency Z n was found in part (b)
to be 50 rad e s The dB amplitude plot is shown in Figure 7.31.
e. From (7.61),
2
A dB Z n e 2 = 10 log ] + 0.5625
2
Then,
Z max = 50 1 2 u 0.04 = 50 0.92 = 47.96 rad e s
733
Then,
A dB Z n = 20 log 2 u 0.2 = 7.96 dB
2Z max
or
Z 0 dB =
20
Point 2
8.14 dB
15
Point 3
7.96 dB
10
Point 1
2.2 dB
Point 4
0 dB
Magnitude in dB
0
5
10
15
20
25
30
35
40
10
Z n e 2 = 25
Z max  48
100
1000
Z 0 dB  68
Z in r/s
Z n = 50
The amplitude plot of Figure 7.32 reveals that the given circuit behaves as a low pass filter.
734
735
can also be used to express voltage and current ratios provided that the voltages and currents have
identical impedances. Then, for voltages we use the expression dB v = 20 log 10 V out e V in , and for
currents we use the expression dB i = 20 log 10 I out e I in
x The bandwidth, denoted as BW , is a term generally used with electronic amplifiers and filters. For
lowpass filters the bandwidth is the band of frequencies from zero frequency to the cutoff frequency where the amplitude fall to 0.707 of its maximum value. For highpass filters the bandwidth is the band of frequencies from 0.707 of maximum amplitude to infinite frequency. For
amplifiers, bandpass, and bandelimination filters the bandwidth is the range of frequencies where
the maximum amplitude falls to 0.707 of its maximum value on either side of the frequency
response curve.
x If two frequencies Z 1 and Z 2 are such that Z 2 = 2Z 1 , we say that these frequencies are separated
by one octave and if Z 2 = 10Z 1 , they are separated by one decade.
x Frequency response is a term used to express the response of an amplifier or filter to input sinusoids of different frequencies. The response of an amplifier or filter to a sinusoid of frequency Z is
completely described by the magnitude G jZ and phase G jZ of the transfer function.
x Bode plots are frequency response diagrams of magnitude and phase versus frequency Z .
x In Bode plots the 3  dB frequencies, denoted as Z n , are referred to as the corner frequencies.
x In Bode plots, the transfer function is expressed in linear factors of the form jZ + z i for the zero
(numerator) linear factors and jZ + p i for the pole linear factors. When quadratic factors with
complex roots occur in addition to the linear factors, these quadratic factors are expressed in the
2
form jZ + j2]Z n Z + Z n .
x In magnitude Bode plots with quadratic factors the difference between the asymptotic plot and the
actual curves depends on the value of the damping factor ] . But regardless of the value of ] , the
actual curve approaches the asymptotes at both low and high frequencies.
x In Bode plots the corner frequencies Z n are easily identified by expressing the linear terms as
z i jZ e z i + 1 and p i jZ e p i + 1 for the zeros and poles respectively. For quadratic factor the cor2
736
Summary
x In both the magnitude and phase Bode plots the frequency (abscissa) scale is logarithmic. The
ordinate in the magnitude plot is expressed in dB and in the phase plot is expressed in degrees.
x In magnitude Bode plots, the asymptotes corresponding to the linear terms of the form
jZ e z i + 1 and jZ e p i + 1 have a slope r 20 dB e decade where the positive slope applies to
zero (numerator) linear factors, and the negative slope applies to pole (denominator) linear factors.
x In magnitude Bode plots, the asymptotes corresponding to the quadratic terms of the form
2
jZ e Z n + j2]Z e Z n + 1 have a slope r 40 dB e decade where the positive slope applies to zero
(numerator) quadratic factors, and the negative slope applies to pole (denominator) quadratic factors.
x In phase Bode plots with linear factors, for frequencies less than one tenth the corner frequency
we assume that the phase angle is zero. At the corner frequency the phase angle is r 45q . For frequencies ten times or greater than the corner frequency, the phase angle is approximately r 90q
where the positive angle applies to zero (numerator) linear factors, and the negative angle applies
to pole (denominator) linear factors.
x In phase Bode plots with quadratic factors, the phase angle is zero for frequencies less than one
tenth the corner frequency. At the corner frequency the phase angle is r 90q . For frequencies ten
times or greater than the corner frequency, the phase angle is approximately r 180q where the pos
itive angle applies to zero (numerator) quadratic factors, and the negative angle applies to pole
(denominator) quadratic factors.
x Bode plots can be easily constructed and verified with the MATLAB function bode(sys) func
tion. With this function, the frequency range and number of points are chosen automatically. The
function bode(sys),{wmin,wmax}) draws the Bode plot for frequencies between wmin and
wmax (in radian/second) and the function bode(sys,w) uses the usersupplied vector w of frequencies, in radian/second, at which the Bode response is to be evaluated. To generate logarithmically spaced frequency vectors, we use the command logspace(first_exponent,last_exponent,
number_of_values).
737
10 s + 5
G s =  s + 100 s + 5000
a. Draw the magnitude Bode plot and find the approximate maximum value of G jZ in dB .
b. Find the value of Z where G jZ = 1 for Z ! 5 r e s
c. Check your plot with the plot generated with MATLAB.
2. For the transfer function of Exercise 1
a. Draw a Bode plot for the phase angle and find the approximate phase angle at Z = 30 r e s ,
Z = 50 r e s , Z = 100 r e s , and Z = 5000 r e s
b. Compute the actual values of the phase angle at the frequencies specified in (a).
c. Check your magnitude plot of Exercise 1 and the phase plot of this exercise with the plots generated with MATLAB.
3. For the circuit of Figure 7.34
a. Compute the transfer function.
b. Draw the Bode amplitude plot for 20 log G jZ
c. From the plot of part (b) determine the type of filter represented by this circuit and estimate the
cutoff frequency.
d. Compute the actual cutoff frequency of this filter.
e. Draw a straight line phase angle plot of G jZ .
f. Determine the value of T Z at the cutoff frequency from the plot of part (c).
g. Compute the actual value of T Z at the cutoff frequency.
L
0.25 H
R
1:
v out t
v in u 0 t
4 u 10
738
Answers to Exercises
7.10 Answers to Exercises
1. a.
5
10 u 5 u 1 + jZ e 5
10 jZ + 5
G jZ =  =  jZ + 100 jZ + 5000
100 u 1 + jZ e 100 u 5000 u 1 + jZ e 5000
1 + jZ e 5
=  1 + jZ e 100 1 + jZ e 5000
20 log G jZ = 20 log 1 + jZ e 5 20 log 1 + jZ e 100 20 log 1 + jZ e 5000
20 log G jZ
20 log 1 + jZ e 5
30
25
Magnitude of G jZ in dB
20
15
10
5
0
5
10
20 log 1 + jZ e 100
20 log 1 + jZ e 100
15
20
25
30
35
40
10
10
10
10
10
Orchard Publications
Z r e s
From this plot we observe that 20 log G jZ max  26 dB for the interval 10 d Z d 5 u 10
10
739
From the phase plot we observe that T 30 r e s  60q , T 50 r e s  53q , T 100 r e s  38q , and
T 5000 r e s  39 q
90
D = tan Z e 5
75
60
T Z
45
30
15
0
15
30
45
60
E = tan Z e 100
1
J = tan Z e 5000
75
90
10
740
10
10
Z r e s
10
10
10
Answers to Exercises
b. We use MATLAB for the computations.
theta_g30=(1+30j/5)/((1+30j/100)*(1+30j/5000));...
theta_g50=(1+50j/5)/((1+50j/100)*(1+50j/5000));...
theta_g100=(1+100j/5)/((1+100j/100)*(1+100j/5000));...
theta_g5000=(1+5000j/5)/((1+5000j/100)*(1+5000j/5000));...
printf(' \n');...
fprintf('theta30r = %5.2f deg. \t', angle(theta_g30)*180/pi);...
fprintf('theta50r = %5.2f deg. ', angle(theta_g50)*180/pi);...
fprintf(' \n');...
fprintf('theta100r = %5.2f deg. \t', angle(theta_g100)*180/pi);...
fprintf('theta5000r = %5.2f deg. ', angle(theta_g5000)*180/pi);...
fprintf(' \n')
ans =
s^2+5100*s+500000
num=[0 10^5 5*10^5];
bode(num,den,w)
den=[1
5.1*10^3
5*10^5];
w=logspace(0,5,10^4);...
741
0.25s
1
V out s
V in s
25 e s
and
V out s
s + 25
4 s + 25
 =  =  (1)
G s = 2
2
V in s
0.25s + s + 25
s + 4s + 100
From (7.53)
742
Answers to Exercises
C
G s = (3)
2
2
s + 2]Z n s + Z n
2
or
1 + jZ e 25 T
G jZ =  (4)
2
1 Z e 10 + j0.4Z e 10 I
The amplitude of G jZ in dB is
2
The asymptote of the first term on the right side of (5) has a corner frequency of 25 r e s and
rises with slope of 20 dB e decade . The second term has a corner frequency of 10 r e s and rises
with slope of 40 dB e decade . The amplitude plot is shown below.
80
3 dB at Z c  13 r e s
60
40
20
0
20 log 1 + jZ e 25
20
20 log G jZ
40
60
80
10
10
10
10
10
10
c. The plot above indicates that the circuit is a lowpass filter and the 3 dB cutoff frequency Z c
occurs at approximately 13 r e s .
743
100 + 4Z c
1
 = 2
2
2
2
100 Z c + 4Z c
2
2 2
Z c 216Z c 10000 = 0
ans =
[
[
2*(27+1354^(1/2))^(1/2)]
2*(271354^(1/2))^(1/2)]
[ 2*(27+1354^(1/2))^(1/2)]
[ 2*(271354^(1/2))^(1/2)]
w1=2*(27+1354^(1/2))^(1/2)
w1 =
15.9746
w2=2*(27+1354^(1/2))^(1/2)
w2 =
15.9746
w3=2*(271354^(1/2))^(1/2)
w3 =
0.0000 + 6.2599i
w4=2*(271354^(1/2))^(1/2)
w4 =
0.0000  6.2599i
744
Answers to Exercises
From these four roots we accept only the first, that is, Z c  16 r e s
e. From (4)
1
T = tan Z e 25
and
0.4Z e 10 I = 2
1 Z e 10
For a first order zero or pole not at the origin, the straight line phase angle plot approximations
are as follows:
I. For frequencies less than one tenth the corner frequency we assume that the phase angle is
zero. For this exercise the corner frequency of T Z is Z n = 25 r e s and thus for
1 d Z d 2.5 r e s the phase angle is zero as shown on the Bode plot below.
180
TZ
135
90
45
T Z n = 25 r e s
M Z n = 10 r e s
G jZ
45
90
135
I Z
180
10
10
10
Z r e s
10
10
10
II For frequencies ten times or greater than the corner frequency, the phase angle is approximately r 90q . The numerator phase angle T Z is zero at one tenth the corner frequency,
it is 45q at the corner frequency, and 90q for frequencies ten times or greater the corner
frequency. For this exercise, in the interval 2.5 d Z d 250 r e s the phase angle is zero at
2.5 r e s and rises to 90q at 250 r e s .
745
ans =
125.0746
This value is about twice as that we observed from the asymptotic plot of the previous page.
Errors such as this occur because of the high nonlinearity between frequency intervals. Therefore, we should use the straight line asymptotes only to observe the shape of the phase angle. It
is best to use MATLAB as shown below.
num=[0 4 100]; den=[1 4 100]; w=logspace(0,2,1000);bode(num,den,w)
746
Chapter 8
Self and Mutual Inductances  Transformers
his chapter begins with the interactions between electric circuits and changing magnetic fields.
It defines self and mutual inductances, flux linkages, induced voltages, the dot convention,
Lenzs law, and magnetic coupling. It concludes with a detailed discussion on transformers.
8.1 SelfInductance
About 1830, Joseph Henry, while working at the university which is now known as Princeton, found
that electric current flowing in a circuit has a property analogous to mechanical momentum which is
a measure of the motion of a body and it is equal to the product of its mass and velocity, i.e., Mv . In
electric circuits this property is sometimes referred to as the electrokinetic momentum and it is equal to
the product of Li where i is the current analogous to velocity and the selfinductance L is analogous
to the mass M . About the same time, Michael Faraday visualized this property in a magnetic field in
space around a current carrying conductor. This electrokinetic momentum is denoted by the symbol
O that is,
O = Li
(8.1)
Newtons second law states that the force necessary to change the velocity of a body with mass M is
equal to the rate of change of the momentum, i.e.,
dv
d
F =  Mv = M  = Ma
dt
dt
(8.2)
where a is the acceleration. The analogous electrical relation says that the voltage v necessary to produce a change of current in an inductive circuit is equal to the rate of change of electrokinetic
momentum, i.e,
di
d
v =  Li = L dt
dt
(8.3)
81
In a loosely wound coil of wire such as the one shown in Figure 8.2, the current through the wound
coil produces a denser magnetic field and many of the magnetic lines link the coil several times.
The magnetic flux is denoted as M and, if there are N turns and we assume that the flux M passes
through each turn, the total flux denoted as O is called flux linkage. Then,
O = NM
(8.4)
By definition, a linear inductor one in which the flux linkage is proportional to the current through it,
that is,
O = Li
(8.5)
(8.6)
di
v = L dt
(8.7)
82
Lenzs Law
8.3 Lenzs Law
Heinrich F. E. Lenz was a German scientist who, without knowledge of the work of Faraday and
Henry, duplicated many of their discoveries nearly simultaneously. The law which goes by his name, is
a useful rule for predicting the direction of an induced current. Lenzs law states that:
Whenever there is a change in the amount of magnetic flux linking an electric circuit, an induced voltage
of value directly proportional to the time rate of change of flux linkages is set up tending to produce a current in such a direction as to oppose the change in flux.
To understand Lenzs law, let us consider the transformer shown in Figure 8.3.
M
Here, we assume that the current in the primary winding has the direction shown and it produces the
flux M in the direction shown in Figure 8.3 by the arrow below the dotted line. Suppose that this flux
is decreasing. Then in the secondary winding there will be a voltage induced whose current will be in
a direction to increase the flux. In other words, the current produced by the induced voltage will tend
to prevent any decrease in flux. Conversely, if the flux produced by the primary winding in increasing,
the induced voltage in the secondary will produce a current in a direction which will oppose an
increase in flux.
(8.8)
(8.9)
83
v1
N1
L1
Figure 8.4. Magnetic lines linking a coil
Next, suppose another coil L 2 with N 2 turns is brought near the vicinity of coil L 1 and some lines
of flux are also linking coil L 2 as shown in Figure 8.5.
M21
i1
+
v1
i2 = 0
N1
N2
ML1
L1
L2
It is convenient to express the flux M 11 as the sum of two fluxes M L1 and M 21 , that is,
M 11 = M L1 + M 21
(8.10)
where the linkage flux M L1 is the flux which links coil L 1 only and not coil L 2 , and the mutual flux
M 21 is the flux which links both coils L 1 and L 2 . We have assumed that the linkage and mutual
fluxes M L1 and M 21 link all turns of coil L 1 and the mutual flux M 21 links all turns of coil L 2 .
The arrangement above forms an elementary transformer where coil L 1 is called the primary winding
and coil L 2 the secondary winding.
84
O 2 = N 2 M 21 = M 21 i 1
where M 21 is the mutual inductance (in Henries) and thus the opencircuit secondary winding voltage
v 2 is
dM 21
di
dO
 = M 21 1v 2 = 2 = N 2 dt
dt
dt
(8.12)
In summary, when there is no current in the secondary winding the voltages are
di 1
di 1
v 1 = L 1  and v 2 = M 21 dt
dt
(8.13)
if i 1 z 0 and i 2 = 0
Next, we will consider the case where there is a voltage in the secondary winding producing current
i 2 which in turn produces flux M 22 as shown in Figure 8.6.
i2
+
N2
v2
L2
Figure 8.6. Flux in secondary winding
(8.14)
(8.15)
85
i2
+
i1 = 0
N2
N1
ML2
L1
v2
L2
Following the same procedure as above we express the flux M 22 as the sum of two fluxes M L2 and
M 12 that is,
(8.16)
M 22 = M L2 + M 12
where the linkage flux M L2 is the flux which links coil L 2 only and not coil L 1 , and the mutual flux
M 12 is the flux which links both coils L 2 and L 1 . As before, we have assumed that the linkage and
mutual fluxes link all turns of coil L 2 and the mutual flux links all turns of coil L 1 .
Since there is no current in the primary winding, the flux linkage in the primary winding is
O 1 = N 1 M 12 = M 12 i 2
(8.17)
where M 12 is the mutual inductance (in Henries) and thus the opencircuit primary winding voltage
v 1 is
dM 12
di
dO
 = M 12 2v 1 = 1 = N 1 dt
dt
dt
(8.18)
In summary, when there is no current in the primary winding, the voltages are
di 2
di 2
v 2 = L 2  and v 1 = M 12 dt
dt
(8.19)
if i 1 = 0 and i 2 z 0
86
M 12 = M 21 = M
The last possible arrangement is shown in Figure 8.8 where i 1 z 0 and also i 2 z 0 .
M21
i1
M12
ML2
v1
i2
+
N1
N2
ML1
L1
v2
L2
Figure 8.8. Flux linkages when both primary and secondary currents are present
(8.21)
(8.22)
(8.23)
O 2 = N 2 M 21 + N 2 M 12
(8.24)
and
Differentiating (8.23) and (8.24) and using (8.13), (8.14), (8.19) and (8.20) we get:
di 2
di 1
v 1 = L 1  + M dt
dt
di 1
di 2
v 2 = M  + L 2 dt
dt
(8.25)
87
i2
v1 L1
L2 v2
i1
i2
v1 L1
L2 v2
di 1
v 2 = M dt
for both
circuits
(a)
(b)
Figure 8.9. Arrangements where the mutual voltage has a positive sign
These designations indicate the condition that a current i entering the dotted (undotted) terminal of
one coil induce a voltage across the other coil with positive polarity at the dotted (undotted) terminal
of the other coil. Thus, the mutual voltage term has a positive sign. Following the same rule we see
that in the circuits of Figure 8.10 (a) and 8.10(b) the mutual voltage has a negative sign.
Example 8.1
For the circuit of Figure 8.11 find v 1 and v 2 if
a. i 1 = 50 mA and i 2 = 25 mA
b. i 1 = 0 and i 2 = 20 sin 377t mA
c. i 1 = 15 cos 377t mA and i 2 = 40 sin 377t + 60q mA
88
v1 L1
i2
L2 v2
(a)
i1
i2
v1 L1
L 2 v2
di 1
v 2 = M dt
for both
circuits
(b)
Figure 8.10. Arrangements where the mutual voltage has a negative sign
i1
M = 20 mH
i2
v1 L1
50 mH
L2 v2
50 mH
Solution:
a. Since both currents i 1 and i 2 are constants, their derivatives are zero, i.e.,
di 1
di
 = 2 = 0
dt
dt
and thus
v1 = v2 = 0
b. The dot convention in the circuit of Figure 8.11 shows that the mutual voltage terms are positive
and thus
di 2
di 1
3
v 1 = L 1  + M  = 0.05 u 0 + 20 u 10 u 20 u 377 u cos 377t
dt
dt
= 150.8 cos 377t mV
di 1
di 2
3
v 2 = M  + L 2  = 20 u 10 u 0 + 0.05 u 20 u 377 u cos 377t
dt
dt
= 377 cos 377t mV
89
Example 8.2
For the circuit of Figure 8.12 find the opencircuit voltage v 2 for t ! 0 given that i 1 0 = 0 .
M = 20 mH
i1
R
t = 0
24 V
5:
i2
v1 L1
50 mH
L2 v2
50 mH
Solution:
For t ! 0
di 1
L  + Ri 1 = 24
dt
di 1
0.05  + 5i 1 = 24
dt
di
1 + 100i 1 = 480
dt
Now,
i1 = if + in
810
Rt e L
= Ae
100t
Then,
i 1 = i f + i n = 4.8 + Ae
100t
we get A = 4.8
Therefore,
i 1 = i f + i n = 4.8 4.8 e
100t
L1
L2
i2
L1
L2
We recall that the direction of the flux M can be found by the righthand rule which states that if the
fingers of the right hand encircle a winding in the direction of the current, the thumb indicates the
direction of the flux. Let us place a dot at the upper end of L 1 and assume that the current i 1 enters
the top end thereby producing a flux in the clockwise direction shown. Next, we want the current in
Circuit Analysis II with MATLAB Applications
Orchard Publications
811
be accomplished if the current i 2 in L 2 enters the lower end as shown and thus we place a dot at that
end.
Example 8.3
For the transformer shown in Figure 8.14, find v 1 and v 2 .
M = 2H
i 2 = 5 cos 377 t A
i 1 = 2 sin 377t A
+
L2
L1
v1
3H
4H
+
v2
Solution:
Let us first establish the dot positions as discussed above. The dotted circuit now is as shown in Figure 8.15.
M = 2H
i 1 = 2 sin 377t A
+
v1
i 2 = 5 cos 377 t A
L1
3H
L2
4H
+
v2
Since i 1 enters the dot on the left side and i 2 leaves the dot on the right side, the fluxes oppose each
other. Therefore,
di 2
di 1
v 1 = L 1  M  = 2262 cos 377t 3770 sin 377t V
dt
dt
di 1
di 2
v 2 = M  + L 2  = 1508 cos 377t + 7540 sin 377t V
dt
dt
812
V1 L1
50 mH
L2
V2
100 mH
R LOAD
500 :
Solution:
The dots are given to us as shown. Now, we arbitrarily assign currents I 1 and I 2 as shown in Figure
8.17 and we write mesh equations for each mesh.
M = 50 mH
R1
0.5 :
+
V in = 120 0q
Z = 377 r e s
V1 L1
I1
50 mH
L2
100 mH
V2
I2
R LOAD
500 :
With this current assignments I 2 leaves the dotted terminal of the right mesh and therefore the
mutual voltage has a negative sign. Then,
Mesh 1:
R 1 I 1 + jZL 1 I 1 jZMI 2 = V in
or
0.5 + j18.85 I 1 j18.85I 2 = 120 0q
(8.26)
813
or
(8.27)
We will find the ratio V 2 e V 1 using the MATLAB code below where V 1 = jZL 1 I 1 = j18.85I 1 and
Z=[0.5+18.85j 18.85j; 18.85j 500+37.7j]; V=[120 0]'; I=Z\V;...
fprintf(' \n'); fprintf('V1 = %7.3f V \t', abs(18.85j*I(1))); fprintf('V2 = %7.3f V \t', abs(500*I(2)));...
fprintf('Ratio V2/V1 = %7.3f \t',abs((500*I(2))/(18.85j*I(1))))
V1 = 120.093 V
V2 = 119.753 V
Ratio V2/V1 =
0.997
That is,
V
119.75
2 =  = 0.997
120.09
V1
(8.28)
and thus the magnitude of V LOAD = V 2 is practically the same as the magnitude of V in . However, we
suspect that V LOAD will be out of phase with V in . We can find the phase of V LOAD by adding the following statement to the MATLAB code above.
fprintf('Phase V2= %6.2f deg', angle(500*I(2))*180/pi)
Phase V2=
0.64 deg
This is a very small phase difference from the phase of V in and thus we see that both the magnitude
and phase of V LOAD are essentially the same as that of V in .
If we increase the load resistance R LOAD to 1 K: we will find that again the magnitude and phase of
V LOAD are essentially the same as that of V in . Therefore, the transformer of this example is an isolation transformer, that is, it isolates the load from the source and the value of V in appears across the
load even though the load changes. An isolation transformer is also referred to as a 1:1 transformer.
If in a transformer the secondary winding voltage is considerably higher than the input voltage, the
transformer is referred to as a stepup transformer. Conversely, if the secondary winding voltage is
considerably lower than the input voltage, the transformer is referred to as a stepdown transformer.
814
(8.29)
i1
i1
i2
v1 L1
i2
L2 v2
v1 L1
L2 v2
di 1
v 2 = M dt
for both
circuits
(a)
(b)
(8.30)
and after replacing M with M 12 and M 21 in the appropriate terms, the instantaneous power delivered
to these terminals are:
di 1
di 2
p 1 = v 1 i 1 = L 1  + M 12  i 1
dt
dt
di 1
di 2
p 2 = v 2 i 2 = M 21  + L 2  i 2
dt
dt
(8.31)
Now, let us suppose that at some reference time t 0 , both currents i 1 and i 2 are zero, that is,
i1 t0 = i2 t0 = 0
(8.32)
(8.33)
Next, let us assume that at time t 1 , the current i 1 is increased to some finite value, while i 2 is still
zero. In other words, we let
i1 t1 = I1
(8.34)
i2 t1 = 0
(8.35)
and
815
t1
(8.36)
p 1 + p 2 dt
and since i 2 t 1 = 0 , then p 2 t 1 = 0 and also di 2 e dt = 0 . Therefore, from (8.31) and (8.36) we
get
W1 =
t1
di 1
2
1
L 1 i 1  dt = L 1 i 1 di 1 =  L 1 I 1
dt
2
t0
t0
t1
(8.37)
Finally, let us at some later time t 2 , maintain i 1 at its previous value, and increase i 2 to a finite value,
that is, we let
i1 t2 = I1
(8.38)
i2 t2 = I2
(8.39)
and
During this time interval, di 1 e dt = 0 and using (8.31) the energy accumulated is
W2 =
t2
p 1 + p 2 dt =
t2
di
di
M I 2 + L i 2 dt
12
1
2
2
dt
dt
t1
(8.40)
t2
2
1
=
M 12 I 1 + L 2 i 2 di 2 = M 12 I 1 I 2 +  L I 2
2 2
t1
Therefore, the energy stored in the transformer from t 0 to t 2 is from (8.37) and (8.40),
t2
2
2
1
1
W t =  L 1 I 1 + M 12 I 1 I 2 +  L I 2
2
2
0
2
(8.41)
Now, let us reverse the order in which we increase i 1 and i 2 . That is, in the time interval t 0 d t d t 1 ,
we increase i 2 so that i 2 t 1 = I 2 while keeping i 1 = 0 . Then, at t = t 2 , we keep i 2 = I 2 while we
increase i 1 so that i 1 t 2 = I 1 . Using the same steps in equations (8.33) through (8.40), we get
t2
2
2
1
1
W t =  L 1 I 1 + M 21 I 1 I 2 +  L I 2
2
2 2
0
(8.42)
Since relations (8.41) and (8.42) represent the same energy, we must have
M 12 = M 21 = M
816
(8.43)
(8.44)
Relation (8.44) was derived with the dot markings of Figure 8.18 which is repeated below as Figure
8.19 for convenience.
i1
i2
v1 L1
L2 v2
i1
i2
v1 L1
L2 v2
di 1
v 2 = M dt
for both
circuits
(a)
(b)
However, if we repeat the above procedure for dot markings of the circuit of Figure 8.20 we will find
that
i1
i2
v1 L1
L2 v2
i1
i2
v1 L1
L2 v2
di 1
v 2 = M dt
for both
circuits
(a)
(b)
Figure 8.20. Transformer circuits with different dot arrangement from Figure 8.19
t2
2
2
1
1
W t =  L 1 I 1 M I 1 I 2 +  L I 2
2
2
0
2
(8.45)
(8.46)
where the sign of M is positive if both currents enter the dotted (or undotted) terminals, and it is
negative if one current enters the dotted (or undotted) terminal while the other enters the undotted
(or dotted) terminal.
817
(8.47)
Obviously, the energy on the left side of (8.47) cannot be negative for any values of i 1 , i 2 , L 1 , L 2 , or
M . Let us assume first that i 1 and i 2 are either both positive or both negative in which case their
product is positive. Then, from (8.47) we see that the energy would be negative if
t2
1 2 1 2
W t =  L 1 i 1 +  L i 2 Mi 1 i 2
2
2 2
0
(8.48)
and the magnitude of the Mi 1 i 2 is greater than the sum of the other two terms on the right side of
that expression. To derive an expression relating the mutual inductance M to the selfinductances L 1
and L 2 , we add and subtract the term
(8.49)
We now observe that the first term on the right side of (8.49) could be very small and could approach
zero, but it can never be negative. Therefore, for the energy to be positive, the second and third terms
on the right side of (8.48) must be such that L 1 L 2 t M or
M d L1 L2
(8.50)
Expression (8.50) indicates that the mutual inductance can never be larger than the geometric mean
of the inductances of the two coils between which the mutual inductance exists.
Note: The inequality in (8.49) was derived with the assumption that i 1 and i 2 have the same algebraic sign. If their signs are opposite, we select the positive sign of (8.47) and we find that (8.50) holds
also for this case.
The ratio M e L 1 L 2 is known as the coefficient of coupling and it is denoted with the letter k , that is,
M
k = L1 L2
818
(8.51)
M = 20 mH
i2
v1 L1
L2 v2
50 mH
50 mH
Solution:
Since the currents enter the dotted terminals, we use (8.45) with the plus (+) sign for the mutual
inductance term, that is,
1 2
1 2
W t =  L 1 i 1 + Mi 1 i 2 +  L i 2
2 2
2
(8.52)
Then,
a.
W t = 0 = 0.5 u 50 u 10
3 2
u 50 u 10 + 20 u 10
+ 0.5 u 50 u 10
3 2
u 50 u 10
u 25 u 10 = 103 u 10
u 25 u 10
J = 103 PJ
b.
Since i 1 = 0 and i 2 = 20 sin 377t
t=0
= 0 , it follows that
819
c.
W t = 0 = 0.5 u 50 u 10
3 2
u 15 u 10 + 20 u 10
+ 0.5 u 50 u 10
u 40 u 10
u 15 u 10
q
u 40 u 10
u sin 60 = 46 u 10
u sin 60
J = 46 PJ
R2
vin
i1
L1
L2
v1
v2
V0
i2
vout
RLOAD
(DC)
Application of KVL around the primary and secondary circuits yields the loop equations
di 2
di 1
v in = R 1 i 1 + L 1  M dt
dt
di 1
di 1
0 = M  + L 2  + R 2 + R LOAD
dt
dt
(8.53)
and we see that the instantaneous values of the voltages and the currents are not affected by the presence of the DC voltage source V 0 since we would have obtained the same equations had we let
V0 = 0 .
820
R1
t=0
vin
200 :
100 :
L2
L1
24 V DC
R2
3H
1 K:
5H
I1
vout
RLOAD
I2
Solution:
The total response consists of the summation of the forced and natural responses, that is,
(8.54)
i 2T = i 2f + i 2n
and since the applied voltage is constant (DC), no steadystate (forced) voltage is produced in the secondary and thus i 2f = 0 .
For t ! 0 the s domain circuit is shown in Figure 8.24.
200
100
2s
v in s
24 e s
3s
I1 s
v out s
5s
I2 s
1000
Figure 8.24. The s domain circuit for the transformer of Example 8.6
(8.55)
821
or
4.36
I 2 s =  s + 340.71 s + 32.02
(8.56)
4.36
= 0.01
r 1 = s + 32.02 s = 340.71
(8.57)
4.36
r 2 = = 0.01
s + 340.71 s = 32.02
(8.58)
from which
(8.59)
32.02t
340.71t
(8.60)
Using the following MATLAB code we get the plot shown on Figure 8.25.
t=0: 0.001: 0.2; i2n=0.01.*(exp(32.02*t)exp(340.71.*t)); plot(t,i2n); grid
Example 8.7
For the transformer of Figure 8.26, find the steadystate (forced) response of v out .
Solution:
The s domain equivalent circuit is shown in Figure 8.27.
We could use the same procedure as in the previous example, but it is easier to work with the transfer
function G s .
822
Figure 8.25. Plot for the secondary current of the transformer of Example 8.6
2H
10 :
v in
5H
3H
v out
100 :
0.1 F
Figure 8.26. Circuit for Example 8.7
2s
10
V in s
3s
5s
170 0q V
I1 s
100
V out s
I2 s
1 e 0.1s
823
(8.61)
V in s
2s + 1 e 0.1s
0
I 2 s =  3s + 10 + 1 e 0.1s
2s + 1 e 0.1s
2s + 1 e 0.1s 5s + 100 + 1 e 0.1s
or
2
2s + 10 V in s
2s + 10 e s V in s
 = I 2 s = 2
3
2
11s + 350s + 1040 + 1100 e s
11s + 350s + 1040s + 1100
2
0.18s + 0.91 V in s
= 3
2
s + 31.82s + 94.55s + 100
0.18s + 0.91 V in s
18s + 91 V in s
V out s = 100 I 2 s = 100  = 3
2
3
2
s + 31.82s + 94.55s + 100
s + 31.82s + 94.55s + 100
(8.62)
and
2
V out s
18s + 91
 = G s = 3
2
V in s
s + 31.82s + 94.55s + 100
(8.63)
and thus
V in s = V in jZ = 170 0q
2.56 u 10 + 91
4.35 u 10 0q
V out jZ =  170 0q = 7
6
4
6
7
j 5.36 u 10 4.52 u 10 + j3.56 u 10 + 100
4.52 u 10 j5.36 u 10
or
824
(8.64)
The expression of (8.65) indicates that the transformer of this example is a stepdown transformer.
VS
L1
V1
L2
V2
I1
I2
Z LOAD
V LOAD
(8.66)
jZL 1 I 1 V S
I 2 = jZM
(8.67)
(8.68)
jZMI 1
I 2 =  jZL 2 + Z LOAD
(8.69)
or
and
or
825
(8.70)
jZM 2
V S = jZL 1  I 1
jZL + Z
(8.71)
LOAD
(8.72)
The first term on the right side of (8.72) represents the reactance of the primary. The second term is
a result of the mutual coupling and it is referred to as the reflected impedance. It is denoted as Z R , i.e.,
Z2 M 2 Z R = jZL 2 + Z LOAD
(8.73)
(8.74)
To express (8.74) as the sum of a real and an imaginary component, we multiply both numerator
and denominator by the complex conjugate of the denominator. Then,
Z 2 M 2 X LOAD + ZL 2
Z 2 M 2 R LOAD
Z R =  j 2
2
2
2
R LOAD + X LOAD + ZL 2
R LOAD + X LOAD + ZL 2
(8.75)
The imaginary part of (8.75) represents the reflected reactance and we see that it is negative. That
is, the reflected reactance is opposite to that of the net reactance X LOAD + ZL 2 of the secondary.
826
(8.76)
Example 8.8
In the transformer circuit of Figure 8.29, Z S represents the internal impedance of the voltage source
VS .
Find:
a. Z in
b. I 1
c. I 2
d. V 1
e. V 2
100 mH
2:
ZS
VS
L1
V1
I1
200 mH
Z = 377 r e s
V S = 120 0q
L2
V2
300 mH
I2
Z LOAD
V LOAD
7540
Z LOAD = 10 j  :
Z
827
and we must add Z s = 2 : to it. Therefore, for the transformer of this example,
142129 u 0.01 + 2
Z 2 M 2  + 2 = j75.4 + Z in = jZL 1 + j113.1 + 10 j20
jZL + Z
2
LOAD
b.
V
120 0q  = 1.98 86.56q A
I 1 = S = 60.42 86.56q :
Z in
c. By KVL
jZMI 1 + jZL 2 + Z LOAD I 2 = 0
or
jZM
j37.7
74.88 3.04q
I 2 =  I 1 =  1.98 86.56q =  = 0.8 80.83 q A
jZL 2 + Z LOAD
j113.1 + 10 j20
93.64 83.87q
d.
V 1 = jZL 1 I 1 jZM I 2 = 75.4 90q u 1.98 86.56q 37.7 90q u 0.8 80.83 q
= 149.29 3.04q 30.15 9.17q = 149.08 + j7.92 30.15 j4.8 = 118.9 1.5 q V
e.
V 2 = Z LOAD I 2 = 10 j20 0.8 80.83 q = 22.36 63.43q u 0.8 80.83 q = 17.89 144.26qV
828
(8.77)
The flux produced in a winding of a transformer due to a current in that winding is proportional to
the product of the current and the number of turns on the winding. Therefore, letting D be a constant of proportionality which depends on the physical properties of the transformer, for the primary
and secondary windings we have:
I 11 = DN 1 i 1
I 22 = DN 2 i 2
(8.78)
The constant D is the same for the primary and secondary windings because we have assumed that
the same flux links both coils and thus both flux paths are identical. We recall from (8.8) and (8.14)
that
O 1 = N 1 M 11 = L 1 i 1
O 2 = N 2 M 22 = L 2 i 2
(8.79)
N 1 M 11 = L 1 i 1 = DN 1 i 1
2
(8.80)
N 2 M 22 = L 2 i 2 = DN 2 i 2
or
2
L 1 = DN 1
L2 =
2
DN 2
(8.81)
Therefore,
2
L2
N 2
2
 =  = a
N
L1
1
(8.82)
jZMI 1
I 2 =  jZL 2 + Z LOAD
(8.83)
I
jZM
2 = I1
jZL 2 + Z LOAD
(8.84)
From (8.69),
or
829
(8.85)
M
k =  = 1
L1 L2
(8.86)
or
M =
(8.87)
L1 L2
L1
L2
(8.88)
(8.89)
N1 I1 = N2 I2
(8.90)
V 2 = Z LOAD I 2
(8.92)
2 2
V
Z M
Z in = s = jZL 1 + I1
jZL 2 + Z LOAD
(8.93)
From (8.72),
830
M = L1 L2
Z L1 L2
Z in = jZL 1 + jZL 2 + Z LOAD
(8.94)
(8.95)
L2 = a L1
Z a L1
Z in = jZL 1 + 2
jZa L 1 + Z LOAD
(8.96)
and if we let jZL 1 o f , both terms on the right side of (8.96) become infinite and we get an indeterminate result. To work around this problem, we combine these terms and we get:
2 2 2
2 2 2
Z a L 1 + jZL 1 Z LOAD + Z a L 1
jZL 1 Z LOAD
 = Z in = 2
2
jZa L 1 + Z LOAD
jZa L 1 + Z LOAD
and as jZL 1 o f ,
Z LOAD
Z in = 2
a
(8.97)
(8.98)
(8.99)
V2
 = a
V1
(8.100)
or
831
V2 I2 = V1 I1
that is, the voltamperes of the secondary and the primary are equal.
An ideal transformer is represented by the network of Figure 8.30.
i1
1:a
i2
v1 L1
L 2 v2
v1 N1
1:a
i2
N2 v2
Speaker
832
(8.102)
or
N
a = 2 =
N1
Z LOAD
 =
Z in
8  =
80000
11  = 100
10000
or
N1
 = 100
N2
(8.103)
that is, the number of turns in the primary must be 100 times the number of the turns in the secondary.
ZS
VS
I1
L1
L2
V1
V2
I2
Z LOAD
V LOAD
From (8.97)
Z LOAD
Z in = 2
a
The input impedance seen by the voltage source V S in the circuit of Figure 8.32 is
Z LOAD
Z in = Z S + a2
(8.104)
and thus the circuit of Figure 8.32 can be replaced with the simplified circuit shown in Figure 8.33.
833
VS
Z LOAD
 V 1 = V 2 e a
a2
I 1 = aI 2
The voltages and currents can now be found from the simple series circuit if Figure 8.33.
ZS
VS
I1
L1
L2
V1
V2
I2
Z LOAD
V LOAD
u
y
Figure 8.34. Circuit for the derivation of Thevenins equivalent
If we open the circuit at points x and y as shown in Figure 8.34, we find the Thevenin voltage as
V TH = V OC = V xy . Since the secondary is now an open circuit, we have I 2 = 0 , and also I 1 = 0
because I 1 = aI 2 . Since no voltage appears across Z S , V 1 = V S and V 2 oc = aV 1 = aV S . Then,
V TH = V OC = V xy = aV S
(8.105)
(8.106)
834
(8.107)
The Thevenin equivalent circuit with the load connected to it is shown in Figure 8.35.
x
u
a2 ZS
aV S
V 2 = aV 1
Z LOAD
I2 = I1 e a
u
y
Figure 8.35. The Thevenin equivalent of the transformer circuit in Figure 8.34
The circuit of Figure 8.35 was derived with the assumption that the dots are placed as shown in Figure 8.34. If either dot is reversed, we simply replace a by a .
Example 8.9
For the circuit of Figure 8.36, find V 2 .
Solution:
We will replace the given circuit with its Thevenin equivalent. First, we observe that the dot in the
secondary has been reversed, and therefore we will replace a by a . The Thevenin equivalent is
I1
10 :
VS
8 0q V
I2
1:10
0.01V 2
L1
L2
V 2 60 + j80 :
835
0.001 V 2
60 + j80 :
V2
or
10
V 2 =  225q = 5 2 135q
2
Other equivalent circuits can be developed from the equations of the primary and secondary voltages
and currents.
Consider, for example the linear transformer circuit of Figure 8.38.
From (8.30), the primary and secondary voltages and currents are:
di 2
di 1
v 1 = L 1  + M dt
dt
di 1
di 2
v 2 = M  + L 2 dt
dt
836
(8.108)
i1
v1 L1
L2 v2
and these equations are satisfied by the equivalent circuit shown in Figure 8.39.
i1
i2
L1
di 2
M dt
v1
L2
di 1
M dt
v2
(8.109)
di 2
di
di
v 2 = M 1 + 2 + L 2 M  dt
dt
dt
i2
L2 M
L1 M
v1
v2
837
current.
x The magnetic field loops are circular in form and are called lines of magnetic flux.
x The magnetic flux is denoted as M and the unit of magnetic flux is the weber (Wb).
x If there are N turns and we assume that the flux M passes through each turn, the total flux
denoted as O is called flux linkage. Then,
O = NM
x A linear inductor one in which the flux linkage is proportional to the current through it, that is,
O = Li
where the constant of proportionality L is called inductance in webers per ampere.
x Faradays law of electromagnetic induction states that
v = dO
dt
x Lenzs law states that whenever there is a change in the amount of magnetic flux linking an electric
circuit, an induced voltage of value directly proportional to the time rate of change of flux linkages
is set up tending to produce a current in such a direction as to oppose the change in flux.
x A linear transformer is a fourterminal device in which the voltages and currents in the primary
di 1
di 1
v 1 = L 1  and v 2 = M 21 dt
dt
if i 1 z 0 and i 2 = 0
x In a linear transformer, when there is no current in the primary winding, the voltages are
di 2
di 2
v 2 = L 2  and v 1 = M 12 dt
dt
if i 1 = 0 and i 2 z 0
x In a linear transformer, when there is a current in both the primary and secondary windings, the
voltages are
838
Summary
di 2
di 1
v 1 = L 1  + M dt
dt
di 1
di 2
v 2 = M  + L 2 dt
dt
x The voltage terms
di 1
di 2
L 1  and L 2 dt
dt
dotted (undotted) terminal of one coil induces a voltage across the other coil with positive polarity
at the dotted (undotted) terminal of the other coil, the mutual voltage term has a positive sign. If a
current i entering the undotted (dotted) terminal of one coil induces a voltage across the other coil
with positive polarity at the dotted (undotted) terminal of the other coil, the mutual voltage term
has a negative sign.
x If the polarity (dot) markings are not given, they can be established by using the righthand rule
which states that if the fingers of the right hand encircle a winding in the direction of the current,
the thumb indicates the direction of the flux. Thus, in an ideal transformer with primary and secondary windings L 1 and L 2 and currents i 1 and i 2 respectively, we place a dot at the upper end of
L 1 and assume that the current i 1 enters the top end thereby producing a flux in the clockwise
direction. Next, we want the current in L 2 to enter the end which will produce a flux in the same
where the sign of M is positive if both currents enter the dotted (or undotted) terminals, and it is
negative if one current enters the dotted (or undotted) terminal while the other enters the undotted
(or dotted) terminal.
x The ratio
M k = L1 L2
839
mary and secondary inductive reactances are very large in comparison with the load impedances.
The primary and secondary coils have many turns wound around a laminated ironcore and are
arranged so that the entire flux links all the turns of both coils.
x In an ideal transformer number of turns on the primary N 1 and the number of turns on the secondary N 2 are related to the primary and secondary currents I 1 and I 2 respectively as
N1 I1 = N2 I2
x An important parameter of an ideal transformer is the turns ratio a which is defined as the ratio of
the number of turns on the secondary, N 2 , to the number of turns of the primary N 1 , that is,
N
a = 2N1
x In an ideal transformer the turns ratio a relates the primary and secondary currents as
I2
 = 1
a
I1
x In an ideal transformer the turns ratio a relates the primary and secondary voltages as
V2
 = a
V1
840
Summary
x In an ideal transformer the voltamperes of the primary and the secondary are equal, that is,
V2 I2 = V1 I1
x An ideal transformer can be used as an impedance matching device by specifying the appropriate
turns ratio N 2 e N 1 = a . Then,
Z LOAD
Z in = a2
x In analyzing networks containing ideal transformers, it is very convenient to replace the trans
former by an equivalent circuit before the analysis. One method is presented in Section 8.11.
x An ideal transformer can be replaced by a Thevenin equivalent as discussed in Section 8.12.
841
L2
L1
2:
v2
2H
1H
i = 4u 0 t A
2. For the transformer circuit of Figure 8.42, find the phasor currents I 1 and I 2 .
M = j1 :
1:
2:
j1 :
10 0q V
j8 :
I1
j10 :
I2
3. For the network of Figure 8.43, find the transfer function G s = V OUT s e V IN s .
1:
0.5 H
1H
1H
1:
0.5 H
V IN s
+
0.5 H
1H
1:
V OUT s
842
Exercises
4. For the transformer of Figure 8.44, find the average power delivered to the 4 : resistor.
8:
2:
1:2
4:
v S = 4 cos 3t V
Figure 8.44. Circuit for Exercise 4
5. Replace the transformer of Figure 8.45 by a Thevenin equivalent and then compute V 1 V 2 I 1
and I 2
I1
I2
1:5
2 + j3 :
V1
V2
12 0q
100 j75 :
6. For the circuit of Figure 8.46, compute the turns ratio a so that maximum power will be delivered
to the 10 K: resistor.
4:
1:a
10 K:
12 0q V
Figure 8.46. Circuit for Exercise 6
843
M = 1H
2:
L2
L1
2H
1H
i = 4u 0 t A
2:
v2
i1
L1
1H
v IN = 8u 0 t V
L2
2H
v2
t ! 0 (1)
The total solution of i 1 is the sum of the forced component i 1f and the natural response i 1n , i.e.,
i 1 = i 1f + i 1n
From (1) we find that i 1f = 8 e 2 = 4 and i 1n is found from the characteristic equation s + 2 = 0
from which s = 2 and thus i 1n = Ae
2t
. Then,
i 1 = 4 + Ae
2t
(2)
Since we are not told otherwise, we will assume that i 1 0 = 0 and from (2) 0 = 4 + Ae or
A = 4 and by substitution into (2)
i 1 = 4 1 4e
2t
and since i 2 = 0 ,
di 1
2t
2t
d
v 2 = 1  =  > 4 1 4e @ = 8e
V
dt
dt
844
Solutions to Exercises
2.
M = j1 :
1:
2:
10 0q V
j1 :
j8 :
I1
I2
j10 :
By Cramers rule,
I1 = D1 e '
I2 = D2 e '
where
' = 1 + j1 j1
= 5
j1 2 j2
= 20 1 j
D 1 = 10 0q j1
0
2 j2
D 2 = 1 + j1 10 0q = j10
j1
0
Thus,
20 1 j
I 1 =  = 4 1 j = 4 2 45q A
5
 = j2 = 2 90q A
I 2 = j10
5
magI1 =
magI2 =
5.66 A
2.00 A
845
1
I2
V IN s
0.5s
I1
+
0.5s
s
I3
V OUT s
We will find V OUT s from V OUT s = 1 : I 3 . The three mesh equations in matrix form are:
s + 1 0.5s 0.5s
1
=
0.5s s + 1 0.5s
0 V IN s
0.5s 0.5s s + 1
0
det_delta =
9/4*s^2+3*s+1
d3=[s+1 0.5*s 0.5*s; 0.5*s s+1 0.5*s; 1 0 0]; det_d3=det(d3)
det_d3 =
3/4*s^2+1/2*s
I3=det_d3/det_delta
I3 =
(3/4*s^2+1/2*s)/(9/4*s^2+3*s+1)
simplify(I3)
ans =
s/(3*s+2)
Therefore,
V OUT s = 1 I 3 V IN s = s e 3s + 2 V IN s
and
G s = V OUT s e V IN s = s e 3s + 2
846
Solutions to Exercises
4.
8:
2:
a = 2
A
1:2
I1
I2
V1
V2
4 0q
4:
I4 :
2
1
For this exercise, P ave 4 : =  I 4: 4 and thus we need to find I 4 : .
At Node A ,
V 2 4 0q
V
 I2 = 0
2 + 8
4
3V 2
1
 I 2 =  (1)
2
8
and
1 3 2
9
P ave 4 : =   4 =  w
24
8
847
1:5
2 + j3 :
VS
12 0q
I2
x
u
V1
V2
x
u
a2 ZS
aV S
100 j75 :
V 2 = aV 1
Z LOAD
I2 = I1 e a
u
y
u
y
j75
a Z S + Z LOAD
and
2
V 2 = Z LOAD I 2 = 125 36.87q u  180q = 50 143.13q V
5
6.
1:a
4:
10 K:
12 0q V
From (8.102)
Z LOAD
Z in = a2
Then,
Z LOAD
 = 2500
a 2 =  = 10000
4
Z in
or
a = 50
848
Chapter 9
One and Twoport Networks
his chapter begins with the general principles of one and twoport networks. The z , y , h , and
g parameters are defined. Several examples are presented to illustrate their use. It concludes
with a discussion on reciprocal and symmetrical networks.
iin 3 :
3:
Ix
12 V
iout
6:
Figures 9.2 and 9.3 show two examples of practical oneport networks.
7:
20Ix
10 :
4:
5:
+ IL
VL
RL
8:
91
iin
8:
2:
+
120 V
4:
6:
iout
10 :
16 :
20 :
A twoport network has two pairs of terminals, that is, four terminals as shown in Figure 9.4 where
i 1 = i 3 and i 2 = i 4
i1 +
i3
i2
i4
(9.1)
}}}}}}}}}}}}}}}
Z n1 i 1 + Z n2 i 2 + Z n3 i 3 + } + Z nn i n = v n
In (9.1) each current can be found by Cramers rule. For instance, the current i 1 is found by
D
i 1 = 1'
(9.2)
where
92
(9.3)
} } } } }
Z n1 Z n2 Z n3 } Z nn
V 1 Z 12 Z 13 } Z 1n
V 2 Z 22 Z 23 } Z 2n
D1 = V Z Z } Z
3 32 33
3n
(9.4)
} } } } }
V n Z n2 Z n3 } Z nn
Next, we recall that the value of the determinant of a matrix A is the sum of the products obtained by
multiplying each element of any row or column by its cofactor*. The cofactor, with the proper sign, is
the matrix that remains when both the row and the column containing the element are eliminated.
The sign is plus (+) when the sum of the subscripts is even, and it is minus () when it is odd. Mathematically, if the cofactor of the element a qr is denoted as A qr , then
A qr = 1
q+r
M qr
(9.5)
where M qr is the minor of the element a qr . We recall also that the minor is the cofactor without a
sign.
Example 9.1
Compute the determinant of A from the elements of the first row and their cofactors given that
A =
1 2 3
2 4 2
1 2 6
Solution:
detA = 1 4 2 2 2 2 3 2 4 = 1 u 20 2 u 10 3 u 0 = 40
2 6
1 6
1 2
93
Z 32 Z 33 } Z 3n
(9.6)
} } } }
Z n2 Z n3 } Z nn
Z 21 Z 23 } Z 2n
C 12 =
Z 31 Z 33 } Z 3n
(9.7)
} } } }
Z n1 Z n3 } Z nn
Z 12 Z 13 } Z 1n
C 21 =
Z 32 Z 33 } Z 3n
(9.8)
} } } }
Z n2 Z n3 } Z nn
(9.9)
C 12 v 1 C 22 v 2 C 32 v 3
C n2 v n
D
 +  +  + } + i 2 = 2 = '
'
'
'
'
(9.10)
Also,
C 21
y 12 = '
C 31
y 13 = '
(9.11)
C 12
y 21 = '
C 22
y 22 = '
C 32
y 23 = '
(9.12)
Likewise,
and so on. By substitution of the y parameters into (9.9) and (9.10) we get:
94
(9.13)
i 2 = y 21 v 1 + y 22 v 2 + y 23 v 3 + } + y 2n v n
(9.14)
If the subscripts of the y parameters are alike, such as y 11 , y 22 and so on, they are referred to as drivingpoint admittances. If they are unlike, such as y 12 , y 21 and so on, they are referred to as transfer
admittances.
R1
R2
i1
i2
the equations of (9.13) and (9.14) will have only two terms each, that is,
i 1 = y 11 v 1 + y 12 v 2
(9.15)
i 2 = y 21 v 1 + y 22 v 2
(9.16)
From Figure 9.5 we observe that there is only one voltage source, v 1 ; there is no voltage source in
Loop 2 and thus v 2 = 0 . Then, (9.15) and (9.16) reduce to
i 1 = y 11 v 1
(9.17)
i 2 = y 21 v 1
(9.18)
Relation (9.17) reveals that the drivingpoint admittance y 11 is the ratio i 1 e v 1 . That is, the drivingpoint admittance, as defined by (9.17), is the admittance seen by a voltage source that is present in the
respective loop, in this case, Loop 1. Stated in other words, the drivingpoint admittance is the ratio of
the current in a given loop to the voltage source in that loop when there are no voltage sources in any other
loops of the network.
Transfer admittance is the ratio of the current in some other loop to the driving voltage source, in this
case v 1 . As indicated in (9.18), the transfer admittance y 21 is the ratio of the current in Loop 2 to the
voltage source in Loop 1.
95
v1
R1
R3
4:
12 :
R2
6:
24 V
Figure 9.6. Circuit for Example 9.2
Solution:
We assign currents as shown in Figure 9.7.
v1
R1
R3
4:
12 :
24 V
R2
6:
i1
i2
(9.19)
6i 1 + 18i 2 = 0
(9.20)
(9.21)
96
(9.22)
The cofactor C 11 is obtained by inspection from the matrix of (9.22), that is, eliminating the first row
and first column we are left with 18 and thus C 11 = 18 . Similarly, the cofactor C 12 is found by eliminating the first row and second column and changing the sign of 6 . Then, C 12 = 6 . By substitution
into (9.20) and (9.21), we obtain
C 11
18
1
 =  = y 11 = '
144
8
(9.23)
C 12
6
1
 =  = y 21 = '
144
24
(9.24)
and
(9.25)
1
i 2 = y 21 v 1 =  u 24 = 1 A
24
(9.26)
Finally, the we observe that the current through the 4 : resistor is 3 A , through the 12 : is 1 A
and through the 6 : is i 1 i 2 = 3 1 = 2A
Of course, there are other simpler methods of computing these currents. However, the intent here
was to illustrate how the drivingpoint and transfer admittances are applied. These allow easy computation for complicated network problems.
(9.27)
Y n1 v 1 + Y n2 v 2 + Y n3 v 3 + } + Y nn v n = i n
97
(9.28)
where
Y 11 Y 12 Y 13 } Y 1n
Y 21 Y 22 Y 23 } Y 2n
' = Y Y Y }Y
31 32 33
3n
(9.29)
} } } } }
Y n1 Y n2 Y n3 } Y nn
V 1 Y 12 Y 13 } Y 1n
V 2 Y 22 Y 23 } Y 2n
(9.30)
D1 = V Y Y } Y
3 32 33
3n
} } } } }
V n Y n2 Y n3 } Y nn
As in the previous section, we find that the nodal equations of (9.27) can be expressed as
v 1 = z 11 i 1 + z 12 i 2 + z 13 i 3 + } + z 1n i n
(9.31)
v 2 = z 21 i 1 + z 22 i 2 + z 23 i 3 + } + z 2n i n
(9.32)
v 3 = z 31 i 1 + z 32 i 2 + z 33 i 3 + } + z 3n i n
(9.33)
C 21
z 12 = '
C 31
z 13 = '
(9.34)
C 12
z 21 = '
C 22
z 22 = '
C 32
z 23 = '
(9.35)
C 13
z 31 = '
C 23
z 32 = '
C 33
z 33 = '
(9.36)
and so on. The matrices C ij represent the cofactors as in the previous section.
98
v1
1
(9.37)
G3
v2
2
i1
G2
vS
v0
Figure 9.8. Circuit to illustrate the definitions of drivingpoint and transfer impedances.
The transfer impedance between nodes 2 and 1 is the ratio of the voltage v 2 to the current at node
1 when there are no other current (or voltage) sources in the network. That is,
v
z 21 = 2i1
(9.38)
Example 9.3
For the network of Figure 9.9, compute the drivingpoint and transfer impedances and the voltages
across each conductance in terms of the current source.
i1
2:
10 :
1:
1:
1:
1:
99
i1
10
1
3 v3
v2 2
1
v0 0
Figure 9.10. Node assignment for network of Example 9.3
(9.39)
1 v 3 v 1 + 1 v 3 v 2 + 1v 3 = 0
(9.40)
v 1 v 2 + 3v 3 = 0
(9.41)
and the transfer impedances z 21 and z 31 from (9.35) and (9.36), that is,
C 12
z 21 = '
(9.42)
C 13
z 31 = '
(9.43)
910
(9.44)
The cofactor C 11 is
4 1 = 12 1 = 11
1 3
C 11 =
(9.45)
(9.46)
C 13 = 2 4 = 2 + 4 = 6
1 1
(9.47)
and
(9.48)
C 12
7
z 21 =  = '
123
(9.49)
C 13
6
z 31 =  = '
123
(9.50)
(9.51)
7
v 2 = z 21 i 1 + z 22 i 2 + z 23 i 3 =  i 1
123
(9.52)
6
v 3 = z 31 i 1 + z 32 i 2 + z 33 i 3 =  i 1
123
(9.53)
Of course, there are other simpler methods of computing these voltages. However, the intent here
was to illustrate how the drivingpoint and transfer impedances are applied. These allow easy computation for complicated network problems.
Circuit Analysis II with MATLAB Applications
Orchard Publications
911
i1
Linear network
(Consists of linear
passive devices and
i3 possibly dependent
sources but no
independent sources
i2 +
v2
i4
Here, we assume that i 1 = i 3 and i 2 = i 4 . We also assume that i 1 and i 2 are obtained by the superposition of the currents produced by both v 1 and v 2 .
Now, we will define the y , z , h , and g parameters.
9.4.1 The y Parameters
The twoport network of Figure 9.11 can be described by the following set of equations.
i 1 = y 11 v 1 + y 12 v 2
(9.54)
i 2 = y 21 v 1 + y 22 v 2
(9.55)
(9.56)
i
y 11 = 1v1
(9.57)
or
and y 11 is referred to as the short circuit input admittance at the left port when the right port of Figure 9.11 is shortcircuited.
Let us again consider (9.54), that is,
i 1 = y 11 v 1 + y 12 v 2
(9.58)
912
(9.59)
TwoPort Networks
or
i
y 12 = 1v2
(9.60)
and y 12 is referred to as the short circuit transfer admittance when the left port of Figure 9.11 is shortcircuited. It represents the transmission from the right to the left port. For instance, in amplifiers
where the left port is considered to be the input port and the right to be the output, the parameter
y 12 represents the internal feedback inside the network.
Similar expressions are obtained when we consider the equation for i 2 , that is,
(9.61)
i 2 = y 21 v 1 + y 22 v 2
In an amplifier, the parameter y 21 is also referred to as the short circuit transfer admittance and represents transmission from the left (input) port to the right (output) port. It is a measure of the socalled forward gain.
The parameter y 22 is called the short circuit output admittance.
The y parameters and the conditions under which they are computed are shown in Figures 9.12
through 9.16.
v1
+
i1
i3
i2
i4
+
+ v2
i 1 = y 11 v 1 + y 12 v 2
i 2 = y 21 v 1 + y 22 v 2
+
v1
i1
i3
i2
i4
i
y 11 = 1v1
v2=0
v2 = 0
913
v1=0
i2
+
i4 v
2
i
y 12 = 1v2
v1 = 0
+
v1
i1
i3
i2
i4
i
y 21 = 2v1
v2=0
v2 = 0
v1=0
i2
+
i4 v
2
i1
i3
i
y 22 = 2v2
v1 = 0
Example 9.4
For the network of Figure 9.17, find the y parameters.
Solution:
a. The short circuit input admittance y 11 is found from the network of Figure 9.18 where we have
assumed that v 1 = 1 V and the resistances, for convenience, have been replaced with conductances in mhos.
10 :
5:
20 :
914
TwoPort Networks
i1
+
0.1 :
0.2 :
v2 = 0
0.05 :
v1 = 1 V
and 0.1 :
conductances. Then,
(9.62)
b. The short circuit transfer admittance y 12 when the left port is shortcircuited, is found from the
network of Figure 9.19.
i1
v1 = 0
0.1 :
0.2 :
0.05 :
+
v2 = 1 V
conductance is in
conductance. The current i 1 , with a minus () sign, now flows through
conductance. Then,
i 1 = 0.1v 2 = 0.1 u 1 = 0.1 A
and
y 12 = i 1 e v 2 = 0.1 e 1 = 0.1 :
(9.63)
915
0.1 :
+
0.2 :
i2
v1 = 1 V
v2 = 0
0.05 :
conductance is in
conductance. The current i 2 , with a minus () sign, now flows through
conductance. Then,
i 2 = 0.1v 1 = 0.1 u 1 = 0.1 A
and
y 21 = i 2 e v 1 = 0.1 e 1 = 0.1 :
(9.64)
d. The short circuit output admittance y 22 at the right port when the left port is shortcircuited, is
found from the network of 9.21.
0.1 :
v1 = 0
0.2 :
i2
0.05 :
+
v2 = 1 V
conductance is shorted out and thus the current i 2 is the is the sum
and 0.1 :
conductances. Then,
916
TwoPort Networks
and
y 22 = i 2 e v 2 = 0.15 e 1 = 0.15 :
(9.65)
Therefore, the twoport network of Figure 9.10 can be described by the following set of equations.
i 1 = y 11 v 1 + y 12 v 2 = 0.3v 1 0.1v 2
i 2 = y 21 v 1 + y 22 v 2 = 0.1 v 1 + 0.3v 2
(9.66)
Note:
In Example 9.4, we found that the short circuit transfer admittances are equal, that is,
y 21 = y 12 = 0.1
(9.67)
This is not just a coincidence; this is true whenever a twoport network is reciprocal (or bilateral). A
network is reciprocal if the reciprocity theorem is satisfied. This theorem states that:
If a voltage applied in one branch of a linear, twoport passive network produces a certain current in any
other branch of this network, the same voltage applied in the second branch will produce the same current
in the first branch.
The reverse is also true, that is, if current applied at one node produces a certain voltage at another,
the same current at the second node will produce the same voltage at the first. An example is given at
the end of this chapter.
Obviously, if we know that the twoport network is reciprocal, only three computations are required
to find the y parameters.
If in a reciprocal twoport network its ports can be interchanged without affecting the terminal voltages and currents, the network is said to be also symmetric. In a symmetric twoport network,
y 22 = y 11
y 21 = y 12
(9.68)
917
i2
i1
10 :
10 :
v1
v2
5:
15 A
4:
20 :
Figure 9.22. Network for Example 9.5
Solution:
We recognize the portion of the network enclosed in the dotted square, shown in Figure 9.23, as that
of the previous example.
1
10 :
i2
i1
10 :
v1
15 A
+
v2
5:
4:
20 :
Figure 9.23. Portion of the network for which the y parameters are known.
(9.69)
i2 = v2 e 4
(9.70)
and at Node 2,
(9.71)
i 2 = y 21 v 1 + y 22 v 2 = 0.1 v 1 + 0.3v 2 = v 2 e 4
or
0.4v 1 0.1v 2 = 15
(9.72)
0.1 v 1 + 0.4v 2 = 0
We will use MATLAB to solve the equations of (9.72) to become more familiar with it.
918
TwoPort Networks
syms v1 v2; [v1 v2]=solve(0.4*v10.1*v215, 0.1*v1+0.4*v2)
v1 = 40
v2 = 10
and thus
v 1 = 40 V
(9.73)
v 2 = 10 V
(9.74)
i 2 = 10 e 4 = 2.5 A
i1
+
v1
+
v2
i2
v 1 = z 11 i 1 + z 12 i 2
v 2 = z 21 i 1 + z 22 i 2
(9.75)
v 2 = z 21 i 1 + z 22 i 2
(9.76)
In twoport network theory, the z ij coefficients are referred to as the z parameters or as open circuit
impedance parameters.
i1
+
v1
+
v2
v
z 11 = 1i1
i2=0
i2 = 0
919
(9.77)
v
z 11 = 1i1
(9.78)
or
and this is the open circuit input impedance when the right port of Figure 9.25 is open.
Let us again consider (9.75), that is,
(9.79)
v 1 = z 11 i 1 + z 12 i 2
This time we assume that the terminal at v 1 is open, i.e., i 1 = 0 as shown in Figure 9.26.
i1=0
+
v2
+
v1
z 12
v
= 1i2
i2
i1 = 0
(9.80)
v
z 12 = 1i2
(9.81)
or
and this is the open circuit transfer impedance when the left port is open as shown in Figure 9.26.
Similar expressions are obtained when we consider the equation for v 2 , that is,
(9.82)
v 2 = z 21 i 1 + z 22 i 2
920
(9.83)
TwoPort Networks
i1
+
v1
+
v2
v
z 21 = 2i1
i2=0
i2 = 0
or
v
z 21 = 2i1
(9.84)
The parameter z 21 is referred to as open circuit transfer impedance when the right port is open as
shown in Figure 9.27.
Finally, let us assume that the terminal at v 1 is open, i.e., i 1 = 0 as shown in Figure 9.28.
i1=0
+
v1
+
v2
z 22
v
= 2i2
i2
i1 = 0
(9.85)
v
z 22 = 2i2
(9.86)
or
921
15 :
Solution:
a. The open circuit input impedance z 11 is found from the network of Figure 9.30 where we have
assumed that i 1 = 1 A .
v1
i1 = 1 A
5:
+
20 :
15 :
v2
i2 = 0
Figure 9.30. Network for computing z 11 for the network of Figure 9.29
We observe that the 20 : resistor is in parallel with the series combination of the 5 : and 15 :
resistors. Then, by the current division expression, the current through the 20 : resistor is 0.5 A
and the voltage across that resistor is
v 1 = 20 u 0.5 = 10 V
z 11 = v 1 e i 1 = 10 e 1 = 10 :
b. The open circuit transfer impedance z 12 is found from the network of Figure 9.31.
We observe that the 15 : resistance is in parallel with the series combination of the 5 : and
20 : resistances. Then, the current through the 20 : resistance is
15
15
i 20: =  i 2 =  u 1 = 3 e 8 A
15 + 5 + 20
40
922
TwoPort Networks
5:
+
i1 = 0
15 :
20 :
v1
v2
i2 = 1 A
Figure 9.31. Network for computing z 12 for the network of Figure 9.29
(9.88)
c. The open circuit transfer impedance z 21 is found from the network of Figure 9.32.
In Figure 9.32 the current that flows through the 15 : resistor is
20
20
i 15: =  i 1 =  u 1 = 1 e 2 A
20 + 5 + 15
40
v1
i1 = 1 A
5:
+
20 :
15 :
v2
i2 = 0
Figure 9.32. Network for computing z 21 for the network of Figure 9.29
(9.89)
923
z 21 = z 12
d. The open circuit output impedance z 22 is found from the network of Figure 9.33.
5:
+
i1 = 0
20 :
v1
+
15 :
v2
i2 = 1 A
Figure 9.33. Network for computing z 22 for the network of Figure 9.29
We observe that the 15 : resistance is in parallel with the series combination of the 5 : and
20 : resistances. Then, the current through the 15 : resistance is
20 + 5
25
i 15: =  i 2 =  u 1 = 5 e 8 A
20 + 5 + 15
40
(9.91)
(9.92)
i 2 = h 21 i 1 + h 22 v 2
(9.93)
924
TwoPort Networks
i1
i2
+
v1
+
v2
v 1 = h 11 i 1 + h 12 v 2
i 2 = h 21 i 1 + h 22 v 2
i1
i2
+
v1
v
h 11 = 1i1
v2=0
v2 = 0
(9.94)
v
h 11 = 1i1
(9.95)
or
i2
+
v1
h 12
v
= 1v2
+
v2
i1 = 0
Figure 9.36. Network for computing h 12 for the network of Figure 9.34
(9.96)
v
h 12 = 1v2
(9.97)
or
925
i1
i2
+
v1
i
h 21 = 2i1
v2=0
v2 = 0
Figure 9.37. Network for computing h 21 for the network of Figure 9.34
or
i
h 21 = 2i1
Therefore, in a twoport network the parameter h 21 represents a current gain (or loss).
Finally, let us assume that the terminal at v 1 is open, i.e., i 1 = 0 as shown in Figure 9.38.
i1=0
i2 +
v2
+
v1
i
h 22 = 2v2
i1 = 0
Figure 9.38. Network for computing h 22 for the network of Figure 9.34
or
i
h 22 = 2v2
926
TwoPort Networks
6:
1:
4:
Solution:
a. The short circuit input impedance h 11 is found from the network of Figure 9.40 where we have
assumed that i 1 = 1 A .
+
4:
v1
i1 = 1 A
6:
1:
i2
v2 = 0
Figure 9.40. Network for computing h 11 for the network of Figure 9.39
From the network of Figure 9.40 we observe that the 4 : and 6 : resistors are in parallel yielding an equivalent resistance of 2.4 : in series with the 1 : resistor. Then, the voltage across the
current source is
v 1 = 1 u 1 + 2.4 = 3.4 V
(9.98)
927
1:
+
v1
i1 = 0
+
4:
v2 = 1 V
Figure 9.41. Network for computing h 12 for the network of Figure 9.39.
(9.99)
6:
1:
4:
i2
v2 = 0
Figure 9.42. Network for computing h 21 for the network of Figure 9.39.
v 2.4: = 2.4 u 1 = 2.4 V
We observe that
928
TwoPort Networks
(9.100)
h 21 = h 12
and this is a consequence of the fact that the given network is reciprocal.
d. The open circuit admittance h 22 is found from the network of Figure 9.43.
+
i1 = 0
6:
1:
v1
4:
i2
+
v2 = 1 V
Figure 9.43. Network for computing h 22 for the network of Figure 9.39.
Since no current flows through the 1 : resistor, the current i 2 is found by Ohms law as
v2
1
 =  = 0.1 A
i 2 = 6+4
10
(9.101)
Note:
The h parameters and the g parameters (to be discussed next), are used extensively in networks consisting of transistors*, and feedback networks. The h parameters are best suited with seriesparallel
feedback networks, whereas the g parameters are preferred in parallelseries amplifiers.
9.4.4 The g Parameters
A twoport network can also be described by the set of equations
i 1 = g 11 v 1 + g 12 i 2
(9.102)
v 2 = g 21 v 1 + g 22 i 2
(9.103)
* Transistors are threeterminal devices. However, they can be represented as largesignal equivalent twoport networks circuits and also as smallsignal equivalent twoport networks where linearity can be applied.
929
i1
+
v2
i2
i 1 = g 11 v 1 + g 12 i 2
v 2 = g 21 v 1 + g 22 i 2
The g parameters, also known as inverse hybrid parameters, represent an admittance, a current gain,
a voltage gain and an impedance.
Let us assume that i 2 = 0 as shown in Figure 9.45.
+
v1
i1
+
v2
i
g 11 = 1v1
i2 = 0
i2 = 0
Figure 9.45. Network for computing g 11 for the network of Figure 9.44
(9.104)
i
g 11 = 1v1
(9.105)
or
i1
+
v2
i
g 12 = 1i2
i2
v1 = 0
Figure 9.46. Network for computing g 12 for the network of Figure 9.44
930
(9.106)
TwoPort Networks
or
i
g 12 = 1i2
(9.107)
Therefore, in a twoport network the parameter g 12 represents a current gain (or loss).
Let us assume that i 2 = 0 as shown in Figure 9.47.
+
v1
i1
+
v2
v
g 21 = 2v1
i2 = 0
i2 = 0
Figure 9.47. Network for computing g 21 for the network of Figure 9.44
(9.108)
v
g 21 = 2i1
(9.109)
or
Therefore, in a twoport network the parameter g 21 represents a voltage gain (or loss).
Finally, let us assume that v 1 is shorted, i.e., v 1 = 0 as shown in Figure 9.48.
v1 = 0
i1
+
v2
v
g 22 = 2i2
i2
v1 = 0
Figure 9.48. Network for computing g 22 for the network of Figure 9.44
(9.110)
v
g 22 = 2i2
(9.111)
or
931
1:
12 :
Solution:
a. The open circuit input admittance g 11 is found from the network of Figure 9.50 where we have
assumed that v 1 = 1 V .
i1
+
4:
1:
12 :
v1 = 1 V
v2
i2 = 0
Figure 9.50. Network for computing g 11 for the network of Figure 9.49.
(9.112)
932
TwoPort Networks
i1
4:
1:
12 :
v1 = 0
i2 = 1 A
Figure 9.51. Network for computing g 12 for the network of Figure 9.49.
(9.113)
1:
+
4:
12 :
v1 = 1 V
+
v2
i2 = 0
Figure 9.52. Network for computing g 21 for the network of Figure 9.49.
Since there is no current through the 4 : resistor, the voltage v 2 is the voltage across the 12 :
resistor. Then, by the voltage division expression,
12
v 2 =  u 1 = 12 e 13 V
1 + 12
We observe that
g 21 = g 12
(9.114)
and this is a consequence of the fact that the given network is reciprocal.
d. The short circuit output impedance g 22 is found from the network of Figure 9.53.
Circuit Analysis II with MATLAB Applications
Orchard Publications
933
4:
1:
12 :
v1 = 0
+
v2
i2 = 1 A
Figure 9.53. Network for computing g 22 for the network of Figure 9.49.
The voltage v 2 is the sum of the voltages across the 4 : resistor and the voltage across the 12 :
resistor. By the current division expression, the current through the 12 : resistor is
1
1
i 12: =  i 2 =  u 1 = 1 e 13 A
1 + 12
13
(9.115)
Then,
1
v 12: =  u 12 = 12 e 13 V
13
and
12
v 2 =  + 4 = 64 e 13 V
13
(9.116)
(9.117)
h 21 = h 12
g 21 = g 12
934
(9.118)
h 11 h 22 h 12 h 21 = 1
g 11 g 22 g 12 g 21 = 1
Z3
Z1
Z2
Z2
Z3
Z1
S
Tee
Z4
Z1
Z3
Z1
Bridged
Bridged Tee
Z3
Z2
Z2
The reverse is also true, that is, if current applied at one node produces a certain voltage at another,
the same current at the second node will produce the same voltage at the first.
It was also stated earlier that if we know that the twoport network is reciprocal, only three computations are required to find the y , z , h , and g parameters as shown in (9.117). Furthermore, if we know
that the twoport network is symmetric, we only need to make only two computations as shown in
(9.118).
Circuit Analysis II with MATLAB Applications
Orchard Publications
935
Z1
Z2
Z2
Z1
Z1
S
Tee
Z3
Z1
Z1
Z1
Bridged
Z1
Bridged Tee
Z2
Z2
Example 9.9
In the twoport network of Figure 9.56, the voltage source v S connected at the left end of the network is set for 15 V , and all impedances are resistive with the values indicated. On the right side of
the network is connected a DC ammeter denoted as A . Assume that the ammeter is ideal, that is, has
no internal resistance.
a. Compute the ammeter reading.
b. Interchange the positions of the voltage source and recompute the ammeter reading.
Z4
Z 1 = 30 :
Z3
Z1
vS
v S = 15 V
Z2
Z 2 = 60 :
Z 3 = 20 :
Z 4 = 10 :
Solution:
a. Perhaps the easiest method of solution is by nodal analysis since we only need to solve one equation.
936
I Z4
Z3
I Z3
Z2
vS
v S = 15 V
Z 1 = 30 :
Z 2 = 60 :
Z 3 = 20 :
Z 4 = 10 :
b
Figure 9.57. Network for solution of Example 9.9 by nodal analysis
By KCL at node a ,
V ab 15 V ab V ab
 +  +  = 0
30
60
20
or
6V = 15
60 ab
30
or
V ab = 5 V
The current through the ammeter is the sum of the currents I Z3 and I Z4 . Thus, denoting the current through the ammeter as I A we get:
V ab V
5 + 15
 = 0.25 + 1.50 = 1.75 A
I A = I Z3 + I Z4 =  +  = 20 10
Z3 Z4
(9.119)
b. With the voltage source and ammeter positions interchanged, the network is as shown in Figure
9.58.
Z4
I Z4
Z1
I Z1
Z2
v S = 15 V
Z 1 = 30 :
Z3
Z 2 = 60 :
vS
Z 3 = 20 :
Z 4 = 10 :
b
Figure 9.58. Network of Figure 9.57 with the voltage source and ammeter interchanged.
937
or
6
 V ab = 15
60
20
or
V ab = 7.5 V
The current through the ammeter this time is the sum of the currents I Z1 and I Z4 . Thus, denoting
the current through the ammeter as I A we get:
V ab V
7.5 15
I A = I Z1 + I Z4 =  +  =  +  = 0.25 + 1.50 = 1.75 A
Z1 Z4
30 10
(9.120)
We observe that (9.119) and (9.120 give the same value and thus we can say that the given network is reciprocal.
9.6 Summary
x A port is a pair of terminals in a network at which electric energy or a signal may enter or leave the
network.
x A network that has only one pair a terminals is called a oneport network. In an oneport network,
the current that enters one terminal must exit the network through the other terminal.
x A twoport network has two pairs of terminals, that is, four terminals.
x For an n port network the y parameters are defined as
i 1 = y 11 v 1 + y 12 v 2 + y 13 v 3 + } + y 1n v n
i 2 = y 21 v 1 + y 22 v 2 + y 23 v 3 + } + y 2n v n
i 3 = y 31 v 1 + y 32 v 2 + y 33 v 3 + } + y 2n v n
and so on.
x If the subscripts of the y parameters are alike, such as y 11 , y 22 and so on, they are referred to as
drivingpoint admittances. If they are unlike, such as y 12 , y 21 and so on, they are referred to as
transfer admittances.
x For a 2 port network the y parameters are defined as
938
Summary
i 1 = y 11 v 1 + y 12 v 2
i 2 = y 21 v 1 + y 22 v 2
x In a 2 port network where the right port is shortcircuited, that is, when v 2 = 0 , the y 11 param
v2 = 0
x In a 2 port network where the left port is shortcircuited, that is, when v 1 = 0 , the y 12 parame
v1 = 0
x In a 2 port network where the right port is shortcircuited, that is, when v 2 = 0 , the y 21 param
v2 = 0
x In a 2 port network where the left port is shortcircuited, that is, when v 1 = 0 , the y 22 parame
v1 = 0
and so on.
x If the subscripts of the z parameters are alike, such as z 11 , z 22 and so on, they are referred to as
drivingpoint impedances. If they are unlike, such as z 12 , z 21 and so on, they are referred to as
transfer impedances.
939
i2 = 0
x In a 2 port network where the left port is open, that is, when i 1 = 0 , the z 12 parameter is
i1 = 0
x In a 2 port network where the right port is open, that is, when i 2 = 0 , the z 21 parameter is
i2 = 0
x In a 2 port network where the left port is open, that is, when i 1 = 0 , the z 22 parameter is
i1 = 0
x A twoport network can also be described in terms of the h parameters with the equations
v 1 = h 11 i 1 + h 12 v 2
i 2 = h 21 i 1 + h 22 v 2
x The h parameters represent an impedance, a voltage gain, a current gain, and an admittance. For
940
Summary
v
h 11 = 1i1
v2 = 0
x In a 2 port network where the left port is open, that is, when i 1 = 0 , the h 12 parameter repre
sents a voltage gain (or loss) in the twoport network. In other words,
v
h 12 = 1v2
i1 = 0
x In a 2 port network where the right port is shorted, that is, when v 2 = 0 , the h 21 parameter
v2 = 0
x In a 2 port network where the left port is open, that is, when i 1 = 0 , the h 22 parameter repre
i1 = 0
x A twoport network can also be described in terms of the g parameters with the equations
i 1 = g 11 v 1 + g 12 i 2
v 2 = g 21 v 1 + g 22 i 2
x The g parameters, also known as inverse hybrid parameters, represent an admittance, a current
i2 = 0
x In a 2 port network where the left port is shorted, that is, when v 1 = 0 , the g 12 parameter rep
resents a current gain (or loss) in the twoport network. In other words,
i
g 12 = 1i2
v1 = 0
941
i2 = 0
x In a 2 port network where the left port is shorted, that is, when v 1 = 0 , the g 22 parameter rep
v1 = 0
x The reciprocity theorem states that if a voltage applied in one branch of a linear, twoport passive
network produces a certain current in any other branch of this network, the same voltage applied
in the second branch will produce the same current in the first branch. The reverse is also true,
that is, if current applied at one node produces a certain voltage at another, the same current at
the second node will produce the same voltage at the first.
x A twoport network is said to be reciprocal if any of the following relationships exists.
z 21 = z 12
y 21 = y 12
h 21 = h 12
g 21 = g 12
x A twoport network is said to be symmetrical if any of the following relationships exist.
z 21 = z 12 and z 22 = z 11
y 21 = y 12 and y 22 = y 11
h 21 = h 12 and h 11 h 22 h 12 h 21 = 1
g 21 = g 12 and g 11 g 22 g 12 g 21 = 1
942
Exercises
9.7 Exercises
1. For the network of Figure 9.59, find the z parameters.
10 :
5:
20 :
15 :
1:
6:
943
i2 +
i1
+
h12 v2
h21 i1
v2
1
h 22 :
h 11 = 1.2 K:
h 12 = 2 u 10
h 21 = 50
h 22 = 50 u 10
944
Solutions to Exercises
9.8 Solutions to Exercises
1.
v
z 11 = 1i1
10 :
i2 = 0
i 5:
v1
v2
20 :
i2 = 0
5:
i1 = 1 A
10 + 20
30
i 5: =  i 1 =  u 1 = 6 e 7 A
5 + 10 + 20
35
v 1 = 5i 5: = 5 u 6 e 7 = 30 e 7 V
v
30 e 7
z 11 = 1 =  = 30 e 7 :
i1
1
v
z 12 = 1i2
10 :
i1 = 0
i1 = 0
i 5:
v1
+
20 :
v2
5:
i2 = 1 A
20
20
i 5: =  i 2 =  u 1 = 4 e 7 A
20 + 5 + 10
35
4
v 1 = 5 u  = 20 e 7 V
7
v
20 e 7
z 12 = 1 =  = 20 e 7 :
i2
1
v
z 21 = 2i1
i2 = 0
v1
i1 = 1 A
10 :
+
5:
20 :
v2
i2 = 0
945
We observe that
z 21 = z 12
v
z 22 = 2i2
10 :
i1 = 0
i1 = 0
20 :
5:
v1
v2
i2 = 1 A
10 + 5
15
i 20: =  i 2 =  u 1 = 3 e 7 A
20 + 10 + 5
35
3
v 2 = 20 u  = 60 e 7 V
7
v
e 7 = 60 e 7 :
z 22 = 1 = 60
1
i2
2.
i
y 11 = 1v1
i1
v2 = 0
+
5:
20 :
15 :
short
v2 = 0
v1 = 1 V
R eq = 5 __ 20 = 4 :
i 1 = v 1 e R eq = 1 e 4 A
1
1e4
y 11 = i 1 e v 1 =  = 1 e 4 :
1
946
Answers to Exercises
i
y 12 = 1v2
i1
5:
v1 = 0
v1 = 0
+
20 :
15 :
short
v2 = 1 V
v 5: = v 2 = 1 V
i 1 = v 5: e 5 = 1 e 5 A
y 12 = i 1 e v 2 = 1 e 5 e 1 = 1 e 5 :
i
y 21 = 2v1
i2
5:
v2 = 0
+
v2 = 0
20 :
v1 = 1 V
15 :
short
v 5: = v 1 = 1 V
i 2 = v 5: e 5 = 1 e 5 A
y 21 = i 2 e v 1 = 1 e 5 e 1 = 1 e 5 :
We observe that
y 21 = y 12
i
y 22 = 2v2
i1
v1 = 0
v1 = 0
i2
5:
+
20 :
short
15 :
v2 = 1 V
i 2 = v 2 e R eq = 1 e 5 __ 15 = 1 e 75 e 20 = 4 e 15 A
y 22 = i 2 e v 2 = 4 e 15 e 1 = 4 e 15 :
Circuit Analysis II with MATLAB Applications
Orchard Publications
947
v
h 11 = 1i1
4 :
i 1:
v2 = 0
1:
v1
i1 = 1 A
v2 = 0
6:
short
4
4
i 1: =  i 1 =  u 1 = 4 e 5 A
1 + 4
5
v 1 = 1 u i 1: = 4 e 5 V
v
4e5
h 11 = 1 =  = 4 e 5 :
i1
1
v
h 12 = 1v2
i1 = 0
i 1:
i1 = 0
v1
i2
4 :
1:
+
6 :
v2
+
v2 = 1 V
v2
1
1
i 2 =  =  =  = 11 e 30 A
__
R eq
6 4 + 1
30 e 11
6
6 11
v 1 = 1 u i 1: = 1 u  u i 2 = 1 u  u  = 1 e 5 V
6 + 4 + 1
11 30
v
1e5
h 12 = 1 =  = 1 e 5 dimensionless
v2
1
i
h 21 = 2i1
i1
4 :
i2
v2 = 0
1:
v2 = 0
6:
i1 = 1 A
short
1
1
i 2 =  u i 1 =  u 1 = 1 e 5 A
5
1 + 4
948
Answers to Exercises
i
1 e 5
h 21 = 2 =  = 1 e 5
i1
1
We observe that
h 21 = h 12
i
h 22 = 2v2
i1 = 0
i2
4 :
i1 = 0
1:
v1
6 :
+
v2
v2 = 1 V
v2
1
1
i 2 =  =  =  = 11 e 30 A
__
R eq
6 4 + 1
30 e 11
i
1
11 e 30
h 22 = 2 =  = 11 e 30 :
v2
1
4.
i
g 11 = 1v1
i1
4 :
i2 = 0
+
v1 = 1 V
v1
i2 = 0
1:
6:
v1
1
1
i 1 =  =  =  = 11 e 10 A
R eq
1 __ 4 + 6 10 e 11
i
1
11 e 10
g 11 = 1 =  = 11 e 10 :
v1
1
i
g 12 = 1i2
i1
i2
4 :
v1 = 0
v1 = 0
short
1:
6 :
v2
i2 = 1 A
949
i1
4 :
i2 = 0
v1 = 1 V
+
v1
1:
i 6: +
v2
6:
i2 = 0
v1
1  = 11 e 10 A
1
 = i 1 =  = 1 __ 4 + 6 10 e 11
R eq
11
1
v 2 = 6 u i 6: = 6 u   = 3 e 5 V
1 + 4 + 6 10
v
3e5
g 21 = 2 =  = 3 e 5
v1
1
We observe that
g 21 = g 12
v
g 22 = 2i2
i1
i2
4 :
i 6:
v1 = 0
v1 = 0
1:
6 :
short
+
v2
i2 = 1 A
4
24
v 2 = 6 u i 6: = 6 u  u i 2 =  u 1 = 12 e 5 V
6+4
10
v
12 e 5
g 22 = 2 =  = 12 e 5 :
i2
1
950
Answers to Exercises
5.
We recall that
v 1 = h 11 i 1 + h 12 v 2 (1)
i 2 = h 21 i 1 + h 22 v 2 (2)
With the voltage source v 1 = cos Zt mV in series with 800 : connected at the input and a 5 K:
load connected at the output the network is as shown below.
800 :
1200 :
i2
+
i1
+
2 u 10 v 2
50 u 10
50i 1
v2
5000 :
1 0q mV
800 + 1200 i 1 + 2 u 10 v 2 = 10
v2
6
50i 1 + 50 u 10 v 2 = i 2 = 5000
or
4
2 u 10 i 1 + 2 u 10 v 2 = 10
50i 1 + 250 u 10 v 2 = 0
We write the two equations above in matrix form and use MATLAB for the solution.
A=[2*10^3 2*10^(4); 50 250*10^(6)]; B=[10^(3) 0]'; X=A\B;...
fprintf(' \n'); fprintf('i1 = %5.2e A \t',X(1)); fprintf('v2 = %5.2e V',X(2))
i1 = 5.10e007 A
v2 = 1.02e001 V
Therefore,
i 1 = 0.51 PA (3)
v 2 = 102 mV (4)
Next, we use (1) and (2) to find the new values of v 1 and i 2
3
v 1 = 1.2 u 10 u 0.51 u 10
+ 2 u 10
u 102 u 10 = 0.592 mV
951
u 50 u 10
u 102 u 10 = 20.4 PA
and the minus () sign indicates that the output voltage in 180q outofphase with the input.
The current gain is
i
20.4 PA
G I = 2 =  = 40
i1
0.51 PA
952
Chapter 10
ThreePhase Systems
Va
Za
Ia
Vb
Zb
Ib
Vc
Zc
Ic
Ib
Ic
101
Va
Ia
Za
Vb
Ib
Zb
Ic
Zc
Vc
Ia + Ib + Ic
This arrangement shown in Figure 10.3 uses only 4 wires instead of the 6 wires shown in Figure
10.1. But now we must find the relative size of the common return wire that it would be sufficient to
carry all three currents I a + I b + I c
We have assumed that the voltage sources are equal in magnitude and 120q apart, and the loads are
equal. Therefore, the currents will be balanced (equal in magnitude and 120q outof phase). These
currents are shown in the phasor diagram of Figure 10.4.
Ic
Ia
Ib
Figure 10.4. Phasor diagram for threephase balanced system
102
ThreePhase Connections
From figure 10.4 we observe that the sum of these currents, added vectorially, is zero. Therefore,
under ideal (perfect balance) conditions, the common return wire carries no current at all. In a practical situation, however, is not balanced exactly and the sum is not zero. But still it is quite small and
in a fourwire threephase system the return wire is much smaller than the other three. Figure 10.5
shows a fourwire, threephase Y system where V a = V b = V c , the three loads are identical,
and I n is the current in the neutral (fourth) wire.
Ia
V a cos Zt V
ZLOAD
Ib
ZLOAD
V b cos Zt 120q V
ZLOAD
V c cos Zt 240q V
Ic
In
V a cos Zt V
ZLOAD
Ib
ZLOAD
V b cos Zt 120q V
ZLOAD
V c cos Zt 240q V
Ic
This arrangement shown in Figure 10.6 could be used only if all the three voltage sources are perfectly balanced, and if the three loads are perfectly balanced also. This, of course, is a physical impossibility and therefore it is not used.
Circuit Analysis II with MATLAB Applications
Orchard Publications
103
V a cos Zt V
ZLOAD
Ib
V b cos Zt 120q V
V c cos Zt 240q V
ZLOAD
ZLOAD
Ic
This arrangement offers the advantage that the 'connected loads need not be accurately balanced.
However, a 'connection with only three voltages is not used for safety reasons, that is, it is a safety
requirement to have a connection from the common point to the ground as shown in Figure 10.5.
104
'
120 V
Neutral wire
240 V
120 V
Industrial facilities need threephase power for threephase motors. Threephase motors run
smoother and have higher efficiency than singlephase motors. A Y ' connection is shown in Figure 10.10 where the secondary provides 240 V threephase power to the motor, and one of the
transformers of the secondary is centertapped to provide 120 V to the lighting load.
105
Vab
Ib
b
Vac
ZLOAD
Vbc
ZLOAD
Ic
Ia
Ib
Figure 10.12. Phasor diagram for Yconnected perfectly balanced load
106
(10.1)
I b = I a 120q
(10.2)
I c = I a +120q
(10.3)
These equations define the balance set of currents of positive phase sequence a b c .
Next, we consider the voltages. Voltages V ab , V ac , and V bc are referred to as linetoline voltages and
voltages V an , V bn , and V cn as phase voltages. We observe that in a Y connected load, the line and
phase voltages are not the same.
We will now derive the relationships between line and phase voltages in a Y connected load.
Arbitrarily, we choose V an as our reference phase voltage. Then,
V an = V an 0q
(10.4)
V bn = V an 120 q
(10.5)
V cn = V an +120q
(10.6)
These equations define a positive phase sequence a b c . These relationships are shown in Figure
10.13.
Vcn
Van
Vbn
Figure 10.13. Phase voltages in a Y connected perfectly balanced load
107
Vab
Ib
b
Vac
ZLOAD
Vbc
ZLOAD
Ic
(10.7)
V ca = V cn + V na = V cn V an
(10.8)
V bc = V bn + V nc = V bn V cn
(10.9)
These can also be derived from the phasor diagram of Figure 10.15.
Vcn
Vca
Vbn
Vab
30q
Van
Van
Vcn
Vbn
Vbc
Figure 10.15. Phasor diagram for linetoline and linetoneutral voltages in Y load
From geometry and the law of sines we find that in a balanced threephase, positive phase sequence
Y connected load, the line and phase voltages are related as
108
3V an 30q
(10.10)
Y connected load
The other two linetoline voltages can be easily obtained from the phasor diagram of the previous
page.
Now, let us consider a ' connected load shown in Figure 10.16.
a
Ia
Iab
ZLOAD
Vab
Ib
ZLOAD
b
Vca
ZLOAD
Vbc
Ic
Ibc
Ica
We observe that the line and phase voltages are the same, but the line and phase currents are not the
same. To find the relationship between the line and phase currents, we apply KCL at point a and we
get:
I ab = I a + I ca
or
I a = I ab I ca
(10.11)
The line currents I b and I c are derived similarly, and the phasetoline current relationship in a ' connected load is shown in the phasor diagram of Figure 10.17.
From geometry and the law of sines we find that a balanced threephase, positive phase sequence ' connected load, the line and phase currents are related as
Ia =
3I ab 30 q
(10.12)
The other two line currents can be easily obtained from the phasor diagram of Figure 10.17.
109
Ica
Ibc
Iab
Iab
30 o
Ib
Ibc
Ia
Ica
Figure 10.17. Phasor diagram for line and phase currents in 'connected load
Za
C
C
Z3
Z1
Zc
Zb
Z2
(10.13)
= Zb + Zc
and in the 'connection, the impedance between terminals B and C is Z 2 in parallel with the sum
Z 1 + Z 3 , that is,
Z BC
1010
'
Z2 Z1 + Z3
= Z1 + Z2 + Z3
(10.14)
(10.15)
Similar equations for terminals AB and CA are derived by rotating the subscripts of (10.15) in a
cyclical manner. Then,
Z3 Z1 + Z2
Z a + Z b = Z1 + Z2 + Z3
(10.16)
Z1 Z2 + Z3
Z c + Z a = Z1 + Z2 + Z3
(10.17)
and
Equations (10.15) and (10.17) can be solved for Z a by adding (10.16) with (10.17), subtracting
(10.15) from this sum, and dividing by two. That is,
2Z 1 Z 3 + Z 2 Z 3 + Z 1 Z 2
Z1 Z3 + Z2 Z3 + Z1 Z2 + Z1 Z3
 = 2Z a + Z b + Z c = Z1 + Z2 + Z3
Z1 + Z2 + Z3
(10.18)
2Z 1 Z 3 + Z 2 Z 3 + Z 1 Z 2 Z 1 Z 2 Z 2 Z 3
2Z a + Z b + Z c Z b Z c = Z1 + Z2 + Z3
(10.19)
2Z 1 Z 3
2Z a = Z1 + Z2 + Z3
(10.20)
Z1 Z3
Z a = Z1 + Z2 + Z3
(10.21)
Similar equations for Z b and Z c are derived by rotating the subscripts of (10.21) in a cyclical manner.
Thus, the three equations that allow us to change any 'connection of impedances into a Y connection are given by (10.22).
Z1 Z3
Z a = Z1 + Z2 + Z3
Z2 Z3
Z b = Z1 + Z2 + Z3
(10.22)
Z1 Z2
Z c = Z1 + Z2 + Z3
' o Y Conversion
Circuit Analysis II with MATLAB Applications
Orchard Publications
1011
IA
Za
C
Zc
Zb
IB
IC
Z3
Z1
Z2
IB
IC
(b)
(a)
(10.23)
V BC = Z b I B Z c I C
(10.24)
V CA = Z c I C Z a I A
(10.25)
If we attempt to solve equations (10.23), (10.24) and (10.25) simultaneously, we will find that the
determinant ' of these sets of equations is singular, that is, ' = 0 . This can be verified with Cramers
rule as follows:
Z a I A Z b I B + 0 = V AB
0 + Z b I B Z c I C = V BC
(10.26)
Z a I A + 0 + Z c I C = V CA
Za Zb 0
' =
Zb Zc = Za Zb Zc Za Zb Zc + 0 + 0 + 0 + 0 = 0
Za 0
(10.27)
Zc
This result suggests that the equations of (10.26) are not independent and therefore, no solution
exists. However, a solution can be found if, in addition to (10.23) through (10.25), we use the equation
IA + IB + IC = 0
1012
(10.28)
(10.29)
V CA = Z c I A Z c I B Z a I A = Z a + Z c I A Z c I B
(10.30)
(10.31)
D
I B = 2'
(10.32)
where
' =
Za
Zb
Za + Zc Zc
= Zc Za Za Zb Zb Zc
(10.33)
= Z c V AB + Z b V CA
(10.34)
and
D1 =
V AB Z b
V CA Z c
Then,
Z c V AB + Z b V CA
Z c V AB Z b V CA
D
I A = 1 =  = '
Za Zb + Zb Zc + Zc Za
Za Zb Zb Zc Zc Za
(10.35)
D
Z a V BC Z c V AB
I B = 2 = '
Za Zb + Zb Zc + Zc Za
(10.36)
Similarly,
(10.37)
Therefore, for the Y connection which is repeated in Figure 10.20 for convenience, we have:
1013
IA
Za
Zc
Zb
IB
IC
(10.38)
Z b V CA Z a V BC
I C = Za Zb + Zb Zc + Zc Za
For the 'connection, which is also repeated in Figure 10.21 for convenience, the line currents are:
IA
Z3
Z1
C
Z2
IC
IB
(10.39)
V CA V BC
I C = Z1
Z2
1014
(10.40)
V BC V AB
Z a V BC Z c V AB
=  Za Zb + Zb Zc + Zc Za
Z2
Z3
(10.41)
V CA V BC
Z b V CA Z a V BC
=  Za Zb + Zb Zc + Zc Za
Z1
Z2
(10.42)
Zb
Zc
1
1=  and = Z3
Za Zb + Zb Zc + Zc Za
Za Zb + Zb Zc + Zc Za Z1
(10.43)
Za
1
= Za Zb + Zb Zc + Zc Za
Z2
(10.44)
From (10.40)
Rearranging, we get:
Za Zb + Zb Zc + Zc Za
Z 1 = Zb
Za Zb + Zb Zc + Zc Za
Z 2 = Za
(10.45)
Za Zb + Zb Zc + Zc Za
Z 3 = Zc
Y o ' Conversion
Example 10.1
For the circuit of Figure 10.22, use the Y o ' conversion to find the currents in the various
branches as indicated.
Solution:
Let us indicate the nodes as a , b , c , and d , and denote the 90 : , 90 : and 90 : resistances as
R a , R b , and R c respectively as shown in Figure 10.23.
Next, we replace the Y connection formed by a , b , c , and d with the equivalent ' connection
shown in Figure 10.24.
1015
I4
60 :
50 :
I7
120V
I6
+
90 :
70 :
80 :
I5
I8
a
I1
I4
Rc
60 :
50 :
I7
120V
I6
+
Ra 90 :
70 :
80 :
I5
Rb
I8
b
a
I1
I4
196 :
60 :
R1
120V
+
174 :
R2
R3
70 :
314 :
I5
1016
and
314 u 70
R ad =   57 :
314 + 70
The circuit of Figure 10.24 reduces to the circuit of Figure 10.25. The circuit of Figure 10.25 can be
further simplified as shown in Figure 10.26.
From the circuit of Figure 10.26,
120
I 2 =  = 0.69 A
174
(10.46)
120
I 3 =  = 1.17 A
103
(10.47)
a
I1
46 :
120V
+
174 :
I2
d
I3
57 :
b
Figure 10.25. Circuit (d) for Example 10.1
1017
I1
120V
174 :
+
I2
103 :
I3
b
Figure 10.26. Circuit (e) for Example 10.1
To compute the other currents, we return to the circuit of Figure 10.25 which, for convenience, is
repeated as Figure 10.27 and it is denoted as Circuit (f).
For the circuit of Figure 10.27, by the voltage division expression
46
V ad =  u 120 = 53.6 V
46 + 57
(10.49)
57
V db =  u 120 = 66.4 V
46 + 57
(10.50)
a
I1
46 :
120V
+
174 :
I2
d
I3
57 :
1018
196 :
60 :
R1
120V
174 :
R2
+
R3
70 :
314 :
I5
(10.51)
V bd
53.6
 =  = 0.95 A
I 5 = 70
60
(10.52)
and
Finally, we return to the circuit of Figure 10.23 which, for convenience, is repeated as Figure 10.29
and denoted as Circuit (h).
a
I1
I4
Rc
60 :
50 :
I7
120V
I6
+
Ra 90 :
70 :
80 :
I1
Rb
I8
I5
b
Figure 10.29. Circuit (h) for Example 10.1
1019
(10.53)
(10.54)
(10.55)
and
Of course, we could have found the branch currents with nodal or mesh analysis.
Quite often, the Y and ' arrangements appear as shown in Figure 10.30 and they are referred to as
the tee (T) and pi (S) circuits. Consequently, the formulas we developed for the Y and ' arrangements can be used with the tee and S arrangements.
A
Za
Zb
A
Z1
Zc
Z3
Z2
C
C
Figure 10.30. T and S circuits
Z1 Z3
Z1
Z
 = Z a =  = 1Z1 + Z2 + Z3
3Z 1
3
1020
(10.56)
ThreePhase Power
IA
A
A
IC
Za
Z3
Z1
C
IA
(a)
Z2
IB
Zc
(b)
IC
Zb
IB
(10.57)
Z1 Z2
Z c = Z1 + Z2 + Z3
' o Y Conversion
(10.58)
where V AN is the linetoneutral voltage, I A is the line current, cos T is the power factor of the load,
and T is the angle between V AN and I A .
If the load is 'connected as in Figure 10.31 (a), the total threephase power is given by
P TOTAL = 3 V AB I AB cos T
' connected load
(10.59)
We observe that relation (10.59) is given in terms of the linetoneutral voltage and line current, and
relation (10.58) in terms of the linetoline voltage and phase current.
Circuit Analysis II with MATLAB Applications
Orchard Publications
1021
3 V AB I A cos T LOAD
(10.60)
It is important to remember that the power factor cos T LOAD in (10.60) refers to the load, that is, the
angle T is not the angle between V AB and I A .
Example 10.2
The threephase generator of Figure 10.32 supplies 100 kW at 0.9 lagging power factor to the threephase load. The linetoline voltage at the load is 2400 V . The resistance of the line is 4 : per conductor and the inductance and capacitance are negligible. What linetoline voltage must the generator supply to the line?
Solution:
The load per phase at 0.9 pf is
1
 u 100 = 33.33 kW
3
Generator
(Yconnected)
Load
(Yconnected)
From (10.10),
V ab =
3V an 30q
Y connected load
(10.61)
(10.62)
1022
(10.63)
(10.64)
and the angle by which the line (or phase) current lags the phase voltage is
1
(10.65)
Next, let us assume that the line current in phase a lies on the real axis. Then, the phasor of the linetoneutral voltage at the load end is
V an load = V an 25.84q = 1386 cos 25.84q + j sin 25.84q = 1247 + j604 V
(10.66)
The voltage drop across a conductor is in phase with the line current since it resistive in nature.
Therefore,
V COND = I LINE u R = 26.7 u 4 = 106.8 V
(10.67)
(10.68)
(10.69)
3 u V an gen =
3 u 1483 = 2569 V
(10.70)
2 V cos Zt
(10.71)
i a = I p cos Zt =
2 I cos Zt
(10.72)
1023
(10.73)
(10.74)
we get
p a = v a i a = 2 V I cos Z t = V I cos 2Zt + 1
The voltage and current in phase b are equal in magnitude to those in phase a but they are 120q outofphase. Then,
vb =
2 V cos Zt 120q
(10.75)
ib =
2 V cos Zt 120q
(10.76)
(10.77)
(10.78)
(10.79)
Recalling that
cos x y = cos x cos y + sin x sin y
(10.80)
we find that the sum of the three cosine terms in (10.79) is zero. Then,
p total = 3 V I
Three phase Balanced System
(10.81)
Therefore, the instantaneous total power is constant and it is equal three times the average power.
The proof can be extended to include any power factor; thus, (10.81) can be also expressed as
p total = 3 V I cos T
(10.82)
Example 10.3
Figure 10.33 shows a threephase feeder with two loads; one consists of a bank of lamps connected
lineto neutral and the rating is given in the diagram; the other load is 'connected and has the
1024
IA
IB
IC
L
Z'
Z'
Z = 18 + j80
Z'
Solution:
To facilitate the computations, we will reduce the given circuit to one phase (phase a ) taken as reference, i.e., at zero degrees, as shown in Figure 10.34
+
IA
IZ
ZY
IL
L
ZL
V L L = 220 0q V
VLN
(Linetoneutral)
Figure 10.34. Singlephase representation of Figure 10.31
V rated
120  = 28.8 :
Z L = R lamp =  = P rated
500
The linetoline voltage is given as V L L = 220 V ; therefore, by (10.10), the linetoneutral voltage
V L N is
1025
For convenience, we indicate these values in Figure 10.34 which now is as shown in Figure 10.35.
+
IA
IZ
IL
L
ZY
ZL
V L N = 127 0q V
Z Y = 27.33 77.32
Z L = 28.8 0q
and
VL N
127 0q = 4.41 0q = 4.41
I L =  = 28.8 0q
ZL
Then,
I Z + I L = 1.02 j4.54 + 4.41 = 5.43 j4.54 = 7.08 39.9q
Finally, the total power delivered to the entire load is three times of P A , that is,
P total = 3 u 690 = 2070 watts = 2.07 Kw
Check:
Each lamp is rated 120 V and 500 w but operates at 127 V . Thus, each lamp absorbs
V oper 2
P oper
=  V rated
P rated
127 2
P oper =  u 500 = 560 w
120
1026
This value is in close agreement with the value on the previous page.
If the load were 'connected, each wattmeter would have its current coil in one side of the ' and its potential
coil from line to line.
1027
Load
n
Wattmeter connections
Figure 10.36. Wattmeter connections in fourwire, threephase system
Load
c
p
Wattmeter connections
Figure 10.37. Wattmeter connections in threewire, threephase system
1028
1
p dt = T
0
0 vi dt
(10.83)
(10.84)
This is the true power absorbed by the load, not power indicated by the wattmeters.
Now, we will compute the total power indicated by the wattmeters. Each wattmeter measures the
average of the line current times the voltage to point p . Then,
1
P wattmeters = T
(10.85)
But
v ap = v an + v np
v bp = v bn + v np
(10.86)
v cp = v cn + v np
(10.87)
and since
ia + ib + ic = 0
(10.88)
(10.89)
This relation is the same as (10.84); therefore, the power indicated by the wattmeters and the true
power absorbed by the load are the same.
Some thought about the location of the arbitrarily selected point p would reveal a very interesting
result. No matter where this point is located, the power relation (10.87) reduces to (10.89). Suppose
that we locate point p on line c . If we do this, the voltage coil of Wattmeter 3 is zero and thus the
reading of this wattmeter is zero. Accordingly, we can remove this wattmeter and still obtain the true
power with just Wattmeters 1 and 2 as shown in Figure 10.38.
1029
Load
c
Wattmeter connections
10.10 Summary
x AC is preferable to DC because voltage levels can be changed by transformers. This allows more
and generators start and run more smoothly since they have constant torque. They are also more
economical.
x If the voltage sources are equal in magnitude and 120q apart, and the loads are also equal, the currents will be balanced (equal in magnitude and 120q outof phase).
x Industrial facilities need threephase power for threephase motors. Threephase motors run
3V an 30q
x In a Y connected load, the line and phase currents are the same.
x In a ' connected system
Ia =
1030
3I ab 30 q
Circuit Analysis II with MATLAB Applications
Orchard Publications
Summary
x In a ' connected load, the line and phase voltages are the same.
x For ' o Y Conversion we use the relations
Z1 Z3
Z a = Z1 + Z2 + Z3
Z2 Z3
Z b = Z1 + Z2 + Z3
Z1 Z2
Z c = Z1 + Z2 + Z3
x For Y o ' Conversion we use the relations
Za Zb + Zb Zc + Zc Za
Z 1 = Zb
Za Zb + Zb Zc + Zc Za
Z 2 = Za
Za Zb + Zb Zc + Zc Za
Z 3 = Zc
x When we want to compute the voltages, currents, and power in a balanced threephase system, it is
very convenient to use the Y connection and work with one phase only.
x If a load is Y connected, the total threephase power is given by
P TOTAL = 3 V AN I A cos T
Y connected load
x If the load is 'connected the total threephase power is given by
P TOTAL = 3 V AB I AB cos T
' connected load
(10.90)
x For any load ( Y or ' connected ) the total threephase power can be computed from
P TOTAL =
3 V AB I A cos T LOAD
and it is important to remember that the power factor cos T LOAD refers to the load, that is, the
angle T is not the angle between V AB and I A .
Circuit Analysis II with MATLAB Applications
Orchard Publications
1031
Z
Z
Z
Wattmeter
Load
c
Figure 10.39. Circuit for Exercise 1
2. In the circuit of Figure 10.40 the lighting load is balanced. Each lamp is rated 500 w at 120 V .
Assume constant resistance, that is, each lamp will draw rated current. The threephase motor
draws 5.0 Kw at a power factor of 0.8 lagging. The secondary of the transformer provides balanced 208 V linetoline. The load is located 1500 feet from the threephase transformer. The
resistance and inductive reactance of the distribution line is 0.403 : and 0.143 : respectively per
1000 ft of the wire line. Compute linetoline and linetoneutral voltages at the load.
1032
Solutions to Exercises
10.12 Solutions to Exercises
1. a
Ia
I ab
I ca
Z
Z
b
V ab
Wattmeter
Z
I bc
Load
Ic
mag = 17.32 A
Thus, I a = 17.32 A
The phase sequence a b c implies the phase diagram below.
From (10.59)
P total =
=
3 V ab I a load pf
3 u 100 u 17.32 u cos 30q = 2 598 w
b.
The wattmeter reads the product V ab u I c where I c is 240q behind I a as shown on the phasor
diagram. Then, the wattmeter reading is
P wattmeter = V ab u I c = 100 0q u 10 3 u cos 60q 240q
= 100 u 17.32 u cos 300q = 866 w
1033
I bc
I ca
Ia
V ab = 100 120 q
and thus
Z LINE = 0.605 + j0.215
Also,
P rated
500
I lamp1 = I lamp2 =  =  = 4.17 A
V rated
120
1500 ft
jX L
R
0.605 :
V an = 208 e 3 0q V
= 120 0q V
I total
j0.215 : Z line
I lamp1
4.17 A 4.17 A
I lamp2
IM
V M = V load
5 e 3 Kw
0.8 pf
We recall that for a single phase system the real power is given by
1034
Answers to Exercises
P real = V RMS I RMS cos T
where cos T = pf
Then, we find the motor current I M in terms of the motor voltage V M as
2083e 3 = I M = 5000
VM
0.8 V M
1
and since cos 0.8 = 36.9q lagging pf , the motor current I M is expressed as
1
2083
I M =  36.9q =  1666 j1251
VM
VM
Next,
1
V an = 120 0q = V line + V M =  > 5.05V M + 1277 + j 1.79V M 398.7 @ + V M
VM
or
2
or
2
ans =
1.0e+002 *
1.0260 + 0.0238i
0.1235  0.0417i
Circuit Analysis II with MATLAB Applications
Orchard Publications
1035
1036
3 u VM =
3 u 102.63 = 177.76 V
Appendix A
Introduction to MATLAB
his appendix serves as an introduction to the basic MATLAB commands and functions, procedures for naming and saving the user generated files, comment lines, access to MATLABs Editor/Debugger, finding the roots of a polynomial, and making plots. Several examples are provided with detailed explanations.
Times Bold Italic: Important terms and facts, notes and file names
When we first start MATLAB, we see the toolbar on top of the command screen and the prompt
EDU>>. This prompt is displayed also after execution of a command; MATLAB now waits for a new
command from the user. It is highly recommended that we use the Editor/Debugger to write our
program, save it, and return to the command screen to execute the program as explained below.
To use the Editor/Debugger:
1. From the File menu on the toolbar, we choose New and click on MFile. This takes us to the Edi* EDU>> is the MATLAB prompt in the Student Version
A1
A2
Roots of Polynomials
One of the most powerful features of MATLAB is the ability to do computations involving complex
numbers. We can use either i , or j to denote the imaginary part of a complex number, such as 34i
or 34j. For example, the statement
z=34j
displays
z = 3.00004.0000i
In the above example, a multiplication (*) sign between 4 and j was not necessary because the complex number consists of numerical constants. However, if the imaginary part is a function, or variable
such as cos x , we must use the multiplication sign, that is, we must type cos(x)*j or j*cos(x) for the
imaginary part of the complex number.
Solution:
The roots are found with the following two statements where we have denoted the polynomial as p1,
and the roots as roots_ p1.
p1=[1 10 35 50 24]
p1 =
1
10
roots_ p1=roots(p1)
35
50
24
roots_p1 =
4.0000
3.0000
2.0000
Circuit Analysis II with MATLAB Applications
Orchard Publications
A3
We observe that MATLAB displays the polynomial coefficients as a row vector, and the roots as a
column vector.
Example A.2
Find the roots of the polynomial
5
p 2 x = x 7x + 16x + 25x + 52
Solution:
There is no cube term; therefore, we must enter zero as its coefficient. The roots are found with the
statements below, where we have defined the polynomial as p2, and the roots of this polynomial as
roots_ p2. The result indicates that this polynomial has three real roots, and two complex roots. Of
course, complex roots always occur in complex conjugate* pairs.
p2=[1 7 0 16 25 52]
p2 =
1
7
16
25
52
roots_ p2=roots(p2)
roots_ p2 =
6.5014
2.7428
1.5711
0.3366+ 1.3202i
0.3366 1.3202i
A4
r3 =
1
poly_r3=poly(r3)
poly_r3 =
1
10
35
50
24
We observe that these are the coefficients of the polynomial p 1 x of Example A.1.
Example A.4
It is known that the roots of a polynomial are 1 2 3 4 + j5 and 4 j5 Find the coefficients
of this polynomial.
Solution:
We form a row vector, say r4 , with the given roots, and we find the polynomial coefficients with the
poly(r) function as shown below.
r4=[ 1 2 3 4+5j 45j ]
r4 =
Columns 1 through 4
1.0000
2.0000
3.0000
4.0000+ 5.0000i
Column 5
4.0000 5.0000i
poly_r4=poly(r4)
poly_r4 =
1
14
100
340
499
246
A5
(A.1)
p 5 x = x 3x + 5x 4x + 3x + 2
at x = 3 .
Solution:
p5=[1 3 0 5 4 3 2]; % These are the coefficients
% The semicolon (;) after the right bracket suppresses the display of the row vector
% that contains the coefficients of p5.
%
val_minus3=polyval(p5, 3) % Evaluate p5 at x=3; no semicolon is used here
% because we want the answer to be displayed
val_minus3 =
1280
Example A.6
Let
5
p 1 = x 3x + 5x + 7x + 9
and
6
p 2 = 2x 8x + 4x + 10x + 12
A6
% The coefficients of p1
p2=[2 0 8 0 4 10 12];
% The coefficients of p2
p1p2=conv(p1,p2)
p1p2 =
2
6
8
34
18
24
74
88
78
166
174
108
Therefore,
p 1 p 2 = 2x
11
6x
10
Example A.7
Let
7
p 3 = x 3x + 5x + 7x + 9
and
6
p 4 = 2x 8x + 4x + 10x + 12
0 5 7
9]; p4=[2 8 0
0 4 10 12]; [q,r]=deconv(p3,p4)
q =
0.5000
r =
0
3
Therefore,
q = 0.5
r = 4x 3x + 3x + 2x + 3
A7
p 5 = 2x 8x + 4x + 10x + 12
d
dx
% The coefficients of p5
der_p5=polyder(p5)
der_p5 =
12
32
10
Therefore,
d
 p 5 = 12x 5 32x 3 + 4x 2 + 8x + 10
dx
n1
n2
bn x + bn 1 x
+ bn 2 x
+ } + b1 x + b0
x = R x = Num
m
m1
m2
Den x
am x + am 1 x
+ am 2 x
+ } + a1 x + a0
(A.2)
where some of the terms in the numerator and/or denominator may be zero. We can find the roots
of the numerator and denominator with the roots(p) function as before.
As noted in the comment line of Example A.7, we can write MATLAB statements in one line, if we
separate them by commas or semicolons. Commas will display the results whereas semicolons will
suppress the display.
Example A.9
Let
5
4
2
p num
x 3x + 5x + 7x + 9R x =  = 6
4
2
p den
x 4x + 2x + 5x + 6
Express the numerator and denominator in factored form, using the roots(p) function.
A8
Rational Polynomials
Solution:
num=[1 3 0 5 7 9]; den=[1 0 4 0 2 5 6]; % Do not display num and den coefficients
roots_num=roots(num), roots_den=roots(den)
% Display num and den roots
roots_num =
2.4186+ 1.0712i
2.4186 1.0712i
0.3370+ 0.9961i
0.3370 0.9961i
1.1633
roots_den =
1.6760+0.4922i
1.67600.4922i
1.9304
0.2108+0.9870i
0.21080.9870i
1.0000
We can also express the numerator and denominator of this rational function as a combination of
linear and quadratic factors. We recall that, in a quadratic equation of the form x 2 + bx + c = 0
whose roots are x 1 and x 2 , the negative sum of the roots is equal to the coefficient b of the x term,
that is, x 1 + x 2 = b , while the product of the roots is equal to the constant term c , that is,
x 1 x 2 = c . Accordingly, we form the coefficient b by addition of the complex conjugate roots and
this is done by inspection; then we multiply the complex conjugate roots to obtain the constant term
c using MATLAB as follows:
(2.4186 + 1.0712i)*(2.4186 1.0712i)
ans = 6.9971
(0.3370+ 0.9961i)*(0.33700.9961i)
ans = 1.1058
(1.6760+ 0.4922i)*(1.67600.4922i)
ans = 3.0512
A9
Thus,
2
2
p num
x 4.8372x + 6.9971 x + 0.6740x + 1.1058 x + 1.1633
R x =  = 2
2
p den
x 3.3520x + 3.0512 x + 0.4216x + 1.0186 x + 1.0000 x + 1.9304
We can check this result with MATLABs Symbolic Math Toolbox which is a collection of tools
(functions) used in solving symbolic expressions. They are discussed in detail in MATLABs Users
Manual. For the present, our interest is in using the collect(s) function that is used to multiply two
or more symbolic expressions to obtain the result in polynomial form. We must remember that the
conv(p,q) function is used with numeric expressions only, that is, polynomial coefficients.
Before using a symbolic expression, we must create one or more symbolic variables such as x, y, t,
and so on. For our example, we use the following code:
syms x % Define a symbolic variable and use collect(s) to express numerator in polynomial
form
collect((x^24.8372*x+6.9971)*(x^2+0.6740*x+1.1058)*(x+1.1633))
ans =
x^529999/10000*x^41323/3125000*x^3+7813277909/
1562500000*x^2+1750276323053/250000000000*x+4500454743147/
500000000000
and if we simplify this, we find that is the same as the numerator of the given rational expression in
polynomial form. We can use the same procedure to verify the denominator.
A10
<
<R C
`L F
Z Ohms
Z (rads/s)
Z Ohms
300
39.339
1700
90.603
400
52.589
1800
81.088
500
71.184
1900
73.588
600
97.665
2000
67.513
700
140.437
2100
62.481
800
222.182
2200
58.240
900
436.056
2300
54.611
1000
1014.938
2400
51.428
1100
469.83
2500
48.717
1200
266.032
2600
46.286
1300
187.052
2700
44.122
1400
145.751
2800
42.182
1500
120.353
2900
40.432
1600
103.111
3000
38.845
Plot the magnitude of the impedance, that is, Z versus radian frequency Z .
Solution:
We cannot type Z (omega) in the MATLAB command window, so we will use the English letter w
instead.
If a statement, or a row vector is too long to fit in one line, it can be continued to the next line by typing three or more periods, then pressing <enter> to start a new line, and continue to enter data. This
is illustrated below for the data of w and z. Also, as mentioned before, we use the semicolon (;) to
suppress the display of numbers that we do not care to see on the screen.
The data are entered as follows:
A11
Of course, if we want to see the values of w or z or both, we simply type w or z, and we press
<enter>. To plot z (yaxis) versus w (xaxis), we use the plot(x,y) command. For this example, we
use plot(w,z). When this command is executed, MATLAB displays the plot on MATLABs graph
screen. This plot is shown in Figure A.2.
1200
1000
800
600
400
200
500
1000
1500
2000
2500
A12
command grid toggles them, that is, changes from off to on or vice versa. The default* is off.
box off: This command removes the box (the solid lines which enclose the plot), and box on
restores the box. The command box toggles them. The default is on.
title(string): This command adds a line of the text string (label) at the top of the plot.
xlabel(string) and ylabel(string) are used to label the x and yaxis respectively.
The amplitude frequency response is usually represented with the xaxis in a logarithmic scale. We
can use the semilogx(x,y) command that is similar to the plot(x,y) command, except that the xaxis
is represented as a log scale, and the yaxis as a linear scale. Likewise, the semilogy(x,y) command is
similar to the plot(x,y) command, except that the yaxis is represented as a log scale, and the xaxis as
a linear scale. The loglog(x,y) command uses logarithmic scales for both axes.
Throughout this text it will be understood that log is the common (base 10) logarithm, and ln is the
natural (base e) logarithm. We must remember, however, the function log(x) in MATLAB is the natural logarithm, whereas the common logarithm is expressed as log10(x), and the logarithm to the
base 2 as log2(x).
Let us now redraw the plot with the above options by adding the following statements:
semilogx(w,z); grid; % Replaces the plot(w,z) command
title('Magnitude of Impedance vs. Radian Frequency');
xlabel('w in rads/sec'); ylabel('Z in Ohms')
A default is a particular value for a variable that is assigned automatically by an operating system and remains
in effect unless canceled or overridden by the operator.
A13
1200
1000
800
600
400
200
0
10
10
10
To display the voltage v in a dB scale on the yaxis, we add the relation dB=20*log10(v), and we
replace the semilogx(w,z) command with semilogx(w,dB).
The command gtext(string)* switches to the current Figure Window, and displays a crosshair
that can be moved around with the mouse. For instance, we can use the command gtext(Impedance
Z versus Frequency), and this will place a crosshair in the Figure window. Then, using the
mouse, we can move the crosshair to the position where we want our label to begin, and we press
<enter>.
The command text(x,y,string) is similar to gtext(string). It places a label on a plot in some specific location specified by x and y, and string is the label which we want to place at that location. We
will illustrate its use with the following example that plots a 3phase sinusoidal waveform.
The first line of the code below has the form
* With MATLAB Versions 6 and 7 we can add text, lines and arrows directly into the graph using the tools provided
on the Figure Window.
A14
This function specifies the number of data points but not the increments between data points. An
alternate function is
x=first: increment: last
and this specifies the increments between points but not the number of data points.
The code for the 3phase plot is as follows:
x=linspace(0, 2*pi, 60);
These three waveforms are shown on the same plot of Figure A.4.
1
0.8
sin(x)
0.6
sin(x+2*pi/3)
sin(x+4*pi/3)
0.4
0.2
0
0.2
0.4
0.6
0.8
1
A15
Color
Symbol
Marker
Symbol
Line Style
blue
point
solid line
green
circle
dotted line
red
xmark
dashdot line
cyan
plus
dashed line
magenta
star
yellow
square
black
diamond
white
triangle down
triangle up
triangle left
triangle right
pentagram
hexagram
For example, plot(x,y,'m*:') plots a magenta dotted line with a star at each data point, and
plot(x,y,'rs') plots a red square at each data point, but does not draw any line because no line was
selected. If we want to connect the data points with a solid line, we must type plot(x,y,'rs'). For
additional information we can type help plot in MATLABs command screen.
The plots we have discussed thus far are twodimensional, that is, they are drawn on two axes. MATLAB has also a threedimensional (threeaxes) capability and this is discussed next.
The plot3(x,y,z) command plots a line in 3space through the points whose coordinates are the elements of x, y and z, where x, y and z are three vectors of the same length.
The general format is plot3(x1,y1,z1,s1,x2,y2,z2,s2,x3,y3,z3,s3,...) where xn, yn and zn are vectors
or matrices, and sn are strings specifying color, marker symbol, or line style. These strings are the
same as those of the twodimensional plots.
A16
(A.3)
z = 2x + x + 3y 1
Solution:
We arbitrarily choose the interval (length) shown on the code below.
x= 10: 0.5: 10;
% Length of vector x
y= x;
z= 2.*x.^3+x+3.*y.^21;
plot3(x,y,z); grid
3000
2000
1000
0
1000
2000
10
5
10
5
5
5
10
10
In a twodimensional plot, we can set the limits of the x and yaxes with the axis([xmin xmax
ymin ymax]) command. Likewise, in a threedimensional plot we can set the limits of all three axes
* This statement uses the so called dot multiplication, dot division, and dot exponentiation where the multiplication,
division, and exponential operators are preceded by a dot. These operations will be explained in Section A.8.
A17
(A.4)
Plot the volume of the cone as r and h vary on the intervals 0 d r d 4 and 0 d h d 6 meters.
Solution:
The volume of the cone is a function of both the radius r and the height h, that is,
V = f r h
The threedimensional plot is created with the following MATLAB code where, as in the previous
example, in the second line we have used the dot multiplication, dot division, and dot exponentiation.
This will be explained in Section A.8.
[R,H]=meshgrid(0: 4, 0: 6);
V=(pi .* R .^ 2 .* H) ./ 3; mesh(R, H, V)
xlabel('xaxis, radius r (meters)'); ylabel('yaxis, altitude h (meters)');
zlabel('zaxis, volume (cubic meters)'); title('Volume of Right Circular Cone'); box on
The threedimensional plot of Figure A.6, shows how the volume of the cone increases as the radius
and height are increased.
A18
Subplots
120
100
80
60
40
20
0
6
4
3
2
1
0
This, and the plot of Figure A.5, are rudimentary; MATLAB can generate very sophisticated threedimensional plots. The MATLAB Users manual contains more examples.
A.8 Subplots
MATLAB can display up to four windows of different plots on the Figure window using the command subplot(m,n,p). This command divides the window into an m u n matrix of plotting areas and
chooses the pth area to be active. No spaces or commas are required between the three integers m, n
and p. The possible combinations are shown in Figure A.7.
We will illustrate the use of the subplot(m,n,p) command following the discussion on multiplication, division and exponentiation that follows.
111
Full Screen
221
223
211
212
221 222
212
211
223 224
Default
222
224
221
223
121
122
122
121
222
224
A19
and
B = > b 1 b 2 b 3 } b n @'
be two vectors. We observe that A is defined as a row vector whereas B is defined as a column vector,
as indicated by the transpose operator (c). Here, multiplication of the row vector A by the column
A20
(A.5)
For example, if
A = >1 2 3 4 5@
and
B = > 2 6 3 8 7 @'
the matrix multiplication A*B produces the single value 68, that is,
A
B = 1 u 2 + 2 u 6 + 3 u 3 + 4 u 8 + 5 u 7 = 68
ans =
68
Now, let us suppose that both A and B are row vectors, and we attempt to perform a rowbyrow
multiplication with the following MATLAB statements.
A=[1 2 3 4 5]; B=[2 6 3 8 7];
A*B
When these statements are executed, MATLAB displays the following message:
??? Error using ==> *
Inner matrix dimensions must agree.
Here, because we have used the matrix multiplication operator (*) in A*B, MATLAB expects vector
B to be a column vector, not a row vector. It recognizes that B is a row vector, and warns us that we
cannot perform this multiplication using the matrix multiplication operator (*). Accordingly, we must
perform this type of multiplication with a different operator. This operator is defined below.
2.ElementbyElement Multiplication (multiplication of a row vector by another row vector)
Let
C = > c1 c2 c3 } cn @
and
D = > d1 d2 d3 } dn @
be two row vectors. Here, multiplication of the row vector C by the row vector D is performed with
the dot multiplication operator (.*). There is no space between the dot and the multiplication sym
A21
c2 d2
c3 d3
(A.6)
cn dn @
This product is another row vector with the same number of elements, as the elements of C and D.
As an example, let
C = >1 2 3 4 5@
and
D = > 2 6 3 8 7 @
Dot multiplication of these two row vectors produce the following result.
C.
D = 1 u 2 2 u 6 3 u 3 4 u 8 5 u 7 = 2 12 9 32 35
12
32
35
Similarly, the division (/) and exponentiation (^) operators, are used for matrix division and exponentiation, whereas dot division (./) and dot exponentiation (.^) are used for elementbyelement division and exponentiation.
We must remember that no space is allowed between the dot (.) and the multiplication, division,
and exponentiation operators.
Note: A dot (.) is never required with the plus (+) and minus () operators.
Example A.13
Write the MATLAB code that produces a simple plot for the waveform defined as
y = f t = 3e
4 t
cos 5t 2e
3 t
t
sin 2t + t+1
(A.7)
A22
y=f(t)
1
0
0.5
1.5
2.5
t
3.5
4.5
Had we, in this example, defined the time interval starting with a negative value equal to or less than
1 , say as 3 d t d 3 MATLAB would have displayed the following message:
Warning: Divide by zero.
This is because the last term (the rational fraction) of the given expression, is divided by zero when
t = 1 . To avoid division by zero, we use the special MATLAB function eps, which is a number
approximately equal to 2.2 u 10
16
The command axis([xmin xmax ymin ymax]) scales the current plot to the values specified by the
arguments xmin, xmax, ymin and ymax. There are no commas between these four arguments. This
command must be placed after the plot command and must be repeated for each plot.
The following example illustrates the use of the dot multiplication, division, and exponentiation, the
eps number, the axis([xmin xmax ymin ymax]) command, and also MATLABs capability of displaying up to four windows of different plots.
Example A.14
Plot the functions
y = sin 2x
z = cos 2x
v = sin 2x e cos 2x
A23
1
0.8
0.8
0.6
0.6
0.4
0.4
0.2
0.2
w=(sinx)2 *(cosx)2
0
0
400
0.25
v=(sinx)2/(cosx)2
300
0.2
0.15
200
0.1
100
0.05
0
z=(cosx)2
0
0
A24
<
10 :
Z ab
< 10 PF
0.1 H
` F
b
10 :
With the given values of resistance, inductance, and capacitance, the impedance Z ab as a function of
the radian frequency Z can be computed from the following expression:
4
10 j 10 e Z Z ab = Z = 10 + 5
10 + j 0.1Z 10 e Z
a. Plot Re ^ Z ` (the real part of the impedance Z) versus frequency Z.
(A.8)
A25
The real, imaginary, and polar plots are shown in Figures A.11, A.12, and A.13 respectively.
Example A.15 clearly illustrates how powerful, fast, accurate, and flexible MATLAB is.
1200
1000
Real part of Z
800
600
400
200
0
0
200
400
600
1400
1600
1800
2000
Figure A.11. Plot for the real part of the impedance in Example A.15
A26
600
Imaginary part of Z
400
200
200
400
600
0
200
400
600
1400
1600
1800
2000
Figure A.12. Plot for the imaginary part of the impedance in Example A.15
90
120
1015
60
812
609
Polar Plot of Z
150
30
406
203
180
210
330
240
300
270
A function file is a userdefined function using MATLAB. We use function files for repetitive tasks.
The first line of a function file must contain the word function, followed by the output argument, the
equal sign ( = ), and the input argument enclosed in parentheses. The function name and file name
must be the same, but the file name must have the extension .m. For example, the function file consisting of the two lines below
A27
a realvalued function of a single real variable. MATLAB searches for a value near a point where the
function f changes sign, and returns that value, or returns NaN if the search fails.
Important: We must remember that we use roots(p) to find the roots of polynomials only, such as
those in Examples A.1 and A.2.
fmin(f,x1,x2) minimizes a function of one variable. It attempts to return a value of x where f x is
minimum in the interval x 1 x x2 . The string f contains the name of the function to be minimized.
Note: MATLAB does not have a function to maximize a function of one variable, that is, there is no
fmax(f,x1,x2) function in MATLAB; but since a maximum of f x is equal to a minimum of f x ,
we can use fmin(f,x1,x2) to find both minimum and maximum values of a function.
fplot(fcn,lims) plots the function specified by the string fcn between the xaxis limits specified by
lims = [xmin xmax]. Using lims = [xmin xmax ymin ymax] also controls the yaxis limits. The
string fcn must be the name of an mfile function or a string with variable x .
Solution:
We first plot this function to observe the approximate zeros, maxima, and minima using the following code.
x=1.5: 0.01: 1.5;
y=1./ ((x0.1).^ 2 + 0.01) 1./ ((x1.2).^ 2 + 0.04) 10;
plot(x,y); grid
A28
100
80
60
40
20
0
20
40
1.5
1
0.5
0.5
1.5
Figure A.14. Plot for Example A.16 using the plot command
The roots (zeros) of this function appear to be in the neighborhood of x = 0.2 and x = 0.3 . The
maximum occurs at approximately x = 0.1 where, approximately, y max = 90 , and the minimum
occurs at approximately x = 1.2 where, approximately, y min = 34 .
Next, we define and save f(x) as the funczero01.m function mfile with the following code:
function y=funczero01(x)
% Finding the zeros of the function shown below
y=1/((x0.1)^2+0.01)1/((x1.2)^2+0.04)10;
This plot is shown in Figure A.15. As expected, this plot is identical to the plot of Figure A.14 that
was obtained with the plot(x,y) command.
A29
100
80
60
40
20
0
20
40
1.5
1
0.5
0.5
1.5
Figure A.15. Plot for Example A.16 using the fplot command
We will use the fzero(f,x) function to compute the roots of f x in (A.20) more precisely. The code
below must be saved with a file name, and then invoked with that file name.
x1= fzero('funczero01', 0.2);
x2= fzero('funczero01', 0.3);
fprintf('The roots (zeros) of this function are r1= %3.4f', x1);
fprintf(' and r2= %3.4f \n', x2)
Whenever we use the fmin(f,x1,x2) function, we must remember that this function searches for a
minimum and it may display the values of local minima* , if any, before displaying the function minimum. It is, therefore, advisable to plot the function with either the plot(x,y) or the fplot(fcn,lims)
command to find the smallest possible interval within which the function minimum lies. For this
example, we specify the range 0 d x d 1.5 rather than the interval 1.5 d x d 1.5
The minimum of f(x) is found with the fmin(f,x1,x2) function as follows.
min_val=fmin('funczero01', 0, 1.5)
min_val = 1.2012
* Local maxima or local minima, are the maximum or minimum values of a function within a restricted range of
values in the independent variable. When the entire range is considered, the maxima and minima are considered
be to the maximum and minimum values in the entire range in which the function is defined.
A30
Display Formats
This is the value of x at which y = f x is minimum. To find the value of y corresponding to this
value of x, we substitute it into f x , that is,
x=1.2012; y=1 / ((x0.1) ^ 2 + 0.01) 1 / ((x1.2) ^ 2 + 0.04) 10
y = 34.1812
To find the maximum value, we must first define a new function mfile that will produce f x We
define it as follows:
function y=minusfunczero01(x)
% It is used to find maximum value from f(x)
y=(1/((x0.1)^2+0.01)1/((x1.2)^2+0.04)10);
We have placed the minus () sign in front of the right side of the last expression above, so that the
maximum value will be displayed. Of course, this is equivalent to the negative of the funczero01
function.
Now, we execute the following code to get the value of x where the maximum y = f x occurs.
max_val=fmin('minusfunczero01', 0,1)
max_val = 0.0999
x=0.0999;% Using this value find the corresponding value of y
y=1 / ((x0.1) ^ 2 + 0.01) 1 / ((x1.2) ^ 2 + 0.04) 10
y = 89.2000
A31
Hexadecimal format.
FORMAT + The symbols +,  and blank are printed for positive, negative and zero elements.
Imaginary parts are ignored.
FORMAT BANK Fixed format for dollars and cents.
FORMAT RAT
Spacing:
FORMAT COMPACT Suppress extra linefeeds.
FORMAT LOOSE Puts the extra linefeeds back in.
Some examples with different format displays age given below.
format short 33.3335 Four decimal digits (default)
format long 33.33333333333334 16 digits
format short e 3.3333e+01 Four decimal digits plus exponent
format short g 33.333 Better of format short or format format short e
format bank 33.33 two decimal digits
format + only + or or zero are printed
format rat 100/3 rational approximation
The disp(X) command displays the array X without printing the array name. If X is a string, the text
is displayed.
The fprintf(format,array) command displays and prints both text and arrays. It uses specifiers to
indicate where and in which format the values would be displayed and printed. Thus, if %f is used,
the values will be displayed and printed in fixed decimal format, and if %e is used, the values will be
displayed and printed in scientific notation format. With these commands only the real part of each
parameter is processed.
A32
Appendix B
Differential Equations
his appendix is a review of ordinary differential equations. Some definitions, topics, and examples are not applicable to introductory circuit analysis but are included for continuity of the
subject, and for reference to more advance topics in electrical engineering such as state variables. These are denoted with an asterisk and may be skipped.
(B.1)
(B.2)
and for our example, C = 1 . Then, by substitution of (B.2) into (B.1) we get
dv C
 = I
dt
(B.3)
(B.4)
B1
Differential Equations
We can find the value of the constant k by making use of the initial condition, i.e., at t = 0 , v C = V 0
and (B.4) then becomes
V0 = 0 + k
(B.5)
v C t = It + V 0
(B.6)
This example shows that when a capacitor is charged with a constant current, a linear voltage is produced across the terminals of the capacitor.
Example B.2
Find the current i L t through an inductor whose slope at the coordinate t i L is cos t and the current i L passes through the point S e 2 ,1 .
Solution:
We are given that
di
L = cos t
dt
(B.7)
di L = cos tdt
(B.8)
i L t = sin t + k
(B.9)
(B.10)
B2
(B.11)
Classification
B.2 Classification
Differential equations are classified by:
1. Type Ordinary or Partial
2. Order The highest order derivative which is included in the differential equation
3. Degree The exponent of the highest power of the highest order derivative after the differential
equation has been cleared of any fractions or radicals in the dependent variable and its derivatives
For example, the differential equation
4
2
d y
d y
dy 8
2x
y
d
y4 + 5 + 6 2 + 3  +  = ye
3
3
dx
dx
dx
dx
x +1
dy
d y
2
2
x 2 2 + x  + x n = 0
dt
dt
2
2w y
w y
2 = a 2
wx
wt
Most of the electrical engineering problems are solved with ordinary differential equations with constant coefficients; however, partial differential equations provide often quick solutions to some practical applications as illustrated with the following three examples.
B3
Differential Equations
Example B.3
The equivalent resistance R T of three resistors R 1 , R 2 , and R 3 in parallel is given by
1 + 1 + 1
1 = R1 R2 R3
RT
Similarly,
R 2
wR T R T 2
wR
 = and T = T
wR 2 R 2
wR 3 R 3
B4
Classification
Solution:
wZ
wZ
We will first find the partial derivatives  and  ; then we compute the change in impedance
wR
wX C
and
R dR + X C dX C
wZ
wZ
dZ =  dR +  dX C = wX C
wR
2
2
R + XC
V
120
R =  =  = 192 :
P
75
and since
2
wP 2V
wP
V
V
P =  then  =  and  = 2
R
wV
R
wR
R
B5
Differential Equations
B.3 Solutions of Ordinary Differential Equations (ODE)
A function y = f x is a solution of a differential equation if the latter is satisfied when y and its
derivatives are replaced throughout by f x and its corresponding derivatives. Also, the initial conditions must be satisfied.
For example a solution of the differential equation
2
d y + y = 0
2
dx
is
y = k 1 sin x + k 2 cos x
since y and its second derivative satisfy the given differential equation.
Any linear, timeinvariant electric circuit can be described by an ODE which has the form
n1
d y
d y
dy
 + } + a 1  + a 0 y
a n n + a n 1 n1
dt
dt
dt
m1
(B.12)
d x
d
x
dx
 + } + b 1  + b 0 x
b m m + b m 1 n1
dt
dt
dt
Excitation Forcing Function x t
If the excitation in (B12) is not zero, that is, if x t z 0 , the ODE is called a nonhomogeneous ODE. If
x t = 0 , it reduces to:
n
n1
d y
d y
dy
 + } + a 1  + a 0 y = 0
a n n + a n 1 n1
dt
dt
dt
(B.13)
The differential equation of (B.13) above is called a homogeneous ODE and has n different linearly
independent solutions denoted as y 1 t y 2 t y 3 t } y n t .
We will now prove that the most general solution of (B.13) is:
yH t = k1 y1 t + k2 y2 t + k3 y3 t + } + kn yn t
(B.14)
where the subscript H on the left side is used to emphasize that this is the form of the solution of the
homogeneous ODE and k 1 k 2 k 3 } k n are arbitrary constants.
B6
d y1
d y1
dy
 + } + a 1 1 + a 0 y 1 = 0
+ a n 1 a n n
n1
dt
dt
dt
(B.15)
d
d
d
a n n k 1 y 1 + a n 1  k 1 y 1 + } + a 1  k 1 y 1 + a 0 k 1 y 1
n1
dt
dt
dt
n1
d n y1
d y1
dy
 + } + a 1 1 + a 0 y 1 = 0
a
= k 1 a n +
n1
n
n1
dt
dt
dt
(B.16)
If y = y 1 t and y = y 2 t are any two solutions, then y = y 1 t + y 2 t will also be a solution since
n
n1
n1
d y1
d y1
dy
 + } + a 1 1 + a 0 y 1 = 0
a n + a n 1 n
n1
dt
dt
dt
and
d y2
d y2
dy
 + } + a 1 2 + a 0 y 2 = 0
+ a n 1 a n n
n1
dt
dt
dt
Therefore,
n
n1
d
d
d
y 1 + y 2 + } + a 1  y 1 + y 2 + a 0 y 1 + y 2
a n n y 1 + y 2 + a n 1 n1
dt
dt
dt
n
n1
d
d
d
y + } + a 1  y 1 + a 0 y 1
= a n n y 1 + a n 1 n1 1
dt
dt
dt
n
n1
d
d
d
y + } + a 1  y 2 + a 0 y 2 = 0
+ a n n y 2 + a n 1 n1 2
dt
dt
dt
(B.17)
In general, if
y = k 1 y 1 t k 2 y 1 t k 3 y 3 t } k n y n t
are the n solutions of the homogeneous ODE of (B.13), the linear combination
y = k1 y1 t + k2 y1 t + k3 y3 t + } + kn yn t
is also a solution.
In our subsequent discussion, the solution of the homogeneous ODE, i.e., the complementary solution, will be referred to as the natural response, and will be denoted as y N t or simply y N . The particular solution of a nonhomogeneous ODE will be referred to as the forced response, and will be
Circuit Analysis II with MATLAB Applications
Orchard Publications
B7
Differential Equations
denoted as y F t or simply y F . Accordingly, we express the total solution of the nonhomogeneous
ODE of (B.12) as:
y t = y Natural + y
Response
Forced
= yN + yF
(B.18)
Response
The natural response y N contains arbitrary constants and these can be evaluated from the given initial conditions. The forced response y F , however, contains no arbitrary constants. It is imperative to
remember that the arbitrary constants of the natural response must be evaluated from the total
response.
d y
d y
dy
 + } + a 1  + a 0 y = 0
a n n + a n 1 n1
dt
dt
dt
(B.19)
be of the form
y = ke
st
(B.20)
a n ks e + a n 1 ks
n 1 st
st
e + } + a 1 kse + a 0 ke
st
= 0
or
n
an s + an 1 s
n1
+ } + a 1 s + a 0 ke
st
= 0
(B.21)
an s + an 1 s
n1
+ } + a 1 s + a 0 = 0 or k = 0
or s = f
(B.22)
but the only meaningful solution is the quantity enclosed in parentheses since the latter two yield trivial (meaningless) solutions. We, therefore, accept the expression inside the parentheses as the only
meaningful solution and this is referred to as the characteristic (auxiliary) equation, that is,
n
n1
+ } + a1 s + a0 = 0
(B.23)
an s + an 1 s
Characteristic Equation
Since the characteristic equation is an algebraic equation of an nthpower polynomial, its solutions are
s 1 s 2 s 3 } s n , and thus the solutions of the homogeneous ODE are:
s1 t
s2 t
s3 t
y 1 = k 1 e y 2 = k 2 e y 3 = k 3 e } y n = k n e
B8
sn t
(B.24)
s1 t
+ k2 e
s2 t
+ } + kn e
sn t
(B.25)
s1 t
+ k2 e
s2 t
= k1 e
s1 t
+ k2 e
s1 t
= k 1 + k 2 e
s1 t
= k3 e
s1 t
and we see that one term of (B.25) is lost. In this case, we express one of the terms of (B.25), say
s1 t
s t
as k 2 te 1 . These two represent two independent solutions and therefore the most general solution has the form:
k2 e
y N = k 1 + k 2 t e
s1 t
+ k3 e
s3 t
+ } + kn e
sn t
(B.26)
If there are m equal roots the most general solution has the form:
yN = k1 + k2 t + } + km t
m1
s1 t
+ kn i e
s2 t
+ } + kn e
sn t
(B.27)
s1 t
+ k2 e
s2 t
= k1 e
Dt + jEt
+ k2 e
Dt j Et
= e
Dt
k1 e
jEt
+ k2 e
= e
Dt
= e
Dt
= e
Dt
k 3 cos Et + k 4 sin E t = e
Dt
k 5 cos Et + M
j Et
(B.28)
B9
Differential Equations
If (B.28) is to be a real function of time, the constants k 1 and k 2 must be complex conjugates. The
other constants k 3 , k 4 , k 5 , and the phase angle M are real constants.
The forced response can be found by
a. The Method of Undetermined Coefficients or
b. The Method of Variation of Parameters
We will study the Method of Undetermined Coefficients first.
B.5 Using the Method of Undetermined Coefficients for the Forced Response
For simplicity, we will only consider ODEs of order 2 . Higher order ODEs are discussed in differential equations textbooks.
Consider the nonhomogeneous ODE
2
dy
d
a 2 + b  y + cy = f x
dt
dt
(B.29)
(B.30)
WRONSKIAN DETERMINANT
If (B.30) is true, we can be assured that all solutions of (B.29) are indeed the linear combination of y 1
and y 2 .
The forced response is, in most circuit analysis problems, obtained by observation of the right side of
the given ODE as it is illustrated by the examples that follow.
B10
d y + 4 dy
 + 3y = 0
2
dt
dt
(B.31)
y t = yN t = k1 e + k2 e
3t
(B.32)
The constants k 1 and k 2 are evaluated from the given initial conditions. For this example,
0
y 0 = 3 = k1 e + k2 e
or
k1 + k2 = 3
(B.33)
Also,
y' 0 = 4 = dy
dt
= k 1 e 3k 2 e
t=0
3t
t=0
or
k 1 3k 2 = 4
(B.34)
Simultaneous solution of (B.33) and (B.34) yields k 1 = 6.5 and k 2 = 3.5 . By substitution into
(B.32), we get
t
y t = y N t = 6.5e 3.5e
3t
(B.35)
B11
Differential Equations
The function y = f t is shown in Figure B.1 plotted with the MATLAB command ezplot(y,[0 10]).
Example B.7
Find the total solution of the ODE
2
dy
2t
dy
+ 4  + 3y = 3e
2
dt
dt
(B.36)
yN t = k1 e + k2 e
3t
(B.37)
(We must remember that the constants k 1 and k 2 must be evaluated from the total response).
To find the forced response, we assume a solution of the form
y F = Ae
2t
(B.38)
We can find out whether our assumption is correct by substituting (B.38) into the given ODE of
(B.36). Then,
4Ae
B12
2t
8Ae
2t
+ 3Ae
2t
= 3e
2t
(B.39)
3t
y t = yN + yF = k1 e + k2 e 3 e
2t
(B.40)
The constants k 1 and k 2 are evaluated from the given initial conditions. For this example,
0
y 0 = 1 = k 1 e + k 2 e 3e
or
k1 + k2 = 4
(B.41)
Also,
y' 0 = 1 = dy
dt
= k 1 e 3k 2 e
3t
+ 6e
t=0
2t
t=0
or
k 1 3k 2 = 7
Simultaneous solution of (B.41) and (B.42) yields k 1 = 2.5 and k 2 = 1.5 . By substitution into (B.40),
we get
t
3t
y t = y N + y F = 2.5e + 1.5e 3 e
2t
(B.42)
d y + 6 dy
 + 9y = 0
2
dt
dt
(B.43)
B13
Differential Equations
Solution:
This is a homogeneous ODE and therefore its total solution is just the natural response found from
the characteristic equation s 2 + 6s + 9 = 0 whose roots are s 1 = s 2 = 3 (repeated roots). Thus, the
total response is
y t = yN = k1 e
3t
+ k 2 te
3t
(B.44)
Next, we evaluate the constants k 1 and k 2 from the given initial conditions. For this example,
0
y 0 = 1 = k 1 e + k 2 0 e
or
k1 = 1
(B.45)
Also,
y' 0 = 1 = dy
dt
= 3k 1 e
3t
+ k2 e
t=0
3t
3k 2 te
3t
t=0
or
3k 1 + k 2 = 1
(B.46)
From (B.45) and (B.46) we get yields k 1 = 1 and k 2 = 2 . By substitution into (B.44),
y t = e
B14
3t
2te
3t
(B.47)
Example B.9
Find the total solution of the ODE
2
dy
2t
dy
+ 5  + 6y = 3e
2
dt
dt
(B.48)
Solution:
No initial conditions are given; therefore, we will express the solution in terms of the constants k 1
and k 2 . By inspection, the roots of the characteristic equation of (B.48) are s 1 = 2 and s 2 = 3
and thus the natural response has the form
yN = k1 e
2t
+ k2 e
3t
(B.49)
2t
(B.50)
B15
Differential Equations
We can find out whether our assumption is correct by substitution of (B.50) into the given ODE of
(B.48). Then,
4Ae
2t
10Ae
2t
+ 6Ae
2t
= 3e
2t
(B.51)
but the sum of the three terms on the left side of (B.52) is zero whereas the right side can never be
zero unless we let t o fand this produces a meaningless result.
The problem here is that the right side of the given ODE of (B.48) has the same form as one of the
terms of the natural response of (B.49), namely the term k 1 e 2t .
To work around this problem, we assume that the forced response has the form
y F = Ate
2t
(B.52)
that is, we multiply (B.50) by t in order to eliminate the duplication of terms in the total response.
Then, by substitution of (B.52) into (B.48) and equating like terms, we find that A = 3 . Therefore,
the total response is
y t = yN + yF = k1 e
2t
+ k2 e
3t
+ 3te
2t
(B.53)
Example B.10
Find the total solution of the ODE
d 2y + 5 dy
 + 6y = 4 cos 5t
2
dt
dt
(B.54)
Solution:
No initial conditions are given; therefore, we will express solution in terms of the constants k 1 and
k 2 . We observe that the left side of (B.54) is the same of that of Example B.9. Therefore, the natural
response is the same, that is, it has the form
yN = k1 e
2t
+ k2 e
3t
(B.55)
Next, to find the forced response and we assume a solution of the form
y F = A cos 5t
B16
(B.56)
but this relation is invalid since by equating cosine and sine terms, we find that A = 4 e 19 and also
A = 0 . This inconsistency is a result of our failure to recognize that the derivatives of A cos 5t produce new terms of the form B sin 5t and these terms must be included in the forced response.
Accordingly, we let
y F = k 3 sin 5 t + k 4 cos 5t
(B.57)
Collecting like terms and equating sine and cosine terms, we obtain the following set of equations
19k 3 + 25k 4 = 0
25k 3 19 k 4 = 4
(B.58)
2t
+ k2 e
3t
38
50
+  sin 5t +  cos 5t
493
493
(B.59)
In most engineering problems the right side of the nonhomogeneous ODE consists of elementary
functions such as k (constant), x n where n is a positive integer, e kx , cos kx , sin kx , and linear combinations of these. Table B.1 summarizes the forms of the forced response for a second order ODE
with constant coefficients.
B17
Differential Equations
TABLE B.1 Form of the forced response for 2nd order differential equations
2
dy
d y
Forced Response of the ODE a 2 + b  + cy = f t
dt
dt
Form of Forced Response y F t
f t
k (constant)
K (constant)
k t ( n = positive integer)
ke
rt
K0 t + K1 t
( r =real or complex)
Ke
n rt
k t e cos Dt or k t e sin D t
n1
+ } + Kn 1 t + Kn
rt
K 1 coaDt + K 2 sin Dt
n
K0 t + K1 t
n1
+ K0 t + K1 t
rt
+ } + K n 1 t + K n e cos Dt
n1
rt
+ } + K n 1 t + K n e sin Dt
We must remember that if f t is the sum of several terms, the most general form of the forced
response y F t is the linear combination of these terms. Also, if a term in y F t is a duplicate of a
term in the natural response y N t , we must multiply y F t by the lowest power of t that will eliminate the duplication.
Example B.11
Find the total solution of the ODE
d 2 y + 4 dy
 + 4y = te 2t e 2t
2
dt
dt
(B.60)
Solution:
No initial conditions are given; therefore we will express solution in terms of the constants k 1 and
k 2 . The roots of the characteristic equation are equal, that is, s 1 = s 2 = 2 , and thus the natural
response has the form
yN = k1 e
2 t
+ k 2 te
2 t
(B.61)
To find the forced response (particular solution), we refer to the table of the previous page and from
the last row we choose the term k t n e r t cos Dt . This term with n = 1 , r = 2 , and D = 0 , reduces to
kte
2 t
B18
2 t
(B.62)
But the terms e 2t and te 2t are also present in (B.61); therefore, we multiply (B.62) by t 2 to obtain a
suitable form for the forced response which now is
3
y F = k 3 t + k 4 t e
2 t
(B.63)
Now, we need to evaluate the constants k 3 and k 4 . This is done by substituting (B.63) into the given
ODE of (B.60) and equating with the right side. We use MATLAB do the computations as shown
below.
syms t k3 k4
f0=(k3*t^3+k4*t^2)*exp(2*t);
f1=diff(f0); f1=simple(f1)
f1 =
t*exp(2*t)*(3*k3*t2*k4+2*k3*t^2+2*k4*t)
f2=diff(f0,2); f2=simple(f2)
f2 =
2*exp(2*t)*(3*k3*t+k46*k3*t^24*k4*t+2*k3*t^3+2*k4*t^2)
f=f2+4*f1+4*f0; f=simple(f)
f = 2*(3*k3*t+k4)*exp(2*t)
Finally, we equate f above with the right side of the given ODE, that is
2 3k 3 t + k 4 e
2t
= te
2t
2t
(B.64)
and we find k 3 = 1 e 6 and k 4 = 1 e 2 . By substitution of these values into (B.64) and combining
the forced response with the natural response, we get the total solution
y t = k1 e
2 t
+ k 2 te
2 t
1 3 2 t 1 2 2 t
+  t e  t e
6
2
(B.65)
B19
Differential Equations
B.6 Using the Method of Variation of Parameters for the Forced Response
In certain nonhomogeneous ODEs, the right side f t cannot be determined by the method of
undetermined coefficients. For these ODEs we must use the method of variation of parameters. This
method will work with all linear equations including those with variable coefficients such as
d 2 y + D t dy
 + E t y = f t
2
dt
dt
(B.66)
(B.67)
du
du
1 y1 + 2 y2 = 0
dt
dt
(B.68)
du1 dy1 du 2 dy 2
  +   = f t
dt dt
dt dt
(B.69)
Simultaneous solution of (B.68) and (B.69) will yield the values of du1 e dt and du 2 e dt ; then, integration of these will produce u 1 and u 2 , which when substituted into (B.67) will yield the total solution.
Example B.12
Find the total solution of
dy
d2 y
2 + 4  + 3y = 12
dt
dt
(B.70)
B20
yN = k1 e + k2 e
3 t
(B.71)
y t = yN + yF = k1 e + k2 e
3 t
+4
(B.72)
b. With the method of variation of parameters we start with the natural response found above as
(B.71) and we let the solutions y 1 and y 2 be represented as
y1 = e
and y 2 = e
3t
(B.73)
or
t
y = u1 e + u2 e
3t
(B.74)
or
du 1 t du 2 3t
 e +  e = 0
dt
dt
(B.75)
or
du
du
1 e t + 2 3e 3t = 12
dt
dt
(B.76)
3t
3t
3t
3t
du
12
3e  =  12e
12e = 6e t
1 = =
t
3t
4t
4t
4t
dt
e
3e + e
2e
e
3e
(B.77)
3t
and
B21
Differential Equations
e
t
du 2
12e = 6 e 3t
e
12 =  =  4t
4t
dt
2e
2e
(B.78)
Now, integration of (B.77) and (B.78) and substitution into (B.75) yields
t
u 1 = 6 e dt = 6e + k 1
3t
y = u1 e + u2 e
3t
3t
u 2 = 6 e dt = 2 e + k 2
3t
3t
= 6e + k 1 e + 2 e + k 2 e
t
3t
= 6 + k1 e 2 + k2 e
t
= k1 e + k2 e
3t
(B.79)
(B.80)
+ 4
We observe that the last expression in (B.80) is the same as (B.72) of part (a).
Check with MATLAB:
y=dsolve('D2y+4*Dy+3*y=12')
y =
(4*exp(t)+C1*exp(3*t)*exp(t)+C2)/exp(t)
Example B.13
Find the total solution of
2
d y + 4y = tan 2t
2
dt
(B.81)
j2t
+ k2 e
j 2t
(B.82)
We let
y 1 = cos 2t and y 2 = sin 2t
B22
(B.83)
(B.84)
or
du
du1
 cos 2t + 2 sin 2t = 0
dt
dt
(B.85)
du dy du dy
du
du
1 1 + 2 2 = f t = 1 2 sin 2t + 2 2 cos 2t = tan 2t
dt dt
dt dt
dt
dt
(B.86)
sin 2t
sin 2t
2
du1
tan
2t
2
cos
2t
cos 2t
sin 2t=
 =  = 2
2
cos 2t
sin 2t
2 cos 2t
dt
2 cos 2t + 2 sin 2t
2 sin 2t
2 cos 2t
(B.87)
and
cos 2t
du 2
sin 2t
2 sin 2t
tan 2t =  = 2
2
dt
(B.88)
Now, integration of (B.87) and (B.88) and substitution into (B.84) yields
2
1 sin 2t
sin 2t 1
u 1 =   dt =   ln sec 2t + tan 2t + k 1
4
2 cos 2t
4
(B.89)
1
cos 2t + k
u 2 =  sin 2t dt = 2
2
4
(B.90)
sin 2t cos 2t 1
sin 2t cos 2t
y = u1 y 1 + u 2 y 2 =   cos 2t ln sec 2t + tan 2t + k 1 cos 2t  + k 2 sin 2t
4
4
4
1
=  cos 2t ln sec 2t + tan 2t + k 1 cos 2t + k 2 sin 2t
4
(B.91)
B23
Differential Equations
B.7 Exercises
Solve the following ODEs by any method.
2
dy
1. dy2 + 4  + 3y = t 1
dt
dt
1
3
2
dy
2. dy2 + 4  + 3y = 4e t
dt
dt
Answer: y = k 1 e t + k 2 e 3t + 2te t
2
dy
1
2
2
3. dy2 + 2  + y = cos t Hint: Use cos t =  cos 2t + 1
dt
dt
cos 2t 4 sin 2t
Answer: y = k 1 e t + k 2 te t + 1 32
50
2
d y
 + y = sec t
4. 2
dt
B24
Appendix C
Matrices and Determinants
or
1 3 1
2 1 5
4 7 6
a 31 a 32 a 33 } a 3 n
} } } } }
a m 1 a m 2 a m 3 } a mn
(C.1)
The numbers a ij are the elements of the matrix where the index i indicates the row, and j indicates
the column in which each element is positioned. Thus, a 43 indicates the element positioned in the
fourth row and third column.
A matrix of m rows and n columns is said to be of m u n order matrix.
If m = n , the matrix is said to be a square matrix of order m (or n). Thus, if a matrix has five rows
and five columns, it is said to be a square matrix of order 5.
In a square matrix, the elements a 11 a 22 a 33 } a nn are called the main diagonal elements. Alternately, we say that the matrix elements a 11 a 22 a 33 } a nn , are located on the main diagonal.
C1
i = 1 2 3 } m
j = 1 2 3 } n
(C.2)
Two matrices are said to be conformable for addition (subtraction), if they are of the same order m u n.
If A = a ij and B = b ij are conformable for addition (subtraction), their sum (difference) will be
another matrix C with the same order as A and B , where each element of C is the sum (difference)
of the corresponding elements of A and B , that is,
C = A r B = > a ij r b ij @
(C.3)
Example C.1
Compute A + B and A B given that
A = 1 2 3 and B = 2 3 0
1 2 5
0 1 4
Solution:
A+B = 1+2
01
2+3
1+2
3+0 = 3 5
4+5
1 3
AB = 12
0+1
2 3 3 0 = 1 1 3
12 45
1 1 1
3
9
and
Henceforth, all paragraphs and topics preceded by a dagger ( ) may be skipped. These are discussed in matrix
theory textbooks.
C2
Matrix Operations
ans =
3
1
5
3
3
9
AB
% Subtract B from A
ans =
1
1
1
1
3
1
If k is any scalar (a positive or negative number), and not [k] which is a 1 u 1 matrix, then multiplication of a matrix A by the scalar k is the multiplication of every element of A by k .
Example C.2
Multiply the matrix
A = 1 2
2 3
by
a. k 1 = 5
b. k 2 = 3 + j2
Solution:
a.
k 1 A = 5 u 1 2 = 5 u 1 5 u 2 = 5 10
2 3
5u2 5u3
10 15
b.
k 2 A = 3 + j2 u 1 2 = 3 + j2 u 1 3 + j2 u 2 = 3 + j2
2 3
3 + j2 u 2 3 + j2 u 3
6 + j4
6 j4
9 + j6
C3
ans =
3.0000+ 2.0000i
6.0000+ 4.0000i
6.0000 4.0000i
9.0000+ 6.0000i
Two matrices A and B are said to be conformable for multiplication A B in that order, only when the
number of columns of matrix A is equal to the number of rows of matrix B . That is, the product
A B (but not B A ) is conformable for multiplication only if A is an m u p matrix and matrix B is
an p u n matrix. The product A B will then be an m u n matrix. A convenient way to determine if
two matrices are conformable for multiplication is to write the dimensions of the two matrices sidebyside as shown below.
Shows that A and B are conformable for multiplication
A
mup
B
pun
A
mup
For matrix multiplication, the operation is row by column. Thus, to obtain the product A B , we
multiply each element of a row of A by the corresponding element of a column of B ; then, we add
these products.
Example C.3
Matrices C and D are defined as
1
C = 2 3 4 and D = 1
2
Compute the products C D and D C
C4
The dimensions for D and C are respectively 3 u 1 1 u 3 and therefore, the product D C is also
feasible. Multiplication of these will produce a 3 u 3 matrix as follows:
1
1 2 1 3 1 4
2 3 4
D C = 1 2 3 4 = 1 2 1 3 1 4 = 2 3 4
2
2 2 2 3 2 4
4 6 8
ans =
7
D*C
% Multiply D by C
ans =
2
2
4
3
3
6
4
4
8
Division of one matrix by another, is not defined. However, an equivalent operation exists, and it
will become apparent later in this chapter, when we discuss the inverse of a matrix.
C5
A =
a 11 a 12 a 13 } a 1 n
0 a 22 a 23 } a 2 n
(C.4)
0 0 } } }
} } 0 } }
0 0 0 } a mn
A square matrix is said to be lower triangular, when all the elements above the diagonal are zero.
The matrix B of (C.5) is a lower triangular matrix.
a 11
B =
0 } 0
0 } 0
a 21 a 22
}
}
am1
}
}
am 2
(C.5)
} 0 0
} } 0
a m 3 } a mn
In a lower triangular matrix, not all elements below the diagonal need to be nonzero.
A square matrix is said to be diagonal, if all elements are zero, except those in the diagonal. The
matrix C of (C.6) is a diagonal matrix.
a 11 0
C =
0 } 0
0 a 22 0 } 0
0
0
0
0 } 0 0
0 0 } 0
0 0 } a mn
(C.6)
A diagonal matrix is called a scalar matrix, if a 11 = a 22 = a 33 = } = a nn = k where k is a scalar. The matrix D of (C.7) is a scalar matrix with k = 4.
4
D = 0
0
0
0
4
0
0
0
0
4
0
0
0
0
4
(C.7)
A scalar matrix with k = 1 , is called an identity matrix I . Shown below are 2 u 2 , 3 u 3 , and 4 u 4
identity matrices.
C6
1 0
0 1
1 0 0 0
0 1 0 0
0 0 1 0
0 0 0 1
(C.8)
ans =
1
0
0
0
0
1
0
0
0
0
1
0
0
0
0
1
Likewise, the eye(size(A)) function, produces an identity matrix whose size is the same as matrix
A . For example, let matrix A be defined as
A=[1 3 1; 2 1 5; 4 7 6]
% Define matrix A
A =
1
2
4
3
1
7
1
5
6
0
1
0
0
0
1
then,
eye(size(A))
displays
ans =
1
0
0
The transpose of a matrix A , denoted as A T , is the matrix that is obtained when the rows and columns of matrix A are interchanged. For example, if
A=
1
1 2 3 then A T =
2
4 5 6
3
4
5
6
(C.9)
C7
% Define matrix A
A =
1
4
2
5
3
6
A'
ans =
1
2
3
4
5
6
A symmetric matrix A is a matrix such that A T = A , that is, the transpose of a matrix A is the same
as A . An example of a symmetric matrix is shown below.
1 2 3
A = 2 4 5
3 5 6
1 2 3
A = 2 4 5 = A
3 5 6
T
(C.10)
If a matrix A has complex numbers as elements, the matrix obtained from A by replacing each
element by its conjugate, is called the conjugate of A , and it is denoted as A
An example is shown below.
A = 1 + j2
3
j
2 j3
A
= 1 j 2
3
j
2 + j3
MATLAB has two builtin functions which compute the complex conjugate of a number. The
first, conj(x), computes the complex conjugate of any complex number, and the second, conj(A),
computes the conjugate of a matrix A . Using MATLAB with the matrix A defined as above, we
get
A = [1+2j j; 3 23j]
A =
1.0000+ 2.0000i
3.0000
conj_A=conj(A)
conj_A =
1.0000 2.0000i
C8
Determinants
3.0000
2.0000+ 3.0000i
A =
0 2
2 0
3 4
3
4 = A
0
1j
3
j
2
1
T
A
=
j
1j
0
2
1+j
3
j
2
1
T*
A
=
j
1j
0
2
1+j
3
j
2
j = A
0
1j
3j
j
2
j
T
A
=
j
1j
0
2
1j
3j
j
2
j
T*
A
=
j
1+j
0
2
1+j
3j
j
2
j = A
0
C.4 Determinants
Let matrix A be defined as the square matrix
a 11 a 12 a 13 } a 1 n
a 21 a 22 a 23 } a 2 n
A = a a a } a
31 32 33
3n
} } } } }
a n 1 a n 2 a n 3 } a nn
(C.11)
(C.12)
C9
a 11 a 12
(C.13)
a 21 a 22
Then,
detA = a 11 a 22 a 21 a 12
(C.14)
Example C.4
Matrices A and B are defined as
A = 1 2 and B = 2 1
3 4
2 0
ans =
2
det(B)
ans =
2
(C.15)
a 31 a 32 a 33
C10
Determinants
detA = a 11 a 22 a 33 + a 12 a 23 a 31 + a 11 a 22 a 33
(C.16)
a 11 a 22 a 33 a 11 a 22 a 33 a 11 a 22 a 33
A convenient method to evaluate the determinant of order 3, is to write the first two columns to the
right of the 3 u 3 matrix, and add the products formed by the diagonals from upper left to lower
right; then subtract the products formed by the diagonals from lower left to upper right as shown
on the diagram of the next page. When this is done properly, we obtain (C.16) above.
a 11 a 12 a 13 a 11 a 12
a 21 a 22 a 23 a 21 a 22
a 31 a 32 a 33 a 31 a 32
This method works only with second and third order determinants. To evaluate higher order determinants, we must first compute the cofactors; these will be defined shortly.
Example C.5
Compute detA and detB if matrices A and B are defined as
2
A = 1
2
3 5
0 1
1 0
2 3 4
0 2
0 5 6
and B = 1
Solution:
2
detA = 1
2
3 5 2 3
0 1 1 0
1 0 2 1
or
detA = 2 u 0 u 0 + 3 u 1 u 1 + 5 u 1 u 1
2 u 0 u 5 1 u 1 u 2 0 u 1 u 3 = 11 2 = 9
Likewise,
2 3 4 2 3
detB = 1 0 2 1 2
0 5 6 2 6
or
detB = > 2 u 0 u 6 @ + > 3 u 2 u 0 @ + > 4 u 1 u 5 @
> 0 u 0 u 4 @ > 5 u 2 u 2 @ > 6 u 1 u 3 @ = 20 38 = 18
C11
ans =
9
B=[2 3 4; 1 0 2; 0 5 6];det(B) % Define matrix B and compute detB
ans =
18
(C.17)
If we remove the elements of its ith row, and jth column, the remaining n 1 square matrix is called
the minor of A , and it is denoted as M ij .
The signed minor 1
i+j
Example C.6
Matrix A is defined as
a 11 a 12 a 13
A = a 21 a 22 a 23
(C.18)
a 31 a 32 a 33
M 12 ,
Solution:
C12
a 22 a 23
a 21 a 23
M 12 =
a 32 a 33
M 11 =
a 31 a 33
a 21 a 22
a 31 a 32
and
D 11 = 1
1+1
M 11 = M 11
D 12 = 1
1+2
M 12 = M 12
D 13 = M 13 = 1
1+3
M 13
M 22
M 23
M 31
M 32
M 33
and cofactors
D 21 D 22 D 23 D 31 D 32 and D 33
1 2 3
2 4 2
1 2 6
(C.19)
Solution:
D 11 = 1
1+1
D 13 = 1
D 22 = 1
D 12 = 1
1+2
2 2 = 10
1 6
(C.20)
2 4 = 0
1 2
D 21 = 1
2+1
2 3 = 6
2 6
(C.21)
1 3 = 9
1 6
D 23 = 1
2+3
1 2 = 4
1 2
(C.22)
2 3 = 8
4 2
D 32 = 1
3+2
1 3 = 8
2 2
(C.23)
1+3
2+2
D 31 = 1
4 2 = 20
2 6
3+1
D 33 = 1
3+3
1 2 = 8
2 4
(C.24)
It is useful to remember that the signs of the cofactors follow the pattern
Circuit Analysis II with MATLAB Applications
Orchard Publications
C13
+
+
+
+
+
+
+
that is, the cofactors on the diagonals have the same sign as their minors.
Let A be a square matrix of any size; the value of the determinant of A is the sum of the products
obtained by multiplying each element of any row or any column by its cofactor.
Example C.8
Matrix A is defined as
A =
1 2 3
2 4 2
1 2 6
(C.25)
ans =
40
We must use the above procedure to find the determinant of a matrix A of order 4 or higher. Thus, a
fourthorder determinant can first be expressed as the sum of the products of the elements of its first
row by its cofactor as shown below.
C14
a 21 a 22 a 23 a 24
a 31 a 32 a 33 a 34
a 22 a 23 a 24
a 12 a 13 a 14
= a 11 a 32 a 33 a 34 a 21 a 32 a 33 a 34
a 42 a 43 a 44
a 41 a 42 a 43 a 44
(C.26)
a 42 a 43 a 44
a 12 a 13 a 14
a 12 a 13 a 14
+a 31 a 22 a 23 a 24 a 41 a 22 a 23 a 24
a 42 a 43 a 44
a 32 a 33 a 34
3
1
2
1
(C.27)
Solution:
Using the above procedure, we will multiply each element of the first column by its cofactor. Then,
>a@
1 0 3
+4 1 0 1
0 0 1
1 0 3
3 1 0 1
0 3 2
1 0 3
1 0 3 2
0 0 1
1 0 1
A=2 0 3 2
0 0 1
>b@
>c@
>d@
and thus
detA = > a @ + > b @ + > c @ + > d @ = 6 3 + 0 36 = 33
C15
4
6
2
1
1
1
(C.28)
then,
detA =
2
3
1
4
6
2
1
1
1
2
3
1
4
6 = 12 + 4 + 6 6 4 12 = 0
2
(C.29)
Here, detA is zero because the second column in A is 2 times the first column.
Check with MATLAB:
A=[2 4 1; 3 6 1; 1 2 1];det(A)
ans =
0
Property 3: If two rows or two columns of a matrix are identical, the determinant is zero. This follows
from Property 2 with m = 1 .
(C.30)
a 31 x + a 32 y + a 33 z = C
and let
a 11 a 12 a 13
'
a 21 a 22 a 23
a 31 a 32 a 33
C16
A a 11 a 13
D1 =
B a 21 a 23
C a 31 a 33
a 11 A a 13
D2 =
a 21 B a 23
a 31 C a 33
a 11 a 12 A
D3 =
a 21 a 22 B
a 31 a 32 C
Cramers Rule
Cramers rule states that the unknowns x, y, and z can be found from the relations
D
x = 1
D
y = 2
'
'
D
z = 3
'
(C.31)
(C.32)
v 2 + 3v 1 4 v 3 = 0
(C.33)
3v 1 + v 2 v 3 = 4
'
D1 =
2 1 3
4 3 2
3 1 1
2 1
4 3 = 6 + 6 12 + 27 + 4 + 4 = 35
3 1
5 1 3
8 3 2
4 1 1
5 1
8 3 = 15 + 8 + 24 + 36 + 10 8 = 85
4 1
C17
D2 =
2
4
3
5 3
8 2
4 1
D3 =
2 1
4 3
3 1
5
8
4
2
4
3
5
8 = 16 30 48 72 + 16 20 = 170
4
2 1
4 3 = 24 24 20 + 45 16 16 = 55
3 1
D
x 2 = 2 = 170
 = 34
7
35
'
D
x 3 = 3 = 55
 = 11
7
35
'
(C.34)
C18
(C.35)
3v 1 + v 2 v 3 = 4
Solution:
As a first step, we add the first equation of (C.35) with the third to eliminate the unknown v2 and we
obtain the following equation.
5v 1 + 2v 3 = 9
(C.36)
Next, we multiply the third equation of (C.35) by 3, and we add it with the second to eliminate v 2 .
Then, we obtain the following equation.
5v 1 5v 3 = 20
(C.37)
(C.38)
Now, we can find the unknown v 1 from either (C.36) or (C.37). By substitution of (C.38) into (C.36)
we get
17
11
5v 1 + 2  = 9 or v 1 =  7
7
(C.39)
Finally, we can find the last unknown v 2 from any of the three equations of (C.35). By substitution
into the first equation we get
Circuit Analysis II with MATLAB Applications
Orchard Publications
C19
(C.40)
(C.41)
We observe that the cofactors of the elements of the ith row (column) of A are the elements of the
ith column (row) of adjA .
Example C.12
Compute adjA if Matrix A is defined as
1 2 3
A = 1 3 4
1 4 3
(C.42)
Solution:
C20
3
4
4
3
2
4
3
3
adjA = 1
1
4
3
1
1
3
3
1
1
3
4
1 2
1 4
2 3
3 4
2
3
3
4
7 6 1
1 0 1
1 2 1
1 2
1 3
2 3
3 4
5 7
(C.43)
1
2
3
2 3
3 4
5 7
1 2
2 3 = 21 + 24 + 30 27 20 28 = 0
3 5
1
=  adjA
detA
(C.44)
Example C.14
Matrix A is defined as
1 2 3
A = 1 3 4
1 4 3
(C.45)
C21
Then,
7 6 1
3.5 3 0.5
1
1
=  adjA =  1 0 1 = 0.5 0 0.5
2
detA
1 2 1
0.5 1 0.5
(C.46)
A =
1
1
1
invA =
3.5000
0.5000
0.5000
2
3
4
3
4
3
3.0000
0
1.0000
0.5000
0.5000
0.5000
Multiplication of a matrix A by its inverse A 1 produces the identity matrix I , that is,
AA
= I or A A = I
(C.47)
Example C.15
Prove the validity of (C.47) for the Matrix A defined as
A = 4
2
3
2
Proof:
detA = 8 6 = 2 and adjA =
C22
2 3
2 4
1
1
1 3 e 2
=  adjA =  2 3 =
2 2 4
detA
1
2
and
AA
3 1 3 e 2 = 4 3
2 1
2
22
= 4
2
6+6 = 1
3+4
0
0 = I
1
(C.48)
where A and B are matrices whose elements are known, and X is a matrix (a column vector) whose
elements are the unknowns. We assume that A and X are conformable for multiplication. Multiplication of both sides of (C.48) by A 1 yields:
1
A AX = A B = IX = A B
(C.49)
or
1
X=A B
(C.50)
Therefore, we can use (C.50) to solve any set of simultaneous equations that have solutions. We will
refer to this method as the inverse matrix method of solution of simultaneous equations.
Example C.16
For the system of the equations
2x 1 + 3x 2 + x 3 = 9
x 1 + 2x 2 + 3x 3 = 6
3x 1 + x 2 + 2x 3 = 8
(C.51)
C23
x1
9
3 1
=
B
=
X
x
6
2 3
2
8
1 2
x3
(C.52)
Then,
1
X = A B
(C.53)
or
x1
x2
x3
2
= 1
3
3 1
2 3
1 2
9
6
8
(C.54)
1 5 7
7 1 5
5 7 1
Therefore,
A
1 5 7
1
1
=  adjA =  7 1 5
detA
18
5 7 1
1 5 7 9
35
35 e 18
1.94
11=
=
= 29 e 18 = 1.61
18 7 1 5 6
18 29
5 7 1 8
5
5 e 18
0.28
(C.55)
To verify our results, we could use the MATLABs inv(A) function, and then multiply A 1 by B.
However, it is easier to use the matrix left division operation X = A \ B ; this is MATLABs solution
of A 1 B for the matrix equation A X = B , where matrix X is the same size as matrix B. For this
example,
A=[2 3 1; 1 2 3; 3 1 2]; B=[9 6 8]';
X=A \ B
C24
Example C.17
For the electric circuit of Figure C.1,
1:
V = 100 v
2:
2:
I1
4:
9:
9:
I3
I2
= 100
9I 1 + 20I 2 9I 3 =
9I 2 + 15I 3 =
(C.56)
Use the inverse matrix method to compute the values of the currents I 1 , I 2 , and I 3
Solution:
For this example, the matrix equation is RI = V or I = R 1 V , where
10 9 0
100
R = 9 20 9 V =
0
0 9 15
0
I1
and I = I 2
I3
1
=  adjR
detR
(C.57)
Therefore, we find the determinant and the adjoint of R . For this example, we find that
C25
(C.58)
Then,
R
219 135 81
1
1
=  adjR =  135 150 90
975
detR
81 90 119
and
I1
We can also use subscripts to address the individual elements of the matrix. Accordingly, the above
code could also have been written as:
R(1,1)=10; R(1,2)=9;
% No need to make entry for A(1,3) since it is zero.
R(2,1)=9; R(2,2)=20; R(2,3)=9; R(3,2)=9; R(3,3)=15; V=[100 0 0]'; I=R\V
I =
22.4615
13.8462
8.3077
Spreadsheets also have the capability of solving simultaneous equations using the inverse matrix
method. For instance, we can use Microsoft Excels MINVERSE (Matrix Inversion) and MMULT
(Matrix Multiplication) functions, to obtain the values of the three currents in Example C.17.
The procedure is as follows:
1. We start with a blank spreadsheet and in a block of cells, say B3:D5, we enter the elements of
matrix R as shown in Figure C.2. Then, we enter the elements of matrix V in G3:G5.
C26
and we press the CrtlShiftEnter keys simultaneously. The values of I then appear in G7:G9.
A
B
C
D
E
F
G
H
1 Spreadsheet for Matrix Inversion and Matrix Multiplication
2
10
9
0
100
3
R=
9
20
9
V=
0
4
0
9
15
0
5
6
0.225 0.138 0.083
22.462
7
1
R = 0.138 0.154 0.092
8
I= 13.846
9
0.083 0.092 0.122
8.3077
10
Example C.18
For the phasor circuit of Figure C.18
170q
85 :
j100 :
R3 = 100 :
L R2
V2
C
IX
V1
VS
R1
j200 :
50 :
C27
(C.59)
and the voltages V1 and V2 can be computed from the nodal equations
V 1 170 0q V 1 V 2 V 1 0
+  +  = 0
85
100
j200
(C.60)
V 2 170 0q V 2 V 1 V 2 0
+  +  = 0
j100
100
50
(C.61)
and
Compute, and express the current I x in both rectangular and polar forms by first simplifying like
terms, collecting, and then writing the above relations in matrix form as YV = I , where
Y = Admit tan ce , V = Voltage , and I = Current
Solution:
The Y matrix elements are the coefficients of V 1 and V 2 . Simplifying and rearranging the nodal equations of (C.60) and (C.61), we get
0.0218 j0.005 V 1 0.01V 2 = 2
(C.62)
V2
2
j1.7
(C.63)
V1
0.0218 j0.005
0.01
0.01
0.03 + j0.01
C28
IX =
0.5149 0.0590i
magIX =
0.5183
thetaIX=angle(IX)*180/pi
thetaIX =
6.5326
Spreadsheets have limited capabilities with complex numbers, and thus we cannot use them to compute matrices that include complex numbers in their elements as in Example C.18
C29
5 9 3
B = 2 8 2
7 4 6
4 6
C= 3 8
5 2
D =
1 2 3
3 6 4
1. Perform the following computations, if possible. Verify your answers with MATLAB.
a. A + B
b. A + C
c. B + D
d. C + D
e. A B
f. A C
g. B D
h. C D
2. Perform the following computations, if possible. Verify your answers with MATLAB.
a. A B
b. A C
c. B D
d. C D
e. B A
f. C A
g. D A
h. D C
3. Perform the following computations, if possible. Verify your answers with MATLAB.
a. detA
b. detB
c. detC
e. det A B
f. det A C
d. detD
4. Solve the following systems of equations using Cramers rule. Verify your answers with MATLAB.
x 1 + 2x 2 3x 3 + 5x 4 = 14
x 1 2x 2 + x 3 = 4
a.
2x 1 + 3x 2 + x 3 = 9
3x 1 + 4x 2 5x 3 = 0
b.
x 1 + 3x 2 + 2x 3 x 4 = 9
3x 1 3 x 2 + 2x 3 + 4x 4 = 19
4x 1 + 2x 2 + 5x 3 + x 4 = 27
C30
2 4 3
b. 2 4 1
1 3 4
2 2 2
x1
2
1
3 x 2 = 10
x3
2
14
1
7
x4
Appendix D
Constructing Semilog Plots with Microsoft Excel
his appendix contains instructions for constructing semilog plots with the Microsoft Excel
spreadsheet. Semilog, short for semilogarithmic, paper is graph paper having one logarithmic
and one linear scale. It is used in many scientific and engineering applications including frequency response illustrations and Bode Plots.
Menu bar
ChartWizard
D1
Menu bar
ChartWizard
Handles
D2
Maximum: 100000
Major Unit: 10
Minor Unit: 10
Make sure that the squares to the left of these values are not checked.
Click on Logarithmic scale to place a check mark, and click on OK to return to the plot.
11. Click anywhere near the yaxis (leftmost vertical line on the plot) and observe that the Chart
Objects Edit Box now displays Value (Y) axis. Click on the Format Chart Area tool which now
displays Format Axis, click on the Scale tab and make the following entries:
Minimum: 80
Maximum: 80
Major Unit: 20
Minor Unit: 20
Make sure that the squares to the left of these values are not checked. Also, make sure that the
Logarithmic scale is not checked. Check on OK to return to the plot.
12. You will observe that the xaxis values appear at the middle of the plot. To move them below the
plot, click on Format Chart Area tool, click on the Patterns tab, click on Tick mark labels (lower
right section), and click on OK to return to the plot area.
13. To expand the plot so that it will look more useful and presentable, make sure that the chart is
selected (the handles are visible). This is done by clicking anywhere in the chart area. Bring the
mouse close to the lower center handle until a bidirectional arrow appears and stretch downwards. Repeat with the right center handle to stretch the plot to the right. Alternately, you may
bring the mouse near the lower right handle and stretch the plot diagonally.
14. You may wish to display the xaxis values in exponential (scientific) format. To do that, click anywhere near the xaxis (zero point), and observe that the Chart Objects Edit Box now displays
Value (X) axis. Click on the Format Chart Area tool which now displays Format Axis, click on the
Number tab and under Category click on Scientific with zero decimal places.
D3
D4
1.E+01
1.E+02
1.E+03
1.E+04
1.E+05
Appendix E
Scaling
his appendix discusses magnitude and frequency scaling procedures that allow us to transform
circuits that contain passive devices with unrealistic values to equivalent circuits with realistic
values.
If we increase the input impedance by a factor k m , we must increase the impedance of each device of
the network by the same factor. Thus, if a network consists of R , L , and C devices and we wish to
scale this network by this factor, the magnitude scaling process entails the following transformations
where the subscript m denotes magnitude scaling.
Rm o km R
Lm o km L
(E.1)
C
C m o km
These transformations are consistent with the timedomain to frequency domain transformations
RoR
L o jZL
(E.2)
1
C o jZC
(E.3)
1
C o sC
new frequency the impedance of each device has the same value as at the original frequency. The freCircuit Analysis II with MATLAB Applications
Orchard Publications
E1
Appendix E Scaling
quency scaling factor is denoted as k f . This factor is also a real positive number and can be greater or
smaller than unity.
The resistance value is independent of the frequency. However, the complex impedance of any
inductor is sL , and in order to maintain the same impedance at a frequency k f times as great, we must
replace the inductor value by another which is equal to L e k f . Similarly, a capacitor with value C must be
replaced with another having a capacitance value equal to C e k f . For frequency scaling then, the following
transformations are necessary where the subscript f denotes magnitude scaling.
Rf o R
L
L f o kf
(E.4)
C
C f o kf
A circuit can be scaled simultaneously in both magnitude and frequency using the scales values below
where the subscript mf denotes simultaneous magnitude and frequency scaling.
R mf o k m R
km
L mf o  L
kf
(E.5)
1
C mf o  C
km kf
Example E.1
L
2.5 :
C
0.5 H
2F
E2
Frequency Scaling
e. the magnitude of the input impedance Z , and using MATLAB sketch it as a function of frequency.
f. Scale this circuit so that the impedance will have a maximum value of 5 K: at a resonant fre6
quency of 5 u 10 rad e s
Solution:
a. The resonant frequency of the given circuit is
1
Z 0 =  = 1 rad e s
LC
e. The magnitude of the input impedance versus radian frequency Z is shown in Figure E.2 and was
generated with the MATLAB code below.
w=0.01: 0.005: 5; R=2.5; G=1/R; C=2; L=0.5; Y=G+j.*(w.*C1./(w.*L));...
magY=abs(Y); magZ=1./magY; plot(w,magZ); grid
E3
Appendix E Scaling
Then,
L m = k m L = 2000 u 0.5 = 1000 H
and
3
C
2
C m =  =  = 10 F
km
2000
After being scaled in magnitude by the factor k m = 2000 , the network constants are as shown in
Figure E.3, and the plot is shown in Figure E.4.
R
5K:
10 3 H
10 3 F
Figure E.3. The network of Figure E.2 scaled by the factor k m = 2000
6
The final step is to scale the above circuit to 5 u 10 rad e s . Using (E.4), we get:
R f = R = 5 k:
6
L f = L e k f = 1000 e 5 u 10 = 200 PH
E4
Frequency Scaling
Figure E.4. Plot for the network of Figure E.2 after being scaled by the factor k m = 2000
C f = C e k f = 10
e 5 u 10 = 200 pF
The network constants and its response, in final form, are as shown in Figures E.5 and E.6 respectively.
Z
5 K:
200 PH 200 pF
Figure E.5. The network of Figure E.2 scaled to its final form
The plot of Figure E.6 was generated with the following MATLAB code:
w=1: 10^3: 10^7; R=5000; G=1/R; C=200.*10.^(12); L=200.*10.^(6); ...
magY=sqrt(G.^2+(w.*C1./(w.*L)).^2); magZ=1./magY; plot(w,magZ); grid
Check:
The resonant frequency of the scaled circuit is
6
1
1
1
Z 0 =  =  =  = 5 u 10 rad e s
6
3
9
LC
0.2 u 10
0.2 u 10 u 0.2 u 10
E5
Appendix E Scaling
Figure E.6. Plot for Example E.1 scaled to its final form
The values of the circuit devices could have been obtained also by direct application of (E.5), that is,
R mf o k m R
km
L mf o  L
kf
km
C mf o  C
kf
R mf = k m R = 2000 u 2.5 = 5 K:
km
2000
L mf =  L = 6 u 0.5 = 200 PH
kf
5 u 10
1
1
 u 2 = 200 pf
C mf =  C = 3
6
km kf
2 u 10 u 5 u 10
E6
Frequency Scaling
Example E.2
and we want the resonant frequency of the scaled circuit to be 500 Hz or 2S u 500 = 3142 rad e s .
Therefore, the frequency scaling factor must be
3142
k f =  = 3142
1
Now, we must compute the magnitude scale factor, and since we want the capacitor value to be 2 PF ,
we use (E.5), that is,
1
C mf =  C
km kf
or
C
1
k m =  =  = 159
6
k f C mf
3142 u 2 u 10
Then, the scaled values for the resistance and inductance are
R m = k m R = 159 u 1 = 159 :
and
km
159
L mf =  L =  u 1 = 50.6 mH
3142
kf
E7
Appendix E Scaling
E.3 Exercises
1. A series resonant circuit has a bandwidth of 100 rad e s , Q 0s = 20 and C = 50 PF . Compute the
new resonant frequency and inductance if the circuit is scaled
a. in magnitude by a factor of 5
b. in frequency by a factor of 5
c. in both magnitude and frequency by factors of 5
2. A scaled parallel resonant circuit consists of R = 4 K: , L = 0.1 H , and C = 0.3 PF . Compute
k m and k f if the original circuit had the following values before scaling.
a. R = 10 : and L = 1 H
b. R = 10 : and C = 5 F
c. L = 1 H and C = 5 F
E8
e 5 = 10 PF
e 5 = 10 PF
u 2 u 10 = 10 or Z 0 NEW = 10000 r e s
= 5 u 10 e 0.3 (2)
2
E9
Appendix E Scaling
NOTES
E10
Index
Symbols and Numerics
A
abs(z) in MATLAB A25
admittance 62, 68, 611, 616, 617
drivingpoint 95
alpha coefficient 13, 115
angle(z) in MATLAB A25
antenna 218
antiresonance 26
asymptotes 76
asymptotic approximations 75
Audio Frequency (AF) Amplifier 218
B
bandwidth 212, 213, 73
beta coefficient 13, 115
Bode Plots 75
bode(sys) in MATLAB 721
bode(sys,w) in MATLAB 721
bodemag(sys,w) in MATLAB 721
box in MATLAB A13
C
clc in MATLAB A2
clear in MATLAB A2
collect(s) in MATLAB 512
column vector in MATLAB A20
command screen in MATLAB A1
command window in MATLAB A1
commas in MATLAB A8
comment line in MATLAB A2
complex conjugate pairs 55, A4
complex numbers A3
complex poles 55
complex roots of
characteristic equation B9
conj(A) in MATLAB C8
conj(x) in MATLAB C8
contour integration 42
conv(a,b) in MATLAB A6
convolution
in the complex frequency domain 412
in the time domain 411
corner frequency  see frequency
Cramers rule C16
P
partial differential equation B3
partial fraction expansion method 52
alternate method 515
PDE  see partial differential equation
plot
magnitude 75
phase 75
polar A25
plot in MATLAB A10
plot3 in MATLAB A16
poles 52, 76
repeated 58
poly(r) in MATLAB A4
polyder(p) in MATLAB A6
polyval in MATLAB A6
port 91
preselector 218
primary winding 84
proper rational function 51, 518
Q
quadratic factors A9
quality factor at parallel resonance 24
quality factor at series resonance 24
quit in MATLAB A2
N
R
NaN in MATLAB A28
natural response B7
critically damped 13
overdamped 13
underdamped 13
negative feedback  see feedback
network
bridged 735
pie 735
nonhomogeneous ODE B6
nthorder delta function 315
M
O
magnetic flux 82
magnitude scaling  see scaling
matrix, matrices
adjoint of C20
cofactor of C12
conformable for addition C2
octave 74
ODE  see ordinary differential equation
onedimensional wave equation B3
oneport network 91
open circuit impedance parameters 919
S
saw tooth waveform 436
scaling
frequency E1
magnitude E1
scaling property in complex
frequency domain 44
script file in MATLAB A26
secordorder circuit 11
semicolons in MATLAB A8
semilog plots
instructions for constructing D1
semilogx in MATLAB A13
semilogy in MATLAB A13
settling time 120
short circuit input admittance 912
short circuit output admittance 913
short circuit transfer admittance 913
signaltonoise ratio (S/N) 218
singlephase threewire system 104
solve(equ) in MATLAB 724
state equations 11
subplot in MATLAB A19
symmetric network 917, 935
symmetric rectangular pulse 36
symmetric triangular waveform 36
T
tee network 935
text in MATLAB A14, A18
Thevenin equivalent circuit 834
threephase
balanced currents 102
computation by reduction
to single phase 1020
Delta to Y conversion 1011
fourwire system 102
fourwire Ysystem 103
equivalent Delta and
Yconnected loads 1010
instantaneous power 1023, 1024
line currents 105
linetoline voltages 107
phase currents 105
phase voltages 107
positive phase sequence 107
power 1021
power factor 1021
systems 101
threewire Ysystem 103
threewire Delta system 104
two wattmeter method of
reading 3phase power 1030
Y to Delta conversion 10.12