Chap14 Frequency Response

Chap14 Frequency Response

金國生 教授
長庚大學電子系
 14.1 Introduction
• In sinusoidal circuit analysis, we have learned how to find
voltages and currents in a circuit with a constant frequency
source. Here, we let the amplitude of the sinusoidal source
remain constant and vary the frequency to obtain frequency
response.
• The frequency response of a circuit is the variation in its
behavior with change in signal frequency.
• The sinusoidal steady-state frequency responses of circuits
are of significance in communications and control systems.
The electric filters eliminate signals with unwanted
frequencies and pass signals of the desired frequencies.

2
 14.2 Transfer Function
• The transfer function H() (also called the network function)
is a useful analytical tool for finding the frequency response of
a circuit.
• A transfer function is the frequency-dependent ratio of an
output function to an input function.
• The transfer function H() of a circuit is the frequency-
dependent ratio of a phasor output Y() (an element voltage
or current) to a phasor input X() (source voltage or current).

Y ( )
H ( ) 
X( )

3
Four Possible Transfer Functions
Vo ( )
H( )  Voltage gain 
Vi ( )
I o ( )
H( )  Current gain 
I i ( )
Vo ( )
H( )  Transfer Impedance 
I i ( )
I o ( )
H( )  Transfer Adimittanc e 
Vi ( )

4
 Zero and Pole
The transfer function H() can be expressed in terms of its
numerator polynomial N() and denominator polynomial D() as

N ( )
H ( ) 
D( )
The roots of N() = 0 are called the zeros of H() and are usually
represented as j = z1, z2,… . Similarly, the roots of D() = 0 are
the poles of H() and are represented as j = p1, p2, …..

A zero, as a root of the numerator polynomial, N(), is a value

that results in a zero value of the function. A pole, as a root of the
denominator polynomial, D(), is a value for which the function
is infinite.
5
 Ex. 14.1
For the RC circuit in Fig. 14.2(a), obtain the transfer function
Vo/Vs and its frequency response. Let vs = Vmcost.

Fig. 14.2(a)

time-domain frequency-domain
Sol:
Vo 1 / jC 1
By voltage division, H( )   
Vs R  1 / jC 1  jRC

6
 Frequency response
Ex. 14.1 (Cont.)
1 1  jRC 1  j /  0 Amplitude
H( )   
1  jRC 1  RC 2 1   / 0 2 response
where 0  1 / RC
1 
H  ,    tan 1

1  ( / 0 ) 2 0

/0 H  /0 H 
0 1 0 10 0.1 -84
Phase
1 0.71 -45 20 0.05 -87
response
2 0.45 -63 100 0.01 -89
3 0.32 -72  0 -90

7
 Ex. 14.2
For the circuit in Fig. 14.6, calculus the gain Io()/Ii()
and its poles and zeros.

Fig. 14.6

8
 Ex. 14.2 (Cont.)
Sol: By current division,

4  j 2 1/(j0.5)
I o ( )  I i ( )
4  j 2  1 / j 0.5 j2
I o ( ) j 0.5 (4  j 2 ) s ( s  2)
  , s  j
I i ( ) 1  j 2  ( j ) 2
s  2s  1
2

The zeros are at

s ( s  2)  0  z1  0, z 2  2
The poles are at
s 2  2s  1  0  ( s  1) 2  0
Thus, there is a double pole at p  1.

9
 14.3 The Decibel Scale
A more systematic way of obtaining the frequency response is to
use Bode plots. Before we begin to construct Bode plots, we
should take care of two important issues: the use of logarithms
and decibels in expressing gain.
1. log P1 P2  log P1  log P2
2. log P1 / P2  log P1  log P2
3. log P n  n log P
4. log 1  0
In communications systems, gain is measured in bels. The bel
is used to measure the ration of two levels of power or power
gain G P2
G  Number of bels  log10
P1
10
 Decibel

The decibel (dB) provides us with a unit of less magnitude. It

is 1/10th of a bel and is given by
P2
GdB  10 log10
P1
– When P1 = P2, there is no change in power and the gain is 0
dB.
– If P2 = 2P1, the gain is GdB  10 log10 2  3 dB
– When P2 = 0.5P1, the gain is GdB  10 log10 0.5  3 dB

11
 Gain in Decibel Scale
The gain G can be expressed in terms of voltage or current ratio. If P1
is the input power, P2 is the out power, R1 is the input resistance and R2 is
the load resistance, then P1= 0.5V12/R1 and P2= 0.5V22/R2.
P2 V22 / R2
GdB  10log 10  10 log 10 2
P1 V1 / R1
2
V  R
 10 log 10  2   10 log 10 1
 V1  R2
V2 R2
GdB  20 log 10  10 log 10
V1 R1
V2
–For R2 = R1, GdB  20 log10
V1
I2
–If P1=I12R1, P2=I22R2, and R2 = R1 GdB  20 log10
I1
12
 Three things are important to note:
1. That 10log10 is used for power, while 20log10 is used for
voltage or current, because of the square relationship
between them (P = V2/R = I2R ).

2. That the dB value is a logarithmic measurement of the ratio

of one variable to another of the same type.
Vo ( ) Vo ( )
H( )  Voltage gain  H( )  Transfer Impedance 

H( )  Current gain 
Vi ( )
I ( )
o

H( )  Transfer Adimittanc e 
I i ( )
I o ( )
I i ( ) Vi ( )
3. We only use voltage and current magnitudes in
V2 I
GdB  20 log10 , GdB  20 log10 2 .
V1 I1
Negative signs and angles will be handled independently.
13
 14.4 Bode Plots
 The frequency range required in frequency response is
often so wide that it is inconvenient to use a linear scale
for the frequency axis.
 Also, there is a more systematic way of locating the
important features of the magnitude and phase plots of
the transfer function.
 For these reasons, it has become standard practice to plot
the transfer function on a pair of semilogarithmic plots:
the magnitude in decibels is plotted against the logarithm
of the frequency; on a separate plot, the phase in degrees
is plotted against the logarithm of the frequency.

14
 Bode Plots
Bode plots are semilog plots (半對數圖) of the magnitude
(in decibels) and phase (in degrees) of a transfer function
versus the logarithm of the frequency.
The transfer function can be written as
H  H  He j
Taking the natural logarithm,
ln H  ln H  ln e j  ln H  j
magnitude phase
The real part of lnH is a function of the magnitude while the
imaginary part is the phase.

15
 Bode Magnitude Plot

In a Bode magnitude plot, the gain

H dB  20 log10 H
is plotted in decibels (dB) versus
frequency.

16
 Standard Form of H()
The standard form (標準型) of H():

which is obtained by dividing out the poles and zeros in H().

Magnitude in log scale:
20 log10 H( )  20 log10 K  20 N log10 j  20 log10 1  j / z1
 20 log10 1  j 21 ( / k )  ( j / k ) 2     
 20 log10 1  j / p1  20 log10 1  j 2 2 ( / n )  ( j / n ) 2    

1  21 /  k 
Phase:   0 o
 ( N 90 o
)  tan 1
 / z  tan    
2 
 1   / k 
1 2

1  2 2 /  n 
 tan  / p1  tan 
1
2 
  
 1   / n 
2

17
 Standard Form of H()

H() may include up to seven types of different factors that

can appear in various combinations in a transfer function.

Seven types of different factors for H and :

1. gain K
2. pole (j)-1 or zero (j) at the origin.
3. Simple pole 1/(1+ j/p1) or simple zero (1+ j/z1)
4. Quadratic pole 1/[1+j22/n+(j/n)2] or
quadratic zero [1+j21/k+(j/k)2]

18
 Construction of Bode Plot
ln H  ln H  j Factors are combined additively
In constructing a Bode plot, we plot each factor separately
and then add them graphically.
The factors can be considered one at a time and then
combined additively because of the logarithms involved.
We find that straight-line plots known as Bode plots
approximate the actual plots to a reasonable degree of
accuracy.

H dB  20 log10 H
Bode plot : 
  in degree

19
 Bode Plot for Gain K
• Constant terms  gain K
– The magnitude is 20log10|K|
– If K is positive, the phase is 0o
– If K is negative, the phase is 180o

Bode plot for gain K

20
Bode Plot for Zero at the Origin
Zero at the origin:
Zero (j)
• Magnitude is 20log10
• The slope is 20 dB/dec
• The phase is 90o
Origin means that  = 1 (or log = 0) and H = 0 dB.

origin

21
 What is Decade?
A decade is an interval between two frequencies with a ratio
of 10; e.g., between 0 and 100, or between 10 and 100 Hz.
Thus, 20 dB/decade means that the magnitude changes
20 dB whenever the frequency changes tenfold or one
decade.

22
 Bode Plot for Pole at the Origin
Pole at the origin:
Pole (j)-1
• Magnitude is 20log10
• The slope is 20 dB/dec
• The phase is 90o

-20 dB/dec -90o

23
Bode Plot for Pole/Zero to the Nth Power
(j)N
• Magnitude is  20Nlog10
• The slope is  20N dB/dec
• The phase is  90N degree

24
 Bode Plot for Simple Zero
Simple zero (1+ j/z1)

• Magnitude is H dB  20 log10 1  j z1

j
  0  H dB  20 log 10 1   20log 101  0
z1
j 
    H dB  20 log 10 1   20log 10
z1 z1
• Phase is tan-1(/z1)
 0,   0
  
  tan 1    45,   z1
 z1  
90,   
25
 Corner Frequency of Simple Zero
Magnitude approach zero when   0. A straight line with
slope 20 dB/dec is for large .
The frequency  = z1 where the two asymptotic lines meet
is called corner frequency or break frequency.
Approximate plot is close to the actual plot except at the
break frequency  = z1. The deviation is 20log10|(1+j1)| =
20log1021/2  3 dB.
slope 20 dB/dec
  0  H dB  0

    H dB  20log10
z1

corner frequency
26
 Phase Plot of Simple Zero
As a straight-line approximation, let   0 for   z1/10,  
45o for  = z1,   90o for   10z1, the straight line with
slope = 45o/dec.

 0,   0
1    
  tan     45,   z1
 z1  
90,   

slope = 45o/dec

27
 Bode Plot for Simple Pole

• Simple pole  1/(1+ j/p1)

• Magnitude is H dB  20 log10 1  j p1

• The corner frequency is at  = p1
• The magnitude slope is 20 dB/dec
• Phase slope is  45o/dec

28
 Bode Plot for Quadratic Pole
• Quadratic Pole  1/[1+j22/n+(j/n)2]
2
j 2 2  j 
• Magnitude is H dB  20 log10 1    
n  n 
2
j 2 2  j 
  0  H dB  20 log10 1      0
n  n 
2
j 2 2  j  
    H dB  20 log10 1      40 log10
n  n  n

2 2 / n
• Phase is    tan 1
1   2 / n2

29
 Magnitude Plot for Quadratic Pole
The amplitude plot consists of two straight asymptotic lines:
One with zero slope for   n, and the other with slope 40
dB/dec for   n. The difference between the actual plot and
the straight-line plot is due to the damping factor 2.

  0  H dB  0

    H dB  40 log 10
n -40 dB/dec

The quadratic pole 1/[1+j22/n+(j/n)2] can be treated as a double

(1+j/n)-2 as far as straight-line approximation is concerned.
30
 Phase Plot for Quadratic Pole
  0 for   n/10,   90o for  = n,   180o for  
10n, and the straight line with slope 90o/dec for n/10   
10n. The difference between the actual plot and the straight-
line plot is due to the damping factor 2.
 0,   0
1 2 2 / n 
   tan    90,   n
1   / n 
2 2

 180 ,   

-90o/dec

31
 Bode Plot for Quadratic Zero
•Quadratic zero  [1+j21/k+(j/k)2]
2
j 21  j 
•Magnitude is H dB  20 log10 1    
k  k 
2  / 
•Phase is   tan 1 1 2 k2
1   / k
•The magnitude slope is 40 dB/dec, phase slop is 90o/dec

  0  H dB  0  0,   0
1 2 2 / n 
   tan   90,   n
    H dB  40 log 10 1   / n 
2 2

n 180 ,   

32
Summary of Bode Straight-Line
Magnitude and Phase Plots

33
34
 Ex. 14.3
Construct the Bode plots for the transfer function.
200 j
H ( ) 
( j  2)( j  10)
Sol:
We first put H() in the standard form by dividing out the poles and zeros.
Thus,
10 j
H ( ) 
(1  j / 2)(1  j / 10)
10 j
 (90  tan 1  / 2  tan 1  / 10)
1  j / 2 1  j / 10
j j
H dB  20 log10 10  20 log10 j  20 log10 1   20 log10 1 
2 10
 
  90  tan 1  tan 1
2 10
35
 Ex. 14.3 (Cont.)
j j
H dB  20 log 10 10  20 log 10 j  20 log 10 1   20 log 10 1 
2 10
There are two corner frequencies at  = 2 and 10 rad/s.

Magnitude plot
H(dB)
20 20 log 10 10

0
0.1 0.2 1 2 10 20 100 200 

-20 20 log 10 j j
 20 log 10 1 
10
j
 20 log 10 1 
2

36
 Ex. 14.3 (Cont.)
 1 
  90  tan
1
 tan
2 10
There are two corner frequencies at  = 2 and 10
rad/s.

Phase plot
H(dB)
90
90

0
0.1 0.2 1 2 10 20 1 100 200 
 tan
 10
 tan 1
-90 2

37
 Ex. 14.4
Obtain the Bode plots for
j  10
H( ) 
j ( j  5) 2
Sol:
Putting H() in the standard form, we get

There are two corner frequencies at  = 5 and 10 rad/s.

0.4(1  j / 10)
H( ) 
j (1  j / 5) 2
j j
H dB  20 log10 0.4  20 log10 1   20 log10 j  40 log10 1 
10 5
 
  0  tan 1
 90  2 tan 1

10 5
38
 Ex. 14.4 (Cont.)
j j
H dB  20 log10 0.4  20 log10 1   20 log10 j  40 log10 1 
10 5

For the pole with corner frequency at  = 5, the slope of the magnitude plot
is 40 dB/decade and that of the phase plot is 90 per decade due to the
power of 2.
39
 Ex. 14.4 (Cont.)
 
  0  tan 1
 90  2 tan 1

10 5

40
 Ex. 14.5
Draw the Bode plots for
s 1
H( s)  2
s  60 s  100
Sol: We express H(s) as
1 / 100 (1  j )
H( ) 
1  j 6 / 10  ( j / 10) 2
j 6  2
H dB  20 log10 100  20 log10 1  j  20 log10 1  
10 100
  6 / 10 
  0  tan   tan 
1 1

1   / 100 
2

For the quadratic pole, n = 10 rad/s, which serves as the corner frequency.

41
 Ex. 14.5 (Cont.)
j 6  2
H dB  20 log10 100  20 log10 1  j  20 log10 1  
10 100

Notice that the quadratic pole is treated as a repeated pole at k, that is,
(1+j/n)-2, which is an approximation.

42
 Ex. 14.5 (Cont.)
  6 / 10 
  0  tan   tan 
1 1

1   / 100 
2

43
 Ex. 14.6
Given the Bode plot in Fig. 14.19, obtain the transfer function
H().
Sol: 1. A zero always causes an upward turn at
a corner frequency, while a pole causes
a downward turn.
2. There is a zero j at the origin ( = 1).
This is indicated by the straight line
with slope 20 dB/dec. Since this line is
shifted by 40 dB  40 dB gain
40  20 log10 K  log10 K  2
K  102  100

Fig. 14.19 3. There are three factors with corner

frequencies at  = 1, 5, and 20 rad/s 
three simple poles
44
 Ex. 14.6 (Cont.)
1. A simple pole at p = 1with slope -20 dB/dec to
cause a downward turn and counteract the zero
at the origin. The pole at p = 1 is determined as
1/(1+j/1).
2. Another simple pole at p = 5 with slope -20
dB/dec causing a downward turn. The pole is
1/(1+j/5).
3. A third simple pole at p = 20 with slope -20
dB/dec causing a further downward turn. The
pole is 1/(1+j/20).
100 j
H ( ) 
(1  j / 1)(1  j / 5)(1  j / 20)
Fig. 14.19 j104

( j  1)( j  5)( j  20)
104 s
H( s)  , s  j
( s  1)(s  5)( s  20)
45
 14.5 Series Resonance
• Resonance occurs in any system that has a complex
conjugate pair of poles, it is the cause of oscillation of
stored energy from one form to another. It is the
phenomenon that allows frequency discrimination in
communications networks.
• Resonance is a condition in an RLC circuit in which the
capacitive and inductive reactances are equal in magnitude,
thereby resulting in a purely resistive impedance.
• Resonant circuit are useful for constructing filters, as their
transfer functions can be highly frequency selective.

46
 Series RLC
Consider a series RLC circuit in the frequency domain. The input
impedance Z is

Vs 1
Z  H( )   R  jL 
I jC
 1 
Z  R  j  L  
 C 

Resonance results when the imaginary part of the transfer

function is zero, or
1
Im( Z )  L  0 0 is the resonant frequency
C
1 1 1
0 L  or 0  rad/s  f 0  Hz
0C LC 2 LC
47
 Resonance

Note that at resonance:

1. The impedance is purely resistive, thus, Z = R. In other
words, the LC series combination acts like a short
circuit, and the entire voltage is across R.
2. The voltage Vs and the current I are in phase, so that
the power factor is unity.
3. The magnitude of the transfer function H() = Z() is
minimum.
4. The inductor voltage and capacitor voltage can be
much more than the source voltage.

48
 Frequency Response of I
The frequency response of the circuit’s current magnitude:

Vm
II
R 2  (L  1 / C ) 2

The average power dissipated by

the RLC circuit is
1 2
P( )  I R
2
The highest power dissipated occurs at resonance, when
1 Vm2
I  Vm R  P(0 ) 
2 R
49
 Half-Power Frequency
When dissipated power is half
the max. value, that is
(Vm / 2 ) 2 Vm2
P(1 )  P(2 )  
2R 4R

1, 2 are called the half-power frequencies.

or
1 2
P( )  I R, I  Vm Z
2
1 Vm2 Vm2
When Z  2 R, P(1 )  2
R  P(1 )  P(2 ) 
2Z 4R
50
 Bandwidth
The half power frequencies are obtained by setting Z  2 R
2
 1 
R   L 
2
  2R
 C 
2
R  R  1
1      
2L  2 L  LC
2
R  R  1
2     
2L  2 L  LC
R
0  12 , B  2  1 
L
Bandwidth
51
 Quality Factor Q
• The “sharpness” of the resonance in a resonant circuit is
measured quantitatively by the quality factor Q.
• The quality factor relates the maximum energy stored to the
energy dissipated in the circuit per cycle of oscillation:

Peak energy stored in the circuit

Q  2
Energy dissipated by the circuit
in one period at resonance

52
 Quality Factor Q
In the series RLC circuit, the peak energy stored is LI2/2 while
the energy dissipated in one period is I2R/21/f0. Hence,

LI 2
1
2f 0 L
Q  2 1 2
2

2 I R (1 / f 0 ) R
0 L1
Q 
R 0CR
R 0 1
B  (0  )
L Q LC

The quality factor of a resonant

circuit is the ratio of its resonant
frequency to its bandwidth.
53
 Half Power Frequencies
For the high-Q circuit (Q  10), the half power
frequencies are symmetrical around the resonant
frequency and can be approximated as

B
1  0 
2
B
2  0 
2

54
 Ex. 14.7
In the circuit of Fig. 14.24, R = 2 , L = 1 mH, and C =
0.4 F. (a) Find the resonant frequency and the half-
power frequencies. (b) Calculus the quality factor and
bandwidth. (c) Determine the amplitude of the current
at 0, 1, and 2.

Fig. 14.24

55
 Ex. 14.7 (Cont.)
Sol:
1 1
(a) 0  
3 6
 50 krad/s
LC 10  0.4 10
2
R  R  1
1      
2L  2 L  LC
2
 3
 (10 3 2
)  (50  10 3 2
)
2 10
 1  1  2500 krad/s  49 krad/s
2  1  1  2500 krad/s  51 krad/s

56
 Ex. 14.7 (Cont.)
(b) B  2  1  2 krad/s
R 2
or B   3  2 krad/s
L 10
0 L 50 10 3 10 3
Q   25
R 2
0 50 10 3
B   2 krad/s
Q 25
B
Since Q  10, 1  0   50  1  49 krad/s
2
B
12  0   50  1  51 krad/s
2

57
 Ex. 14.7 (Cont.)

(c) At  = 0 Vm 20
I   10 A
R 2

Vm 10
At  = 1, 2 I   7.071 A
2R 2

58
 14.6 Parallel RLC Resonance
The parallel RLC circuit is the dual of the series RLC circuit.
The admittance is

I 1 1 1  1 
Y  H ( )    jC  or Y   j C  
V R jL R  L 
1 1
Resonce  C   0 or 0  rad/s
L LC
Notice that at resonance, the parallel LC combination acts like an open
circuit, so that the entire current flows through R. The inductor and
capacitor current can be much more than the source current at resonance.
59
 Parallel RLC
Series RLC Duality Parallel RLC
 1  1  1 
Z  R  j L   Y   j C  
 C  R  L 
From the duality, by replacing R, L, and C in the expressions for
the series circuit with 1/R, C, and L respectively, we obtain for
the parallel circuit:
Series RLC Parallel RLC
2 2
R  R  1 1 1  1
1       1      
2L  2 L  LC 2 RC  2 RC  LC
2 2
R  R  1 1  1  1
2      2     
2L  2 L  LC 2 RC  2 RC  LC
60
 Parallel RLC
1
B  2  1 
RC
0 R
Q  0 RC 
B 0 L

The half power frequencies in terms of the quality factor is:

2
 1  0 For high-Q circuits (Q  10)
1  0 1    
 2Q  2Q
2 B B
 1   1  0  , 2  0 
2  0 1     0 2 2
 2Q  2Q
61
62
 Ex. 14.8
In the parallel RLC circuit of Fig. 14.27, let R = 8 k, L
= 0.2 mH, and C = 8 F. (a) Calculus 0, Q, and B. (b)
Find 1 and 2. (c) Determine the power dissipated at
0, 1, and 2.

63
 Ex. 14.8 (Cont.)
Sol:
1 1 10 5
(a) 0     25 krad/s
LC 3
0.2 10  8 10 6 4
R 8 10 3
Q   1600
0 L 25 10  0.2 10
3 3

0
B  15.625 rad/s
Q
(b) Due to the high value of Q, we can regard this as a high-Q
circuit, Hence,
B
1  0   25000  7.812  24992 rad/s
2
B
2  0   25000  7.812  25008 rad/s
2
64
 Ex. 14.8 (Cont.)
(c) At   0 , Y  1/R or Z  R  8 k
V 10  90
Io    1.25  90 mA
Z 8000

1 2 1
P  I o R  (1.25 10 3 ) 2 (8 10 3 )  6.25 mW
2 2
Vm2 100
or P    6.25 mW
2 R 2  8 10 3

At   1 , 2
Vm2
P  3.125 mW
4R

65
 Ex. 14.9
Determine the resonant frequency of the circuit.

1 1 2  j 2
Sol: Y  j 0.1    0.1  j 0.1 
10 2  j 2 4  4 2
At resonant, Im(Y )  0, and
20
0 0.1   0  0  2 rad/s
4  40 2

66
 14.7 Passive Filters
Filters are the circuits used in radio and TV receivers to allow us to
select one desired signal out of a multitude of broadcast signals in the
environment. As a frequency-selective device, a filter can be used to
limit the frequency spectrum of a signal to some specified band of
frequencies.
Lowpass Bandpass

Highpass Bandstop

67
Passive Filters

68
 Lowpass Filter
A lowpass filter is designed to pass only frequencies from dc up
to the cutoff frequency c.

A typical lowpass filter is formed when the output of an RC

circuit is taken off the capacitor. The transfer function is
Vo 1 / jC 1
H( )   or H( ) 
Vi R  1 / jC 1  jRC
69
 Lowpass Filter
Note that H(0) = 1 and H() = 0. The half-power frequency,
which is equivalent to the corner frequency on the Bode plots
but in the context of filters is usually known as the cutoff
frequency c, is obtained by setting the magnitude of H()
equal to 1 / 2 , thus,

1 1
H (c )  
1  c R C
2 2 2
2
1
or c 
RC

70
 Highpass Filter
A highpass filter is designed to pass all frequencies above its
cutoff frequency c.

A highpass filter is formed when the output of an RC circuit is

taken off the resistor. The transfer function is
Vo R jRC 1
H( )    H( )  , c 
Vi R  1 / jC 1  jRC RC
71
 Bandpass Filter
A bandpass filter is designed to pass all frequencies within a
band of frequencies, 1 <  < 2.

The RLC series resonant circuit provides a bandpass filter when

the output is taken off the resistor. The transfer function is
Vo R 1
H( )   , 0 
Vi R  j (L  1 / C ) LC
72
Note that H(0) = 0 and H() = 0. The bandpass filter passes
a band of frequencies (1 <  < 2) centered on 0, the
center frequency.

1
0 
LC

73
 Bandstop Filter
A bandstop filter is designed to stop or eliminate all frequencies
within a band of frequencies, 1 <  < 2.

A bandstop filter is formed when the output RLC series resonant

circuit is taken off the LC series combination. The transfer function is
Vo j (L  1 / C ) 1
H( )   , 0 
Vi R  j (L  1 / C ) LC
74
 Ex. 14.10
Determine what type of filter is shown in Fig. 14.39.
Calculus the corner or cutoff frequency. Take R = 2
k, L = 2 H, and C = 2 F.

Fig. 14.39

75
 Ex. 14.10 (Cont.)
Sol:

Vo R 1 / sC
H( s)  
Vi sL  R 1 / sC
where s  j
1 R / sC R
R  
sC R  1 / sC 1  sRC
R /(1  sRC ) R
H( s)   2 , s  j
sL  R /(1  sRC ) s RLC  sL  R
or
R
H ( ) 
  2 RLC  jL  R
76
 Ex. 14.10 (Cont.)
Since H(0) = 1 and H() = 0, we conclude that the circuit is a
second-order lowpass filter. The magnitude of H is
R 1
H   c
( R   RLC)   L
2 2 2 2
2
2
1 R
H2  
2 ( R  c2 RLC) 2  c2 L2
2
 L 
or 2  (1  c2 LC ) 2   c 
 R 
2  (1  c2 4 10 6 ) 2  (c 10 3 ) 2 , let c  c  10 3
2  (1  4c2 ) 2  c2 or 16c4  7c2  1  0
c2  0.5509 and  0.1134  c  0.742 krad/s  742 rad/s
77
 Ex. 14.11
If the bandstop filter in Fig. 14.37 is to reject a 200 Hz
sinusoid while passing other frequencies, calculus the values
of L and C. Take R = 150  and the bandwidth as 100 Hz.

Sol:
B  2 (100 )  200 rad/s
R R 150
B  L   0.2387 H
L B 200
0  2f 0  2 (200 )  400
Since 0  1 / LC
1 1
C 2   2.653 F
0 L (400 ) (0.2387 )
2

78
 14.8 Active Filters
Passive filters are useful at high frequencies. Three major
limitations to the passive filters:
(1) Cannot generate gain greater than 1
(2) May require bulky and expensive inductors.
(3) Perform poorly at frequencies below the audio frequency
range (300 Hz < f < 3 kHz).

Active filters consist of combinations of resistors, capacitors, and

op amps, offering advantages over passive filters.
(1) Smaller and less expensive. Do not require inductors.
(2) Provide amplifier gain.
(3) Combined with buffer amplifiers to isolate each stage of the filter,
allowing designing the stages independently and then cascading
them to realize the desired transfer function.
(4) Practical limit of active filters is ~100 kHz
79
 First-Order Lowpass Filter
Filters are often classified according to their order (or number of poles). One
type of first-order filter is shown in Fig. 14.41. The components selected for
Zi and Zf determine whether the filter is lowpass or highpass, but one of the
components must be reactive.

Fig. 14.41

Vo Zf 1 R f / jC f Rf
H( )    , where Z i  Ri and Z f  R f  
V1 Zi jC f R f  1 / jC f 1  jC f R f
Rf 1 1 At low frequencies (0) the gain
 H( )   , c 
Ri 1  jC f R f Rf C f tends to -Rf/Ri.
80
 First-Order Highpass Filter
Figure 14.43 shows a typical highpass filter. For this filter,
the transfer function is
Figure 14.43
Vo Zf
H( )  
V1 Zi
where Z i  Ri  1 / jCi and Z f  R f
Rf jCi R f
H( )   
Ri  1 / jCi 1  jCi Ri
1
c 
Ri Ci

At very high frequencies () the gain tends to Rf/Ri.

81
 Bandpass Filter
By cascading a unity-gain lowpass filter, a unity-gain highpass filter, and an
inverter with gain –Rf/Ri, we can construct a bandpass filter.

82
 Bandpass Filter (Cont.)
The composite transfer function is obtained by multiplying the transfer
functions of lowpass and highpass filters with the gain of the inverter; that is,
Vo  1  jC2 R  R f  R 1 jC2 R
H( )  
  

       f
Vi  1  jC1 R  1  jC2 R  Ri  Ri 1  jC1 R 1  jC2 R
The lowpass section sets the upper corner frequency 2 and the highpass
section sets the lower corner frequency 1.
1 1 0
2  , 1  ,  0  1 2 , B   2  1 , Q 
RC1 RC2 B
To find the passband gain K ,
Rf j / 1 Rf j2
H( )   
Ri (1  j / 1 )(1  j / 2 ) Ri (1  j )(2  j )
At the center frequency 0  12 ,
Rf j02 R f 2 R f 2
H(0 )    K
Ri (1  j0 )(2  j0 ) Ri 1  2 Ri 1  2
Passband gain 83
 Bandreject Filter
A bandreject filter may be constructed by parallel combination of
a lowpass filter and a highpass filter and a summing amplifier.

The circuit is designed such that the lower cutoff frequency 1

is set by the lowpass filter while the upper cutoff frequency 2
is set by the highpass filter. The gap between 1 and 2 is the
bandwidth of the filter.
84
Bandreject Filter (Cont.)
Vo Rf  1 jC2 R 
H ( )      
Vi Ri  1  jC1 R 1  jC2 R 
Rf  1 j /  2 
H ( )    
Ri  1  j / 1 1  j / 2 
R f (1  j 2 / 2  ( j ) 2 / 12 )

Ri (1  j / 1 )(1  j / 2 )
Rf
K
Ri
At the center frequency 0  12 ,
R f (1  j 20 / 2  ( j0 ) 2 / 12 )
H (0 ) 
Ri (1  j0 / 1 )(1  j0 / 2 )
Rf 21

Ri 1  2
85
 Ex. 14.12
Design a lowpass active filter with a dc gain of 4 and a corner
frequency of 500 Hz.
Sol:
1
c  2f c  2 (500 ) 
Rf C f
Rf
H (0)    4
Ri
If C f  0.2F, then
1
Rf   1.59 k
2 (500 )0.2 10 6

Rf
Ri   397 .5 
4
86
 Ex. 14.13
Design a bandpass filter in the form of Fig. 14.45 to pass
frequencies between 250 Hz and 3,000 Hz and with K = 10.
Select R = 20 k.
Sol:
1 1
C2  
R1 R 2f1
1
  31.83 nF
20 10  2  250
3

1 1 1
C1     2.65 nF
R2 R 2f 2 20 10  2  3000
3

Rf 1  2 f f 10(3250 )
K K 1 2   10.83
Ri 2 f2 3000
If Ri  10 k, then R f  108 .3 k.
87
 14.9 Scaling
In designing and analyzing filters and resonant
circuits or in circuit analysis in general, it is
sometimes convenient to work with element
values of 1  , 1 H, or 1 F, and then transform the
values to realistic values by scaling.
There are two ways of scaling a circuit:
magnitude or impedance scaling, and frequency
scaling.

88
Element Values for Butterworth LPF

89
 Magnitude Scaling
Magnitude scaling is the process of increasing all impedances in
a network by a factor, the frequency response remaining
unchanged.
The impedances of individual elements R, L, and C are
1
Z R  R, Z L  jL, Z C 
jC
Multiply the impedance of each circuit element by a factor Km
and let the frequency remain constant,

Z R  K m Z R  K m R, Z L  K m Z L  jK m L,
1
Z C  K m Z C 
jC / K m

90
 Magnitude Scaling (Cont.)
Thus, in magnitude scaling, the new values of the elements
and frequency are
C
R  K m R, L  K m L, C   ,   
Km
The primed variables are the new values and the unprimed
variables are the old values.
1 1 1
0     0
LC  K m LC / K m LC
The resonant frequency has not changed. Similarly, the quality
factor and the bandwidth are not affected by magnitude scaling.

91
 Frequency Scaling
Frequency scaling is the process of shifting the frequency
response of a network up or down the frequency axis while
leaving the impedance the same.

We achieve frequency scaling by multiplying the frequency

by a factor Kf while keeping the impedance the same.
Since the impedances of L and C are frequency-dependent, if
we apply frequency scaling to ZL() and Zc(), we obtain
L
Z L  j (K f ) L  jL  L 
Kf
1 1 C
ZC    C 
j (K f )C  jC Kf

92
 Frequency Scaling (Cont.)
The value of R is not affected, since its impedance does not
depend on frequency. Thus, in frequency scaling, the new
values of the elements and frequency are

L C
R   R , L  
, C  ,   K f 
Kf Kf

For the resonant frequency and the bandwidth,

1 1 Kf
0     K f 0
LC  ( L / K f )(C / K f ) LC
B  K f B

93
 Magnitude and Frequency Scaling
If a circuit is scaled in magnitude and frequency at the
same time, then

Km 1
R  K m R, L  L, C   C
Kf Km K f
  K f 

We set Km = 1 when there is no magnitude scaling

or Kf = 1 when there is no frequency scaling.

94
 Ex. 14.14
A fourth-order Butterworth lowpass filters is shown in
Fig. 14.48(a). The filters is designed such that the cutoff
frequency c = 1 rad/s. Scale the circuit for a cutoff
frequency of 50 kHz using 10-k resistors.

Fig. 14.48(a)

95
 14.14 (Cont.)
Sol:
mH

c 2  50 10 3 R 10 10 3

Kf     10 , K m  
5
 10 4
c 1 R 1
Km 10 4 (1.848) Km 10 4 (0.765 )
L1  L1   58.82 mH, L2  L2   24.35 mH
Kf  10 5
Kf  10 5

C1 0.765 C2 1.848
C1    243 .5 pF, C     588 .2 pF
K m K f  10 K m K f  10
9 2 9

96
Ex. 14.14

97
 14.12 Applications
 Resonant circuits and filters are widely used, particularly in
electronics, power systems, and communications systems.
 For example, a Notch filter with a cutoff frequency at 60 Hz
may be used to eliminate the 60-Hz power line noise in
various communications electronics.
 Filtering of signals in communications systems is necessary
in order to select the desired signal from a host of others in
the same range and also to minimize the effects of noise and
interference on the desired signal.
 Applications: Radio Receiver, Touch-Tone Telephone, and
Crossover Network

98
 Radio Receiver
 Modern receivers use a frequency mixer circuit, which always produces the
same IF signal but retains the audio frequencies carried on the incoming signal.
A local oscillator produces an signal that is combined with the incoming wave
by the frequency mixer to produce an output signal that contains the difference
frequency of the two signals.
 If the resonant circuit is tuned to receive an 800-kHz incoming signal, the LO
must produce a 1255-kHz signal, so that the difference (1255-800 = 455 kHz)
of frequency is available at the output of the mixer. The original audio signal is
extracted in the detector stage. The detector removes the IF signal, leaving the
audio signal. The audio signal is amplified to drive the loudspeaker.

Block diagram of an AM radio receiver

99
 Ex. 14.17
The resonant or turner circuit of an AM radio is portrayed in
Fig. 14.63. Given that L = 1 H, what must be the range of
C to have the resonant frequency adjustable from one end of
the AM band to another? (AM: 540~1600 kHz)

Fig. 14.63

100
 Sol:
Ex. 14.17 (Cont.)
Since the resonant circuit is a parallel type,
1 1
0  2f 0  or C  2 2
LC 4 f 0 L
For the high end of the AM band, f0 = 1600 kHz, and the
corresponding C is
1
C1   9.9 nF
4 1600 10 10
2 2 6 6

For the low end of the AM band, f0 = 540 kHz, and the
corresponding C is
1
C2   86.9 nF
4  540 10 10
2 2 6 6

Thus, C must be an adjustable capacitor varying from 9.9 nF to 86.9 nF.

101
Touch-Tone Telephone
The keypad has 12 buttons arranged in four
rows and three columns. The arrangement
provides 12 distinct signals by using seven
tones divided into two groups: the low-
frequency group (697 to 941 Hz) and the
high-frequency group (1,209 to 1,477 Hz).
Pressing a button generates a sum of two
sinusoids corresponding to its unique pair of
frequencies. For example, pressing the
number 6 button generates sinusoidal tones
with frequencies 770 Hz and 1,477 Hz.
When a caller dials a telephone number, a
set of signals is transmitted to the telephone
office, where the touch-tone signals are
decoded by detecting the frequencies they
contain.

102
 Touch-Tone Telephone (Cont.)
The signals are first amplified and separated into their respective groups by
the lowpass and highpass filters. The limiters (L) are used to convert the
separated tones into square waves. The individual tones are identified using
seven bandpass filters, each filter passing one tone and rejecting other tones.
Each filter is followed by a detector (D). The outputs of the detectors provide
the required dc signals needed by the switching system to connect the caller to
the party being called.

Fig 14.65

103
 Ex. 14.18
Using the standard 600- resistor used in telephone circuits and
a series RLC circuit, design the bandpass filter BP2 (697~852Hz)
in Fig 14.65.
Sol:
Since BP2 passes frequencies 697 Hz to 852 Hz and is centered
at f0 = 770 Hz, its bandwidth is

B  2 ( f 2  f1 )  2 (852  697)  973.89 rad/s

R 600
L   0.616 H
B 973.89
1 1 1
C 2  2 2  2  69.36 nF
0 L 4 f 0 L 4  770  0.616
2

104
 Crossover Network
The crossover network couples an audio amplifier to woofer and tweeter
speakers, which consists of one highpass RC filter and one lowpass RL filter. It
routes frequencies higher than a prescribed crossover frequency fc to the
tweeter (high-frequency loudspeaker) and frequencies below fc into the woofer
(low-frequency loudspeaker). A woofer is designed to reproduce the lower part
of the frequency range, up to about 3 kHz. A tweeter can reproduce audio
frequencies from about 3 kHz to about 20 kHz. The two speaker types can be
combined to provide the optimum in frequency response.

Equivalent model
高音: high pass

低音: low pass

105
 Crossover Network (Cont.)

The transfer function of the highpass filter is

V1 jR1C
H1 ( )  
Vs 1  jR1C
The transfer function of the lowpass filter is given by
V2 R2
H 2 ( )  
Vs R2  jL
The values of R1, R2, L, and C may be selected such that the two filters
have the same cutoff frequency, known as the crossover frequency.
106
 Ex. 14.19
In the crossover network of Fig. 14.66, suppose each speaker
acts as a 6- resistance. Find C and L if the crossover frequency
is 2.5 KHz.
Sol:
Fig. 14.66

1 1 1
c  2f c  C    10.61 F
R1C 2f c R1 2  2.5 10  6
3

R2 R2 6
c  2f c  L   382 H
L 2f c 2  2.5 10 3

107
108