Chapter 3

Switching Circuits, Power Computations,

and Component Concepts

3.1 Introduction

In this chapter, several general concepts will be discussed to provide necessary

background material in power electronics. The review material will cover switching
diode circuits, power computation, harmonic analysis, and component concepts.

3.2 Switching Diode Circuits

Switching power electronic circuits process power more efficiently than linear
power electronic circuits. This is why power electronic design engineers must
understand the analysis switching circuits. In this section, we will discuss the
analysis of switching circuits that include diodes, SCRs, and ideal switches under
both dc and ac excitations.
Diode circuits are nonlinear and their analysis is normally not as straightforward
as the analysis of linear circuits. By adding switches to diode circuits, additional
nonlinearity is introduced, making the analysis even more complex. These circuits
are encountered frequently in power electronic circuits, such as in diode and SCR
rectifier circuits, pulse width modulation (PWM), and resonant converters. To
simplify the analyses, we will assume throughout this chapter that the diodes,
SCRs, and the switches are ideal.

3.2.1 Switching Diode Circuits Under dc Excitation

We begin by analyzing diode resonant networks under dc excitation, which are

often encountered in resonant type dc-dc and dc-ac power electronic circuits. It will
Fig. 3.1 Series resonant
RLC circuit with dc source

be shown later that resonance is used in power electronics to achieve several

important functions including filtering and soft switching. Let us first consider a
series RLC circuit as shown in Fig. 3.1. Assume the switch is closed at t ¼ 0, and
vC(0) and iL(0) are the capacitor and inductor initial values, respectively, just
before the switch is closed at t ¼ 0.
The analysis of this circuit is straightforward and is given here as a review. Using
KVL around the loop yields the following integral-differential equation in terms of
the resonant current iL(t):
ðT
diL 1
L þ RiL þ iL dt ¼ V dc
dt C 0

By taking the first derivative of this equation, we obtain a second-order differ-

ential equation:

d 2 iL R diL 1
þ þ iL ðtÞ ¼ 0 ð3:1Þ
dt2 L dt LC
Since the excitation is a dc source, there exists only a transient (natural)
solution,1 whose roots are obtained from the characteristic equation given by:

R 1
s2 þ sþ ¼0 ð3:2aÞ
L LC
The roots of Eq. (3.2a) are given by:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2
R R 1
s1, 2 ¼   ð3:2bÞ
2L 2L LC

Depending on the different values of these roots, there exist three different well-
understood general solutions for the transient responses:

1
Also known as the homogeneous solution.
Case I: [Equal Real Roots]
When the roots are real and equal, we obtain what is called a critically damped
circuit, which occurs when the following circuit condition is met:
 2
R 1
¼
2L LC

and the roots of Eq. (3.2b) become:

R
s1, 2 ¼ 
2L
Under this condition, the general solution for the inductor current is given by:

iL ðtÞ ¼ ðA1 þ A2 tÞeðR=2LÞt ð3:3Þ

and the capacitor voltage is given by:

diL
vc ðtÞ ¼ V dc  L  RiL
dt
    ð3:4Þ
R R ðR=2LÞt
¼ V dc þ A2 1  t  A1 e
2L 2L

The constants A1 and A2 are obtained from the given initial conditions.
Case II: [Unequal Real Roots]
Under this condition, the circuit is known as overdamped and it occurs when:
 2
R 1
>
2L LC

The general solution is given as follows:

iL ðtÞ ¼ A1 es1 t þ A2 es2 t ð3:5Þ

and the capacitor voltage is given by:

vC ðtÞ ¼ V dc  ðLs1  RÞA1 es1 t  ðLs2  RÞA2 es2 t ð3:6Þ

where
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2
R R 1
s1 ¼  þ  ð3:7aÞ
2L 2L LC
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 2
R R 1
s2 ¼    ð3:7bÞ
2L 2L LC
Case III: [Complex Pair of Roots]
Under this case, the circuit is known as underdamped and it occurs when:
 2
R 1
<
2L LC

The responses are oscillatory with the general solution for iL(t) as given in
Eq. (3.8):

iL ðtÞ ¼ eαt ðA1 cos ωd t þ A2 sin ωd tÞ ð3:8Þ

While the voltage equation is given by:

vC ðtÞ ¼ V dc  Leαt ½ðA2 ωd  A1 αÞ cos ðωd tÞ  ðA2 α þ A1 ωd Þ sin ðωd tÞ

 Reαt ½A2 sin ðωd tÞ þ A1 cos ðωd tÞ

And the complex roots are given by:

s1 ¼ α þ jωd
s2 ¼ α  jωd

The parameters ωd, α, and ω0 are given by:

qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ωd ¼ ω20  α2
R
α¼
r2Lffiffiffiffiffiffi
1
ω0 ¼
LC
where ω0 is known as the resonant frequency, α is called the damping factor, and ωd
is known as the damped resonant frequency. The ratio α/ω0 is defined as the
damping ratio, δ:

α R
δ ¼ pffiffiffiffiffiffiffiffiffi
ω0 2 L=C

It can be shown that the constants A1 and A2 for a given initial capacitor
voltage, vC(0), and an initial inductor current, iL(0), for the series resonant RLC
circuit of Fig. 3.1, can be found using Table 3.1.
For the critically damped case, we have δ ¼ 1, and for a purely capacitive-
inductive circuit (R ¼ 0), we have δ ¼ 0. In the latter case, the response is purely
oscillatory. Such a response is encountered frequently in dc-dc soft-switching
power electronic circuits. Another parameter that is normally given in the RLC
circuit is the quality factor, Q0, which is defined as:
Table 3.1 Three possible cases and their response constants for the series resonant RLC circuit
Circuit type A1 A2
Case I: critically damped iL(0) V dc  vC ð0 Þ R
þ iL ð0 Þ
L 2L
Case II: overdamped V dc  vC ð0 Þ s1 iL ð0 Þ V dc  vC ð0 Þ s2 iL ð0 Þ
þ  þ
Lðs1  s2 Þ s1  s2 Lðs1  s2 Þ s1  s2
Case III: underdamped iL(0) V dc  vC ð0 Þ R
 iL ð 0  Þ
ωd L 2ωd L
s
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
R R 2 1 R R 2 1 R
s1 ¼  þ  , s2 ¼  þ  , ωd ¼ ω20  α2 , α ¼ ,
2L 2L LC 2L 2L LC 2L
qffiffiffiffiffi
ω0 ¼ LC 1

pffiffiffiffiffiffiffiffiffi
ω0 L L=C 1
Q0  ¼ ¼ ð3:9Þ
R R 2δ
The higher the Q0, the more oscillatory the current response becomes.
Another important parameter of particular interest in power electronics is what is
commonly referred to as the circuit characteristic impedance, Z0, which is defined
as:
rffiffiffiffi
L
Z0 ¼ ð3:10Þ
C
If R in the RLC circuit represents the load, then Q0 is known as the normalized
load and is given by:

Zo
Q0 ¼ ð3:11aÞ
R
To give the same measure of oscillation in the series case for parallel resonant
RLC circuits, the normalized load is defined as:

R
Q0 ¼ ð3:11bÞ
Z0
The higher the Q0, the more oscillation the voltage response becomes.
Normalized loads will be studied in dc-dc resonant converters in Chap. 6. Notice
that Q0 is defined the same as the quality factor of the resonant circuit. However,
one should distinguish between the Q0 of the resonant circuit in which R represents
the losses in the resonant circuit that can be ignored and when R represents the load,
where Q0 becomes a normalized load, i.e., a design parameter.
Fig. 3.2 (a) Circuit for
Example 3.1 and (b) its
current and voltage
waveforms

Example 3.1
Consider the circuit of Fig. 3.2a with R ¼ 200 Ω, L ¼ 2 mH, C ¼ 0.01 μF, and
Vdc ¼ 20 V. Derive the expressions for iL(t) and vC(t) for t > 0. Assume the
initial inductor current and the initial capacitor voltage are zeros, i.e., iL(0) ¼ 0
and vC(0) ¼ 0.
Solution
Here we have R/L ¼ 100  103 rad/s and 1/LC ¼ 50  109 rad2/s2. Since (R/2L)2
< 1/LC, the circuit is underdamped; hence the roots of the characteristic equation
are s1 ¼ (50  103  j218  103) rad/s and s2 ¼ (50  103  j218  103) rad/s.
The general solution for the inductor current is given by:
    
iL ðtÞ ¼ e5010
3
t
A1 cos 218  103 t þ A2 sin 218  103 t A ð3:12Þ

The constants A1 and A2 can be obtained from Table 3.1. However, for illustra-
tion purposes we will show how to find A1 and A2.
Since iL(0+) ¼ iL(0) ¼ 0, then A1 ¼ 0. To solve for A2, we use the capacitor
initial condition. Applying KVL to the circuit, we obtain:
 diL
V dc þ L þ vC þ RiL ¼ 0
dt
Evaluating this equation at t ¼ 0+, we obtain the first derivative at t ¼ 0+:

diL ð0þ Þ V dc  vC ð0þ Þ

¼ ¼ 10 A=ms
dt L
Taking the first derivative of Eq. (3.12) and evaluating it at t ¼ 0+ and setting it to
10 A/ms, we obtain the following equation:
 
diL ðtÞ 50103 t
dt t¼0þ ¼ iL ðtÞ ¼ 50e A2 sin 218  103 t þ 218  103 e50t A2 ð cos 218tÞ
t¼0
¼ 10 A=ms

Solving for A2, we obtain A2 ¼ 0.046.

The inductor current and capacitor voltage expressions for t > 0 are given below:
 
iL ðtÞ ¼ 0:046e5010 t sin 218  103 t A
3

diL
vC ðtÞ ¼ V dc  RiL  L
dt
Substituting for iL(t), the capacitor voltage is given by:
3     
vC ðtÞ ¼ 20  e5010 t 4:6 sin 218  103 t  20:056 cos 218  103 t V

The diode switches off when iL(t) ¼ 0 which occurs when 218  103t ¼ π, or
t ¼ π/(218  103) ¼ 14.4 μs, at which the capacitor voltage equals 29.76 V.
The sketches for iL and vC are given in Fig. 3.2b.
Example 3.2
Repeat Example 3.1 by replacing the diode across R as shown in Fig. 3.3a. Again,
assume iL(0) ¼ 0 and vC(0) ¼ 0.
Solution
At t ¼ 0+, the diode begins conducting, because the inductor current just after t > 0
is positive (assuming an ideal diode). The equivalent circuit for t  0 is shown in
Fig. 3.3b. This circuit has a damping ratio of zero, resulting in a purely sinusoidal
response with iL(t) given by:

iL ðtÞ ¼ A1 cos ω0 t þ A2 sin ω0 t

where ω0 ¼ 2.23  105 rad/s. Since the initial inductor current is zero, the constant
A1 ¼ 0, and from Table 3.1, A2 ¼ 45  103 ¼ 0.045, then the inductor current and
capacitor voltage become:

iL ðtÞ ¼ 45 sin ω0 t mA
vC ðtÞ ¼ 20ð1  cos ω0 tÞ V
Fig. 3.3 (a) RLC diode circuit for Example 3.2. (b) Equivalent circuit for 0  t  π/ω0. (c) The
inductor and capacitor waveforms. (d) Equivalent circuit for t  π/ω0

The inductor and capacitor waveforms are shown in Fig. 3.3c.

At t ¼ π/ω0, the diode becomes reverse biased since iL starts becoming negative.
At this time the equivalent circuit model changes to a series RLC circuit with a new
capacitor initial value, vC(π/ω0) ¼ 40 V, and iL(π/ω0) ¼ 0. Figure 3.3d shows the
equivalent circuit for t > π/ω0. The solution for iL and vC is similar to that in
Example 3.1 except that when iL becomes zero again, the diode will conduct and
the equivalent circuit mode of Fig. 3.3b becomes valid again. This process con-
tinues until the diode current decreases to zero.
Exercise 3.1
Repeat Example 3.2 with the diode direction reversed. Sketch the capacitor voltage
and inductor current waveforms for 0  t < π/ωd.
 
Answer: For 0  t < π/ωd, iL ðtÞ ¼ 0:046e5010 t sin 2:18  105 t A
3

 
vC ðtÞ ¼ 20  4:6 sin 2:18  105 t þ 20:1 cos 2:18  105 t e5010 t v
3

Exercise 3.2
Solve for iL(t) and vC(t) after the switch is closed at t ¼ 0 in Fig. E3.2. This circuit is
frequently encountered in dc-dc resonant converters. Assume the initial values are
vC(0) ¼ 0 and iL(0) ¼ 1.5 Ig. (The circuit that established this initial inductor
current is not shown.)

Fig. E3.2 Circuit for

Exercise 3.2

Answer: iL(t) ¼ 1.5Ig cos w0tA, vC(t) ¼ 1.5IgLω0 sin ω0t V,

pffiffiffiffiffiffiffiffiffiffiffi
ω0 ¼ 1=LC for ω0 t  π
iL(t) ¼  1.5Ig, vC(t) ¼ 0, for ω0t > π
Exercise 3.3
Repeat Exercise 3.2 by assuming the initial conditions are vC(0) ¼ 0 V and
iL(0) ¼ 0.5 Ig A.
Answer: iL(t) ¼ 0.5Ig cos w0t A, vC(t) ¼ 0.5IgLω0 sin ω0t V
pffiffiffiffiffiffiffiffiffiffiffi
ω0 ¼ 1=LC for ω0 t  π

iL(t) ¼  0.5Ig, vC(t) ¼ 0, for ω0t > π

Exercise 3.4
Give switch implementation (unidirectional or bidirectional) for Exercises E3.2
and E3.3.
Exercise 3.5
Consider the resonant circuit in Fig. E3.5. These types of circuits are known as
resonant circuits that exhibit “soft-switching” phenomenon to be discussed in
Chap. 6. By definition, the soft-switching converter employs resonance to bring
voltage across the off switch to zero before the switch is turned on (known as
zero-voltage switching (ZVS)) or to bring the current through the on switch to zero
before the switch is turned off (known as zero-current switching (ZCS)). For the
circuit shown, what is the earliest time switch S1 can be turned off and still achieve
ZCS. Assume before t ¼ 0, S1 has been closed while S2 has been open. As t ¼ 0, S2
closes and S1 remains closed. Given parameters are Lr ¼ 50 μH, Cr ¼ 20 μF,
Iin ¼ 10 A, VCr(0-) ¼ 20 V, and ILr(0-) ¼ 0 A, and assume all components are
ideal.

Fig. E3.5 Resonant circuit

for Exercise 3.5
i sw1 i sw2

Iin S1 t=0 S2

Cr
VCr
VLr
iCr
Lr iLr

Answer: T ¼ 28.82 μs

3.2.2 Switching Diode Circuits with an ac Source

Large classes of switching converter circuits use an ac source excitation rather than
a dc source. The analysis of diode switching circuits with ac sources are carried out
in two steps: first obtain the transient response (also known as the natural response)
by setting the ac source to zero, and then obtain the steady-state response (also
known as the forced response) by converting the circuit to phasor domain. The final
solution is the sum of both the natural and the forced responses. Figure 3.4a shows
an RLC circuit with an ac source, vs(t) ¼ Vs sin ωt V. Assume the switch is turned on
at t ¼ 0, and we wish to solve for the inductor current.

Fig. 3.4 RC circuit with ac excitation: (a) time-domain circuit and (b) equivalent phasor-domain
circuit
 For t > 0, KVL yields the following differential-integral equation in terms
of iL(t):
ðt
diL 1
V s sin ωt þ L þ iL ðtÞdt þ RiL ¼ 0
dt C 1

Taking the first derivative of the above equation we obtain:

d2 iL R diL 1 Vsω
þ þ iL ðtÞ ¼ cos ωt ð3:13Þ
dt2 L dt LC L
The right-hand side of Eq. (3.13) represents the forced excitation. First we obtain
the natural response component of the complete solution by setting the forced
function (source function vs and its derivatives) to zero, to yield:

d 2 iL R diL 1
þ þ iL ðtÞ ¼ 0 ð3:14Þ
dt2 L dt LC
The natural response of iL is the same as the response done for the dc source RLC
circuit shown in Table 3.1 assuming the overdamped case, i.e., the transient
response, iL, tran, is given by:

iL, tran ðtÞ ¼ A1 es1 t þ A2 es2 t ð3:15Þ

The steady-state response component is obtained by transferring the circuit into

the phasor domain as shown in Fig. 3.4b. The steady-state response, iL, ss, is easily
obtained:

V s ∠0 V ∠θ
I L, ss ¼  ¼ s ð3:16Þ
R þ j ωL  ωC
1 jZ j

where
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
   1 
1 2 ωL  ωC
jZ j ¼ R2 þ ωL  and θ ¼ tan 1
ωC R

In time domain, iL, ss is given by:

Vs
iL, ss ðtÞ ¼ sin ðωt  θÞ ð3:17Þ
jZ j

The total response is obtained by adding Eqs. (3.15) and (3.17) to yield:

iL ðtÞ ¼ iL, tran ðtÞ þ iL, ss ðtÞ ð3:18Þ

Vs
¼ A1 es1 t þ A2 es2 t þ sin ðωt  θÞ
jZ j
 The constants A1 and A2 are obtained from the initial condition of the inductor
current and capacitor voltage at t ¼ 0+.
Example 3.3
Consider the ac-diode circuit shown in Fig. 3.5a with a dc source in the load side
representing either a charged battery or a back electromotive force (emf) to excite
the armature circuit in a dc motor. Sketch the waveforms for io, vD, and vo. Assume
vs ¼ 100 sin 377t. What is the average value of vo?

Fig. 3.5 (a) Ac-diode a + vD – io

circuit for Example 3.3.
(b) The voltage and current
waveforms +
8Ω

vs + vo
– +
60 V = Vdc
–
–

b
vs
100

wt2
p
wt
0 2p
wt1

vo

100

60
wt
36.87° 36.87°
io

wt

vD

wt

–60

–160
Solution
The diode will turn on when vD > 0 which occurs at t ¼ t1 when vs(t1) ¼ 60 V, i.e.:

100 sin 377t1 ¼ 60 V

resulting in t1 ¼ 1.7 ms. For 0  t  t1 the diode is off and for the interval t1  t < t2
the diode is on, where t ¼ T/2  t1 ¼ 6.63 ms.
Between t1  t < t2, the output voltage, vo, equals vs, and the output current is
given by:

vs  V dc
io ðtÞ ¼ ¼ 12:5 sin 337t  7:5 A
R
For all other times, namely, 0  t  t1 and t2  t < T, io ¼ 0 and vo ¼ VDC ¼ 60 V.
The average output voltage is obtained from the following equation:
ð t2 
1
V o, ave ¼ V s sin ωt dt þ V dc ðT=2 þ 2t1 Þ
t t1
ð 6:63 ms 
1
¼ 100 sin 377t dt þ 60ð8:33 ms þ 3:4 msÞ ¼ 67:79 V
16:67 ms 1:7 ms

The voltage and current waveforms are shown in Fig. 3.5b.

Exercise 3.6
Repeat Example 3.3 by including a freewheeling diode across the load as shown in
Fig. E3.6.
Answer: 67.79 V

Fig. E3.6 Circuit for

Exercise 3.6

Exercise 3.7
Repeat Exercise 3.6 by reversing the polarity of the 60 V dc source.
Answer: 31.83 V
Example 3.4
Consider the circuit shown in Fig. 3.6a with a dc source in the load side. Assume
ideal diodes, zero initial inductor current and Vdc < Vs. Find the expressions for
iD1, iD2, io, and vo for 0 < t < 2T, where T is the period of vs given by T ¼ 2π/ω.
Assume vs ¼ Vs sin ωt with Vs ¼ 100 V, ω ¼ 377 rad/s., T ¼ 16.67 ms.
Solution
At t ¼ 0 the switch closes and vs is switched into the circuit. However,
when vs  Vdc, D1 remains OFF and so does D2.
At t ¼ t1, vs ¼ Vdc, which force D1 to turn ON and D2 to turn OFF, resulting in
the equivalent circuit of Mode 1 for t > t1 as shown in Fig. 3.6b. t1 can be
determined by the following equation:
 
1 V dc
t1 ¼ sin 1 ¼ 0:67 ms
ω Vs

a D1 io

t=0 + R=3Ω
+
vs D2 vo L = 10 mH
–
+ Vdc = 25 V
–
–

b c
D1 D1

R R

vs + vs +
D2 L D2 L
– –
+ +
Vdc Vdc
– –

d
iL

16.2 A

T
t
t1 = 0.67 ms T/2 t2 = 11.48 ms 2T

Fig. 3.6 (a) Circuit for Example 3.4. (b) Mode 1: 0  t < T/2. (c) Mode 2: T/2 < t < T. (d) Sketch
for iL(t) for 0 < t < 2T
 The differential equation for Mode 1 is given by:

diL R vs  V dc
þ iL ¼
dt L L
The transient response of iL(t) is:

iL, tran ðtÞ ¼ I 1 eðtt1 Þ=τ

where τ ¼ L/R ¼ 3.33 ms, and the steady-state response is given by:

Vs V dc
iL, ss ðtÞ ¼ sin ðωt þ θÞ  ¼ 20:75 sin ð377t  51:5∘ Þ  8:33 A
jZ j R
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
where jZj ¼ R2 þ ðωLÞ2 ¼ 4:82 Ω and θ ¼ tan1(ωL/R) ¼ 51.5∘.
The overall response is given by:

Vs V dc
iL ðtÞ ¼ I 1 eðtt1 Þ=τ þ sin ðωt1  θÞ þ ð3:19Þ
j zj R

The constant I1 is obtained by setting iL ¼ 0 at t ¼ t1 in Eq. (3.19) which yields:

Vs V dc
I1 ¼  sin ðωt1  θÞ 
jzj R
¼ 20:82 A

The general solution for iL(t) is given by:

iL ðtÞ ¼ 24:75 eðtt1 Þ=τ þ 20:75 sin ð377t  51:5∘ Þ  8:33 0  t < T=2 ð3:20Þ

where

iL ðT=2Þ ¼ 10:03 A ð3:21Þ

At t ¼ T/2 the source voltage becomes negative, forcing D2 to turn ON and D1 to

turn OFF. The equivalent circuit for Mode 2 is shown in Fig. 3.6c. The inductor
current at t ¼ T/2 becomes the initial inductor current for the next cycle at t ¼ T/2.
This mode remains until the inductor current become zero again and the transient
response of iL(t) at Mode 2 is given by:

iL, tran ðtÞ ¼ I 2 eðtT=2Þ=τ

and the steady-state response is given by:

V dc
iL, ss ðtÞ ¼  ¼ 8:33 A
R
resulting in the following expression for iL(t):

iL ðtÞ ¼ I 2 eðtT=2Þ=τ  8:33 A

The constant I2 is obtained by setting iL ¼ 10.03 at t ¼ T/2 which yields:

I 2 ¼ 10:03 þ 8:33 ¼ 18:36 A

and the overall response for Mode 2 can be given by:

iL ðtÞ ¼ 18:36 eðtT=2Þ=τ  8:33 A T=2  t < t2

where t2 is the time when iL(t) decreases to zero. It can be found setting iL ¼ 0 at
t ¼ t2 which yields:
8:33
t2 ¼ τ ln þ T=2 ¼ 10:97 ms
18:36
After that iL(t) will remain zero until t ¼ T + t1. The sketch for iL(t) for 0 < t < 2T
is shown in Fig. 3.6d.

3.3 Controlled Switching Circuits

Consider the SCR circuit shown in Fig. 3.7a driven by an ac voltage

source vs ¼ Vs sin ωt. Assume the switch is turned on at t ¼ 0, and the SCR is
triggered by applying a gate current, ig, at t ¼ t1.
At t ¼ 0 when the switch is closed, it is still an open circuit and io ¼ 0. At t ¼ t1,
the SCR is triggered and since vs(t1) ¼ Vs sin ωt1 > 0, the SCR turns on resulting in
the equivalent circuit shown in Fig. 3.7b.
The circuit is equivalent to the case of the ac-diode circuit for t  0. The
differential equation representing this circuit is given by:

d2 iL R diL 1 Vsω
2
þ þ iL ðtÞ ¼ cos ωt t  t1
dt L dt LC L
The general solution for io(t) for t > t1 is given by:

Vs
io ðtÞ ¼ sin ½ωt  θ þ A1 es1 ðtt1 Þ þ A2 es2 ðtt1 Þ ð3:22Þ
jZ j

where
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 
1 2
jZj ¼ R2 þ ωL 
ωC
ωL  ωC 1
θ ¼ tan1
R
Fig. 3.7 (a) SCR circuit. a ig@t = t1
(b) Equivalent circuit S io
for t > t1. (c) Output voltage
waveform L
t = 0 + vAK – + C

vs + vo R
–
–

b
SCR io

L
+ vAK – + C
+ vo
vs R
–
–

c
vo
Vs

t2
0 t1 T/2 T t

The constants A1 and A2 are obtained by applying the initial conditions of the
inductor and the capacitor.
The output voltage, vo, for t > 0 is given by:
8
>
> 0 0  t < t1
<
vo ¼ V s sin ωt t1  t < t2
>
>
:
0 t2  t < T

where t1 is the time at which SCR is first turned on and t2 is when the SCR is turned
off because io(t) is zero for t2 > T/2. If the SCR gate current is applied at t ¼ t1 into
the new cycle, vo becomes a periodical waveform as shown in Fig. 3.7c whose
average value is given by:
ð t2
1 Vs
vo, ave ¼ V s sin ωtdt ¼ ð cos α  cosβÞ ð3:23Þ
T t1 2π
where α ¼ ωt1 and β ¼ ωt2. The angle α is known as the firing angle, and (β  α) is
known as the SCR conduction angle. Notice that by varying α, we can vary the
average output voltage. Such circuits will be studied in details in Chap. 8.
Exercise 3.8

Assume the SCR circuit in Fig. 3.7a is triggered at α ¼ 30 (i.e. t ¼ 1.39 ms) after the
switch is closed at t ¼ 0. Derive the expression for iL(t) and vc(t) for 0 < ωt < 2π,
assuming zero initial condition. Use R ¼ 200 Ω, L ¼ 2 mH, C ¼ 0.01 μF,
and vs(t) ¼ 20 sin 377t.
Answer:
 
iL ðtÞ ¼ eαðtt1 Þ ½A1 cos ωd ðt  t1 Þ þ A2 cos ωd ðt  t1 Þ þ 75:4  106 sin ωt þ 89:9
t1 < t < t2

where
α ¼ 50  103, ωd ¼ 217,945, t1 ¼ 1.39 ms, t2 ¼ 1.404 ms, A1 ¼  65.35  106,
A2 ¼ 22.92  103
Exercise 3.9
Repeat E3.8 by placing an ideal diode as shown in Fig. E3.9.

Fig. E3.9 Circuit for

Exercise 3.9

Answer:

iL ðtÞ ¼ eαðtt1 Þ ½A1 cos ωd ðt  t1 Þ þ A2 cos ωd ðt  t1 Þ þ 75:4

 
 106 sin ωt þ 89:9

where
α ¼ 50  103, ωd ¼ 217,945, t1 ¼ 4.27 ms, A1 ¼  75.4  106, A2 ¼  17.3  103
3.4 Basic Power and Harmonic Concepts

In this section we will review some basic power concepts applied to sinusoidal and
non-sinusoidal current waveforms that are of particular importance in power
electronic circuits.

3.4.1 Average, Reactive, and Apparent Powers

Power Flow
As stated earlier, the function of the power electronic circuit is to process power by
performing some conversion function through a set of switching actions dictated by
some control circuit. The direction of power flow in a power electronic circuit is an
important concept since it relates to identifying the input and output ports in the
power electronic circuit. Normally, the situation exists as shown in Fig. 3.8a where
the power flow is from the input side to the output side. Here Pin and Pout represent

Io
a
Source side Load side
(input) (output)
is io
+ II I
Power-
+ processing
vs vo Load Vo
– circuit
Pin (Ploss) Pout
–
III IV

b
Io

Source side Load side

(input) (output)
is io
+ II I
Power-
+ processing
vs vo Load Vo
– circuit
Pin (Ploss) Pout
–
III IV

Fig. 3.8 Power flow: (a) unidirectional (input-to-output) (b) bidirectional

the average input power from the source side (input ac or dc) and the average output
power at the load side (output ac or dc), respectively. One may conclude from
Fig. 3.8a that the average power flow is from the input terminal (source side) to the
output terminal (load side); hence, the direction of power flow becomes the basis for
defining the power circuit port. However, since some power electronic circuits have
a dc source in the load side, it is possible to have the power flow in the opposite
direction, i.e., the circuit is capable of bidirectional power flow, as shown in
Fig. 3.8b. As a result, one has to be careful in identifying the source and load
sides. A good discussion on this issue, supported by several examples, is given in
Kassakian et al.
The efficiency of the power processing circuit is an extremely important param-
eter since it has a direct impact on the cost, performance, size, and weight of the
system as discussed in Chap. 1. For Fig. 3.8a, the efficiency, η, is defined by:

Pout
η¼  100%
Pin
Pout
¼
Pout þ Ploss
If the power circuit consists of ideal switching devices that operate in either the
on- or the off-states and lossless energy storage elements like capacitors, inductors,
and transformers, then the overall efficiency of the power processing circuit is
100%.
Average Values and rms
For a given periodical voltage signal, v(t), with period T, its average value is
defined by:
ðT
1
V ave ¼ vðtÞdt ð3:24Þ
T 0

and the root-mean-square (rms) value is given by:

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ð
1 T 2
V rms ¼ v ðtÞdt ð3:25Þ
T 0

Instantaneous Power
The instantaneous power delivered to a load element that has v(t) across it and i(t)
through it, as shown in Fig. 3.9a, is given by:

PðtÞ ¼ vðtÞ iðtÞ ð3:26Þ

The voltage and current waveforms can be sinusoidal, periodical, or constant as

shown by the arbitrary waveforms in Fig. 3.9c. Notice that the instantaneous power
in this figure can be either positive, zero, or negative. For example,
Fig. 3.9 (a) Time-domain and (b) phasor-domain elements and (c) arbitrary current and voltage
waveforms

between 0  t  t1, the circuit element absorbs or dissipates power (positive),

between t1  t  t2, the power transfer to the element is zero, and finally
between t2  t  t3, the load generates power (positive) or returns power (negative)
to the source.
Average Power
If the voltage and current repeat periodically, then we can define the total average
power either generated or dissipated by the circuit element, which is given by:
ð
1 T
Pave ¼ pðtÞdt
T 0
ð
1 T
¼ iðtÞvðtÞdt
T 0

The average power is also known as real power, which comes from the complex
number representation. In terms of phasor representation, the linear time-domain
circuit element of Fig. 3.9a is redrawn as a phasor-domain circuit element as shown
in Fig. 3.9b.
Apparent Power
The phasor magnitudes |Is| and |Vs| represent the peak of the source current and
voltage, respectively, and θi, θv represent the current and voltage phase shift,
respectively. If we let the total impedance of the element be given by:

Z ¼ R þ jX ¼ jZ jejθ

where X is the total reactance of a capacitive circuit, 1/ωC , or a total reactance of

an inductive circuit, ωL, then the total input complex power may be defined by:

VI ∗ jV jjI j jðθv θi Þ

PT ¼ ¼ e
2 2
¼ V rms I rms ejθ
¼ Sejθ

where I∗ is the complex conjugate of I. The parameter S is known as apparent

power, and θ represents the total phase shift between i(t) and v(t) and is also known
as power factor angle.
Normally, PT is expressed in terms of real and imaginary complex numbers as
follows:

PT ¼ P þ jQ ð3:27Þ

The real part, P, is the average power, which is given by:

P ¼ S cos θ
ð3:28Þ
¼ V rms I rms cos θ

and Q is the reactive power, which is given by:

Q ¼ S sin θ
ð3:29Þ
¼ V rms I rms sin θ
 From these expressions, S can also be expressed mathematically as:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
S¼ P2 þ Q 2 ð3:30Þ

The units of P are watts, representing the power being dissipated. Units for
Q 2are volt-amperes reactive (var), representing the reactive power being stored in
the inductor or/and capacitor, and the units for S is volt-amp (VA), representing the
rms product of the voltage and current values. The reactive power is not a useful
parameter, and it is normally desired to make Q equal to zero, which means the total
power is equal to the real or average power. Knowing the Q of the circuit helps the
designer to compensate so that the load always draws real power. This case
corresponds to a unity power factor to be discussed shortly.
Exercise 3.10
Consider the one-port network of Fig. E3.10 with i(t) a triangular waveform.
Determine the average and rms current values and the average power absorbed by
the network in steady state under the following cases: (i) purely resistive
with R ¼ 0.5 Ω, (ii) purely inductive L ¼ 1 mH, (iii) purely capacitive C ¼ 1 μF,
and (iv) resistive-inductive with R ¼ 0.5 Ω and L ¼ 1 mH. Assume zero initial
conditions.

Fig. E3.10 A network and its input current waveform for E3.10

Answer: 2 A, 3.1 A, 4.66 W, 0 W, 0 W, 4.66 W.

3.4.2 Sinusoidal Waveforms

3.4.2.1 Instantaneous and Average Powers

First let us consider the case for linear, one-port network shown in Fig. 3.10a. Since
the network consists of linear components, its instantaneous source current and
voltage expression may be represented as follows:

2
This Q has nothing to do with resonant circuit parameter Q discussed earlier and the one to be
discussed in Chap. 6.
Fig. 3.10 (a) Linear a is(t )
one-port circuit. (b) Circuit
waveforms
a
Linear
vs(t ) +
one-port
–
network
a’

Vs
Is is
vs

qi wt

–Is
–Vs
qv qv – qi

p(t )

wt

is ðtÞ ¼ I s sin ðωt  θi Þ ð3:31aÞ

vs ðtÞ ¼ V s sin ðωt  θv Þ ð3:31bÞ

where θi and θv are the port current and voltage phase shift, respectively.
The instantaneous power p(t) is given by:

pðtÞ ¼ is vs
ð3:32Þ
¼ I s V s sin ðωt  θi Þ sin ðωt  θv Þ

The waveforms for is(t), vs(t), and p(t) are shown in Fig. 3.10b.
The input average power can be calculated from the following integral:
ðT
Pave ¼ pðtÞdt
0
ðT ð3:33Þ
1
¼ is ðtÞvs ðtÞdt
T 0
 Substituting for is and vs from Eqs. (3.31) and by using the following trigono-
metric identities:

cos ðθ1  θ2 Þ ¼ cos θ1 cos θ2  sin θ1 sin θ2

sin ðθ1  θ2 Þ ¼ sin θ1 cos θ2  cos θ1 sin θ2

the instantaneous power may be expressed as follows:

Is V s
pð t Þ ¼ ½ cos ðθv  θi Þ  cos ð2ωt þ θv þ θi Þ ð3:34Þ
2
Substitute Eq. (3.34) in the integral of Eq. (3.33), and the average power
becomes:

Is V s
Pave ¼ ½ cos ðθv  θi Þ ð3:35Þ
2
In terms of the rms parameters, the average power is given by:

Pave ¼ I s, rms V s, rms cos ðθv  θi Þ ð3:36Þ

3.4.2.2 Power Factor

Power factor is a very important parameter in power electronics because it gives a

measure of the effectiveness of the real power utilization in the system. It also
represents a measure of distortion of line voltage and line current and the phase shift
between them. Let us consider Fig. 3.10a in providing the basic definition of power
factor.
Power factor (pf) is defined as the ratio of the average power measured at the
terminals a-a’ of Fig. 3.10a and the rms product of vs and is as given in Eq. (3.37):

Real power ðaverageÞ

Power factor ¼ ð3:37Þ
Apparent power

For purely sinusoidal current and voltage waveforms, the average power is given
in Eq. (3.36) and the apparent power is given by Is, rmsVs, rms. As a result, Eq. (3.37)
yields:

I s, rms V s, rms cos θ

Power factor ¼
I s, rms V s, rms ð3:38Þ
¼ cos θ

Hence in linear power systems, when the line voltage and line currents are purely
sinusoidal, the power factor is equal to the cosine of the phase angle between the
current and voltage. However, in power electronic circuits, due to the switching of
Fig. 3.11 (a) Leading a b
power factor. (b) Lagging I I
power factor

+ +
R R

V V

–j/wC jwL
– –

1
Z = R– j
wC
V Z = R + jwL
= Z –q
I
V
= Z –q
2 I
1
Z = R2 +
wC Z = R2 + (w L) 2

–1 wL
q = tan–1 < 0° q = tan–1 > 0°
w RC R

active power devices, the phase angle representation alone is not valid. This is why
we will shortly define power factor for terminals, whose currents and/or voltages
are non-sinusoidal (distorted).
The angle θ is known as the power factor angle; therefore, the power factor
varies between 0 and 1, depending on the type of the network. For θ > 0, the current
lags the voltage, representing inductive-resistive load as shown in Fig. 3.11b. The
network load is said to be having a lagging power factor. Similarly, for θ < 0, the
current leads the voltage, representing capacitive-resistive load with leading power
factor as shown in Fig. 3.11a.
Let us calculate the power factor for resistive, inductive, and capacitive
two-terminal networks:
1. Resistive network
The voltage and current relation is given by:

vs ¼ i s R

and the power factor angle is θ ¼ θv  θi ¼ 0, resulting in a power factor equal to 1.

2. Capacitive network
The capacitor current and voltage relation is given by:

dvs
is ¼ C
dt
 In phasor domain, we have:

jV s j∠θv 1
¼ j
jI s j∠θi ωC
jV s j 1
∠θv  θi ¼ ∠  90
jI s j ωC

Therefore, the power factor angle is θv  θi ¼ θ ¼  90 , resulting in zero
power factor. This means the purely capacitive circuit has no average power
delivered, (as expected from an ideal capacitor). This is a leading power factor

because current leads voltage by 90 .
3. Inductive network

jV s j∠θv
¼ jωL
jI s j∠θi
jV s j
∠θv  θi ¼ ωL∠90
jI s j

The power factor angle is θ ¼ + 90 , resulting in lagging power factor

because current lags voltage by 90 .
Example 3.5
Determine L in the circuit of Fig. 3.12 so that the power factor becomes unity.
Solution
The total impedance seen by the source is given by:

1
Z in ¼ jωL  j þ RL
ωC
For unity power factor, the phase angle of Zin must be zero, i.e., Zin is a real
number. Setting the imaginary part to zero yields:

1
ωL  ¼0
ωC

Fig. 3.12 Circuit for

Example 3.5
and solving for the inductor value, we obtain:

1
L¼
ω2 C

Exercise 3.11
Consider the circuit of Fig. E3.11 with 10 kVA load and vs ¼ 100 sin ωt.

Fig. E3.11 Circuit for

Exercise 3.11

It is required to compensate the load to improve power factor to 0.97. Determine

the type and value of the load impedance Zload that must be added in parallel to
achieve pf ¼ 0.97, and repeat the process to achieve pf ¼ 1.0.
Answer: For pf ¼ 0.97, Zload ¼  j1.056Ω (C ¼ 2.5 mF for ω ¼ 60 Hz) and for
pf ¼ 1.0, Zload ¼  j0.756Ω (C ¼ 3.5 mF for ω ¼ 60 Hz)

3.4.3 Non-sinusoidal Waveforms

Harmonics cause serious problems of interference with sensitivity measurements

and communication systems. To reduce or eliminate line voltage and current
harmonics, it is necessary to add filters in the ac input side. Ac and dc filter design
is a very specialized topic, and the literature is rich in analysis, design, and
implementation of such filters. For this reason the topic will not be addressed in
this textbook. The voltage source of an ideal electrical power system supplies
energy at a constant and single frequency with constant voltage amplitude under
all load conditions. However, in practical power systems, single and constant
frequency and fixed amplitude voltage sources are not available.
The importance of studying current and voltage harmonics has grown recently
because of the widespread use of power switching devices in various power
electronics applications.
The topic of studying a system’s harmonics is specialized and cannot be fully
addressed in a textbook like this. Nevertheless, some important harmonics issues
will be addressed here. The reader is encouraged to see the references listed at the
end of this chapter
 Because of the switching nature of the majority of power electronic circuits, the
line (source) current is highly distorted while the line voltage remains nearly
sinusoidal. Only in a limited number of circuits we do have almost sinusoidal line
voltage and current. In steady state, the non-sinusoidal nature of the line current
produces unwanted oscillatory components at different frequencies. Such signals
are called harmonics or harmonic components. Under some load conditions, these
harmonics have high amplitudes that result in highly undesirable effects. These
harmonics must be removed or at least significantly reduced. As a result, it becomes
necessary to study the harmonics and power factor values in non-sinusoidal current
waveforms. We will assume that the distorted waveforms are in steady state with a
given fundamental frequency. The existence of these harmonics affects the overall
efficiency, performance, and cost of the power electronic system. It will be shown
that because of this distortion, the apparent power rating (volt-amp) of the source
must be higher than the actual real power needed by the load.
The best tool available to study harmonics is the use of the Fourier analysis
method. The basis for harmonic calculations is the Fourier theorem introduced by
the French mathematician Jean Baptiste Joseph Fourier in 1822. First we review the
Fourier analysis technique and harmonic components.

3.4.3.1 Fourier Analysis

The Fourier theorem states that, in steady state, for any given periodic function f(t),
one can represent it by the sum of a constant F0 and infinite sine and cosine
functions ( f2, f2, . . .) defined by the following formula:

f ðtÞ ¼ F0 þ f 1 ðtÞ þ f 2 ðtÞ þ . . . þ f n ðtÞ

X
1
ð3:39Þ
¼ F0 þ ðan cos n ωt þ bn sin n ωtÞ
n¼1

The constant coefficient can be obtained by taking the integral of both sides of
Eq. (3.39) from 0 to T to give the following expression for F0:
ðT
1
FO ¼ f ðtÞdt ð3:40Þ
T 0

From Eq. (3.40), we see that F0 represents the average (dc) value of f(t).
Also it can be shown that the coefficient an and bn are evaluated from the
following integrals:
ð
2 T
an ¼ f ðtÞ cos nωt dt n ¼ 1, 2, 3, . . . , 1 ð3:41aÞ
T 0
ðT
2
bn ¼ f ðtÞ sin nωt dt n ¼ 1, 2, 3, . . . , 1 ð3:41bÞ
T 0
Fig. 3.13 Phase
representation for Eq.(3.42)

These equations constitute a frequency domain representation of f(t)3 and sug-

gest that any non-sinusoidal waveform with frequency ω, (the waveform repeated
every time period T ¼ 2π ⁄ w) can be expressed as the sum of a constant and infinite
sinusoidal waveforms at multiples (nω, with n ¼ 1, 2, . . .) of the original frequency.
The frequency of the original waveform (at n ¼ 1) is known as the fundamental
frequency. The values of f(t) at n ¼ 1, 2, 3, . . ., 1 are known as the Fourier compo-
nents of the waveform f(t) or the harmonic components of f(t). At n ¼ 1, the
component is usually called the fundamental component or fundamental harmonic,
denoted by the subscript 1.
By defining the triangle shown in Fig. 3.13, it is also possible to represent f(t) in
the following form:
X1
f ðtÞ ¼ Fo þ n¼1
Fn sin ðn ωt þ θn Þ ð3:42Þ

where Fn is the peak value of the nth harmonic and θn is the phase shift, which are
given by:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Fn ¼ a2n þ b2n

 
an
θn ¼ tan 1
bn

If the waveform has any form of symmetry shown in Table 3.2, calculation of the
integrals for an and bn can be significantly reduced. In power electronic circuits, odd
symmetry waveforms are more frequently encountered than the even symmetry
waveforms.

3
It is also common to represent an and bn in terms of frequency as follows:
ð 2π ð 2π ð 2π
1 1 1
an ¼ f ðωtÞ sinðnωtÞ dðωtÞ bn ¼ f ðωtÞ cosðnωtÞ dðωtÞ and F0 ¼ f ðωtÞ d ðωtÞ
π 0 π 0 2π 0
Table 3.2 Simplification of Fourier coefficients due to function symmetry
Odd symmetry an ¼ 0 ð T=2
f(t) ¼  f(t) (for all n) bn ¼ T4 f ðtÞ sin n ωt dt
0
Even symmetry ð bn ¼ 0
4 T=2
f(t) ¼ f(t) an ¼ f ðtÞ cos n ωt dt (for all n)
T 0
8 ð 8 ð
Half-wave symmetry < 4 T=2 < 4 T=2
f(t) ¼  f(t + T/2) an ¼ T 0 f ðtÞ cos n ωt d n odd bn ¼ T 0 f ðtÞ sin n ωt d n odd
: :
0 n even 0 n even
8 ð
Odd and half-wave symmetry an ¼ 0 < 8 T=2
f(t) ¼  f(t + T/2) (for all n) bn ¼ f ðtÞ sin n ωt d n odd
f(t) ¼  f(t) : T 0
0 n even
8 ð
Even and half-wave symmetry < 8 T=2 bn ¼ 0
f(t) ¼  f(t + T/2) an ¼ T 0 f ðtÞ cos n ωt d n odd (for all n)
f(t) ¼ f(t) :
0 n even
Fig. 3.14 Frequency
spectrum representation
for f(t)

Fig. 3.15 Current

waveform for Example 3.6

Another useful representation of f(t) is in terms of its frequency spectrum as

shown in Fig. 3.14. It is a plot of the magnitude of the harmonic component against
its frequency. As n increases the peak of the higher harmonic components
decreases. The peak value at F1, i.e., the fundamental, is normally referred to as
the “wanted” harmonic or component, whereas the higher components represent the
“unwanted” components.
Example 3.6
Consider a half-wave power electronic circuit whose line current is(t) waveform is
shown in Fig. 3.15. Calculate the harmonics of is(t).
Solution
The average value is clearly Is/2, so we have:

Is
F0 ¼
2
The coefficients an are given by:
ð 2π
1
an ¼ f ðωtÞ cos n ωtd ðωtÞ
π 0
ðπ π
1 Is
¼ I s cos ðn ωtÞdt ¼ sin ðn ωtÞ ¼ 0 n ¼ 1, 2,
π 0 nπ 0
 The coefficients bn are given by:
ð 2π
1
bn ¼ f ðωtÞ sin n ωtdðωtÞ
π 0
ðπ π
1 I s 2I s
¼ I s sin ðn ωtÞdt ¼ cos ðn ωtÞ ¼ n ¼ 1, 3, 5
π 0 nπ 0 nπ

Exercise 3.12
Determine the average and rms values for the output voltage waveform for a half-
wave rectifier circuit shown in Fig. E3.12:

Fig. E3.12 Circuit for

Exercise 3.12

Answer: Vs/π, Vs/2

3.4.3.2 Line Current Harmonics

If we apply the Fourier equations to the line current is and the line voltage vs, which
are periodical non-sinusoidal waveforms with period T and zero dc, then is(t) and
vs(t) are given as follows:

X
1
is ðtÞ ¼ I dc þ I sn sin ðn ωt þ θni Þ
n¼1
ð3:43Þ
X
1
¼ I dc þ I s1 sin ðωt þ θ1i Þ þ I sn sin ðn ωt þ θni Þ
n¼2
X
1
vs ðtÞ ¼ V dc þ V sn sin ðn ωt þ θnv Þ
n¼1
ð3:44Þ
X
1
¼ V dc þ V s1 sin ðωt þ θ1v Þ þ V sn sin ðn ωt þ θnv Þ
n¼2
 The fundamental current and voltage components are:

is1 ðtÞ ¼ I s1 sin ðωt þ θ1i Þ ð3:45aÞ

vs1 ðtÞ ¼ V s1 sin ðωt þ θ1v Þ ð3:45bÞ

where Is1 and Vs1 are the peak values of the current and voltage fundamental
components, respectively.
The rms values of is(t) and vs(t) can be easily obtained using Eqs. (3.43) and
(3.44):
ðT
1
I 2s, rms ¼ i2s ðtÞdt
T 0
 2  2  2
I s1 I s2 I sn ð3:46Þ
I 2s, rms ¼ I 2dc þ pffiffiffi þ pffiffiffi þ . . . þ pffiffiffi
2 2 2
¼ I 2dc þ I 2s1, rms þ I 2s2, rms þ . . . þ I 2sn, rms n ¼ 1, 2, . . . , 1

Similarly,
 2  2  2
V s1 V s2 V sn
V 2s, rms ¼ V 2dc þ pffiffiffi þ pffiffiffi þ . . . þ pffiffiffi
2 2 2 ð3:47Þ
¼ V 2dc þ V 2s1, rms þ V 2s2, rms þ . . . þ V 2sn, rms n ¼ 1, 2, . . . , 1

These equations are obtained since the integration of the product of two different
frequency components over T is zero.
The instantaneous power is given by:

pðtÞ ¼ is ðtÞvs ðtÞ

X
1 X
1
¼ Idc V dc þ I dc V sn sin ðn ωt  θnv Þ þ V dc I dc V sn sin ðn ωt  θnv Þ
n¼1 n¼1 ð3:48Þ
! !
X
1 X
1
þ I dc V sn sin ðn ωt  θnv Þ V sn sin ðn ωt  θnv Þ
n¼1 n¼1

Evaluating the average value of p(t) of Eq. (3.48) over the fundamental fre-
quency, ω, shows that the second and third terms in Eq. (3.48) are zeros and the
fourth term is simply the sum of expressions similar to Eq. (3.46) but evaluated at
each harmonic component as shown below:

Pave ¼ I dc V dc þ I s1, rms V s1, rms cos ðθv1  θi1 Þ þ I s2, rms V s2, rms cos ðθv2  θi2 Þ þ . . .
X
1
¼ I dc V dc þ I sn, rms V sn:rms cos θn
n¼1
ð3:49Þ
where θn ¼ θvn  θin(n ¼ 1, 2, . . ., 1), which represents the phase shift between the
nth voltage and current harmonics.
It is clear from the above equation that true dc power can be obtained only if both
the line current and voltage have dc components. The second term represents the
average power at the source terminal obtained from the rms value of the harmonic
components.

3.4.3.3 Total Harmonic Distortion

Since the wanted portion of the distorted waveform is(t) is the fundamental com-
ponent, then the difference between the “desired” rms value of is(t) and the
“wanted” value is appropriately called the distorted portion of is(t) defined as:
X1
is, dist ¼ is ðtÞ  is1 ðtÞ ¼ i ðtÞ
n¼2 sn
ð3:50aÞ

X1
vs, dist ¼ vs ðtÞ  vs1 ðtÞ ¼ v ðtÞ
n¼2 sn
ð3:50bÞ
The relative measure of the distortion is defined through an index called the total
harmonic distortion (THD), which is the ratio of the rms value of the distorted
waveform and the rms value of the fundamental component. The THD in the
current and voltage is given in Eqs. (3.51a) and (3.51b), respectively, assuming
no dc components.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
I dist, rms I s2, rms 2 þ I s3, rms 2 þ I s4, rms 2 þ . . .
THDi ¼ ¼
I s1, rms I s1, rms
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi ð3:51aÞ

2
2
I s2, rms I s3, rms
¼ I s1, rms þ Is1, rms þ . . .

pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
V dist, rms V s2, rms 2 þ V s3, rms 2 þ V s4, rms 2 þ . . .
THDv ¼ ¼
V s1, rms V s1, rms
r
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi ð3:51bÞ
2
2
V s2, rms V s3, rms
¼ V s1, rms þ V s1, rms þ . . .

In terms of the rms of the original waveform, Eqs. (3.51a) and (3.51b) may be
rewritten as:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
  ffi
I s, rms 2
THDi ¼ 1 ð3:52aÞ
I s1, rms
 s
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

V s, rms 2
THDv ¼ 1 ð3:52bÞ
V s1, rms

It is also common to refer to THD as a percentage.

Exercise 3.13
Calculate the THDi for is(t) given in Example 3.6 and THDv for vs of E3.12.
Answer: 121 % , 100 % .

3.4.3.4 Power Factor

The equation to calculate power factor for distorted waveforms are more complex
when compared to the sinusoidal case discussed earlier. Applying the definition of
the power factor given in Eq. (3.37), to the distorted current and voltage waveforms
of Eqs. (3.43) and (3.44) and the average power given in Eq. (3.49) (with zero dc
components), pf may be expressed as:
P1 P1
n¼1 I s, rms V sn, rms cos θn n¼1 I s, rms V sn, rms cos θ n
pf ¼ ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P1 2 P1 2 ð3:53Þ
I s, rms V s, rms
n¼1 I sn, rms V sn, rms n¼1 vsn, rms

The above expression for pf can be significantly simplified if we assume the line
voltage is purely sinusoidal and distortion is only limited to is(t); thus it can be
shown that pf can be expressed as:

I s1, rms
pf ¼ cos θ1 ð3:54Þ
I s, rms

where θ1 is the phase angle between the voltage vs(t) and the fundamental
component of is(t). This assumption is valid in many power electronics applica-
tions. The line voltage is normally undistorted, and the line current is what gets
distorted, i.e.:

vs ðtÞ ¼ V s sin ωt ð3:55aÞ

is ðtÞ ¼ distorted ðnonsinusoidalÞ ð3:55bÞ

The current is expressed in terms of the Fourier series as follows:

is ðtÞ ¼ I 1 sin ðωt þ θ1 Þ þ I 2 sin ðωt þ θ2 Þ þ . . . þ I n sin ðωt þ θn Þ ð3:56Þ

resulting in the average power given by:
ðT
1
Pave ¼ vs is dt
T 0
ð T 
1
¼ ðV s sin ωtÞI1 sin ðωt þ θ1 Þ þ I 2 sin ðωt þ θ2 Þ þ . . . þ I n sin ðωt þ θn ÞDT
T 0
ð T 
1
¼ V s I 1 sin ωt sin ðωt þ θ1 Þdt
T 0

V s I1
¼ cos θ1
2
Pave ¼ I s, rms V s, rms cos θ1

Hence the power factor is given by:

I s1, rms V s, rms cos θ1 I s1, rms

Power factor ¼ ¼ cos θ1 ð3:57Þ
I s, rms V s, rms I s, rms

The expression Is1, rms/Is, rms is caused by the distortion of the line current and
appropriately called the distortion power factor, kdist, and the expression cosθ1, is
caused by the displacement angle between the line voltage and the fundamental
current component and commonly known as the displacement power factor, kdisp.
Hence, the power factor in power electronics is more useful if it is represented as a
product of the kdist and kdisp:

pf ¼ kdist kdisp ð3:58Þ

where

kdisp ¼ cos θ1
kdisp ¼ I s1, rms =I s, rms

In terms of kdist, it can be shown that the current THDi can be expressed as:
sffiffiffiffiffiffiffiffiffiffi
1
THDi ¼ 1 ð3:59Þ
kdist 2

Example 3.7
Consider a source with terminal voltage, vs(t), and with its terminal current, is(t), in
a nonlinear circuit given by the following expressions:

v ðtÞ ¼ 4 cos ðωo tÞ

 s   
is ðtÞ ¼ 3 cos ωo t  30 þ 2
cos 3ωo t  50
(a) Calculate the total rms values of the terminal voltage, vs(t), and current, is(t).
(b) Calculate PAVG (real power) delivered by source.
(c) Calculate the rms values of the fundamental components of the terminal
voltage, vs(t), and current, is(t).
(d) Calculate the current THD and voltage THD.
(e) Calculate power factor.
Solution
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

2
2
(a) V S rms ¼ p4ffiffi2 ¼ 2:83 V, I S rms ¼ p3ffiffi
2
þ p2ffiffi3 ¼ 2:55 A
 
(a) PAve ¼ 43
2 cos 0  30 ¼ 5:2 W
(c) Vs1 rms ¼ 2.83 V, I s1 rms ¼ p3ffiffi2 ¼ 2:12 A
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

2 ffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

2 ffi
(d) THDv ¼ V s rms
V s1 rms  1 ¼ 0%, THDi ¼ I s rms
I s1 rms 1¼ 66:8%
(e) Power Factor ¼ Apparent Power ¼ 2:832:55 ¼ 71.8%
Real Power 5:2 W

Or ¼ kdisp kdist ¼ cos 30  IIs1s rms

rms
¼ 0:866  2:12
2:55 ¼ 71:99%

Example 3.8
Calculate kdist, kdisp, THDii, and THDv for the waveforms shown in Fig. 3.16.
Solution
Since the voltage is purely sinusoidal, then THDv ¼ 0%.
To obtain the current THDi, first we obtain the rms value of is(t) and its
fundamental component:
ð 2πθ ð 2πθ ð 2πθ 
1 1  
I 2s, rms ¼ i2s dωt ¼ I 2o dωt þ I 20 dωt
2π θ 2π θ πθ

¼ I 20

Fig. 3.16 Waveforms for Example 3.8

 The fundamental component of is(t) is given by:

I s1 ¼ I 0s1 sin ωt þ I 00s1 cos ωt

where
ð 2πθ ð 2πθ 
1
I 0s1 ¼ I o sin ωtdωt þ I o sin ωtdωt
2π θ πθ

If we let ωτ ¼ ωt + θ, the above integral becomes:

ð π ð 2π 
1
I 0s1 ¼ I o sin ðωτ  θÞdωt  I o sin ðωτ  θÞdωt
2π θ π

Io
¼ ½ cos ðπ  θÞ þ cos ðθÞ þ cos ð2π  θÞ þ cos ðθÞ
π
4I o
¼ cos θ
π
Similarly, I 00S1 ¼ ð4I o =π Þ sin θ. Therefore, the peak fundamental component is
given by:
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2  2 4I o
I s1 ¼ I 0s1 þ I 00S1 ¼
π
Hence, the fundamental component of is(t) is given by:

is1 ðtÞ ¼ I s1 sin ðωt þ θÞ

and the rms of Is1 is given by:

pffiffiffi
4 I0 2 2
I s1, rms ¼ pffiffiffi ¼ Io
2π π

kdist, kdisp,and THD are given by the following expressions:

pffiffiffi
2 2
kdist ¼
π
kdisp ¼ cos θ
rffiffiffiffiffiffiffiffiffiffiffiffiffi!
π2
THD ¼  1  100%
8
¼ 48:34%

Example 3.9
The circuit diagram shown in Fig. 3.17a shows a conventional light-dimming
scheme used with incandescent lamps (incandescent lamps can be assumed purely
Fig. 3.17 (a) Conventional a
light-dimming circuit for
Example 3.9. (b) The +
voltage waveform. (c) The vs(t ) Conventional is(t )
current waveform + Light
vo
Dimming
–
Circuit
Incandescent
is(t ) –
Lamp
120Vrms
100W

b vo(t)

vs(t )
p
a=
2
a 2·p
0 a p t

p
a=
2

c
120Ö2 Vo(t )

120Ö2 V’s(t )
144
3p
2 2p
0 p p t
2
120Ö2
– V’s(t )
144

–120Ö2
Vo(t )

resistive). The voltage waveform of Fig. 3.17b shows the input and output voltages
for the dimming circuit shown at an instant the dimmer is reducing light intensity.
For this lighting control application, assuming that the incandescent lamp acts like a
pure resister, determine the following:
(a) The Fourier series coefficients for the fundamental component of the source
current is(t)
(b) The THDi of the source current is(t)
(c) The power factor as seen by the source
 pffiffiffi
Assume the vs ðtÞ ¼ V s, peak sin ωt ¼ 120 2 sin ð2:π:60tÞ V, and the rated lamp
power is 100 W at 120Vrms.
Solution
(a) Since it is assumed that the incandescent lamp acts like a pure resister, the
current is(t) will be the same shape as the output voltage as shown in Fig. 3.17c
and given by:

is ðtÞ ¼ I s, peak sin ωt

with its peak Is, peak is given by:

pffiffiffi
V s, peak 120 2
I s, peak ¼ ¼ ¼ 1:2 A
Rlamp 144

where Rlamp is equivalent lamp resistance at the full rated power, which is
obtained from:

V rms 2 1202
Rlamp ¼ ¼ ¼ 144 Ω
Po 100
By inspection, the current is(t) is half-wave symmetric with:
 
T
is ðtÞ ¼ is t 
2

Hence, its dc value is zero and coefficients an & bn ¼ 0 for n even and for
n odd:
ðT
4 2
an ¼ f ðtÞ cos nωtdt
T 0
ðT
4 2
bn ¼ f ðtÞ sin nωtdt
T 0

For n ¼ 1, the Fourier series coefficients for the fundamental component of

is(t) are obtained, where θ ¼ ωt:
ðπ
2
a1 ¼ I s, peak sin θ cos θdθ
π π
2
ðπ
2
b1 ¼ I s, peak sin 2 θdθ
π π
2

Using the trigonometric relation sin θ cos θ ¼ 12 sin 2θ and

sin θ ¼ 2 ð1  cos 2θÞ and evaluating the above two integrals, it can be
2 1

shown that the coefficients a1 and b1 are given by:

 I s, peak
a1 ¼  ¼ 0:38
2
I s, peak
b1 ¼ ¼ 0:59
2
Therefore
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
I s1, peak ¼ a1 2 þ b1 2 ¼ 0:7 A
I s1, peak
I s1, rms ¼ pffiffiffi ¼ 0:49 A
2
and the fundamental source current, is1(t), may be expressed by:
 
is1 ðtÞ ¼ I s1, peak sin ωt þ φ
 
is1 ðtÞ ¼ 0:7 sin 2:π:60t  32:78

(b) THDi for is(t) is given by:

sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
  ffi
I s, rms 2
THDi ¼  1  100%
I s1, rms

The rms value for the current is(t), Is, rms, is obtained from:

V o, rms 84:85
I s, rms ¼ ¼ ¼ 0:59 A
Rlamp 144

Therefore:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 
0:59 2
THD1 ¼  1  100% ¼ 62:6%
0:5

(c) The power factor is given by:

I s1, rms 0:49  

PF ¼ cos ðθv  θi Þ ¼ cos 32:78 ¼ 0:7
I s, rms 0:59

Therefore, current (fundamental) lags voltage by 32.78 .
Exercise 3.14
Calculate the THDi and the pf for the current waveform shown in Fig. 3.15, by using
up to the fifth harmonic.
Answer: 117 % , 0.64
Exercise 3.15
Calculate the THDi and the pf for the current waveform shown in Fig. E3.15.

Fig. E3.15 Waveform for Exercise 3.15

Answer: 136 % , 0.51.

3.5 Capacitor and Inductor Responses

The transient and steady-state values of the capacitor and inductor voltage and
current are well understood by undergraduate electrical engineering students.
However, a brief review of such responses might be useful to some readers at this
point.

3.5.1 Capacitor Transient Response

Consider the RC circuit of Fig. 3.18a with a dc excitation and an ideal switch.
Assume the switch was open for t < t0, and at t ¼ t0, the switch is closed. The
capacitor voltage for t < t0 is equal to the dc source, Vdc.
For t > t0 the time-domain capacitor current is given in terms of its voltage:

dvC
iC ¼ C ð3:60Þ
dt
Substituting for iC ¼ (0  vC(t))/R and solving for vC(t), we obtain the following
general solution:
 a +
R1
vc C

Vdc + – t = t0
–
R2
ic

b
vc (t )

Vdc

t
0 t0 ≅ 5τ + t0

ic (t )

Vdc/R2

t
0 t0

Fig. 3.18 (a) RC circuit with dc excitation and an ideal switch. (b) The current change through the
capacitor

 
vC ðtÞ ¼ vC, f þ vC, i  vC, f eðtt0 Þ=τ ð3:61Þ

where
vC, f ¼ final value at t ¼ 1.
vC, i ¼ initial value at t ¼ t0þ .
τ ¼ circuit time constant with τ ¼ R2C
Since the capacitor voltage does not change instantaneously, we have:

vC ðt0 Þ ¼ vC ðt0þ Þ

From the circuit diagram, at t ¼ 1 the capacitor becomes fully discharged

through R2 with a time constant R2C. The capacitor’s final value is given by:

vc, f ¼ vc ð1Þ ¼ 0
 The final expression for vC(t) and iC(t) for t > t0 is given by:

vC ðtÞ ¼ V dc eðtt0 Þ=τ ð3:62aÞ

V dc ðtt0 Þ=τ
iC ðtÞ ¼ e ð3:62bÞ
R
The sketch is shown in Fig. 3.18b. The steady state is reached at
approximately 5τ.
The capacitor current shows that at t ¼ t0 it was zero, but at t ¼ t0þ , it suddenly
becameVdc/R2. This brings us to another important statement about the capacitor
current: The current through the capacitor can change instantaneously in switching
circuits as shown in Fig. 3.18b.
Example 3.10
Consider the circuit of Fig. 3.19a where the switch has been closed for t < 0. At
t ¼ t0þ , the switch is opened. Sketch the capacitor voltage and current for t > 0.
Assume the capacitor was initially uncharged.
Solution
The capacitor voltage equation is given by:

dvc
C ¼ iC
dt
¼ Ig

Fig. 3.19 (a) Circuit for Example 3.10. (b) The capacitor current and voltage waveforms
Fig. 3.20 (a) Series combination of capacitor and diode. (b) Capacitor current and voltage
waveforms

Therefore, vC(t) is:

Ig
vC ðtÞ ¼ t
C
The capacitor current and voltage waveforms are shown in Fig. 3.19b.
Since the capacitor voltage does not change instantaneously, the switch in this
example is not allowed to close again unless a way is found by which the switch
voltage is prevented from appearing across the capacitor (i.e., voltage diversion).
This can be accomplished by adding a diode in series with the capacitor as shown in
Fig. 3.20a. As long as the switch is closed, the capacitor voltage remains constant.
At t ¼ t1, the switch is closed again, forcing D to turn OFF, since its anode
voltage is pulled to the ground. The capacitor voltage at t ¼ t1 is given by:

Ig
vc ðt1 Þ ¼ t1
C
In practice, the energy stored in the capacitor dissipates through the capacitor’s
equivalent resistance and the diode’s leakage current, as illustrated in Exercise 3.16.
Exercise 3.16
Consider the circuit of Fig. 3.20a with the switch waveform shown in Fig. 3.21.
Assume an ideal switch and ideal diode except it has a 100 μA leakage current.
What are the capacitor voltages at t1 ¼ 5 μs, 10 μs, 0.5 s? At what time does the
capacitor voltage become zero again?
Answer: 5 V, 10 V, 5 V, 1 s
Fig. 3.21 Switch
waveform

3.5.2 Capacitor Steady-State Response

Now let us consider the circuit of Fig. 3.18a with the switch turned ON and OFF
repeatedly according to the waveform shown in Fig. 3.22a. After a few switching
cycles, the capacitor voltage reaches steady state at which its value at the beginning
of each switching cycle is the same. Mathematically, this can be shown as follows:

vC ðnT þ t0 Þ ¼ vC ððn þ 1ÞT þ t0 Þ ð3:63Þ

Figure 3.22b, c show the steady-state capacitor voltage and current waveforms,
respectively. Now we come to another important property of the capacitor. In
steady state, the average capacitor current is zero.
ð ðnþ1ÞT ð ðnþ1ÞT
1 dvC C
I c, ave ¼ iC dt ¼ C dt ¼ ½vC ððn þ 1ÞT  vC ðnT ÞÞ ¼ 0
T nT nT dt T
ð3:64Þ

This is illustrated by the equal negative and positive shaded areas for iC(t) shown
in Fig. 3.22c.

3.5.3 Inductor Transient Response

Similarly, let us look at the transient response for the inductor current by consid-
ering the dual circuit of Fig. 3.18a, which is shown in Fig. 3.23. Assume the switch
was open for a long time before it is closed at t ¼ t0. Assume that initially the
inductor current is zero.
For t > t0, the time-domain inductor voltage is given by:

diL
vL ¼ L ð3:65Þ
dt
Substituting for vL ¼ R(Idc  iL(t)), where R ¼ R1 j j R2,and solving for iL(t), we
obtain the following general solution for iL(t):
 
iL ðtÞ ¼ iL, f þ iL, i  iL, f eðtt0 Þ=τ ð3:66Þ
Fig. 3.22 (a) Switch waveform, (b) steady-state capacitor voltage, (c) steady-state capacitor
current

Fig. 3.23 Inductor

switching circuit
where
iL, f ¼ final value as t ¼ 1
iL, i ¼ initial value at t ¼ t0þ
τ ¼ circuit time constant with τ ¼ L/R
Since the inductor current does not change instantaneously, we have:

I L ðt0 Þ ¼ I L ðt0þ Þ ð3:67Þ

From the circuit, it is clear that at t ¼ 1, iL(t ¼ 1) ¼ iL, f ¼ Idc, and its voltage
becomes zero.
The final expression for iL(t) is given by:


iL ðtÞ ¼ I dc 1  eðtt0 Þ=τ

and the inductor voltage is given by:

vL ðtÞ ¼ RI dc eðtt0 Þ=τ

Again at t ¼ t0 , vL ðt0 Þ, and at t ¼ t0þ , the inductor voltage is given:

vL ðt0þ Þ ¼ RIdc

We conclude that the voltage across the inductor can change instantaneously
under switching action, as shown in the inductor current and voltage waveforms in
Fig. 3.24.
Exercise 3.17
Consider the switching circuit given in Fig. E3.17 with an ideal diode. Assume the
initial inductor current is zero and the switch is turned on at t ¼ 0 and turned off
at t ¼ 10 μs. Sketch the inductor current waveform for t > 0. If the diode has a 5 Ω
forward resistance, what is iL at t ¼ 0.5 μs, 10 μs, and 100 μs? At what time does the
inductor becomes discharged?

Fig. E3.17 Switching

circuit for Exercise 3.17

Answer: 50 μA, 1 mA, 0.95 mA, 10 ms

Fig. 3.24 Inductor current and voltage waveforms

Notice in the above exercise that when the switch is opened again at t ¼ 10 ms,
the inductor current gets trapped and the diode turns on and diverts the current from
the switch branch to the diode branch. This is known as current commutation and
will be used extensively in Chaps. 5 and 6. In practice, trapped energy in the
inductor would dissipate through the inductor and diode resistances. To improve
the converter’s efficiency, normally the trapped energy is allowed to be reconnected
by returning it to the source through a feedback circuit along with the use of
transformers, to be illustrated in Chap. 5.

3.5.4 Inductor Steady-State Response

Let us reconsider the same circuit of Fig. 3.23 except that the switch turns ON and
OFF repeatedly according to the waveform shown in Fig. 3.25a.
After a few switching cycles, the inductor reaches steady state at which its value
at the beginning of each switching cycle is the same. Mathematically, we express
the steady-state condition as:
 iL ðnT þ t0 Þ ¼ iL ððn þ 1ÞT þ t0 Þ ð3:68Þ

The above condition makes it possible to show that the average value of the
steady-state inductor voltage is zero. This is shown as follows:
ð ðnþ1ÞT ð ðnþ1ÞT
1 1 diL L
V L, ave ¼ vL dt ¼ L dt ¼ ½iL ððn þ 1ÞT Þ  iL ðntÞ ð3:69Þ
T nT T nT dt T

The steady-state current and voltage waveforms are shown in Fig. 3.25b, c,
respectively. The zero average inductor voltage is illustrated by the equal negative
and positive shaded areas for vL(t) shown in Fig. 3.25c.

a
Switch

On Off On

t
0 T nT (n + 1)T

b
iL(t )

t
0 nT + t0 (n + 1)T + t0

c
vL(t )

+
t
0 –

Fig. 3.25 Inductor current and inductor voltage waveforms for Fig. 3.22 when the switch is
repeatedly turned on and off according to waveform (b) and (c), repectively
Problems

In all the problems, assume ideal diodes unless stated otherwise.

3.1 Derive the expressions for iL(t) and vc(t) in switching circuits shown in
Fig. P3.1. Assume all initial conditions are zero. Use Vdc ¼ 20 V, Idc ¼ 2 A,
R ¼ 200 Ω, C ¼ 0.01 μF, and L ¼ 2 mH, sketch iL and vC.

Fig. P3.1

3.2 Assume the switch in Fig. P3.2 is opened at t ¼ 0+, derive the expressions for iL,
vc and isw and sketch them for t > 0. Assume iL(0) ¼ Ig and vc(0) ¼ Vo.

Fig. P3.2

3.3 Consider the switching diode circuits of Fig. P3.3 for t > 0 when the switch is
turned on. (a) Derive expressions for iL(t), isw(t), and vc(t). Assume iL(0) ¼
1.0Ig and vc(0) ¼ Vo. Suggest possible switch implementations (unidirectional
or bidirectional).
Fig. P3.3

3.4 Consider the transistor switching circuits of Fig. P3.4. At t ¼ 0 the transistor is
turned on by a signal to the base. Derive and sketch the waveforms for iQ, iL, vc,
and vo. Assume vc(0) ¼ 20 V and iL(0) ¼ 0. Compare the two circuits.

Fig. P3.4

3.5 Consider the diode circuit of Fig. P3.5 with switch S1 opened and S2
closed at t ¼ 0. Assume the initial conditions are given as vc1(0) ¼ 0
and iL(0) ¼  0.5 Ig. Derive the expressions for vc1 and iL and sketch them.
Use Ig ¼ 2 A, C1 ¼ C2 ¼ 0.01 μF, and L ¼ 2 mH.

Fig. P3.5
3.6 Assume the switch in Fig. P3.6 is opened at t ¼ 0+. Derive the expressions
and sketch the waveforms for vc1, vc2, iL, and iD. Assume vc1(0) ¼
vc2(0) and iL(0) ¼ 0.

Fig. P3.6

3.7 Consider the capacitor current shown in Fig. P3.7 for a given switch-mode
power supply with C ¼ 1 μF, and assume vc(0) ¼ 100 V. (a) Sketch vC showing
the peak values and times, and (b) determine the ripple voltage across the
capacitor.

Fig. P3.7

3.8 Consider the SCR-diode switching circuit shown in Fig. P3.8 with the switch
closing at t ¼ 0. This circuit is a voltage-commutation circuit known as the
impulse-commuted chopper used to force the turnoff of the SCR by additional
switch S. Assume ideal diode and the capacitor were initially charged toVo
and the SCR was on for t < 0.
(a) Sketch the waveforms of iSCR, vc, iD, and vo.
(b) Show that the time, Δt, it takes for iC to reach zero again after t > 0 is
given by:
ðV dc þ V o ÞC
Δt ¼
Io
Discuss the drawback of such an arrangement.
Fig. P3.8

3.9 For the SCR to turn off in the circuit of Fig. P3.8, the capacitor voltage must
first be charged to a large negative value. This is done through an external
circuit consisting of a diode and an inductance as shown in Fig. P3.9. The
purpose of D1 and L is that when the switch is turned off, the capacitor
voltage returns to its original negative value, Vo. Assume S and SCR were
open for a long time with a capacitor initial value equal to+Vo. At t ¼ 0, SCR
is triggered. Sketch the waveforms for iD1, iD0, iSCR, vC, and vo. Compare this
circuit with the circuit given in Problem 3.8.

Fig. P3.9

3.10 Consider the SCR circuit shown in Fig. P3.10. Because the SCR turns OFF
naturally due to the fact that its current becomes zero, the circuit is known as a
self-commutated circuit. At t ¼ 0, the SCR is turned ON by applying ig. Derive
the expressions for vc(t), iL(t), and the SCR commutation time. Let L ¼ 0.1 mH,
C ¼ 47 μF, and Vdc ¼ 120 V. Assume the initial capacitor voltage is (i) zero
and (ii) Vi (where |Vi| < Vdc).
Fig. P3.10

D3.11 Design for L and C in Problem 3.10 so that the SCR commutation time
is 100 μs, and the peak capacitor voltage does not exceed 200 V.
3.12 Consider a source terminal connected to a power electronic circuit whose
current and voltage waveforms are shown in Fig. P3.12, with t1 ¼ 4 μs
and T ¼ 10 μs.
(a) Calculate the average and rms source current and source voltage values.
(b) Calculate the average input power.

Fig. P3.12
 3.13 Consider the circuit of Fig. P3.13 where vs is a train of pulses as shown.
Assume RC  T so that Vo is assumed constant and equal to 75 V. Sketch
the steady-state inductor current waveform, iL.

Fig. P3.13

3.14 The switch in Fig. P3.14 is closed at t ¼ 0. Assume ideal diodes and all
inductor initial conditions are zero. Obtain the expressions for the inductor
current and sketch them for 0  t  2T. Assume vs(t) is sinusoidal with
ω ¼ 377 rad/s and Vs ¼ 100 V. Let L ¼ 1 mH, R ¼ 1 Ω, and Vdc ¼ 20 V.
Fig. P3.14
 3.15 Consider a power source with its terminal voltage and current given as:
   
vs ðtÞ ¼ 100 þ 80 sin ωt  100 þ 70 cos 2ωt þ 120 þ 25 sin 3ωt V
   
is ðtÞ ¼ 12 þ 10 sin ωt þ 25 þ 5 sin 2ωt  30 þ 2 cos 3ωt A

where ω is the fundamental angular frequency.

Calculate:
(a) The rms value of is(t) and vs(t)
(b) The average input power supplied by the source
(c) The rms values of the fundamental components of is(t) and vs(t)
(d) THDi and THDv
(e) kdist, kdisp, and pf

3.16 Consider the phasor circuit of Fig. P3.16 with ZL ¼ 20 ∠  36 Ω and

Vs ¼ 80 ∠ 0 V.
(a) Determine the circuit’s real, apparent, and reactive powers and the input
power factor.
(b) Determine the type and value of the load needed to be connected between
0
a and a to achieve unity power factor.

Fig. P3.16

3.17 Prove the following integrals:

ðT
Fn ¼ sin n ωt sin m ωtdt
ðT 0

¼ cos n ωt cos m ωtdt

0(
0 n 6¼ m
¼ T
n¼m
2
and
 ÐT
Fn ¼o sin ðnωtÞ cos ðmωtÞdt
8
<0 n 6¼ m
¼
: T n¼m
2
Where ω ¼ 2π/T, and n ¼ 1, 2, . . . 1 ; m ¼ 1, 2, . . . 1 .
3.18 Sketch the steady-state iL and vL waveforms by assuming the switch in
Fig. P3.18 is repeatedly opened and closed as shown. Use T ¼ 10 ms.

Fig. P3.18

3.19 Show that the Fourier series for the half-wave rectifier output voltage of
Fig. P3.19 is given by:

Vs 2Vs 2Vs X
1
1
v o ðt Þ ¼ þ sin ωt  cos ωt
π π π n¼2, 4, 6...
n 21

Fig. P3.19
3.20 Show that the Fourier series for the output voltage waveform of a full-wave
rectifier of Fig. P3.20 is given by:

2V s 2V s X
1
1
vo ðtÞ ¼  cos ωt
π π n¼2, 4, 6...
n 21

Fig. P3.20

3.21 By considering only the first four terms of the Fourier series given in Problem
3.19, verify the answers given in Exercise 3.11.
3.22 Derive the Fourier coefficient equations an and bn given in Eq. (3.41).
3.23 The phase current waveform in a three-phase full-wave rectifier SCR circuit
under highly inductive load is shown in Fig. P3.23. Show that the harmonics
of i(t) are given by the following expression:
pffiffiffi  
2 3I o cos 5ωt cos 7ωt cos 11ωt cos 13ωt
iðtÞ ¼ cos ωt  þ  þ þ ...
π 5 7 11 13

No triple harmonics in i(t) and n ¼p6k

ffiffiffi  1(k ¼ 0, 1, 2, . . .). Show that the
rms magnitude of the nth harmonic is 6Io =πn.

Fig. P3.23
3.24 Show that the Fourier series for the phase current of a six-pulse SCR converter
shown in Fig. P3.24 is given by:

6Io cos 5ωt cos 7ωt cos 11ωt cos 13ωt
iðtÞ ¼ cos ωt þ   þ
π 5 7 11 13
cos 17ωt cos nωt
þ þ ... þ
17 n

Fig. P3.24

3.25 (a) Figure P3.25a shows a typical ac line current waveform in a single-phase,
full-wave SCR controlled rectifier circuits under resistance load and
sinusoidal ac source. Show that the fundamental line current component
is given by:

is1 ðtÞ ¼ I s1 sin ðωt þ θÞ

where

2π
ω¼
T
cos 2α  1
θ ¼ tan 1
2π  2α þ sin 2α
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
I s1 ¼ 1 þ 2ðπ  αÞ2 þ 2ðπ  αÞ sin 2α  cos 2α
Is
 pffiffiffiffiffi
2π
(b) Find is1(t) for the half-wave SCR controlled line current shown in
Fig. P3.25b.
Fig. P3.25

3.26 Show that the high harmonic Fourier coefficients for Fig. P3.26 are given by:

an ¼ 0
X
1 
1 2
bn ¼ ð2α þ sin 2αÞ þ sin αð1  nÞ
π n¼3, 5, 7
π ð 1  nÞ

What value of h will achieve 95% input power factor assuming the
waveform represents a line current with a sinusoidal line voltage
Fig. P3.26

3.27 Figure P3.27 shows the ac line current for a full-wave, single-phase SCR
controlled converter under highly inductive load. Show that the an coefficients
are given by:

4I o nð π  α Þ
an ¼ sin
nπ 2
for n ¼ 1, 3, 5, 7, . . .

Fig. P3.27

3.28 Show that the Fourier components of is(t) of Fig. P3.28 are given by:

4I o X
1
i s ðt Þ ¼ cos nα sin n ωt
nπ n¼1, 3, 5, ...
Fig. P3.28

3.29 Determine the Fourier components for the waveforms given in Fig. P3.29.

Fig. P3.29

General Problems

3.30 In the circuit of Fig. P3.30, the switch is opened at t ¼ 0 after it has been
closed for a long time. Derive the expression for iL(t), iD(t) and vC(t) for t > 0,
and sketch them.
Fig. P3.30

3.31 Show that the load current after the switch is closed at t¼0 in the diode
circuit given in Fig. P3.31 is given by:

io ðt Þ ¼ 0 0  t < t1
Vs V dc t
io ðtÞ ¼ sin ðωt  θÞ  þ Aeτ t1  t < t2
jZ j R
i o ðt Þ ¼ 0 t2  t < T

where

τ ¼ L=R
ωl
θ ¼ tan 1
R
V dc V s
A¼  sin ðωt1  θÞ
R jZ j

Assuming the inductor is not initially charged, and here t1 is the time
when the diode turns on and is given by ωt1 ¼ sin1(Vdc/Vs), and t2 is the
time at which the diode turns off before the next cycle starts at t ¼ T. Find the
expression for t2.

Fig. P3.31
 3.32 The circuit of Fig. P3.32a is known as a single-phase bridge inverter whose
purpose is to convert the dc input voltage, Vdc, to an ac output voltage, vo. If
the switching sequence of S1, S2, S3, and S4 is done in such a way that the
output voltage is shown in Fig. P3.32b,
(a) show that the Fourier series for v0 is given by:

V o ¼ V 1 sin ωt þ V 3 sin 3 ωt þ V 5 sin 5 ωt þ . . . þ V n sin n ωt

where

4V in nα
Vn ¼ sin n ¼ 1, 3, 5, 7,
nπ 2

Use the above results to find the Fourier series for io(t).

Fig. P3.32

3.33 Consider the two-phase switching commutation circuit shown in Fig. P3.33.
Assume S1 has been on for a long time prior to ωt ¼ 150o when S2 is turned
on. (a) Sketch the waveforms for ia(t) and ib(t), and (b) determine the time
during which both D1 and D2 were on. Assume va(t) ¼ 100 sin(377t) and

vb(t) ¼ 100 sin(377t  120 ) V. Use LS ¼ 1 mH and Rs ¼ 2 Ω.
Fig. P3.33

3.34 The circuit shown in Fig. P3.34 is known as a forced commutation circuit
whose voltage waveform over one switching period, TS, is shown for t > 0.
Derive the expression for vC(t) over one switching cycle. The conduction
states of SCR1 and SCR2 are shown on the waveform.

Fig. P3.34
 3.35 Figure P3.35 shows a DIAC-TRIAC switching circuit used in a heater
controller, motor speed variation and light dimmer applications. Design for

R and RL so that the TRIAC triggers at 30 and 210 during the positive and
negative cycles, respectively. Assume vs(t) ¼ 110 sin 2 π60t and the DIAC
breakover voltage is 24 V.

Fig. P3.35

3.36 Assume the switch is turned on at t ¼ 0 and the capacitor was initially
discharged in the circuit of Fig. P3.36. Derive the expression for vC(t) and
sketch it for 0 < t < 40 ms, where vs(t) ¼ 110 sin 2 π50t, R ¼ 10 kΩ, and
C ¼ 1 μF.

Fig. P3.36

3.37 Figure P3.37 shows a self-oscillating LC circuit that allows the SCR to turn
off naturally without using an additional auxiliary SCR. This circuit is
known as a series resonant turnoff circuit. (a) Derive the expression for
vC(t) and iL(t). (b) Determine the power rating of the SCR. Assume
the SCR is first triggered at t ¼ 0 and repeatedly every T ¼ 2 ms. Assume
vC(0) ¼ 30 V and iL(0) ¼ 0.
Fig. P3.37

3.38 The circuit given in Fig. P3.38 represents one possible implementation of a
family of dc-dc converters known as soft-switching converters.
(a) If we assume the switch is turned ON at t ¼ 0, show that the capacitor
voltage for t > 0 is given by:

vc1 ¼ V o þ ðV dc  V o Þ cos ωo t
pffiffiffiffiffiffiffiffiffiffi
where ωo ¼ Ceq L and Ceq ¼ C1 + C2.
(b) In order to turn S back on while the capacitor voltage across C1 is zero,
the voltage across it must be allowed to reach zero again during the
off-time of the switch. Show that for zero-voltage switching to occur, the
following condition must be met:

V dc
Vo <
2

Fig. P3.38
3.39 Figure P3.39 shows output and input waveforms in a cycloconverter-type
power electronic circuit, where the output frequency is one-half of the input
voltage frequency. Find the Fourier series representation for such waveform.

Fig. P3.39

3.40 Figure P3.40 shows a typical inductor current waveform in a switch-mode

power supply when operating at the boundary of continuous and discontinu-
ous mode operations. (a) Determine the Fourier components for i(t), (b) cal-
culate its average and rms values, and (c) calculate the THDi.

Fig. P3.40
3.41 Repeat Problem 3.40 for the inductor current shown in Fig. P3.41, which
represents a discontinuous conduction mode of operation in a switch-mode
power supply.

Fig. P3.41

3.42 Figure P3.42 shows a half-wave rectifier waveform with a sine-squared pulse
represented mathematically for 0 < t < T/2 by the following equation:

is ðtÞ ¼ I o sin 2 ωt

(a) Determine the Fourier coefficients.

(b) Calculate the rms and THDi values.

Fig. P3.42

3.43 Consider a sinusoidal line current shown in Fig. P3.43 that can be represented
mathematically for 0 < t < T by the following equation:

is ðtÞ ¼ I o sin n ωt

where n is an odd integer. By assuming that the line voltage is given by vs(t) ¼
Vs sin ωt V, derive the expression for the power factor the THDi.
Fig. P3.43

3.44 The simplified equivalent circuit of a dc-dc converter known as a buck

converter is shown in Fig. P3.44a. Assume the switch is turned on and off
according to the waveform shown in Fig. P3.44b Sketch the steady-state
waveforms for vsw, isw, and io and derive the expression for the average output
power.

Fig. P3.44

3.45 Figure P3.45 shows eight different switch-mode topologies of which only one
topology is valid. Identify such topology and state which circuit law (KVL or
KCL) is being violated for each of the other topologies.
Fig. P3.45

3.46 Sketch the waveforms for iD1, iD2, vo and find the average output voltage for
the circuit of Fig. P3.46a, where vs1(t) and vs2(t)are shown in Fig. P3.46b.
Fig. P3.46 a
iD1
D1 +

vs1 +
vo Io
–

vs2 +
–

D2
iD2

b
vs1

Vs

t
0 T/2 T

–Vs

vs2

Vs

t
0 T/2 T

–Vs

3.47 Repeat Problem 3.46 by replacing D1 and D2 by two switches S1 and S2 as

shown in Fig. P3.47. Assume that S1 is conducting only when |vs1(t)| > Vs/3
and S2 is the complement of S1.
Fig. P3.47

3.48 Consider the full-bridge rectifier of Fig. P3.48 with a current source is(t) ¼
Is sin ωt and a constant output voltage, Vo. Sketch vs and io, and find the
average output power.

Fig. P3.48

3.49 Consider the full-bridge rectifier of Fig. P3.49 with a voltage source vs(t) ¼
Vs sin ωt and a constant output current, Io. Sketch is and vo and find the average
output power.
Fig. P3.49

3.50 Figure P3.50 shows four different switch-mode topologies of which only one
topology is valid. Identify such topology and state the circuit law (KCL or
KVL) that is being violated for each of the other topologies.

Fig. P3.50

3.51 Figure P3.51a shows an equivalent circuit for a switch-mode converter—Cuk

(pronounced chook). Show that the ratio between the average output voltage,
Vo, and the average input voltage, Vin, is given by the following relation:

V0 D
¼
Vin 1  D
The switch waveform is shown in Fig. P3.51b.
Hint: Use the average voltage Vx as given by Vin  Vo.
Fig. P3.51

3.52 Figure P3.52a shows a switch-mode power electronic converter known as the
buck-boost converter. This circuit is used to convert a DC input voltage to
another dc level at the load. If we assume that the inductor, L, is large and the
average output current is represented by a constant current source, then the
equivalent circuit becomes as shown in Fig. P3.52b. Assume the switch is
turned on and off as shown in Fig. P3.52c. Figure P3.52a show that the
capacitor current is given in Fig. P3.52d, b sketch the steady-state average
capacitor voltage.
Fig. P3.52 a S iD

–
Vin +
L C R Vo
–
+

b
S iD

ic
Vin +
IL C Io
–

c
Switch

On Off On
t
0 DT T (1 + D)T

d
ic

IoD/(1 – D)

t
0 DT T
–Io

3.53 Determine the power factor for the circuits given in Problems 3.48 and 3.49.
3.54 Consider a transformer with a turn ration N1/N2 ¼ 200, and rated at 240 kVA
(44 kV ⁄ 220 V) used to step down a 60 Hz voltage in a distributed system.
(a) Determine the rated primary and secondary currents.
(b) Determine the load impedance seen between the primary terminals when
the load is fully loaded.
3.55 Consider the periodical current waveform shown in Fig. P3.55.
Determine its RMS expression. If you can only reduce one of the duty
ratios D1, D2, and D3 by 10%, determine which one will give you the smallest
reduction in RMS value.

Fig. P3.55