Lecture # 7 Active Filter Circuits

BASS TREBLE
 Introduction: Categories
Filters

Digital Filters Analog Filters

Analog Continuous Analog sampled

Filters data Filters

Passive Filters Active Filters Switched Switched

(RC & RLC) (R + C + Op-Amp) Capacitor Filters Current Filters

2
Introduction: Definitions & Applications
A filter is a device that passes electric signals at certain frequencies or frequency ranges while
preventing the passage of others.
Filter Circuits Applications:
Filters are basic building blocks in all Communication Systems such as RF, IF and LF filters
❖ Low Frequency band-pass filters are used in the audio frequency range (0 kHz to 20 kHz) for
modems and speech processing.
❖ High-frequency band-pass filters (several hundred MHz) are used for channel selection in
telephone central offices.
❖ Data acquisition systems usually require anti-aliasing low-pass filters as well as low-pass noise
filters in their preceding signal conditioning stages.
❖ System power supplies often use band-rejection filters to suppress the 60 (or 50)-Hz line
frequency and high frequency transients.
❖ In addition, there are filters that do not filter any frequencies of a complex input signal, but just add
a linear phase shift to each frequency component, thus contributing to a constant time delay. These
are called all-pass filters.
3
 Introduction
Passive filter Active filter

❖ Filter circuits consist of R,L,C ❖ Filter circuits consist of op-amps

❖ Inductors are usually large, heavy and ❖ Filter circuits without using inductor
costly, and may introduce electromagnetic
field effects → disadvantage

❖ Maximum magnitude does not exceed 1 ❖ Filter circuits provide a control over
amplification.
❖ Cut-off frequency and passband magnitude
were altered with the addition of a resistive ❖ Does not affected → use active filter circuits
load at the output of the filter. to implement filter designs when gain, load
variation and physical size are important
parameters in the design specifications.

4
Introduction: Ideal Filter Response Characteristics
|H| |H|

|H| |H|

Figure: Ideal transmission characteristics of the four major filter types: (a) low-pass (LP), (b) high-
pass (HP), (c) bandpass (BP), and (d) bandstop (BS).
5
 Introduction: Filter Specification
• Passband edge: p Stopband edge : s
• Maximum allowed variation in passband transmission : Amax
• Minimum required stopband attenuation : Amin
|H|, |H|,


Fig.: Spec. of the transmission characteristics of a LPF Fig.: Transmission specifications for a BPF

• The first step of filter design is to determine the filter specifications.

• Then find a transfer function H(s) whose magnitude meets the specifications.
• The process of obtaining a transfer function that meets given specifications is called filter approximation.
6
Introduction
Filter Transfer Function
✓ A filter is a linear 2-port network represented by the ratio of the output to input voltage.
✓ Transfer function H(s) = Vo(s) / Vi(s).
✓ Transmission: evaluating H(s) for physical frequency s = j → H(j) = |H(j)|.ej().
Gain function: G()  20 log|H(j)| (dB)
Attenuation function: A()  - 20 log|H(j)| (dB)
✓ Output frequency spectrum : |Vo(s)| = |H(s)||Vi(s)|.

Figure: The filters studied in this chapter are linear circuits represented by the general
two-port network shown. The filter transfer function H(s) ≡ Vo(s) ⁄ Vi(s).

7
Introduction: FILTER TRANSFER FUNCTION
am s m + am −1s m −1 + ... + ao
❖ The filter TF is expressed as the ratio of 2 polynomials: H (s) =
an s n + an −1s n −1 + ... + bo
✓ The degree of the denominator → filter order.
✓ To ensure the stability of the filter → n ≥ m
✓ The coefficients ai and bj are real numbers.
am ( s − z1 )( s − z2 ) ... ( s − zm )
❖ The TF can be factored and expressed as: H (s) =
( s − p1 )( s − p2 ) ... ( s − pn )
✓ Zeros: z1, z2, …, zm and (n - m) zeros at infinity.
✓ Poles: p1, p2,…, pn
✓ Zeros and poles can be either a real or a complex number.
✓ Complex zeros and poles must occur in conjugate pairs.

Poles values of s that cause the denominator to become zero (roots of the denominator polynomial).
Zeros values of s that cause the numerator to become zero.

8
 |H|

Introduction: TF Examples
➢ Low-Pass Filter (with Stopband Ripple):
a4 ( s 2 + l21 )( s 2 + l22 )
H (s) =
s 5 + b4 s 4 + b3 s 3 + b2 s 2 + b1s + b0
|H|

➢ Low-Pass Filter (without Stopband Ripple):

am
H (s) =
s 5 + b4 s 4 + b3 s 3 + b2 s 2 + b1s + b0

|H|
➢ Band-Pass Filter:
a5 s ( s 2 + l21 )( s 2 + l22 )
H (s) =
s 6 + b5 s 5 + b4 s 4 + b3 s 3 + b2 s 2 + b1s + b0

9
 Fundamentals of Low-Pass Filters
The most simple low-pass filter is the passive RC low-pass network.
Its transfer function is:
1
1
H (s) = sC = where the complex frequency variable s = j𝜔
R+
1 1 + sRC Fig.: 1st-Order RC Low-Pass
sC
For a normalized presentation of the transfer function, s is referred to the filter’s corner (cut-off)
frequency, or –3 dB frequency, ωC, and has these relationships
s j  f 
s= = = j   = j
C C  fC 
With the corner frequency of the low-pass in Fig. being fC = 1/2πRC, s becomes s = sRC and the
transfer function H(s) results in:
1 1
𝐻 𝑠 = ; The magnitude of the gain response is: H =
1+𝑠 1+ ()
2

For frequencies  >> 1, the rolloff is 20 dB/decade.

10
 First-order low-pass filter
(or Single-pole active low-pass filter)

1st-Order Noninverting Low-Pass Filter 1st-Order Noninverting Low-Pass Filter with Amplification
𝑅2
The amplifier is configured as a voltage- The DC voltage gain for the filter: 𝐷𝐶 𝑔𝑎𝑖𝑛 = 1 +
𝑅1
follower (Buffer) giving it a DC gain: Av = +1
as opposed to the previous passive RC filter The voltage gain:
Vout AF AF = the pass band gain of the filter,
which has a DC gain of less than unity. Av = =
Vin  f 
2
(1 + R2/R1)
1+   ƒ = the freq. of the input signal in Hz
 fC  ƒC = the cut-off freq. in Hz

11
 First-order low-pass filter
(or Single-pole active low-pass filter)

Transfer function of the low-pass filter:

Zf R2 || (1 / sC )
H (s ) = − =−
Zi R1
c
H (s) = −A
R2 1
A= c =
s + c R1 R2 C
First order Inverting Low-passfilter
The gain in the passband, A, is set by the ratio R2/R1
Feedback path

The transfer function has the same form as for

passive low-pass filter except the gain A in the
pass-pand.

A general op amp circuit.

12
 Review The Decibel
pout
Number of decibels = 10log 10
pin
vout iout
Number of decibels = 20log 10 = 20log 10
vin iin
Bode plots differ from the frequency response plots in two important ways:
✓ First, instead of using a linear axis for the frequency values, a Bode plot uses a logarithmic
axis ➔ plot a wider range of frequencies of interest.
✓ Second, instead of plotting the absolute magnitude of the transfer function versus
frequency, the Bode magnitude is plotted in decibels (dB) versus the log of the frequency.
AdB = 20log10|H(j)|

When AdB = 0, the transfer function magnitude is 1, since 20log10(1) = 0.

When AdB< 0, the transfer function magnitude is between 0 and 1, and
when, AdB > 0, the transfer function magnitude is greater than 1.

13
 Review Logarithmic axis
A logarithmic scale (or log scale) is a way of displaying
numerical data over a very wide range of values in a compact
way. As opposed to a linear number line in which every unit of
distance corresponds to adding by the same amount, on a
logarithmic scale, every unit of length corresponds to
multiplying the previous value by the same amount. Hence,
such a scale is nonlinear.

Fig. (a) showing log scale on x axis.

Fig. (c) Semilog plane showing
log scale on the x-axis and linear
scale on the y-axis.

Fig. (b) frequency is showing log scale on x axis.

14
 Review Bode Plots
Vout 1
=
Vin 1 + j
C

Vout
➢ If  << C = 1 = 0 dB
Vin

Vout C   ( C )
➢ If  >> C = = −20 log  
Vin   C 
➢ If  = C Vout 1
=
Vin 1 + j
Vout 1
= = −3 dB
Vin 2

15
Review Bode Plots
Vout 1
=
Vin 1 + j
C
Vout
➢ If  << C = 1 = 0 dB
Vin
(/C)
Vout C  
➢ If  >> C = = −20 log  
Vin  
 C
(/C) Vout 1
➢ If  = C =
Vin 1 + j
Vout 1
= = −3 dB
Vin 2

−1
𝜔
(/C) 𝜃 = − tan ( )
𝜔𝐶
16
 Review Bode Plots
1
𝐻 𝜔 =
(1 + 𝑗𝜔/𝜔1 )(1 + 𝑗𝜔/𝜔2 )

→ log(|H()|) = - log(|1 + j/1|) - log(|1 + j/2|)

The first term on the right is zero for  << 1

The second term on the right is zero for  << 2

The magnitude of H() is zero for  << 1, then decreases by 20 dB per
decade after 1. After 2 the second term on the right kicks in and the
magnitude decreases by another 20 dB per decade to 40 dB per decade.

17
Review Bode Plots 1
𝐻 𝜔 =
(1 + 𝑗𝜔/𝜔1 )(1 + 𝑗𝜔/𝜔2 )
1 2
−20log(/2)

20log(/1)
−20log(/1) − 20log(/2)

dB
1 2 The bode plots for
101 102
each factor add to
give the bode plot
for the function.

18
Example
Designing a low-pass Op Amp Filter

Using the circuit shown in Fig., calculate values for C and R2 that, together with
R1 = 1Ω, produce a low-pass filter having a gain of 1 in the passband and a
cutoff frequency of 1 rad/s. Construct the transfer function for this filter and use
it to sketch a Bode magnitude plot of the filter’s frequency response.

Solution
Calculate the required value of R2:
R2 = AR1 = 1 × 1 = 1Ω
Calculate C to meet the specified cutoff frequency
1 1
C= = = 1 (F )
R2C 1 1

The transfer function for the low-pass filter is

C −1
H (s) = − A =
s + C s + 1

19
 Example
Designing a low-pass Op Amp Filter

The Bode plot of |H(j)| is shown in

Fig. This is the so-called prototype
low-pass op amp filter, because it uses
a resistor value of 1 Ω and a capacitor
value of 1 F, and it provides a cutoff
frequency of 1 rad/s.
>> w=0.1:.1:10;
>> h=20*log10(abs(1./(1+j*w)));
>> semilogx(w,h);
>> grid on;
>> xlabel('\omega(rad/s)');
>> ylabel('|H(j\omega)|dB');

AdB = 20log10|H(j)|
20
First-order high-pass filter

Transfer function of the low-pass filter:

Zf R2
H (s) = − =−
Zi R1 + (1/ sC )
s
H (s) = −A
s + c

It also has the same form as 1

R2 c =
passive high-pass filter, except for A=
R1 R1C
the gain.

21
First-order high-pass filter
Example
Designing a high-pass Op Amp Filter
We have the Bode magnitude plot of a
high-pass filter. Calculate values of R1
and R2 that produce the desired
magnitude response.
Given C = 0.1 μF.

22
Example
Designing a high-pass Op Amp Filter
Solution:
Note that the gain in the passband is 20 dB; therefore,
A = 10. Also note that the 3 dB point is 500 rad/s.
The transfer function that has the magnitude response
shown in the Bode plot is given by
−10 s
H (s) =
s + 500
so −10s − ( R2 / R 1 ) s
H (s) = =
s + 500 s + (1/ R1C )

Equating the numerators and denominators and then

simplifying we get two equations:
R 1
10 = 2 , 500 =
R1 R1C
 R1 = 20 k , R2 = 200 k .

23
 Example
Designing a high-pass Op Amp Filter Solution:

>> w=1:10000; The circuit is shown in Fig.

>> h=20*log10(10*(abs((j*w/500)./(1+j*w/500))));
>> semilogx(w,h);
>> grid on;
>> xlabel('\omega(rad/s)');
>> ylabel('|H(j\omega)|dB');

Because we have made the assumption that the op amp in this high-pass filter
circuit is ideal, the addition of any load resistor, regardless of its resistance, has no
effect on the behavior of the op amp. Thus, the magnitude response of a high-pass
filter with a load resistor is the same as that of a high-pass filter with no load
resistor.
24
Problem 1
Designing a high-pass Op Amp Filter
Compute the values for R2 and C that yield a high-pass filter with a passband
gain of 1 and a cutoff frequency of 1 rad/s if R1 is 1. (Note: This is the
prototype high-pass filter.)

25
Problem 1
Designing a high-pass Op Amp Filter

Solution:
− ( R2 / R1 ) s
H (s) =
s + (1/ R1C )
1
= 1 rad / s; R1 = 1 , C = 1 F
R1C
R2
= 1; R2 = R1 = 1 
R1
−s
H prototype ( s ) =
s +1

26
Problem 2
Designing a low-pass Op Amp Filter

Compute the resistor values needed for the low-pass filter circuit in
Fig. to produce transfer function
− 20000
H (s) =
s + 5000
Assume that C = 5 F

27
Problem 2
Designing a low-pass Op Amp Filter

Solution:
− (1/ R1C ) −20000
H (s) = =
s + (1/ R2C ) s + 5000
1
= 20000; C = 5 F
R1C
1
 R1 = −6
= 10 ()
20000  (5 10 )
1 1
= 5000  R2 = −6
= 40 ()
R2C 5000  (5 10 )

28
Scaling

The designer can transform the convenient values into realistic values using the
process known as scaling.
There are two types of scaling magnitude and frequency.
Magnitude scaling: Scale a circuit in magnitude by multiplying the impedance at a
given frequency by the scale factor km. Thus we multiply all resistors and inductors
by km and all capacitors by 1/km.
R’ = kmR; L’ = kmL; C’ = C/km
Frequency scaling: we change the circuit parameters so that at the new frequency,
the impedance of each element is the same as it was at the original frequency. If we
let kf denote the frequency scale factor. Thus for frequency scaling:
R’ = R; L’ = L/kf; C’ = C/kf

29
Scaling
A circuit can be scaled simultaneously in both magnitude and frequency. The
scaled values (primed) in terms of (unprimed) are
km 1
R  = k m R; L  = L; C  = C
kf km k f

Example: Scaling a Series RLC passive Circuit

30
Scaling

Solution:
Let computing the frequency scale factor that will shift the center frequency from 1
rad/s to 500 Hz. The unprimed variables represent values before scaling, whereas the
primed variables represent values after scaling
o 2 (500)
kf = = = 3141.59
o 1
1 C 1
The magnitude scale factor: km = = = 159.155
k f C  (3141.59)(2 10 )
−6

R = km R = 159.155 (),
km
L = L = 50.66 (mH )
kf

31
Scaling
Problem 3
What magnitude and frequency scale factor will transform the prototype
high-pass filter into a high pass filter with a 0.5 pF capacitor and a cutoff
frequency of 10 kHz?

32
Scaling

Solution:
c = 2 f c = 2 104 = 20000 ( rad / s)
 k f = 20000 = 62831.85
C 1
C =  0.5  10 = −6

k f km km k f
1
 km = = 31.83
( 0.5 10 ) ( 62831.85 )
−6

33
 First-order band-pass filter
The band-pass filter consists of 3 separate components:
1. ωc2 – cut-off frequency of unity-gain low-pass filter
2. ωc1 – cut-off frequency of unity-gain high-pass filter
3. Gain component to provide the desired level of gain in the
passband.

These three components are cascaded in series. So:

Requirement for two cut-off frequencies in broadband
bandpass filter:
c 2
2
 c1

Bode magnitude plot of a bandpass filter

34
First-order band-pass filter
Construct the bandpass filter
We can construct a circuit that provides each of the three components by cascading a low-pass
op amp filter, a high-pass op amp filter, and an inverting amplifier, as shown in Fig. below.

Transfer function of the band-pass filter is the product of the transfer functions of the three
cascaded components:
Vo  −c 2  − s  − R f  − Ac 2 s 1
H (s) = = 
Rf 1
   = A =  =  =
Vi  s + c 2  s + c1  Ri  ( s + c1 )( s + c 2 )
c2 c1
Ri RL C L RH C H

Standard form for the transfer function of a bandpass filter

s
H ( s) = Converting into the form of the standard transfer function for a
s 2 +  s + o2 bandpass filter → assumption c2 >> c1
35
First-order band-pass filter

The bandpass filter circuit

Then, (c2 + c1)  c2

− Ac 2 s
➔ the transfer function for the cascaded bandpass filter: H ( s ) = 2
s + c 2 s + c1c 2
Compute the values of RL and CL in the low-pass filter to give us the desired upper cutoff
frequency, c2:
1
c 2 =
RLCL
36
First-order band-pass filter

Compute the values of RH and CH in the high-pass filter to give us the desired lower cutoff
frequency, c1: 1
c1 =
RH CH

Compute the values of Ri and Rf in the inverting amplifier to provide the desired passband gain. To
do this, we consider the magnitude of the bandpass filter's transfer function, evaluated at the center
frequency o
− Ac 2 ( jo ) Ac 2
H ( jo ) = = =A
( jo ) + c 2 ( jo ) + c 2c1
2
c 2

Rf
Besides, the gain of the inverting amplifier H ( jo ) = (*)
Ri

Any choice of resistors that satisfies Eq. (*) will produce the desired passband gain.

37
First-order band-pass filter
Example
Design a bandpass filter for a graphic
equalizer to provide an amplification of
2 within the band of frequencies between
100 and 10000 Hz. Use 0.2 µF
capacitors.

Solution
We design each subcircuit in the cascade and meet the specified cutoff frequency values. In
this case c2 = 100c1 so we can say that c2 >> c1
1
Begin with the Iow-pass stage. c 2 = = 2 (10000),
RLCL
1
RL =  80 ()
2 10000  0.2 10 −6

38
First-order band-pass filter
Solution
Next, calculate for the high-pass stage
1 1
c1 = = 2 (100), RH =  7958 ()
RH CH 2 100  0.2 10 −6

For the gain stage: one of the resistors can be selected arbitrarily. Let's select a 1 kΩ resistor
for Ri . So with A = Rf/Ri ➔ Rf = ARi = 2(1000) = 2 kΩ
The resulting circuit is shown in Fig.

39
First-order band-reject filter
We can use a component approach to the design of op amp bandreject filters too.

The band-reject filter consists of 3 separate components:

1. ωc1 – cut-off frequency of unity-gain low-pass filter
2. ωc2 – cut-off frequency of unity-gain high-pass filter
3. Gain component to provide the desired level of gain in
the passband.

These 3 components cannot be cascaded

in series, but we have to use a parallel
connection and a summing amplifier to
construct band-reject filter.

Bode magnitude plot of a band-reject filter 40

 First-order band-reject filter
Construct the band-reject filter
Block diagram

The most important difference is A parallel op amp

that these three components bandreject filter
cannot be cascaded in series
because they do not combine
additively on the Bode plot.
Instead, we use a parallel
connection and a summing
amplifier
circuit

41
First-order band-reject filter
Assume that the two cutoff frequencies are widely separated → the resulting design is a broadband
bandreject filter, and c2 >> c1. Then each component of the parallel design car be created
independently, and the cutoff frequency specifications will be satisfied. The transfer function of the
resulting circuit is the sum of the low-pass and high-pass filter transfer functions

 R f   − c1 −s 
H (s ) =  −   + 
 Ri  s + c1 s + c2 

R f  s 2 + 2 c1s +  c1 c 2 
H (s ) =  
Ri  (s +  c1 )(s +  c 2 ) 


The cutoff frequencies are

1 1
c1 = c 2 =
RLC L RH C H

In the two passbands (as s → 0 and s → ∞), the gain of

the transfer function is Rf/Ri.
Therefore K = Rf/Ri
42
First-order band-reject filter
Example
Design a circuit based on the parallel band-reject op-amp
filter as in the figure. Use 0.5 µF capacitors for the
design.

Solution
From the Bode figure, we see that the bandreject filter has
cutoff frequencies of 100 rad/s and 2000 rad/s and a gain of 3
in the pass bands → c2 = 20c1
So, assume that: c2 >> c1
From low-pass filter model and use scaling to meet the
specifications for cutoff frequency and capacitor value. The
frequency scale factor kf = 100 → shifts the cutoff frequency
from 1 rad/s to 100 rad/s. The magnitude scale factor fm is
20000, which permits the use of a 0.5 F capacitor

43
First-order band-reject filter
Solution
RL = 20 k , CL = 0.5  F .
1 1
Cutoff frequency of the low pass filter c1 = = −6
= 100 (rad / s)
RLCL 20 10  0.5 10
3

Similarly, to design high pass filter, from the high-pass op amp filter model. We
have kf = 2000, and km = 1000, resulting in the following scaled component values
RH = 1 k , CH = 0.5  F .

Finally, because the cutoff frequencies are widely separated, we can use the ratio Rf/Ri to establish
the desired passband gain of 3.
Choose Ri = 1 kΩ, (as for RH). ➔ Rf = 3kΩ, and K = Rf/Ri = 3000/1000 = 3
The resulting parallel op amp bandreject filter circuit is shown in Fig.

44
First-order band-reject filter
Solution
 −c1 − j   R f 
H ( j ) =  + − 
 j + c1 j + c2   Ri 

AdB = 20 log10 H ( j )

>> f=10:80000;
>> w=2*pi*f;
>> H=(((-2*pi*100)./(j*w+2*pi*100))+((-j*w)./(j*w+2*pi*2000)))*(-5);
>> A=20*log10(abs(H));
>> semilogx(f,A);
>> grid on;
>> xlabel ('F (Hz)');
>> ylabel ('A_{dB}');

45
First-order band-reject filter
Solution

The resulting bandreject filter circuit design example

46
High Order Op Amp Filters

This figure shows the Bode magnitude plots of a

cascade of identical prototype low pass filters and
includes plots of just one filter, two in cascade,
three in cascade, and four in cascade.

The more poles filter has, the faster its roll-off rate.

As the order of the low-pass filter is increased by

adding prototype low-pass filters to the cascade, the
cutoff frequency also change

The Order (number of poles) of the filter

A pole (single pole) is simply one resistor & one capacitor.

47
High Order Op Amp Filters
➢ The number of poles determines the roll-off rate of the filter. For example, a
Butterworth response produces -20dB/decade/pole.
This means that:
✓ One-pole (first-order) filter has a roll-off of -20 dB/decade
✓ Two-pole (second-order) filter has a roll-off of -40 dB/decade
✓ Three-pole (third-order) filter has a roll-off of -60 dB/decade

48
High Order Op Amp Filters

➢ The number of filter poles can be increased by cascading. To obtain a filter

with three poles, cascade a two-pole with one-pole filters.

Three-pole (third-order) low-pass filter.

49
High Order Op Amp Filters

In general, an n-element cascade of identical low-pass filter will transition from the
passband to the stopband with a slope of 20n dB/decade. Both the block diagram and
the circuit diagram for such a cascade are shown in Fig.

50
High Order Op Amp Filters

The transfer function for a cascade of n prototype low-pass filters

The order of a filter is determined by the number of poles in its transfer function

51
High Order Op Amp Filters

For example, n = 4: c 4 = 4
2 − 1  0.435 rad / s

Thus we can design a fourth order low-pass filter with any arbitrary cutoff frequency by starting
with a 4th-order cascade consisting of prototype low-pass filters and then scaling the
components by kf = c/0.435 to place the cutoff frequency at any value of c desired.

Example: Designing a 4th-order low-pass op amp fitter with a cutoff frequency

of 500 Hz and a passband gain of 10. Use 1 F capacitors. Sketch the Bode
magnitude plot for this filter.

52
High Order Op Amp Filters
Example Solution

 rad 
We have already used Eq. c 4 = 4
2 − 1  0.435   to calculate the cutoff frequency for the resulting
 s 
fourth-order low-pass filter as 0.435 rad/s.
A frequency scale factor of kf = 7222.39 will scale the component values to give a 500 Hz cutoff
frequency. A magnitude scale factor of km = 138.46 permits the use of 1 F capacitors. The scaled
component values are thus: R = 138.46 ; C = 1 F
Finally, add an inverting amplifier stage with a gain of Rf/Ri = 10. As usual, we can arbitrarily select one
of the two resistor values. Because we are already using 138.46  resistors, let Ri = 138.46 ; then,
Rf = 10Ri = 1384.6 
The circuit for this cascaded the 4th-order low-pass filter is shown in Fig. (next slide).
It has the transfer function:

53
Fig.: The cascade circuit
for the fourth order low-
pass filter design in
example
54
>> f=10:1:1000;
>> w=2*pi*f; The Bode magnitude plot for this transfer function
>> wc=7222.39;
>> h=20*log10(10)-80*log10(abs(1+j*w/wc));
>> semilogx(f,h);
>> grid on;
>> xlabel(‘F[Hz]');
>> ylabel('|H(jf)| dB');

4
 
 −c 
4
 −1 
H (j ) = ( −10)   = ( −10)  
 j + c   1 + j 
 c 
 

j
AdB = 20 log10 (10) + 20 log10 1 +
c

55
 Filter response characteristics
➢ There are 3 characteristics of filter
response :
i) Butterworth characteristic
ii) Chebyshev characteristic
iii) Bessel characteristic.

➢ Each of the characteristics is

identified by the shape of the
response curve Comparative plots of three types of
filter response characteristics.

56
 Filter response characteristics
Butterworth characteristic

➢ Filter response is characterized by flat amplitude

response in the passband.

➢ Provides a roll-off rate of -20 dB/decade/pole.

➢ Filters with the Butterworth response are

normally used when all frequencies in the
passband must have the same gain.

57
 Filter response characteristics
Chebyshev characteristic

➢ Filter response is characterized by overshoot or ripples in

the passband.

➢ Provides a roll-off rate greater than -20 dB/decade/pole.

➢ Filters with the Chebyshev response can be implemented

with fewer poles and less complex circuitry for a given roll-
off rate.

58
 Filter response characteristics
Bessel characteristic

➢ Filter response is characterized by a linear

characteristic, meaning that the phase shift
increases linearly with frequency.
➢ Filters with the Bessel response are used
for filtering pulse waveforms without
distorting the shape of waveform.

59
 Damping factor (DF)
The damping factor (DF) of an active filter determines which response characteristic
the filter exhibits.
➢ This active filter consists of an amplifier, a
negative feedback circuit & RC circuit.
➢ The amplifier and feedback & connected in a
non-inverting configuration.
➢ DF is determined by the negative feedback
and defined as :
R1
DF = 2 −
R2

Fig.: General diagram of active filter

60
Butterworth Filters
A unity-gain Butterworlh low-pass filter has a transfer function whose magnitude is given by

(*)

where n is an integer that denotes the order of the filter

1. The cutoff frequency is c rad/s for all values of n.
2. If n is large enough, the denominator is always close to unity when  < c
3. In the expression for |H(j)|, the exponent of /c is always even.
Given an equation for the magnitude of the transfer function, how do we find H(s)?
Set c equal to 1 rad/s in Eq. (*)
We use scaling to transform the prototype filter to a filter that meets the given filtering
specifications. Note that s = j
|H(j)|2 = H(j)H(-j) = H(s)H(-s)

61
Butterworth Filters
1 1 1 1
H (j ) = = = =
2
With s2 = 2 ➔
1 +  2n 1 + ( 2 ) n 1 + (− s 2 ) n 1 + (−1) n s 2 n
1
H ( s) H (− s) =
1 + (−1) n s 2 n

The procedure for finding H(s) for a given value of n is as follows:

1. Find the roots of the polynomial
1 + (-1)ns2n = 0
2. Assign the left-half plane roots to H(s) and the right-half plane
roots to H(-s).
3. Combine terms in the denominator of H(s) to form 1st- and 2nd-
order factors.

62
Butterworth Filters
Example: Find the Butterworth transfer functions for n = 2 and n = 3.

Solution
For n = 2, we find the roots of the polynomial: 1 + (-1)2s4 = 0
Rearranging terms we find s 4 = −1 = 1180
Therefore, the four roots are:

s1 = 145 = 1/ 2 + j / 2 Roots s2 and s3 are in the left-half plane. So:

s2 = 1135 = −1/ 2 + j / 2 H (s) =
1
( s + 1/ 2 − j / 2)( s + 1/ 2 + j / 2)
s3 = 1225 = −1/ 2 + − j / 2
1
s4 = 1315 = 1/ 2 + − j / 2 = 2
s + s 2 +1
63
Butterworth Filters
Solution For n = 3, we find the roots of the polynomial 1 + (-1)3s6 = 0
s 6 = 10 = 1360
Therefore, the six roots are
s1 = 10 = 1 s4 = 1180 = −1 + j 0
s2 = 160 = 1/ 2 + j 3 / 2 s5 = 1240 = −1/ 2 + − j 3 / 2
s3 = 1120 = −1/ 2 + j 3 / 2 s6 = 1300 = 1/ 2 + − j 3 / 2

Roots s3, s4, and s5 are in the left half plane. Thus
1 1
H (s) = =
( s + 1)( s + 1/ 2 − j 3 / 2)( s + 1/ 2 + j 3 / 2) ( s + 1)( s + s + 1)
2

Note that in passing that the roots of the Butterworth polynomial are always equally
spaced around the unit circle in the s plane.
64
Butterworth Filters
Table 1 Normalized (so that c = 1 rad/s) Butterworth polynomials up to
8th order.
n Nth-Order Butterworth Polynomial
1 (s + 1)
2
3
4
5
6
7
8
To assist in the design of Butterworth filters, Table 1 lists the Butterworth
polynomials up to n = 8.

65
Butterworth Filters
The Butterworth polynomials in Table 1 are the product of 1st- and 2nd-order factors ➔ we
can construct a circuit whose transfer function has a Butterworth polynomial in its denominator
by cascading op amp circuits, each of which provides one of the needed factors. A block
diagram of such a cascade is shown in Fig. below, using a 5th-order Butterworth polynomial as
an example (a cascade of 1st-and 2nd-order circuits yielding a fifth-order low-pass Butterworth
filter).

All odd-order Butterworth polynomials include

the factor (s + 1) → all odd-order Butterworth
filter circuits must have a subcircuit that provides
the transfer function H(s) = 1/(s + 1). This is the
transfer function of the prototype low-pass op
amp filter
66
Butterworth Filters
Now, to find a circuit that provides a transfer function of the form
H(s) = 1/(s2 + b1s + 1)

The s-domain nodal equations at the noninverting

terminal of the op amp and at the node labeled Va:
Va − Vi V −V
+ (Va − Vo ) sC1 + a o = 0
R R
V −V
Vo sC2 + o a = 0 Fig.: circuit that provides the 2nd-order
R
transfer function for the Butterworth filter cascade

(2 + RC1s )Va − (1 + RC1s )Vo = Vi 2 + RC1s Vi

− Va + (1 + RC2 s )Vo = 0 −1 0 Vi
Vo = = 2
Using Cramer's rule 2 + RC1s −(1 + RC1s) R C1C2 s 2 + 2 RC2 s + 1
−1 1 + RC2 s

67
Butterworth Filters
The transfer function for the circuit:
1 1
Vo R 2C1C2 Finally, set R = 1 C1C2
H ( s) = = H ( s) =
Vi s 2 + 2 s + 1 2
s2 + s +
1
RC1 R 2C1C2 C1 C1C2

This Eq. required for the 2nd-order circuit in the Butterworth cascade or the form
1 2 1
H (s) = Choose capacitor values so that b1 = & 1=
s 2 + b1s + 1 C1 C1C2

The procedure for designing an nth-order Butterworth low-pass filter circuit with a cutoff
frequency of c = 1 rad/s and a gain of 1 in the passband: Use frequency scaling to calculate
revised capacitor values that yield any other cutoff frequency; and use magnitude scaling to
provide more realistic or practical component values in our design.
➔ We can cascade an inverting amplifier circuit to provide a gain other than 1 in the passband.

68
 Butterworth Filters
Example: Design a 4th-order Butterworth low-pass filter with a cutoff frequency of 500 Hz and
a passband gain of 10. Use as many 1 k resistors as possible. Compare the Bode magnitude plot
for this Butterworth filter with that of the identical cascade filter in previous example

Solution:
From the Table 1 (slide #65), we find that the 4th-order butterworth polynomial is
(s2 + 0.765s + 1)(s2 + 1.848s + 1)
Need a cascade of two 2nd-order filters to yield the 4th-order transfer function plus an inverting
amplifier circuit for the passband gain of 10. The circuit is shown in Fig. below

69
 Butterworth Filters
Solution
Let the first stage of the cascade implement the transfer function for the polynomial
(s2 + 0.765s+ 1).
The preceding values for C1, C2, C3, & C4
2 1 C1 = 2.61 F
Use Eq. b1 = & 1= yield a 4th-order Butterworth filter with a c
C1 C1C2 C2 = 0.38 F of 1 rad/s. kf = 3141.6 will move fc to 500 Hz.
Let the second stage of the cascade A magnitude scale factor of km = 1000 will
implement the transfer function for the permit the use of 1k resistor in place of 1
polynomial (s2 + 1.848s + 1). C3 = 1.08 F resistors.
C4 = 0.924 F

The resulting scaled component values:

R = 1 k; C1 = 831 nF; C2 = 121 nF; C3 = 344 nF; C4 = 294 nF
Finally, we need to specify the resistor values in the inverting amplifier stage to yield a
passband gain of 10. Let R1 = 1 k; then Rf = 10R1 = 10 k
70
Butterworth Filters

Figure compares the magnitude

responses of the 4th-order identical cascade
filter from previous Example and the
Butterworth filter we just designed. Note that
both filters provide a passband gain of 10 (20
dB) and a cutoff frequency of 500 Hz, but the
Butterworth filter is closer to an ideal low-pass
filter due to its flatter passband and steeper
rolloff at the cutoff frequency. Thus, the
Butterworth design is preferred over the
identical cascade design.

71
Narrow Bandpass & Bandreject Filters
So far, we mentioned only about broadband filters or low Q filters that constructed from simple
high-pass & low-pass filters.
This sort of broadband filter with the transfer functions for cascaded bandpass & parallel bandreject
fillers that have discrete real poles. The synthesis techniques work best for cutoff frequencies that
are widely separated and therefore yield the lowest quality factors.
Consider the transfer function that results
 −c  − s  sc 0.5 s
H (s) =    = =
 s + c  s + c  s 2
+ 2 c s + c
2
s 2
+  s +  2
c

This is standard form of the transfer function of a bandpass filter

The bandwidth and center frequency:
 = 2c   1
Q= o = c =
2 = 2  2c 2
o c

Thus with discrete real poles, the highest quality bandpass filter (or bandreject filter) we can
achieve has Q = 1/2.
72
 Narrow Bandpass & Bandreject Filters
To build active filters with high quality factor values, we need an op amp circuit that can produce
a transfer function with complex conjugate poles.
Va −Vo −Vo
@ inverting input = Va =
1/ sC R3 sR3C
Vi − Va Va − Vo V V
@ node a: = + a + a
R1 1/ sC 1/ sC R2
Vi = (1 + 2sR1C + R1 / R2 )Va − sR1CVo
−s
R1 R2
H (s) =
R1C
(transfer function V0 /Vi) Req = R1 || R2 =
2 1 R1 + R2
s2 + s+
R3C Req R3C 2
−K  s
The standard form of the transfer function for a bandpass filter, H (s) =
s 2 +  s + o2
2 1 1
= ; K = ; o2 =
R3C R1C Req R3C 2
73
Narrow Bandpass & Bandreject Filters

Consider the prototype version of the circuit

o = 1 rad / s & C = 1 F
R1 = Q / K
R2 = Q / (2Q 2 − K )
R3 = 2Q

Finally, use the scaling is used to specify practical values for the circuit
components.

74
Narrow Bandpass & Bandreject Filters

Example: design a bandpass filter, using the circuit in Fig. below which has a
center frequency of 3000 Hz, a quality factor of 10, and a passband gain of 2.
Use 0.01 F capacitors in your design. Compute the transfer function of your
circuit, and sketch a Bode plot of its magnitude response.

75
Narrow Bandpass & Bandreject Filters
Solution In the prototype circuit, we easily calculate:
R1 = 10 / 2 = 5, R1 = 26.5 k 
The scaling factor are
R2 = 10 / (200 − 2) = 10 /198, R2 = 268.0 
R3 = 2(10) = 20 k f = 6000 km = 108 / k f R3 = 106.1 k 

The transfer function:

−3770 s
H (s) = 2
s + 1885.0 s + 355 106
Bode plot of its magnitude response
76
Narrow Bandpass & Bandreject Filters
The parallel implementation of a bandreject filter that combines low-pass and high-pass filter
components with a summing amplifier
2(Va −  Vo ) a
@ node a (Va − Vi ) sC + (Va − Vo ) sC + =0
R
b
Va  2sCR + 2 − Vo  sCR + 2  = sCRVi (1)
Vb − Vi Vb − Vo
@ node b + + (Vb −  Vo ) 2 sC = 0
R R
Vb  2 + 2RCs  − Vo 1 + 2 RCs  = Vi (2)

Summing the currents away from the noninverting input A high-Q active bandreject filter aka. twin-T notch filter
terminal of the top op amp gives
Vo − Vb
( o a)
V − V sC + = 0  − sRCVa − Vb + ( sRC + 1)Vo = 0 (3)
R
To solve the set of 3 equations (1), (2), and (3): use Cramer's rule to solve for V0:

77
Narrow Bandpass & Bandreject Filters
2( RCs + 1) 0 sCRVi
0 2( RCs + 1) Vi

Vo =
− RCs −1 0 (
= 2 2 2
R 2C 2 s 2 + 1) Vi
2( RCs + 1) 0 −( RCs + 2 ) R C s + 4 RC (1 −  ) s + 1
0 2( RCs + 1) −(2 RCs + 1)
− RCs −1 RCs + 1

The transfer function: The standard form for the transfer function of a
 2 1 
 s +  bandreiect filter
H (s) =
Vo
=  R 2C 2  s 2 + o2
Vi  2 4(1 −  ) 1  H (s) = 2
s + + s +  s + o2
 RC R 2C 2 
Equating both Equ. 1 4(1 −  )
o2 =
2 2
,  = .
RC RC
In this circuit, we have three parameters (R, C, and ) and two design constraints (0 and β).
Thus one parameter is chosen arbitrarily. Usually, we select capacitor value.

78
Narrow Bandpass & Bandreject Filters
1  1
R= ,  = 1− = 1−
o C 4o 4Q

Example: Design a high-Q active

bandreject filter (based on the circuit in
Fig.) with a center frequency of 5000
rad/s and a bandwidth (β) of 1000
rad/s. Use 1F capacitors in your
design.

79
Narrow Bandpass & Bandreject Filters
Solution
In the bandreject prototype filter, 0 = 1 rad/s, R = 1 Ω, and C = 1 F.
With Q = 0/β = 5000/1000 = 5, we found that:
R = 1 / (0C) = 1 / (5000×10-6) = 200 Ω;
 = 1- β/40 = 0.95
Thus: we need resistors:
200  (R), 100  (R/2), 190  (R), and
10  [(1 - )R].

80
Narrow Bandpass & Bandreject Filters
Solution

Bode magnitude plot is:

81
Narrow Bandpass & Bandreject Filters
Problem4: Design an active bandpass filter (the figure below) with Q = 8, A = 5,
and 0 = 1000 rad/s. Use 1 F capacitors and specify the values of all resistors.

82
Narrow Bandpass & Bandreject Filters
Solution:  1 
− s
 R1C  A s
The transfer function: H (s) = = 2
 2   R + R  s +  s +  2

s2 +  s+
1 2
2 
o

 3   1 2 3 
R C R R R C

2 2
With  =  R3 =
R3C C 2 106
R3 = = 16 (k )
o 1000 125 1
= = = 125 ( rad / s )
Q 8
1 1 1
From A =  R1 = = = 1.6 (k )
R1C A C 5 125 110 −6

R1 + R2 1600 + R2
From o2 =  106 =
R1 R2 R3C 2 1600 R2 16000  (10 )
−6 2

83
Narrow Bandpass & Bandreject Filters
Solution:
Solving for R2: R2 =
(1600 + R2 )106
 246 R2 = 16000
256 10 5

 R2 = 65.04 (  )

84
Narrow Bandpass & Bandreject Filters
Problem 5:
Design an active unity-gain bandreject filter with 0 = 1000 rad/s and Q = 4.
Use 2 F capacitors, and specify the values of R and 

85
Narrow Bandpass & Bandreject Filters
Solution:
Transfer function, center frequency, and bandwidth:
s 2 + (1/ R 2C 2 ) s 2 + o2
H (s) = = 2

s2 + 
4 (1 −  )   1  s +  s +  2

s+ 2 2 
o

 RC  R C 
1 4 (1 −  )
o = ; =
RC RC
1 1
R= = = 500 ()
oC 1000  2 10 −6

o1000 4 (1 −  )
= = = 250  = 250
Q 4 RC
4 (1 −  ) = 250 RC = 250  500  2 10 −6 = 0.25
0.25
(1 −  ) = = 0.0625;  = 0.9375
4
86
Narrow Bandpass & Bandreject Filters
Exercise 1
Find the transfer function V0/Vi for the circuit shown in Fig. if Zf is the equivalent
impedance of the feedback circuit, Zi is the equivalent impedance of the input
circuit, and the operational amplifier is ideal.

87
Narrow Bandpass & Bandreject Filters
Exercise 1
Sol.: summing the currents at the inverting input node yields

0 − Vi 0 − Vo
+ =0
Zi Zf
Vo V
 =− i
Zf Zi
Vo Zf
 H (s) = =−
Vi Zi

88
Narrow Bandpass & Bandreject Filters
Exercise 2
a) Use the results of exercise 1 to find the transfer function of the
circuit showing fig.
b) What is the gain of the circuit as  → 0?
c) What is the gain of the circuit as  → ∞?
d) Do your answers to (b) and (c) make sense in terms of known
circuit behavior?

89
Narrow Bandpass & Bandreject Filters
Exercise 2 Sol.:
R2 (1/ sC2 ) R2
(a) Z f = =
R2 + (1/ sC2 ) R2C2 s + 1
1/ C2
=
s + 1/ R2C2
1/ C1
Likewise Zi =
s + 1/ R1C1

− (1/ C2 )  s + 1/ R1C1  C1  s + (1/ R1C1 ) 

H (s) = =−
 s + (1/ R2C2 )  (1/ C1 ) C2  s + (1/ R2C2 ) 

90
Narrow Bandpass & Bandreject Filters
Exercise 2
−C1  j + (1/ R1C1 )  C1  j C1
Sol.: ( b ) H ( j ) =   = −
C1 R2C2
= −
R2
(c) H ( j ) = −   = −
C2  j + (1/ R2C2 )  C2 R1C1 R1 C2  j C2

(d) as  → 0 the 2 capacitor branches become open and the circuit reduces to a resistive
inverting amplifier having a gain of –R2/R1.
as  → ∞ the 2 capacitor branches approach a short circuit and in this case we encounter an
indeterminate situation; namely vn → vi but vn = 0 because of the ideal op amp. At the same
time the gain of the ideal op amp is infinite so we have the indeterminate form 0 → ∞.
Although  = ∞ is indeterminate we can reason that for finite large values of , H(j) will
approach – C1/C2 in value. In other words, the circuit approaches a purely capacitive
inverting amplifier with a gain of
( −1/ jC2 ) / (1/ jC1 ) or − C1 / C2

91