ac circuit

analysis
with illustrative
problems

Phillip Cutler
Education Research Associates
and
Orange Coast
College
Costa Mesa
California

McGraw-Hill Book Company

New York
St. Louis
San Francisco
Dusseldorf
Johannesburg
Kuala Lumpur
London
Mexico
Montreal
New Delhi
Panama
Rio de Janeiro
i
Singapore
Sydney
Toronto
 A need frequently arises to examine the frequency re-
sponse (both amplitude and phase) of a circuit or ampli-
fier. This is particularly true in more advanced courses
where attention is focused on the behavior of feedback
amplifiers or control systems. Hence it is advisable to
introduce the subject at an elementary level so that we
will be on firmer ground in the future.

11-1 The transfer function

Most any circuit or system that we might consider has
some kind of input and output. Generally we are in-
terested in output voltage versus input voltage or voltage
11 the Bode plot gain as a function of frequency, but this need not be the
case. We could just as well be interested in current gain,
or output voltage versus input current, and so on. When
dealing with control systems we might be interested in
motor speed versus armature voltage, or height of liquid
in a tank versus a potentiometer setting, etc. In any
event, the output versus input relationship as a function
of frequency is, for our purposes, called the transfer
function. The transfer functions of most concern to us
are exclusively electrical in nature and generally a
voltage gain.
There are a few different methods of plotting transfer
functions versus frequency. One of the simplest and most
useful is called the Bode plot, after a famous scientist who
did some outstanding research in feedback theory. It
turns out that to conveniently construct a Bode plot, the
transfer function is best expressed in certain algebraic
forms which are readily conveyed by example.

11-2 The low-pass filter

The first circuit we will develop the transfer function
for is the low-pass filter of Fig. 11-1. We seek the voltage

j
E! 1

c?
'•"'-'i
i E2 , 1
•" ~ RC

&
Low-pass filter
FIGURE 11-1

gain Ay = E;/Ei as a function of frequency. Clearly we

have a voltage divider here and hence
E
! -JXc l/Jtl)C
1) A,
Ei R - jXc
R +
jcoC

the Bode plot 149

Remember, we can transfery from numerator to denomi- As / increases, we would expect Ay to decrease and
nator and vice versa simply by changing its sign, since exhibit some phase lag since E; is taken off the capacitor
this corresponds to the similar transfer of an angle in and capacitor voltage lags capacitor current by 90'.
polar form. Inspection of Eq. (11-lfe) [or Eq. (11-la)] indicates that
Finding a common denominator and simplifying (1), when/ the frequency of the input signal £\, is the same as
we obtain /,,, the quantity ///,, = 1 and hence the magnitude of
A, = 1\ 2 = 0.707 and the phase angle is 0 = - t a n ' '
2) ..,- •
I + JDRC 1 + jw,, 1 = -45s.
As/becomes very large, and in particular for/>/,,,
where t h e y term in Eq. (11-la) approaches infinity, so that
3) T, = RC A,, s I'y'x = -jO, which means the gain magnitude
approaches zero and the gain phase is —90°. This is to be
Now it turns out that the product RC is called a time expected, since at high frequencies Xc -^ R. so that the
constant and it has the dimensions of seconds, since 1 circuit impedance to E, is essentially resistive and hence
V A / sec the current tends to be in phase with Ei. However, this
4, RC^.^ approximately zero-degree phase angle of the current is
A'^"'^
added to the 90° phase lag (- 90°) of the capacitor to
Now it is convenient to reason that the time constant T^ yield the phase of E;. Hence E.; lags E i by approximately
is the period corresponding to a fictitious radian fre- 90° at/ ?> /,. Of course, with Xc approaching zero in the
quency co,, so that high-frequency spectrum (/>/»,), Xc < R and hence the
magnitude of £^ approaches zero. Thus the gain magni-
5) (^ = — tude approaches zero.
Now work through problems PS 11-1, PS 11-2, and
or
PS 11-3.
6) TA = —
^h
11-3 The high-pass filter
The use of the h subscript will be clarified later. Substitut-
Another circuit commonly encountered is the high-pass
ing (6) into (2) yields, for the low-pass filter,
filter of Fig. 11-2. This circuit is commonly used as a
-^ coupling network in many amplifier circuits where E,
7) (11-lfl)
E,
1^ l^ »
c->h h —H^-T
c
Notice how simple this expression is and, in particular. . f
note its form. In polar form we should express (7) as R e. E, . 'T
: 2 El -
^'T
E,
8) 2
E> y l +(///J /tan-•///, &
1 High-pass filter
2 /-tan- 1 ///, (11-lfe)
v'l + (//A) FIGURE 11-2
For reasons to become apparent later, the frequency /,,
is called the comer or break frequency and it is extremely contains a dc as well as an ac component. Usually we
important in constructing Bode plots. want only the ac component to appear at the output.
It might be interesting to correlate the mathematical The capacitor C, also called a blocking capacitor,
form of 11-la (or 11-lfc) with a qualitative analysis of the assures this, since only a time-varying voltage can cause
circuit. Surely at low frequencies Xc is so much larger capacitor current to flow.2
than R that the magnitude of E^ s £1. Correspondingly We can develop the voltage transfer function by the
in (11-la) for/approaching zero, or more significantly voltage divider relationship and writing
for/^/t, they term approaches zero and A,, s 1/0°. R
1) A, -^
i - y x^_
' In a similar manner, the time constant L / R in an inductive circuit has E, R - jXc
the dimensions of seconds, since
V R wCR
L H A/sec 2
Recall that i, = C-^.
J " 'a = ~v7A = sec
at

150 ac circuit analysis

 If we let Now work through problems PS 11-4, PS 11-5. and
I PS 11-6.
2) ^
C~R Note that, in all the problems, we have tried to ex-
we can rewrite ( 1 ) as press the transfer function so that it contains terms
of the form j x and'or 1 + j x . This will be possible
3» A, (11-2^) whenever we deal with circuits containing a single-
1-7 energy storage element such as just one capacitor or one
inductor.
Now Eq. (3) is perfectly valid, but it is more desirable
(also for reasons to become apparent later) to get rid of 11-4 The decibel
the negative denominator term. This may be accom-
The transfer functions encountered in practice will in
plished by multiplying the numerator and denominator
general contain factors in the numerator and denomi-
of (3) by j f f i . If you do this and rearrange terms, there
results
nator. Therefore evaluation of the transfer function at a
particular frequency requires the process of multiplica-
tion and/or division, neither of which is conveniently
^ performed by graphic techniques. Since a graphic display
4) A, = (ii-2fr) of the transfer function is our goal. we seek an alternative
approach.
^'7, You will recall that with the aid of logarithms, the
which may also be written as operations of multiplication and division are converted
to addition and subtraction, since
5) A, = — j ' 11 —= 90-' - tan"' /• /,
1) log afr = log a + log h (ll-3a)
v 1 +(/^) 2/ ——————————
(H-2c) 2) log — = log a — log b (1 l-3fc)
b
Although (4) appears more complicated than (3). it
is somewhat easier to Bode plot, which after all is our and raising a number to a power converts to multiplica-
goal. tion
To correlate (5) with a qualitative analysis of Fig. 11-2, 3) log ax = x log a (ll-3c)
we may reason as follows. In the low-frequency spectrum,
say for/<§./(, the numerator magnitude s 0 and the Now graphic addition and subtraction is no problem
denominator magnitude s 1. Hence for/<$/(, A,, s 0. since we need only count divisions or lay off distances
Physically we should expect this. since A"c > R at low with a pair of dividers. For this reason, among others,
frequencies. The question is what constitutes low fre- the vertical scale in our graphic display will be expressed
quencies'? The answer is that / in the high-pass or /,, in a logarithmic unit called the decibel, abbreviated dB.
in the low-pass filter forms a convenient spot frequency The decibel is formally defined as a unit of power gain
to separate the low and high frequency spectrum. given by
Since at low frequencies A'c > R, the current leads
EI by almost 90°; hence the output voltage which is taken 4) A, = 10 log - f l dB (ll-4a)
off the resistor should be in phase with the current. 'i
Inspection of the phase portion of (5) indicates that for Equation (ll-4a)T is read as, "the power gain Ap, ex-
/<0i the tangent is s 0° and hence the net angle is pressed in dB, is equal to 10 times the log of the power
90° - 0° or S 90°. ratio P^/PI." A more common form of expressing Eq.
At / = f t where / is the corner or break frequency, (ll-4a) is shown in Eq. (H-4fc):
| A, = l/v'2 and the phase is 90° - 45° = 45°.
In the high-frequency spectrum (/ > j\) we see from 5) A p(dB)= 10 log ^ (11-4&)
Eq. (4) that the reactive term in the denominator is
predominant, and hence A,, approaches 1/0'''. This is Although Eq. (ll-4a) is in better mathematical form,
in agreement with the physical facts, for X,- < R. E^\ S Eq. (H-4fc), which is read and interpreted in the same
£"1 . Furthermore, the current in this case is essentially way, is more prevalent and hence both forms will be
determined by R and hence is almost in phase with Ep used here.
Since E^ is taken off the resistor which has no phase Symbols other than A are commonly used to designate
angle associated with it, it follows that E; is essentially the gain or ratio between two quantities. For example,
in phase with Ep the symbol K or G is often used, and as a further refine-

the Bode plot 151

merit the symbol G is used if P^ > Pi or x if Py, < P\- A = 20 log \ dB = 20 (log 1 - log 2) dB
We will, however, stick to the symbol A irrespective of
Since the logarithm of 1 to any base is zero, the above
whether P^ or Pi is the greater number. reduces to
Since electrical power is given by P = £2//? or I 2 R ,
we may express Eq. (4) as A = 20(0 - l o g 2 ) d B = -20 log 2 dB
= -20 (0.301) dB = -6.02dB
6) A,= 10108^(18= lOlogl^-^dB
c
•i •\ ll••\ Now work through problems PS 11-7, PS 11-8. and
PS 11-9.
^Olog^dB
E^R:
^E, R, 11-5 The Bode plot
= 10 log -2 dB + l O l o g — d B
V^l/ ^2
A most useful means of displaying the amplitude and
E* n phase response of a circuit or system is to plot the
= 20 log — dB + 10 log — dB magnitude (amplitude response) in decibels versus fre-
—' i "^ quency as one curve and the phase versus frequency as
Similarly a separate curve. The same frequency axis is used for
P I^R, both curves, and semilog graph paper is used so that a
7) A,= 1 0 1 o g - d B = l O l o g — — — d B wide frequency range can be displayed. Curves with this
'i 'i "i type of display are known as Bode plots, and they are
most useful in electronic circuit and control system
= 10 log
^ analysis.
To familiarize ourselves with Bode plots, we will first
= 10 log dB + l O l o g^2
—dB construct the Bode plots of some basic transfer function
/?, factors. For the moment we are not concerned with how
^
these functions are physically realized or even if they
= 20 log -2- dB + 10 log — dB can be. We just want to learn the mechanics of con-
/i R^
structing them. Once these basic forms are mastered,
Equations (6) and (7) are particularly useful if one we can move on to more complicated forms which are
wishes to determine power gain (or loss) by measuring essentially combinations of the basic forms.
£ and R, or / and R, which is usually easier to do than a Bode also showed that the phase and amplitude
direct measurement of power. However, many electronic response for most transfer functions are uniquely related
circuits are primarily designed from a standpoint of and that the slopes of the amplitude response are
voltage or current gain, and hence we drop the resistive indicative of the phase response. Since it is usually
terms in (5) and (6) so that the definition of voltage and easier to measure amplitude rather than phase, it be-
current gain in decibels are respectively, hooves us to see how we can predict the phase shift by
observing the amplitude response.
E
8) A, = 20 log ! dB = 20 log |/1,| dB (ll-5a)
£1
11-6 Basic forms
9) /4, = 20 log dB = 201og|-4i|dB (ll-5fr) The first transfer function form (factor) we will consider
is that of A = k where A is the general symbol we are
The ratio of the output over the input is correctly using to denote a transfer function and k is some arbi-
called gain when two quantities are of similar units. trary constant, a number which is independent of fre-
However, it is common practice to refer to the output- quency. Thus if A = 100, we have A in dB given by
input ratio as gain or system gain even when they are A = 20 log 100 dB = 40 dB. Similarly if A = 0.2, we
different dimensionally. For example, we might consider have A =10 log 0.2 dB s -14 dB. These functions
the gain of an optical system as being the output light Bode plot as straight lines, as shown in Fig. 11-3. Note
intensity of a lamp versus applied voltage. It should that there is no phase shift when the amplitude response
also be clear that gain may be a number more or less is flat {A = a constant).
than unity. If the gain is less than unity, it represents Next assume a transfer function of the form
attenuation or loss within the network or system.
Note that if the gain is less than unity, the gain ex- 1) A=f./~f (ll-6a)
pressed in dB will be negative, since the logarithm of a \ "c/
fraction is negative. For example, where a>c is some arbitrary constant which we will refer

152 ac circuit analysis

 Bode plot of a constant
A =*

FIGURE 11-3

to as the comer frequency,-and n is an integral exponent These values are plotted for the n = 1 curve in Fig.
1, 2, 3, 4 , . . . , n.3 To Bode plot this function, let us first 11-4, with the dashed line for the phase and solid line
consider n = 1. In this case the function is simply for the amplitude.
Note and memorize the following, as evidenced in
. . co
2) A =]— Fig. 11-4. For a transfer function of the form in Eq. (1)
CO,
with n = 1, the slope of the Bode plot is positive and a
Clearly the j indicates a constant phase shift of 90° constant equal to 20 dB/decade s 6 dB/octave where a
irrespective of the ratio CO/CD,;. Hence the phase response decade is a frequency ratio of 10, and an octave is a
plots as a straight line (dashed) at 90 ° as shown in Fig. frequency span of 2. The plot crosses 0 dB at co = co,..
11-4 for n = 1. The phase angle associated with this slope is 90°.
The only way we can plot the magnitude (in dB) of Now let n == 2 in Eq. (1) so that the specific function we
A is to assume various values of CD relative to our arbi- are dealing with is
trary co... Constructing a table of values we have
/ co' I CO ' CO I CO 1

cu/cu; 0.1 0.5 1 2 5 10

3) A= 7
CO:. ^, <("c. ^
dB = 20 log (cu/to,) -20 -6.02 0 6.02 13.98 20 Clearly the phase angle in this case is a constant equal to
6 c^f\o 180°. The magnitude and phase of (3) may also be
tabulated in order to sketch the Bode plot. Thus we
•'A function of (his form is sometimes called a derivative function. obtain

the Bode plot 153

 ^FE^ ' , ! -'Tft^ H r4 frif ^1 :;I; fit: i-i^t"-^
-H--4——•'*»1r'^ Am plitudel-
1
111
—"-'"'

db
\\ \ \'.
4+4—<
, "H
T^
•i i i 1 +-
•p '^ "ft "ri ^ -::-^....ll^ii 1 ; [ ! !;]| i ' 1,1 ,^! 1
-^-- .-..^. • -phase4*^^-"^ 315°

4^-L-t
-^ ^1' 4t*l -i-u **T' "tT"''^;'"'——•'"T"-'
M >w* •M> •jt" -1. «•-LM-^——.««^. ^ " - 3 ^ S27no
60 ^ ,. i • i . l
, ,
,,
\: ' i '
1 1 1 1
•-"H
;i
^-^
ff
BUdH/aec i ' , ; l ^ ; j :i rTT"^'
•' ' t ' •y =
18dR/oct^———-r-
;
' 1
]

>-. ^. ^ 2^5,
^^
1

50 ^~. >
!•'' • i 1! i "^ n.iiii,;iii...>,
:

..1 1 1 i 'i i ]! i li i if ii.

^ T-' y
."» 4 ——— 77 - 2^::
40
i ;
-»<
—7 .,^"-2-^
•*UUO/UBt' ,: : ' 1 ; • ^-•''31''"'': 180°
^'"^ S
—
' \. ' • i '
1 1 ' '-y-
4
^
lope ,2^/^:S;a:S::
i
! ' i \\ . ^ , \ , , ,.,,v.,i ,1
30 ', i i
i
.

r <-3> -^-t-4-^——"-^u-——^ ^in^r^w

lil
———^
^i i^ ^." ^^
20
i
—v<
^—a .-8^; ,-^d" = it——~t"——^—^
POrlB/decT"*" '^""TT" *
1 900

'^ -^
4t ———1——————t————^ q,

^ i—3- Pe - ,i
-,—j-y "~r^ ^ 1
W 6dB/oct ——
10 —^ ^ M— fff r w* -j- [ | ; i ; +T- l l ? ' " ! 1 -t-^i- ^^^E^s-
i .^ ' ^' •; ^— Vs.'- "ss
ff* 1
i'' . li 1 ; 1 1 1 . . ,

1 i• .[
^r-^rT
! 1
, i
l i . 1
; j 1 ! ! ; ,
j j ' 1
^^-—— ' l''
0
. 1' ^ff:*41"""^
1
0

' l i t
, , ; jff
•^^ • •
i ! :i , i '
i i ; .. .. ... i i • i 1 j i '. 11 . —--
^ u 450
-10 1—~~ • ' ' • t ••'•ta ^ T»»
ff1-
'* T*' l 4
' j ' ! 1 ' " l ^ " i ! l t !! ; 4** - SlEi::^^^"
' iiii' i ^ ; '' ' 'i' ^\
..^yy
-^*< -y ~3^
^-
i ' ' 1
i
,
i '
i, i
. ;
! •
; i | ' • i ii
' 1 ; i1

-20 ^^T ^-^

r' —T"^+h"'—^-+^
t+ + ^^"-^ffi-900

-30
—'~ i^
)*"-;— ^
—4"»c
i^I;:
-^
yr~
^
. i,
i •
^a
S:
3-
f
\
"\ »t--*T^^~ ^Hi
"<• 1 " I lIi It ' ^TTT
. 1 • •^ ttt
n<20dB/dec)+.. -4w-|—
~: t+ 4 t r
; 1 "^ ' T7 tt " ^" 't
4 +J
350
setea:®!"'
1' , J1
—,^T'''' 2 •

.z - .,, i''
^- slope n(6d8/oct) SSffla-
i i i 11 'i' 1 iii * T 'i
-40 ea^-
^h: ill |il| ^ [•-0-n90°-- i l l lU 1 i ll ' | Ii t*" •
i .
^4 i, ' ttn^
i '• i -'-t——
j
"i~H'
i ^ r rf+tr
i i 1[ 3""
'^ ' '
-+
'tti t[
T^ " ^ 11 i ' i i
' -:i::::.:::::: :!::
.' 1

-50 -,—-y 'f— 11 • * ' 11,'" ' ' 11 ' ' i •.' ' t ± '
—yf- 1 ' ! ! I'
i
,. !

'! , T '

^ -1 —-.-
^ • 1

t l l
-60 r^,——t "Ttt -l- -+i-^t••••H••• •• "r- :StsaS
0.1 10 100
ai/<*'c

FIGURE 11-4

ii)'(DC 0.1 0.5 1 -)

5 10 and the phase is given by

ilB = 40 log (fj o^.) -40 -12.04 0 12.04 27.96 40 <A = n90° (ll-6c)
0 t ono
Another commonly encountered form of transfer
function term is4
Note that because this time the transfer function is
the square of the previous case (n = 1), the slope is 1) A = (ll-7a)
exactly twice as much and the phase angle is doubled. / u)}
Therefore for n = 2 and a transfer function of the form [j^
given in Eq. (1), you should think in terms of a positive
Inspection of(l) reveals that an increase ofco causes the
slope of 40 dB/dec or about 12 dB/oct and associate a
denominator to increase, since u>c is some fixed value,
180° phase shift with this slope. These results are also
and hence it causes A to decrease. Therefore, the ampli-
shown in Fig. 11-4.
tude response should have a negative slope. We also see
A Bode plot for n = 3 is also shown in Fig. 11-4. As
that for any value of n, the amplitude must be 0 dB at
might be expected, this curve also goes through 0 dB
cu = co,.. Since the j represents 90° of phase in the
at o> = (»c with a positive slope equal to 60 dB/dec or
denominator of A, the net phase angle in the denominator
approximately 18 dB/oct and the associated phase
must be n 90°. Consequently, the net phase of A must
angle is 270°. be -n 90°.
As you can see from Fig. 11-4, the slope in general is Although we will not take the time to derive the data
given by
slope = n (20 dB/dec) S n (6 dB/OCt) (ll-6fc) •'This transfer function is sometimes referred to as the integral function.

154 ac circuit analysis

 for plotting A. the results are given in Fig. 11-5. The
magnitude response from (1) is as follows: 1) 1 +J
(.0,

A = Let us first consider the case where n = 1 and calculate

(WiWcf the magnitude and phase. This yields the data in the
following table:
3) A^ == 20 log 1 - n log ^
C) (';,. 0.1 0.5 1 ^
5 10
Since the log 1 to any base is zero. we have PC
0 5.71' 26.6' 45 63.4 78.71 84.3
cu
10°
4) .<IB = -n 20 log A 1.00 1.12 1.414 2.24 5.1 10.0 1-

Inspection of Fig. 11-5 indicates that except for being

0>,
''•JB 0.043 0.969 3.01 6.99 14.1 20.0 ^
negative, the slopes and phase for various values of n
A Bode plot of the above data is in the solid line
are identical with Fig. 11-4. Therefore, for a function of
curves of Fig. 11-6. Figure 11-6 also shows some dashed
the form in (1).
lines which represent straight-line approximations to
slope = - n (20 dB/dec) or - n (6 dB/oct) (1 l-7fc) the actual curves. The beauty of the Bode plot (among
other things) lies in the ease with which the straight-line
(f) = -n90° (ll-7c)
approximations to the actual characteristics may be
Now work through problems PS 11-10 and PS 11-11. constructed. For example, consider the upper magnitude
The next form of a transfer function factor we might response curve. Note that it may be approximated by
familiarize ourselves with is the two dashed straight lines which intersect at ai,,. We

..11;: , M , | . . >
———-——•—wW-t Amplitude^ m ft
i ^
—i——
^" . ..i i . i , , : , :;^;;' : Hi ! ; , ;'
1 i——-*. n , i -»^.,^.^ Pha?- ill,,i,l',,,l,l',ll,."
1 1
S^lnrl^ai I I !'! '
1

' i 111 ^!
I n, !! , . i , i.
60 • • 1

^——
^—— 1 ; 1 ——^~t4^+^4
^•fi \-
-s;,— i
/' "" ^:ts:"^:n-:m i n" 1 1 1
"c ^titltttttti^
^^
S ' i ^ !i i

-^>—
^. .... , 1
^

'i.
f-n(20dB/dec)1 T:- ^ 1 4
't^ <t'°
I \
c:— — 1 i——- •^ • slope- \-n{ 6dB/oct) J^.I.^T—^—|M|,,||,,|
\\\ 1 , , M ,'

40
s^—. ^c"'—
-'s^.' — ...^... f~Sit— 1 1
^ , >-.^ . . . :
*s ( = —n qn°' "i i !" | j
\
! ' ' :
'[

-
' • 1
, .

--'L. : • ' ' 1 t 1 \\ \ ' 44 E l

7*> .t-t- . , . .
S^—i•—Si—— \ • 1
i
•' i :''
' ;'
' ! T '• !
1i 1
,, !i
[''
' ! ' ' i ' , ' ' ^ !! i! i
;,,-^,,....:. \^.. "'-Hw——^-^ i
'' ' ' ' '',
1 1 ' • ''' i
'!'
i, \i} I ' ' I 1 > i

-^- «^ -s
20 t -^trr-r 90
- -^— SIT-*
v
' T
.
; , ' i. i ' i ill Ai^
1
-S k—m^Uu
— " "• ' •^
^is- '^FSffi? ^-^——^^^•l+^^^^rrt^s0
ttt
ss •"'s! iri— : ^—
' 1
^:-!-.
^
i i ! 1 ! \ 1 ] [ t '
it-'^S P
•i j ' 'E'^ tii'S ^
• i>
1 , 1 1 1 1 ' 1 1

"x'tn °
, 1, t-*.i
''
, If , ' . , , , 1 , ',., , i

-14-
-*-.4+
tT
^
"4 ill
-t+'T^-"

"^i^ ^^ Sd-
sf^
^^^i-LU
»^. . 1
^.1

•«c +u*
7"!
?1
1 ' """^t*tf "w'
--i-f-^-j----^'^^ ••4"' •4'!
• • +""""
; -45°
•^-satttr;
.i<tf^
T
F ---^
^
t T*
. '' ^ •^ +..-S
\\ TT" +^S ^-
m
i ^^4 » * « i ! • • "4 :5:: t2
1 1 •
n-i +H*
'TT •»*•
^tH
**f*t-
t^^- B
l-^--^.4^•-L4t•,,f4ti••u+*t+rtt^= if -90"
|.|! 1 ; . 1 1
1 i . ,^».,. ,'
< (
"It-

" - 1' ! i i il 1 i
M1

ll -*w- 1
-^-— -n-r-t ^a "^
>+». 4 -4

-135
lE v °
"y- &"^ ; 3S^ !" {d: -20dB/dec - - - - -
+t+J• )4->4.t^U „'t ^ ! ' ' ' '^'e1""' Sl'
^
P?
^-i- Slopes ... •-•4
S
^^
^ •-**-*

•^ ^ [-- -6dB/oct •--++

''' i ' ; r ,! ^ , N( i , .i
I):
!.
11 !; 1 i i( ' i i' agu-i , U, . ' i11 ^ '1 -180
-40
' ! ''' '^i ! l ' ! ^ *l
i i \^
i>
^^

t~ f j ' ^^>. n = 2-T-S"£iI•I+1 \'±t®t

! t S' l ''
• ' ' 1 ' It ''i
1 ^ 1

\ ; ii . '
'

' , l; i ; ; ' ! " [t

----- I
r :
i ' 1" -40dB/dec
-^i-
+4* •w —^ ^t- %
Slop ...... . j -225"
———^ -"-"- .1. v^ ttE^l 1 ; i! -60dB/dec
-^ -12dB/oct [ )•---•

-60
' 1 1 L—Lj-
-11,—...,,:
•I-** ii^ t+tf ^s ^
^^-^--s- ope=
-18dB/oct
-•-yi
^S n= 3 :::i±ttiiiiiiffl : ±::::-:::t
.^. , t e . . ^ . . — . . ' — ^ . w ^ u . - ^ ^ . — . — - . . - .^-^—•k.-J.u - .^10 •
3
270°
.1 1 10
tJ/(^C

FIGURE 11-5

the Bode plot 155

 25
01
A /'
= 1 +
'^c
t
20
f
t
dB
Amplitude 15
.
/
10 ' dB
^
/
3d8 ^^ /
,J
5 < -^''
i ^\ l^''
1 d ^^ <r /
0 a^
•l^
"c ax; 2fc•^c 6° 1QtJc
10 >
t
90
i•5 9( ° - ....

^^
I •
75
6° ^
J?

,- 60
.r-^
Phase
lead ^
——f K
^
J
30 7<i ^
'> ^f
.
^
^,

^6 - ^
15
•*
^
0 i
"c ^[f CJc 2 LJC 1 0^
10

FIGURE 11-6

call u)c the break, corner, or cutoff frequency because of For (D ~s> u>c the straight-line approximation and actual
the abrupt change of slope exhibited by these lines at amplitude curve merge for all practical purposes and
We- For u) < cue the approximated magnitude response the phase error is again about 6°. As you can see, the
of this transfer function is flat. At a decade below co,. or amplitude and—to a somewhat lesser extent—the phase
u> = 0.1 <y<- the error in this approximation is negligible response can be readily approximated by merely draw-
since the above table of values indicates an actual value ing straight lines with well-defined slopes. If greater
of only 0.043 dB. In the lower figure the dashed-line accuracy is desired, we can use the easily remembered
approximation to the phase curve is only about 6° below amplitude correction factors of the following table:
the true value.
At an octave below &>,., or u> = 0.5 u>^ the error (U/CU, 0.1 0.5 1 2 10
between the actual and approximate amplitude curves dB 0 1 3 1 0
is only about 1 dB and the phase error is only about 6°, 6 6
<f>° 0 6 6
which is essentially the largest phase error at any point.
At o»c we have the largest amplitude error, which is
only 3 dB, and the phase error is zero since we draw the The above correction factors and the general shape of
phase approximation curve (dashed line) so that it the phase and amplitude curves should be committed
intersects 0° at (Uc/10 and 90° at 10 a»c. A straight line to memory.
connecting these two points will intersect the point Now let us consider other values of n in the function
((o/a>, = 1, <f> = 45°).
At 0)^ the approximation to the amplitude response A- (ll-8fl)
changes slope abruptly from 0° dB/dec to 20 dB/dec or
1)
v 1 + 7wj
"
6 dB/oct. At an octave above a>c, or 2 cu,., the amplitude which has an amplitude given by
error is such that the true response is only about 1 dB
high and the phase error is about 6° low. 2) A^ = n 20 log ,/! + (ftJ/Mc)2 (ll-W

156 ac circuit analysis

and a phase angle You will notice that this expression is the reciprocal of
CtJ Eq. (ll-8a), which we just studied. The amplitude
3) 4> = n tan (ll-8c) response is given by
&>,
From (2) we see that the amplitude in dB and phase in 2) -4dB = -" 20 log ^' 1 + ((y,oj,) 2 (ll-9fc)
degrees will be n times the values we just obtained for
The phase response is given by
plotting Fig. 11-6. Therefore, the correction factors in
the previous table must also be multiplied by n.
A Bode plot of straight-line approximations to ( 1 ) 0 = —n tan \-9c)
0),
for different values of n is shown in Fig. 11-7. Since the
amplitude slope for (y < ^c was zero, it will still be zero A Bode plot for n = 1 is shown in Fig. 11-8 and, as you
for any value of n. However, for u> > o)c tne slope must can see, the only difference in the amplitude response in
be multiplied by the appropriate n. Thus, for n = 3. this case and that of Fig. 11-6 is that for CD > o.»c ^s gain
the slope for a> > c'j,. is 3 (20 dB/dec) = 60 dB/dec or decreases. The slopes for co > (L>^ in Eq. (ll-9a) for any
18 dB/oct. The corresponding phase asymptotically value of n are simply the negative of the slopes for Eq.
approaches 3 {90') = 270". For n = 3 at o> = cu,.- ^e (ll-8a). The correction factors are also identical in
phase is n (45-) = 135' and the amplitude error is n magnitude except that now they lie below the straight-
(3 dB) = 9 d B . line approximations. The phase characteristic and its
Now work problem PS 11-12. correction factors for any value of n is also identical
Another commonly encountered form of transfer with the single exception that we now have phase lag
function is rather than lead. Thus for co t> (D,. and n = 1, the phase
angle approaches —90°.
.-1 — / \ ,, Now
1 N U W work
W U I K through problem
UllUUgll p l U U H - l l l PS 11-13.
I J 11-1J.

n + j -"-) (ll-9a) Transfer functions occasionally contain factors fa of

\ "^ the form

d8
Annplitude in - 3
60 Ph ase f
. CJ
55 A d +
^ /

e n t an -' "-y.
|
\
"c 1

50 AM n 10 log \/\ + (ul ^> 2

45

40
/
/
r
i
0" )
(lead)
f
315 r 35 ^

/ ^ f r7 =

270 30
i <-. y
225

180
25

20 /
,--// 'f
t /•
:'

»
4

•>
'

^
t
7 n= 1

./// //
135 15 ^ y-
- »•
,' ' /

^A
90 10 ^ n=1
^ ." . <• -
45

0
5

0 —————————————————
0.1
^ f

--
0.5 1 2 10
tJ/OJc

FIGURE 11-7

the Bode plot 157

 1 1 1
0. 1 ^ C0.5 uC C^ 2 "C 1C0 t^c

0
/
N /
^
^ >.
/ ^
s\
- -5 —^
00
3 1 dE
OJ
3 B\?
"§ -10
C
A =
1
1d1
s\
1 -15 1 +/
^C
^q ^—
-20 s

0
\^
-15
^1
0
6
^S > s
1'
2 -45
-30

6
^
-S
S
^ ^

S'

£
-75
ks- ^ \
6°
^ •k,

i ^
6
"

0.1 0.5 1 2 10
fc>/CJc

FIGURE 11-8

/ a>^ denominator form of Eq. (2) is more common, we will

A = + 7-2.5 -" + I discuss it first.
[j^ &>.
Inspection of (2) indicates that for &) <^ co,. or (co/cu,:)
^ cy 0} <^ I, the amplitude of A approaches 0 dB and the phase
+ j2S (11-lOa)
^<u.. &), approaches 0°. For co > co,., the real term being squared
will overwhelm the imaginary term, so that
1
3) A s
A =
/ cy' w
+ j2S - + 1
(U
\.u>,
f? I tO

7 {We. ^"c.
^ CO.
Thus the phase angle approaches 180° and the amplitude
approaches zero. In general, for co > co,;, the amplitude is
' CO
1 - +j28to- (11-lOfc) given by
I.C"r. co,
CO
These forms are called quadratic factors and they 4) A.IB == 20 log • 40 log —
generally arise due to interaction between two energy W CO,

storage elements in a circuit or system.5 If the circuit w

contains inductance and capacitance, so that some form If we let co/co,. = 1 in Eq. (4), we have A^y = 0. and
of resonance phenomenon occurs, we can expect to for co/coc = 2 we obtain A^y S—12. Thus the slope in
encounter quadratic factors in their analysis. Since the the region of co > co,; tends to approach -12 dB/oct,
which is essentially the same as —40 dB/dec. Exactly
'The energy storage elements can be two capacitors, two inductors, or the same conclusions can be drawn for (1) except that
an inductor and capacitor. there the amplitude response rises and the phase leads

ac circuit analysis
 Frequency, radians per sec Frequency, radians per sec

FIGURE 1 1-9

as co increases, whereas in (2) the amplitude decreases sketching denominator quadratic factors, we have Fig.
and the phase lags. In both cases the phase asymptotically 11-9. Note that as 6 (sometimes called damping factor)
approaches 180°, though from different directions. increases, the peak or resonant point shifts below a);,.
Now what happens at CD = o><. in (2)? Clearly the func- The same curves may be used for numerator quad-
tion reduces to ratic factors if you remember that for o» > o»c the
amplitude curve slopes upwards at 40 dB/dec; and at
5) A = -1- = ^-/-90°
i
o»c, instead of a resonant rise, we may see a resonant dip
j20 20 ——— as shown in Fig. 11-lOa. The phase response is also
similar except that it approaches 180° of phase lead for
Thus the phase at w = co, is -90° in (2) or +90° in (I). u> > coc. Fig. 11-lOfc illustrates the phase response in
The corresponding amplitude in (2) is this case.
6) A^ = -20 log 26 (11-lOc) Now work through problem PS 11-14.
Inspection of (6) indicates that i f < $ = ^ , log 25 = log 1 =
0, so that A^ =0. If S < ^, say 6 = j for example, 180°
log 26 will be the log of a number less than 1 which is a
negative quantity, and in this case
Slope
Ads = -201og2(^) = -20(- 0.3979) s +7.96dB 40dB/dec 90°

Thus we see that the amplitude response actually rises

above the 0 dB level for 6 < 7 in (2). For 6 > ?, say
5 = 1 , the log is positive and hence the number ofdB is
now negative. Both the phase and magnitude of(l) and
(2) strongly depend upon the value of i5 in the vicinity of
&» = u>c, and it is probably best to plot a few points for (a> W
an accurate representation. However, as an aid to FIGURE 11-10

the Bode plot 159

 11-7 Constructing the Bode plot (JojC.R, + 1)(1//^)
In general, a transfer function will contain various factors [juC,(R^ + R,) + l](l/ycyC,)
in the numerator and denominator which consist of the __1 + jwC^R^
basic forms we just learned to plot. We can use this 1 +juC,{R^ + R,)
background to construct transfer function Bode plots of
Let
practical networks. There are various ways of doing
this, all of which are best illustrated by example. T, = C^i + R^) = 1/cdi, T, = C^R^ = l / w ^ .
Assume we wish to develop and Bode plot the voltage w^ = lOc^i
gain transfer function of the so-called lag network of
Fig. 11-1 la. First let us develop the transfer function E^ _ 1 + j w / W j A/
1)
and inspect it for time constants {RC products or L / R E,,, 1 +JUJ/0)i ^

ratios) whose reciprocals denote corner frequencies. If Equation (1) may be resolved into two factors
several time constants arise in the analysis, we can
correspondingly expect to have several corner fre- -E" 1 +7
U)

quencies to consider. E.
Starting with the voltage divider relationship, we may 1 +7 QJ,
derive
The first factor was studied in conjunction with Eq.
(11-8) and the second with Eq. (11-9). A Bode plot of the
R. J^c, R^ + l / j c o C i amplitude response of the first factor (numerator) is
/?! 4- Ri - ] X c , PI + P2 + I/.AOC;, shown as curve .V in Fig. l l - l l f r . The amplitude response

160 ac circuit analysis

 of the second factor is shown as curve D. Since the PROBLEMS WITH SOLUTIONS
amplitude scale is logarithmic (dB). multiplication of the
PS 11-1 What is the corner frequency of the low-pass
amplitude of these factors is equivalent to addition.
filter in Fig. PS l l - l ? What is the voltage gain transfer
Hence, the composite amplitude response is simply the
function at the corner frequency and at 1,10, 1,2, twice.
point-by-point addition of curves N and D as shown in
and ten times the corner frequency?
curve .4^g = .V + D. A pair of dividers is most useful in
this regard. Correction factors can be inserted so that a 10 kn
smooth curve can be drawn, but generally the straight- -VSAr
line approximations are adequate.
The phase of the first and second factors has also been
approximated with straight-line segments as 4>\ and 4>o 100V 0.01 AF .

respectively. The phase angles may also be algebraically

added with a pair of dividers or by counting squares in a
manner similar to the amplitude response.
With a little practice you will in most cases be able to FIGURE PS l l - l
construct the amplitude response of A without having
to construct the individual factors. For example, with SOLUTION The time constant is
c^ > c'j, in (1). we know the first break (corner) occurs at 1) r,, = RC = 10 kQ(0.01 ^F) = 1 x 10--4 sec
MI as we go from the low to the high-frequency range.
Since 01, is in the denominator, ,4^3 will break downward The corresponding radian corner frequency is
at cy, with a slope of 20 dB/dec and continue doing so
until c'j;. whereupon the numerator starts climbing at 2) o),, = — = 104 rad/sec
^h
20 dB/dec. Thus the positive numerator slope cancels the
negative denominator slope so that the slope of A^y The corresponding corner frequency is
levels off at 0 dB/dec.
^h
Although it is somewhat trickier, the phase response 3) /„ = -^ = 1592 Hz
may be similarly reasoned. To a first approximation the
denominator phase breaks at O.lcji and decreases at The general expression for the transfer function by
about 45' dec. Thus at w, the denominator angle is Eq.(11-1) is
45'. but note the numerator phase is coming into play,
since fc>, = O.lcu, and the numerator phase is increasing . Eo 1

at 45" dec. Thus at o»i the numerator and denominator E,, ,/

1 +7
phase slopes cancel, so that the phase remains constant A
at the -45° level it acquired at a»i. This holds true until Plugging in the given values of///,, = 0.1, 0.5, 1, 2, and
a»2. at which point the denominator phase has gone 10 into (4), we have
through its maximum possible shift of —90°, so that the
slope of the denominator phase angle is zero. The numer- ->
/ />, 0.1 0.5 1 10
ator phase slope, however, continues to climb at 45°/dec
1-4>-1 0.995 0.894 0.707 0.447 0.0995
for a decade above a>t. This pulls the net phase slope
(cf)) up at 45°/dec above co^ until 10a^, at which point [Q -5.71° -26.6° -45° -63.4° -84.3°
the numerator has gone through its maximum possible
phase shift of ±90°. Thus for aj > w^ both numerator PS 11-2 Prove that the corner frequency of a low-pass
and denominator phase slopes are zero and the net filter such as shown in Fig. 11-1 is that frequency at
phase slope levels off at 0° due to the -90° in the which Xc = R.
denominator canceling the +90° due to the numerator. SOLUTION Applying the voltage divider relationship
From a physical viewpoint we can reason in Fig. to Fig. 11-1, we have
11-1 la that at low frequencies C^ looks like an open
£2 -JXc
circuit and hence Eg = E,n or A^ = 0. As o» increases. 1) A,=
C; comes into play so that it starts to load Ry, which E, R -jXc
decreases Eg. This accounts for the break at (a,. As a> Dividing numerator and denominator by — j X c and
increases further, C^ eventually looks like a short and rearranging terms, we obtain
Eg = E,n/?2/(^i + ^2)- This accounts for co;, where C;
I 1
fades out of the picture. A,. =
Now work through problems PS 11-15, PS 11-16, R R
1
--^ 1+Jv.
PS 11-17. ~]AC ^C

the Bode plot 161

If we let/? = .Vein Eq. (2). .f_

1 fi
= 0.707/-45° 4) A,. =
3) A,. =
•+!
which is precisely what we showed in the previous
Setting/,7, equal to 0.1. 0.5, 1, 2, and 10 yields the
problem when we set/=yi,.
following values
PS 11-3 What would be the corner frequency of the
circuit in Fig. PS ll-3 9 / /, 0.1 0.5 1 2 10

A-, 0.9950 0.447 0.707 0.894 0.995

s
L° 84.3 63.4' 45- 26.6 s 5.71-

PS 11-5 Determine the voltage transfer function of

the coupling network shown in Fig. PS 11-5. Resistor
R^ is analogous to the output resistance of a voltage
source driving a high-pass filter.
FIGURE PS 11-3

SOLUTION By inspection we see that R^ and R^ may ^

be thevenized with the voltage source to simplify the
equivalent circuit that the capacitor looks back into.
Thus, by analogy with Fig. 11-1, intuition, or both, fl;.
T), = R,|,C = (/?i | R^}C. The radian corner frequency is
then u>i, = 1/T,,.
PS 11-4 What is the corner frequency of the high-pass
filter in Fig. PS 11-4? What is the voltage gain transfer FIGURE PS 11-5
function at the corner frequency and at I/10, 1/2, twice,
and ten times the corner frequency?
SOLUTION Applying the voltage divider relationship
25>iF we have
-1C 1) A,. = la R.
E, RI + R^ - j X c

ioov nj)E,n 1^^ E' Dividing numerator and denominator by Ri + R; yields

RI
RI+RI___
2) ^-v = n , D V
/?> + Pi Xr
R^ + R^ - JJ R^ + R
FIGURE PS 11-4

SOLUTION The time constant from Eq. (11-2) is R.

1) T, = RC = 1 kQ(25 ^F) = 25 x 10-3 sec R, + R,
\-j
The corresponding radian comer frequency is (oC{Ri + Ri)
If we let the corner frequency be
2) (o, = — = 4 0 rad/sec
T
! 1
3) co, =
And the corresponding corner frequency is C(Ri + Rz)
we can rewrite (2) as
3) y, = -"L = 6.37 Hz
271 ^
4) A,, =
The general expression for the voltage transfer func- R, + R^ , ",
1-7 CO
tion by Eq. (ll-2b) is

162 ac circuit analysis

Multiplying numerator and denominate) r o f ( 4 ) by jaj/cOt Therefore
and introducing/yields
7) A, = 2O l o g .^ dB = 20 log 11 5.6 dB
./ -in
jl — •> 0(2.063 D d B = 41.26 dl3
51 \ - i ^
+R x
"' . -",^ PS 11-8 Exp ress th e voltage gain o the netwo rk in
Fig. PS 11-8 in dB.
Note that for /'> /, we just see the R I . R, voltage i 8 kP.
divider, since X^ ^ 0. ^A/——— •———">
PS 11-6 Prove that the corner fre quency is the
frequency at which \c = Ri + P^ ln F ig. PS 11-5.
SOLUTION If we divide numerator anid deno minator 12 k^
of Eq. ( 1 ) by R[ + R^ in the solution of the srevious
problem, we obtain
c3—————
^ 1 FIG,L'RE PS 1 1-8
R, + / ? , ' . . Xc
1 SOLUTION
R i 4^2
-» ktl
When .Y^ = R i + ^2 the above equatio n reduc es to .4, = 20 log ° dB
.10 = 20 log dB
''•in 12 k0 + 18 kQ (
R2 1 /
^ -• 070"'( PI )/45- = 20 log (0.4) dlD ^0.3979)
1
— »LH c1B
0 .L.
K I + /<2 0
1—71 1
\r< i + PI
;1 \ P

. ,, = -7.9588 dB
This is exactly the same result that would be obt ained by
setting/=/, in (5) of the previous probi em. Th us when PS 11-9 The voltage gain in c1B o f t h e four-ter minal
/ = / , . X c = ^ + RZ. network in Fig PS 11 -9 is A, = -5.67 cIB. What is the
voltage gain rat io E-i/lr-i ••)•
PS 11-7 Express the voltage gain o the circuit in

r"
Fig. PS 11-7 in dB a t / = / , , the resona nt freq uency.
?
270 n 39 mH
'
E i dBA, =
-5.67
?
E2

1t j
E,n(^\J 40pF^.S; E0
dSA, = -5.6 7
FieSURE PS 1 1-9

» SOLUTION
FIGURE PS 11-7
1) /4, = 20 lo
^ t
• -'l
iE i

SOLUTION
2) -5.67d B = 20 lo
^
'-'i
iE }

1) /, - ——f== - 127.4 kH z -5.67\

In^LC 3) ^ = l c si-^' ( -0.2835)
log- 1 \
£•1 [ 20 )
2) CD, = 800.6 x 103 rad/sec
Therefore we seek the rnumber whlich re]3r esents the ratio
3) X L = co,L = 31.22kn E ^ / E i and wh(ose logarithm is -0.28 3J5 (the not ation
/. ^ 31.22 kn ,, log"' implies the ant ilog). You will rec:all that in log
4) Q H 5.6 tables positive mantis sa values only a r<e usually tabu-
R 0.27 kQ
. lated, and hence to ob5tain a pos itive nn antissa w e add
5) E,, = eE,^ and subtract 1, so that
ir
fi^
"' A!
A
- p ° n
-V 4) £2
- = log- ' ( 1 - 0.2835 - 1) = lo g -'(0.7165 - 1)
^in ^i

the Bode plot 163

Looking in a table of logarithms (or on a slide rule) for The phase angle is always zero, since A equals a constant.
the number corresponding to 0.7165, we have 5.206 and Remember that a constant amplitude response implies
with a characteristic of - 1.* zero phase shift.
E
! = 0.5206 2) A = 20 log 0.004 ^ -48 dB
£, The phase angle is again zero at all <y.
PS 11-10 Construct a Bode plot for the amplitude 3) The corner frequency is u>c = 50 rad/sec. Since n = 1,
response for each of the following transfer functions: this Bode plot will have a positive slope of 20 dB/dec
(or 6 dB/oct) and pass through 0 dB at cu = a»,.. The
. . co phase shift is a constant 90°.
A = 3200 3)
^50 4) The corner frequency is &>,; = 300 rad/sec. Hence the
Bode plot will pass through 0 dB at a>^ = 300. Since the
A = 0.004 4) A= cu term is in the denominator and n = 2, the slope will be
(U
negative and equal to —40 dB/dec (or — 12 dB/oct). For
300; n = 2, (j) = -n 90° = -2(90°) = - 180°.
Determine the phase response in each case by inspection The resultant curves are plotted in Fig. PS 11-10.
of the above transfer functions.
SOLUTION PS 11-11 What is the transfer function corresponding
to the amplitude response curves labeled a, b, c. d in
I) A = 20 log 3200 = 70.1 dB Fig. PS 11-II?
•Negative mantissas for negative logarithms may be conveniently SOLUTION Curve a crosses 0 dB at a corner fre-
evaluated on a slide rule with the L and Cl scales. quency ^ = 40 Hz and a positive slope s 6 dB/oct.

FIGURE PS 11-10

164 | ac circuit analysis

 -———————
U. -— i——
+4
--

— — — — — — • — --" ' ' —

60 ^
•^ ^ ii. , .——
,
-——-———————.——————-

...— ——
40
,
—— —[
"t ^~ ---- ®
cV-^ ———\
[
——-^————
Mf-— -/
JT f - ^==
————^,——
in. ,.(
20 -^ . , .. \
.... - - \-/-
<^
——————\t
: ; : ; ::"'L,—
^ -^r — i '
V
—\
-——- I--..;^ ^
d8 0 ^ ^T* r^
———Tf
^-^
^ ^-^
h
,<*"•.»•'•' ——.yc— Sr
——
©
^"-4
^- *-\ ' ' !
• i
—j^"'
i—^— \.- •::
^r—— -\-
-20
^r — — ^ —-^-
^ -T-.\
^
i
t-j ) : ' ! ';
T-'""
1 ..
i '• 1 '•' 1 1
j • : 1i
i ;
• 'i^
'. : 1

i
;
! ;w
-40 —h- i] 1 1 •!
—————(^ : i ; ! * ;!'

, - • ,i
' '—/

i . . ; , .1 , „
. i ' '
L>•'; i11 1n
•

,•
'
i i
1 '
!
• 1 . !
l' i
i L1': i
; ,•
i' ^ , * ;
| ; 1 '
[
• ! ! ' 'i ' 1 ! '' "!
1 i ;
I
'' | • i , . i 1 1. i' 1 '] ^ j
-60
0.01 0.1 1.0 10
^(kHz)

FIGURE PS 11-11

Since the 6 dB/oct slope implies 90° of phase shift, PS 11-12 Construct a Bode plot for the function
2
^•io ^O^'^) -
First show the straight-line approximations for the
Curve b is constant at —43 dB. Therefore amplitude and phase response and then insert the correc-
-43dB = 201ogAb tion factors so that a smooth curve may be constructed.
SOLUTION The amplitude response will be flat from
43 dB' / = 0 Hz until/ = /, = 500 Hz, For/ > f c , the slope rises
A,, = log == 0.00708
20 at 40 dB/dec or 12 dB/oct. Since n = 2, the amplitude
correction factors are 2, 6, and 2 dB at 0.5/,/, and 2 f c
Curve c has fc = 200 Hz and a positive slope of 18 respectively. The phase corrections are approximately
dB/oct. Hence the phase shift is 3(90°) as n = 3. Therefore + 12° and - 12° at O.I/ and 10/ respectively and -12°
and 4-12° at 0.5 / and 2/ respectively. The phase curve
. /V
Ac = [ j will go through 90° at / and approach 0° and 180°
200.
asymptotically for / <S / and / S> / respectively. The
Curve d has f , = 3000 Hz and a negative slope of 24 resultant plot is shown in Fig. PS 11-12.
dB/oct. Therefore PS 11-13 Construct a Bode plot of the transfer
_1____
A,= function A =
/ ^3•
3000/
1 + 7 500 Hzy

the Bode plot 165

 0.5 /, ^

FIGURE PS 11-12

SOLUTION The approximated amplitude response

will be flat for/> /<. until/c. At/^ the slope changes from jwCR + j^LC - 1)
zero to — 60 dB/dec. Correction factors for sketching the
Now the trick is to algebraically wrestle the final form of
actual response are 3, 9, and 3 dB at 0.5 f ^ f c , and 2 fc
respectively and + 18° and - 18° at 0.5 /^ and 2/c respec- (1) into the quadratic form of Eq. (11-lOfc), repeated
here as (2).
tively. The phase curve will go through —135° at f c
and approach 0° and - 270° asymptotically for / < fc 2,
and f ^ > f c respectively. The resultant plot is shown in I CD a)
1 - +j2S
Fig. PS 11-13. ^ co-

PS 11-14 Determine the voltage gain transfer func- if we let

tion and construct a Bode plot for the circuit of Fig.
3) co,2 = —— = 83.333 x 106 (rad/sec)2 or co = 9129
PSIl-Ha. .n-/
SOLUTION First we must develop the transfer func- rad/sec and therefore obtain
tion. Thus
4) c-
co^L
1) A ^^^-^^——^
" £,„ R + j ( X ^ - Xc) and then substitute (3) and (4) into (1), we have
l//coC 1
—
5) A,,=
O)R CO
K +J\WL -
wC) co^L co-

166 ac circuit analysis

 FIGURE PS 11-13

for co < u>c a straight line along 0 dB until co,;, at which

point it breaks downward with a slope of -40 dB/dec
\0>,7 CO.L CD.
for o» > u>c- The response in the vicinity of ca,. is deter-
mined by the damping factor 6. Turning to Fig. 11-9 we
Since
see that for 6 = 0.043 we are off the plot. Hence we turn
6) e-^-n.65 toEq.(ll-lOc).
10) dB peak = -20 log 26 = 21.3 dB
we may substitute (6) into (5) to obtain
which yields the height of the resonant peak at co,;.
1 For a more accurate plot, we might as well evaluate the
7) A,, =
, ,(o\2 1 u transfer function. Thus we have
'-W ^Q^
Equation (7) is in the same form as (2), with 11) A =
1-(^9 ^2(0.0429) "
1 9129
8) 25 =
Q
or (U-
+7(9 399 x l(r6)(o
83.33x10- -
9) S = —— = 0.0429 s 0.043
26 and plugging in various values of <u yields the following
In constructing the Bode plot, we know we will have table of data, which is plotted in Fig. PS 11-14&:

the Bode plot 167

 0.1 (o. ^ ' 0.5o, 0.8(0, "C 1.5(u, 2 oi, 4o)(. lOc.^

co rad/sec 912.9 4564 7303 9129 18258 36516 91290

if> -0.497° -3.27° -10.8' -90° -174' - 177" -178.7° -179.5°
-1,18 0.087 2.48 8.72 21.3 -1.98 -9.56 -23.5 - 39.9

R L
470 n 0.6 H

C
E,n
0.02 M?

FIGURE PS 11-14

168 | ac circuit analysis

PS 11-15 Develop the transfer function and construct Now let R i C , = TI = l/fc>iI and R^C^I
""" "eq'- — '•2 [/(.•)-,:
a Bode plot of the amplitude response of the network
_Eo_ _ ^2 1 +j(•)|ti)^
shown in Fig. PS ll-15a. Assume Ri = R^. 2)
E,,, /?i + R^ 1 +;&V(U^

2. Analyze the transfer function to ascertain what con-

tributions the various factors will make in the Bode plot.
Equation (2), the transfer function, is of the form
Eo b
=a
^ ^
where a = RI,{R\ + Ri). b = 1 4- j u / o } ^ and c = 1 +
jto/ui^. Therefore,
dB = 20 (log a + log b - log c)
The log a term will contribute a component which is
independent of frequency, while the b and c terms are
frequency-dependent. We have shown in previous ex-
c*Ji CJ2
amples that the b term will break at u}^ and slope up-
FIGURE PS 11-15 ward at 20 dB per decade, while the c term breaks at w^
and slopes downward at 20 dB per decade. Inspection
of the circuit indicates that at low frequencies, where
SOLUTION A'c, is very large compared with R i, the output will equal
I. First develop the transfer function. ^R^(R^ + /?2)]£m. At frequencies where Xc, is small
compared with R^, the output will approach the input.
1)
R: 3. Construct the Bode plot. For the purpose of sketch-
R^-jXc,) ing, let us assume that R^ = Rf Then R^/(Ri + R^) =
R. + ^, or — 6 dB. This fixes the ratio of TI to T;:
RI -7^c,
R. = R,C,
R, R^RZ C,
R.2 + julC^ Ri 4- l/ju>C^ 1 or R^)C, =0.5^C\
R,+R,
R. ^ RiCi "2
R, ti 0.5RlC^ (i)t
Ri1 + 7-cuC\ {ju)R^C^ + l)//cuCi T-2

The Bode plot is sketched in Fig. PS 11-15&.

^2
PS 11-16 Bode plot the transfer function
PI
Rl2 + o»
1 +jwRtCi 10 1 + j
20^
RZ A =
. cy V . w }
R^ + j w R t R ^ C i + R i l+J
1 +jo}RiCi
^ ^loo,
- ^2(t +7'tt>^iCi)
SOLUTION In the very low frequency spectrum or for
-
/?i + R; +• JCDR^R^CI o» approaching zero, we see that A = 10, or 20 dB. This
means we have a network capable of providing an
Dividing numerator and denominator by R[ + R^ amplification of ten times down to dc or zero frequency,
yields as shown in Fig. PS 11-16. Electronic amplifiers can do
this. As cu increases, the first break occurs at <u = 5
R^ 1 +7'£u/?iCi rad/sec, and since this is in the denominator, the gain
RI + R^ 1 4-yco^Ci (amplitude) rolls off at -20 dB/dec from the 20-dB
where level. The next break at CD = 20 rad/sec is in the numera-
tor, which tries to pull A up at 4-20 dB/dec. However,
/?,/?
1^2 this is canceled by the previous —20 dB/dec slope, so
R,. =
eq that the net slope is now zero and it stays that way until
^1 + R;

the Bode plot 169

 FIGURE PS 11-16

CD = 100, at which it breaks down again at —20 dB/dec SOLUTION The above transfer function may be re-
and remains so. If correction factors are inserted or written as
Ada is actually computed, the smoothed amplitude CO
response curve results.
COi
The phase response is approximated by lines sloping 2) A =
plus or minus 45°/dec, depending upon whether or not . d} . W
1 +J— 1 + ; —
the term is in the numerator or denominator. The phase (Ui COi
approximations start at 0° at 0.1 a>c, go through 45° Inspection of (2) indicates that the first factor is that of a
at a>,, and terminate at + 90° at 10 u>c. The resultant phase high-pass RC type filter and the second factor is that of a
approximation is illustrated by the stepped phase curve. low-pass filter with identical comer frequencies. There-
For more accuracy, the various correction factors can be fore, in the frequency range co < <Ui, we have the low-
inserted or the actual phase angle of A may be computed pass response climbing at 20 dB/dec while the high-pass
to yield the smoothed phase curve of Fig. PS 11-16. response is flat at 0 dB. At <u = coi, the high-pass re-
PS 11-17 Bode plot the transfer function sponse breaks to zero slope while the low-pass response
breaks downward at —20 dB/dec. This yields the ap-
(U
proximated amplitude response of Fig. PS 11-17.
CD, The approximate phase response may be sketched by
1) A =
0) first drawing the high-pass filter phase response as a line
1 +J
0», starting at co/coi = 0.1, sloping down 45°/dec from +90°

170 ac circuit analysis

to 0° at a>/'(0t = 10' ^d then sketching the low-pass high- and low-pass phase curves. Should greater accuracy
filter phase with the same slope and <u values but starting be required, correction factors may be inserted or actual
at 0' and terminating at —90°. The approximate phase values computed in (2). The actual phase and magnitude
may then be sketched by algebraically adding both the are also shown in Fig. PS 11-17.

the Bode plot

 PROBLEMS WITH ANSWERS

PA 11-1 Deve lop the transfer f unction and Bode plot P»A 11-4 Develop the transfer fu notion and Bode plot
the amplitude re sponse of the cir cuit in Fig. PA I I - l a . t ie amplitude response of the circ uit in Fig. PA ll-4a.

r
L , I/ -
"c 2cji;

1 . T ^
".
Y .[
c
• t 0
1

(a)
0
-dB -20dB\.
dec

16)
'
<0——————
(a>
+• • -6
-dB

(6)

ANSWER See Fig. P' A l l - l f r . AiNSWEP? See Fig. P/\ 11-4b.

PA 11-2 Deve lopth e transfer function and Bode plot F»A 11-5 Develop the transfer fu notion and Bode plot
the amplitude re sponse of the circuit in Fig. PA 1 l-2a. the ampli tude respomse of the circ uit in Fig. PA ll-5a.

60 kn
o—WSr-t———' i——c ft /? (^ c- RC
50 i
]t °- - I -9.54

x
?s
| ^30 kn ^. -40dB
-9.54 c? ^ c? ^ E0
>>.dec
1 nF ? -dB
^ -dB
0——————•—————« 0
(a) W
(a) w
•

ANSWER See Fig. P>A ll-2b. ANSWER? See Fig. PA 11-5b.

PA 11-3 Deve op th e transfer function and Bode plot P*A 11-6 Develop the transfer fu notion and Bode plot
the amplitude re sponse of the circuit in Fig. PA 11-3a. t he amplitude respo)nse of the circ uit in Fig. PA ll-6a.

3———V^———•—— > <o— i—<0 c

,5tJc "c ^ we
t R
<
0 c ,I ft
-9.54
>fl

' T E -20d^<!
-6.02 - dec ^————— I
c E-0
-'-~-20dB--^

r
(a)
-dB

?1 (a)
C?
0
-dB
dec

tol

ANSM^EJ? See Fig. P'A n-3fc. /iNSWEf^ See Fig. PA 11.6ft.

172 | ac circuit analysis

 PA 11-7 Evaluate A at &» = &»,. i f ( 5 = ^ i n the follow-
ing transfer function:
[ + 7 ' " - ) 100
co, '
A =
/ u) \ /<y\ -. M
1 + 7 - 1 - - +7-2<)—
\ "c/L V^c/ M,
3
ANSWER 43dB/-45

PA 11-8 Bode plot A = ————————— Show

\_l +jc'A0.5 x 10 "^J
both phase and amplitude.

-20

ANSWER See Fig. PA 11-8.

the Bode plot | 173

 PA 11-9 Estimate the transfer function applicable to
Fig. PA 11-9.

0.01 0.1 1.0 10 100

ANSWER
0} cy
j 0.01 1 +J 100
A = 10
1 +3
0.01
(0
f[ ll ++j /I ^(- ^
l(V

174 | ac circuit analysis

PA 11-10 Bode plot

. / . • a) V, . &)
' ^T ^oT
A =
/ . *" \ / < • (y
^^- ^ 100;

-90

ANSWER See Fig. PA 11-10.

the Bode plot | 175

PROBLEMS WITHOUT ANSWERS

P 11-1 Bode plot the transfer function A = 28,000. P 11-8 Develop the transfer function and construct
the Bode plot for the circuit of Fig. P 11-8.

93 kn
o—VA-

7kn<

P 11-2 What is the corner frequency of a transfer

function expressed as 1 •+- j u j T where T is a time con-
stant?
P 11-9 Construct a Bode plot for the circuit of Fig.
P 11-9.
1 kn
o—WSr

CO
P 11 -3 Bode plot the transfer function A = j I^F'
CL>,

P 11-10 Construct a Bode plot for the circuit of Fig.

P 11-10.

/ /-v 1fF
P 11 -4 Bode plot the transfer function A = [ j — I .
\ Jc / -ie
100 kn'

P 11-11 Develop the transfer function and construct a

P 11 -5 Bode plot the transfer function A = , ^ ^ , . Bode plot for the circuit of Fig. P 11-11.
I cu, I
10 mH

1 kn<

P 11 -6 Bode plot the transfer function

A* _
= .1, +. j. / P 11-12 Develop the transfer function and construct a
5kHz Bode plot for the circuit of Fig. P 11-12.

3n

P 11 -7 Bode plot the transfer function 4H-

M
*=0.5;L ^
A =
1 •I I ]

\2 • 1 H-
f
1 +J 80kHz;

176 | ac circuit analysis

P 11-13 Sketch the Bode plot for the circuit of Fig. P 11-15 Estimate the transfer function of the Bode
P 11-13. plot in Fig. P 11-15.
100 kn
40 d8
«2
i < i—i '——o
ff, i
i kn
H?
c
> E
1tJF
/.3<
1 kn^

-20

P 11-14 With regard to the transfer function

100(1 +7'CL>1)
A = P 11-16 Estimate the transfer function of the Bode
1 +^J5) 2 0•(ol)
plot in Fig. P 11-16.
fill in the blank spaces.
(a) The factor 100 in the numerator represents _____
dB of gain. 58 dB
(b) The 1 •+- ;&>1 factor will break at o> = ______and
slope ______ dB/dec or ______ dB/oct. It will 40 dB
,-dB.
contribute degrees of phase at •dec
the corner frequency. .20-

(c) The ( 1 + jwS)2 factor breaks at u> = . and !?A

slopes ______ dB/dec. It contributes 1000 l^sec"
degrees of phase . at the corner frequency and
the corresponding amplitude response is
(d} The j<jj I factor contributes degrees of
phase ______ and slopes

the Bode plot 177