8

!"# $%&'()*+&$(%
Filters are of two types—FIR and IIR. The type of filters which make use of feedback
connection to get the desired filter implementation are known as recursive filters. Their
impulse response is of infinite duration. So they are called IIR filters. The type of filters
which do not employ any kind of feedback connection are known as non-recursive filters.
Their impulse response is of finite duration. So they are called FIR filters. IIR filters are
designed by considering all the infinite samples of the impulse response. The impulse
response is obtained by taking inverse Fourier transform of ideal frequency response. There
are several techniques available for the design of digital filters having an infinite duration
unit impulse response. The popular methods for such filter design uses the technique of first
designing the digital filter in analog domain and then transforming the analog filter into an
equivalent digital filter because the analog filter design techniques are well developed. In this
chapter, we discuss various methods of transforming an analog filter into a digital filter and
methods of designing digital filters.

!", '-.*$'-/-%&01 2('1 &'3%02('/3&$(%

The system function describing an analog filter may be written as:
M
bk s k
Y ( s) k 0
H a (s) = = N
X (s )
ak s k
k 0

548
 Infinite-duration Impulse Response (IIR) Filters ! 549

where {ak} and {bk} are filter coefficients. The impulse response of these filter coefficients
is related to Ha(s) by the Laplace transform

H a (s) = h(t ) e st dt

The analog filter having the rational system function Ha(s) given above can also be described
by the linear constant coefficient differential equation.

N M
d k y(t ) d k x (t )
ak k
= bk
k 0 dt k 0 dt k

where x(t) is the input signal and y(t) is the output of the filter.
The above three equivalent characterizations of an analog filter leads to three
alternative methods for transforming the analog filter into digital domain. The restriction on
the design is that the filters should be realizable and stable.
For stability and causality of analog filter, the analog transfer function should satisfy
the following requirements:
1. The Ha(s) should be a rational function of s, and the coefficients of s should be real.
2. The poles should lie on the left half of s-plane.
3. The number of zeros should be less than or equal to the number of poles.
For stability and causality of digital filter, the digital transfer function should satisfy the
following requirements:
1. The H(z) should be a rational function of z and the coefficients of z should be real.
2. The poles should lie inside the unit circle in z-plane.
3. The number of zeros should be less than or equal to the number of poles.
We know that the analog filter with transfer function Ha(s) is stable if all its poles lie
in the left half of the s-plane. Consequently for the conversion technique to be effective, it
should possess the following desirable properties:
1. The imaginary axis in the s-plane should map into the unit circle in the z-plane.
Thus, there will be a direct relationship between the two frequency variables in the
two domains.
2. The left half of the s-plane should map into the interior of the unit circle centred at
the origin in z-plane. Thus, a stable analog filter will be converted to a stable
digital filter.
The physically realizable and stable IIR filter cannot have a linear phase. For a filter to
have a linear phase, the condition to be satisfied is h(n) = h(N – 1 – n) where N is the length
of the filter and the filter would have a mirror image pole outside the unit circle for every
pole inside the unit circle. This results in an unstable filter. As a result, a causal and stable
IIR filter cannot have linear phase. In the design of IIR filters, only the desired magnitude
550 ! Digital Signal Processing

response is specified and the phase response that is obtained from the design methodology is
accepted.
The comparison of digital and analog filters is given in Table 8.1.

!"#$%& '() !"#$%&'(")* "+* ,'-'.%/* %)0* 1)%/"-* 2'/.3&(

Digital filter Analog filter

1. It operates on digital samples (or sampled 1. It operates on analog signals (or actual
version) of the signal. signals).
2. It is governed (or defined) by linear 2. It is governed (or defined) by linear differ-
difference equations. ential equations.
3. It consists of adders, multipliers, and delay 3. It consists of electrical components like
elements implemented in digital logic resistors, capacitors, and inductors.
(either in hardware or software or both).
4. In digital filters, the filter coefficients are 4. In analog filters, the approximation problem
designed to satisfy the desired frequency is solved to satisfy the desired frequency
response. response.

!"#$%&$'()* +,* "-'-&$.* ,-.&(/)

1. The values of resistors, capacitors and inductors used in analog filters change with
temperature. Since the digital filters do not have these components, they have high
thermal stability.
2. In digital filters, the precision of the filter depends on the length (or size) of the
registers used to store the filter coefficients. Hence by increasing the register bit
length (in hardware) the performance characteristics of the filter like accuracy,
dynamic range, stability and frequency response tolerance, can be enhanced.
3. The digital filters are programmable. Hence the filter coefficients can be changed
any time to implement adaptive features.
4. A single filter can be used to process multiple signals by using the techniques of
multiplexing.

0-)$"#$%&$'()* +,* "-'-&$.* ,-.&(/)

1. The bandwidth of the discrete signal is limited by the sampling frequency. The
bandwidth of real discrete signal is half the sampling frequency.
2. The performance of the digital filter depends on the hardware (i.e., depends on the
bit length of the registers in the hardware) used to implement the filter.

123+/&$%&* ,($&4/()* +,* 115* ,-.&(/)

1. The physically realizable IIR filters do not have linear phase.
2. The IIR filter specifications include the desired characteristics for the magnitude
response only.
 Infinite-duration Impulse Response (IIR) Filters ! 551

!"4 )-0$5%1 (21 $$'1 2$6&-'1 781 399'(:$/3&$(%1 (21 )-'$;3&$;-0

The analog filter having the rational system function H(s) can also be described by the linear
constant coefficient differential equation.
N M
d k y(t ) d k x (t )
ak k
= bk
k 0 dt k 0 dt k

In this method of IIR filter design by approximation of derivatives, an analog filter is

converted into a digital filter by approximating the above differential equation into an
equivalent difference equation.
dy(t )
The backward difference formula is substituted for the derivative at time t = nT.
dt
Thus,
dy(t ) y(nT ) y(n 1) T
=
dt t nT T

or dy(t ) y(n) y(n 1)

=
dt t nT T

where T is the sampling interval and y(n) = y(nT).

The system function of an analog differentiator with an output dy(t)/dt is H(s) = s, and
the digital system which produces the output [y(n) – y(n – 1)]/T has the system function
H(z) = [1 – z–1]/T. Comparing these two, we can say that the frequency domain equivalent
dy(t ) y(n) y(n 1)
for the relationship = is:
dt t nT T
1
1 z
s=
T
Thus, this is the analog domain to digital domain transformation.
d 2 y (t )
Also, the second derivative can be replaced by the second backward difference:
dt 2

d 2 y(t ) d dy(t )
=
dt 2 t nT
dt dt t nT

[ y(nT ) y(nT T )] / T [ y(nT T) y( nT 2T )] / T

=
T
y(n) 2y(n 1) + y(n 2)
=
T2
552 ! Digital Signal Processing

The equivalent expression in frequency domain is:

1 2
2 1 2z +z
s =
T2

2
1
or 2 1 z
s =
T

The ith derivative of function y(t) results in the equivalent frequency domain relationship as:
i
1
i 1 z
s =
T

As a result, the digital filter’s system function H(z) can be obtained from the analog filter’s
system function Ha(s) by the method of approximation of the derivatives as:

H (z ) = H a (s ) s
1 z 1
T

The outcomes of the mapping of the z-plane from the s-plane are discussed below.

1 z 1 1
We have s= , i.e. z=
T 1 sT
Substituting s = j! in the expression for z, we have

1
z=
1 j T
1 T
= 2 2
+j 2
1+ T 1+ T2
Varying ! from – " to " the corresponding locus of points in the z-plane is a circle
with radius 1/2 and with centre at z = 1/2, as shown in Figure 8.1.
It can be observed that the mapping of the equation s = (1 – z–1)/T, takes the left half
plane of s-domain into the corresponding points inside the circle of radius 0.5 and centre at
z = 0.5. Also the right half of the s-plane is mapped outside the unit circle. Because of this,
this mapping results in a stable analog filter transformed into a stable digital filter. However,
since the location of poles in the z-domain are confined to smaller frequencies, this design
method can be used only for transforming analog low-pass filters and band pass filters which
are having smaller resonant frequencies. This means that neither a high-pass filter nor a
band-reject filter can be realized using this technique.
The forward difference can be substituted for the derivative instead of the backward
difference.
 Infinite-duration Impulse Response (IIR) Filters ! 553

*+,-./& '() 4%$$')-* "+* !5$/%)3* ')."* "5$/%)3* 67* .83* 6%9:;%&0* 0'++3&3)93* #3.8"0<

This provides
dy(t ) y( nT + T ) y(nT )
=
dt T
y( n + 1) y(n)
=
T
The transformation formula would be
z 1
s=
T
or z = 1 + sT
The mapping of the equation z = 1 + sT is shown in Figure 8.2. This results in a worse
situation than the backward difference substitution for the derivative. When s = j!, the
mapping of these points in the s-domain results in a straight line in the z-domain with
co-ordinates (zreal, zimag) = (1, !T). As a result of this, stable analog filters do not always
map into stable digital filters.
The limitations of the mapping methods discussed above can be overcome by using
more complex substitution for the derivatives. An Nth order difference is proposed for the
derivative, as shown
N
dy(t ) 1 y(nT + kT ) y(nT kT )
= ak
dt t nT T k 1
T

Here {ak} are a set of parameters selected so as to optimize the approximation. The transfor-
mation from the s-plane to the z-plane will be
N
1
s= ak ( z k z k)
T k 1
554 ! Digital Signal Processing

*+,-./& '(0 4%$$')-* "+* !5$/%)3* ')."* "5$/%)3* 67* .83* +"&;%&0* 0'++3&3)93* #3.8"0<

Thus, if we choose proper values for {ak}, then the j! axis can be mapped into the unit
circle and the left half of the s-plane can be mapped into points inside the unit in the z-plane.

EXAMPLE 8.1 Convert the analog low-pass filter specified by

2
H a (s) =
s +3
into a digital filter making use of the backward difference for the derivative.
Solution: We know that the mapping formula for the backward difference for the derivative
is given by
1
1 z
s=
T
2
For the given analog filter function H a (s) = , the corresponding digital filter function
s +3
is:

2
H ( z) = H a (s) 1
s=
1 z
(1 z 1 )
T +3
T
2T
=
1 z 1 + 3T

2 2
If T = 1 s, H (z ) = 1
= 1
1 z +3 4 z
 Infinite-duration Impulse Response (IIR) Filters ! 555

EXAMPLE 8.2 Making use of the backward difference for the derivative, convert the
analog filter function given below to a digital filter function.
4
H a (s) = 2
s +9

Solution: The mapping formula for the backward difference by the derivative is:
1
1 z
s=
T
Therefore, for the given Ha(s), the corresponding digital filter function is:

4
H ( z ) = H a ( s) 1 z 1 = 2
1
s
T 1 z
+9
T

4T 2
= 1
1 2z + z 2 + 9T 2
If T = 1 s, then
4 4
H (z ) = 1 2
=
1 2z +z +9 10 2z 1 + z 2

EXAMPLE 8.3 Convert the analog filter given below into a digital filter using the
backward difference for the derivative:
3
H a ( s) =
(s + 0.5) 2 + 16

Solution: For the given Ha(s), the system function of the corresponding digital filter is:

3
H ( z ) = H a (s ) 1 =
s
1 z
( s + 0.5) 2 + 16 1 z 1
T s=
T

3
= 2
1
1 z
+ 0.5 + 16
T

3T 2
=
[(1 + 0.5T ) z 1 ]2 + 16 T 2
556 ! Digital Signal Processing

3T 2
=
(1 + 0.5 T ) 2 + z 2
2(1 + 0.5 T ) z 1
+ 16 T 2
If T = 1 s, then
3 3
H ( z) = 2 1
= 1 2
2.25 + z 3z + 16 18.25 3z +z

!"< )-0$5%1 (21 $$'1 2$6&-'1 781 $/9*60-1 $%;3'$3%&1 &'3%02('/3&$(%

In this technique, the desired impulse response of the digital filter is obtained by uniformly
sampling the impulse response of the equivalent analog filter. The main idea behind this is to
preserve the frequency response characteristics of the analog filter. For the digital filter to
possess the frequency response characteristics of the corresponding analog filter, the
sampling period T should be sufficiently small (or the sampling frequency should be
sufficiently high) to minimize (or completely avoid) the effects of aliasing.
Let ha(t) = Impulse response of analog filter
T = Sampling period
h(n) = Impulse response of digital filter
For impulse invariant transformation,
h(n) = ha(t)|t = nT = ha(nT)
The Laplace transform of the analog filter impulse response ha(t) gives the transfer
function of analog filter.
# L[ha(t)] = Ha(s)
The transformation technique can be well understood by first considering a simple
distinct poles case for the analog filter’s system function as shown below.
N
Ai
H a (s) =
i 1
s pi

The impulse response ha(t) of the analog filter is obtained by taking the inverse Laplace
transform of the system function Ha(s).
N
# ha (t ) = L 1 [H a ( s)] = Ai e pit ua (t )
i 1

where ua(t) is the unit step function in the continuous-time case.

The impulse response h(n) of the equivalent digital filter is obtained by uniformly
sampling ha(t), i.e.,
N
h(n) = ha (nT ) = Ai e pi nT ua (nT )
i 1
 Infinite-duration Impulse Response (IIR) Filters ! 557

The system function of the digital system of above expression can be obtained by
taking z-transform, i.e.

n
H (z ) = h(n) z
n 0

Using the above equation for h(n), we have

N
H ( z) = A i e pi nT ua (nT ) z n

n 0 i 1

Interchanging the order of summation, we have

N
H ( z) = A i e pi nT ua (nT ) z n

i 1 n 0

N
Ai
=
i 1 1 e piT z 1

Comparing the above expressions for Ha(s) and H(z), we can say that the impulse invariant
transformation is accomplished by the mapping.

1 1
is tranformed to
s pi 1 e piT z 1

5(.$&-+%* 6(&7((%* $%$.+'* $%"* "-'-&$.* 3+.()

The above mapping shows that the analog pole at s = pi is mapped into a digital pole at
z = e pcT . Therefore, the analog poles and the digital poles are related by the relation.
z = esT
The general characteristic of the mapping z = esT can be obtained by substituting s = + j!
and expressing the complex variable z in polar form as z = re j .
# re j = e( + j )T
= e Tej T

That means
T
z =r =e
and z= = T
So the relationship between analog frequency ! and digital frequency is

= T or = .
T
558 ! Digital Signal Processing

As a result of this, < 0 implies that 0 < r < 1 and > 0 implies that r > 1 and
= 0 implies that r = 1. Therefore, the left half of s-plane is mapped into the interior of the
unit circle in the z-plane. The right half of the s-plane is mapped into the exterior of the unit
circle in the z-plane. This is one of the desirable properties for stability. The j!-axis is
mapped into the unit circle in z-plane. However, the mapping of j!-axis is not one-to-one.
The mapping = !T implies that the strip of width 2 /T in the s-plane for values of
s in the range – /T $% !% $% /T maps into the corresponding values of – $% % $% , i.e., into
the entire z-plane. Similarly, the strip of width 2 /T in the s-plane for values of s in the
range /T $% !% $% 3 /T also maps into the interval – $% % $% , i.e., into the entire z-plane.
Similarly, the strip of width 2 /T in the s-plane for values of s in the range – /T $%!%$%–3 /
T also maps into the interval – $% % $% , i.e., into the entire z-plane. In general, any
frequency interval (2k – 1) /T $ ! $ (2k + 1) /T, where k is an integer, will also map into
the interval – $% % $% in the z-plane, i.e., into the entire z-plane. Hence the mapping from
the analog frequency ! to the digital frequency by impulse invariant transformation is
many-to-one which simply reflects the effects of aliasing due to sampling of the impulse
response. Figure 8.3 illustrates the mapping from the s-plane to z-plane.

(a) (b)
*+,-./& '(1 4%$$')-* "+* =%>* !5$/%)3* ')."* =6>* "5$/%)3* 67* '#$?/(3* ')@%&'%).* .&%)(+"&#%.'")<

The stability of a filter (or system) is related to the location of the poles. For a stable
analog filter the poles should lie on the left half of the s-plane. That means for a stable
digital filter the poles should lie inside the unit circle in the z-plane.

8)(,4.* -234.)(* -%#$/-$%&* &/$%),+/2$&-+%)

Some of the useful impulse invariant transformations are given below. The first one can be
used when the analog real pole has a multiplicity of m. The second and third equations can
be used when the analog poles are complex conjugate.

1 ( 1)m 1 d m 1 1
1. m 1 1
; s = pi
(s + pi )m (m 1)! ds 1 e sT z
 Infinite-duration Impulse Response (IIR) Filters ! 559

s+a 1 e aT (cos bT ) z 1
2.
( s + a) 2 + b 2 1 2e aT (cos bT ) z 1 + e 2 aT z 2

b e aT (sin bT ) z 1
3.
( s + a) 2 + b 2 1 2e aT (cos bT ) z 1 + e 2 aT
z 2

EXAMPLE 8.4 For the analog transfer function

2
H a (s) =
( s + 1) ( s + 3)
determine H(z) if (a) T = 1 s and (b) T = 0.5 s using impulse invariant method.

2
Solution: Given, H a (s) =
( s + 1)( s + 3)
Using partial fractions, Ha(s) can be expressed as:

A B
H a (s) = +
s +1 s +3

2
A = (s + 1) H a ( s) s 1
= =1
s+3 s 1

2
B = (s + 3) H a (s) s 3
= = 1
s +1 s 3

# 1 1 1 1
H a (s) = =
s +1 s +3 s ( 1) s ( 3)

By impulse invariant transformation, we know that

Ai Ai
(is transformed to) pi T 1
s pi 1 e z
Here Ha(s) has two poles and p1 = –1 and p2 = –3.
Therefore, the system function of the digital filter is:
1 1
H ( z) = p1T
1 e z 1
1 e p2T z 1

1 1
= 1 3T 1
1 e Tz 1 e z
560 ! Digital Signal Processing

(a) When T = 1 s
1 1
H (z ) =
1 e 1z 1
1 e 3z 1

1 1
= 1 1
1 0.3678 z 1 0.0497 z

(1 0.0497 z 1 ) (1 0.3678 z 1 )
=
(1 0.3678 z 1 )(1 0.0497 z 1 )

0.3181 z 1
=
1 0.4175 z 1 + 0.0182 z 2

(b) When T = 0.5 s

1 1
H (z ) = 0.5 1 3 0.5 1
1 e z 1 e z

1 1
= 1 1
1 0.606 z 1 0.223 z

(1 0.223 z 1 ) (1 0.606 z 1 )
=
(1 0.606 z 1 )(1 0.223 z 1 )

0.383 z 1
=
1 0.829 z 1 + 0.135 z 2

EXAMPLE 8.5 Convert the analog filter with transfer function

s + 0.1
H a (s ) =
(s + 0.1)2 + 9

into a digital filter using the impulse invariant transformation.

Solution: Observe that the given system function of the analog filter is of the standard
s+a
form H a (s) = , where we are given a = 0.1 and b = 3.
( s + a) 2 + b2
By the impulse invariant transformation, we know that

s+a 1 e aT (cos bT ) z 1
(s + a) 2 + b2 (is transformed to)
1 2e aT (cos bT ) z 1 + e 2 aT z 2
 Infinite-duration Impulse Response (IIR) Filters ! 561

Therefore, for the given Ha(s), we can write the system function of the digital filter
1 e 0.1T (cos 3T ) z 1
H (z ) =
1 2 e 0.1T (cos 3T ) z 1 + e 2(0.1)T
z 2

Assuming T = 1 s, we have

1 e 0.1 (cos 3) z 1
H ( z) =
1 2 e 0.1 (cos 3) z 1 + e 0.2 z 2

1
1 0.9048 ( 0.9899) z
= 1 2
1 2(0.9048) ( 0.9899) z + 0.8187 z
1 + 0.8956 z 1
=
1 + 1.7913 z 1 + 0.8187 z 2

EXAMPLE 8.6 The system function of an analog filter is expressed as:

s + 0.5
H a (s) =
(s + 0.5)2 + 4
Convert this analog filter into a digital filter using the impulse invariant transformation.
Assume T = 1 s.
Solution: Observe that the given system function of the analog filter is of the standard
s+a
form H a (s) = , where we are given a = 0.5 and b = 2.
( s + a) 2 + b 2
By the impulse invariant transformation, we know that
s+a 1 e aT (cos bT ) z 1
(s + a) 2 + b2 (is transformed to)
1 2e aT (cos bT ) z 1 + e 2 aT z 2

Therefore, for the given Ha(s), we can write the system function of the digital filter

1 e 0.5T (cos 2T ) z 1
H ( z) =
1 2 e 0.5T (cos 2T ) z 1 + e 2(0.5)T
z 2

Given T = 1s, we have

1 e 0.5 (cos 2) z 1
H (z ) =
1 2 e 0.5 (cos 2) z 1 + e 1 z 2

1 0.606 ( 0.416) z 1
=
1 2(0.606) ( 0.416) z 1 + 0.3678 z 2

1 + 0.252 z 1
=
1 + 0.504 z 1 + 0.3678 z 2
562 ! Digital Signal Processing

EXAMPLE 8.7 The system function of an analog filter is expressed as:

2
H a (s ) =
s(s + 2)
Find the corresponding H(z) using the impulse invariant method for a sampling frequency of
4 samples per second.
Solution: Given sampling rate = 4 samples/second
1
# Sampling period T = = 0.25 s
4
Expressing the given Ha(s) in terms of partial fractions, we have

2 1 1 1 1
H a (s) = = =
s( s + 2) s s+2 s (0) s ( 2)
By the impulse invariant transformation, we know that
A A
(is transformed to)
s pi 1 e piT z 1

Here Ha(s) has two poles and p1 = 0 and p2 = –2.

Therefore, the system function of the digital filter is:

1 1
H ( z) = p1T
1 e z 1
1 e p2T z 1

1 1
=
1 e(0)T z 1
1 e( 2) T
z 1

1 1
= 1 2(0.25) 1
1 z 1 e z
1 1
= 1 1
1 z 1 0.606 z
(1 0.606 z 1 ) (1 z 1 )
=
(1 z 1 ) (1 0.606 z 1 )
0.394 z 1
=
1 1.606 z 1 + 0.606 z 2

EXAMPLE 8.8 Convert the analog filter with system transfer function
2
H a (s ) =
(s + 0.4)2 + 4
into a digital filter using the impulse invariant transformation.
 Infinite-duration Impulse Response (IIR) Filters ! 563

Solution: Observe that the given system function of the analog filter is of the standard
b
form H a (s) = , where we are given a = 0.4 and b = 2.
(s + a) 2 + b2
By the impulse invariant transformation, we know that

b (sin bT ) z 1
e aT

( s + a) 2 + b 2 (is transformed to)

1 2e aT (cos bT ) z 1 + e 2 aT
z 2

Therefore, for the given Ha(s), we can write the digital filter function as:
(0.4)T
e(sin 2T ) z 1
H (z ) =
1 2 e (0.4)T (cos 2T ) z 1 + e 2(0.4)T
z 2

For T = 1 s,
0.4 1
e (sin 2) z
H ( z) = 0.4 1 0.8 2
1 2e (cos 2) z +e z
0.909 z 1
=
1 + 0.5578 z 1 + 0.449 z 2

EXAMPLE 8.9 Determine H(z) using the impulse invariant technique for the analog
system function

1
H a ( s) =
(s + 1)( s 2 + s + 2)

Solution: Using partial fractions, the given Ha(s) can be written as

1 A Bs + C
H a (s ) = = +
(s + 1) ( s2 + s + 2) s + 1 s 2 + s + 2
Therefore, we can write
A(s 2 + s + 2) + ( Bs + C ) (s + 1) = 1

i.e., ( A + B) s 2 + (A + B + C ) s + (2A + C ) = 1

Comparing the coefficients of s2, s and the constants on either side of the above expression,
we get
A + B = 0, i.e., B = –A
A + B + C = 0, # C=0
2A + C = 1, # A = 0.5 and B = – 0.5
564 ! Digital Signal Processing

So the system response can be written as:

0.5 0.5 s
H a (s) =
s + 1 s2 + s + 2
0.5 s
= 0.5
s +1 (s + 0.5) 2 + (1.3228)2

0.5 s + 0.5 0.5

= 0.5
s +1 ( s + 0.5) 2 + (1.3228)2 ( s + 0.5)2 + (1.3228)2

0.5 s + 0.5 0.25 1.3228

= 0.5 +
s +1 ( s + 0.5)2 + (1.3228)2 1.3228 ( s + 0.5)2 + (1.3228) 2

0.5 (s + 0.5) 1.3228

= 0.5 2 2
+ 0.1889
s +1 (s + 0.5) + (1.3228) (s + 0.5) 2 + (1.3228)2
Using the impulse invariant transformation, this analog system function Ha(s) can be
transformed into digital system function as:

0.5 1 e 0.5T (cos 1.3228 T ) z 1

H (z ) = 1
0.5
1 e Tz 1 2 e 0.5T (cos 1.3228 T ) z 1 + e 2(0.5) T
z 2

e 0.5T (sin 1.3228) z 1

+ 0.1889
1 2 e 0.5T (cos 1.3228) z 1 + e 2(0.5)T
z 2

Let T = 1 s, we have

0.5 1 0.606 (cos 1.3228) z 1

H (z ) = 1
0.5
1 0.3678 z 1 1.213(cos 1.3228) z 1 + 0.3678 z 2

0.606(sin 1.3228) z 1
+ 0.1889
1 1.213(cos 1.3228) z 1 + 0.3678 z 2

0.5 1 0.1487 z 1
= 1
0.5
1 0.3678 z 1 0.2977 z 1 + 0.3678 z 2

0.5874 z 1
+ 0.1889
1 0.2977 z 1 + 0.3678 z 2

0.5 0.5 0.6109 z 1

= 1
1 0.3678 z 1 0.2977 z 1 + 0.3678 z 2

1 2
0.646 z 0.0407 z
= 1 2 3
1 0.6655 z + 0.4773 z + 0.1352 z
 Infinite-duration Impulse Response (IIR) Filters ! 565

!"= )-0$5%1 (21 $$'1 2$6&-'1 781 &>-1 7$6$%-3'1 &'3%02('/3&$(%1 /-&>()
In the previous sections, we have studied the IIR filter design using (a) approximation of
derivatives method and (b) Impulse invariant transformation method. However the IIR filter
design using these methods is appropriate only for the design of low-pass filters and band pass
filters whose resonant frequencies are small. These techniques are not suitable for high-pass
or band reject filters. The limitation is overcome in the mapping technique called the
bilinear transformation. This transformation is a one-to-one mapping from the s-domain to
the z-domain. That is, the bilinear transformation is a conformal mapping that transforms the
imaginary axis of s-plane into the unit circle in the z-plane only once, thus avoiding aliasing
of frequency components. In this mapping, all points in the left half of s-plane are mapped
inside the unit circle in the z-plane, and all points in the right half of s-plane are mapped
outside the unit circle in the z-plane. So the transformation of a stable analog filter results in
a stable digital filter. The bilinear transformation can be obtained by using the trapezoidal
formula for the numerical integration.
b
Let the system function of the analog filter be H a ( s) =
s +a
The differential equation describing the above analog filter can be obtained as:

Y (s) b
H a (s) = =
X ( s) s + a

or sY(s) + aY(s) = bX(s)

Taking inverse Laplace transform on both sides, we get

dy(t )
+ a y(t ) = bx (t )
dt
Integrating the above equation between the limits (nT – T) and nT, we have
nT nT T
dy(t )
dt + a y(t ) dt = b x (t ) dt
dt
nT T nT T nT T

The trapezoidal rule for numeric integration is expressed as:

nT
T
a(t ) dt = [ a(nT ) + a(nT T )]
2
nT T

Therefore, we get
T T T T
y(nT ) y(nT T) + a y(nT ) + a y( nT T) = b x (nT ) + b x( nT T)
2 2 2 2
566 ! Digital Signal Processing

Taking z-transform, we get

T T
Y (z)[1 z 1 ] + a [1 + z 1 ] Y (z ) = b [1 + z 1 ] X ( z)
2 2
Therefore, the system function of the digital filter is:

Y ( z) b
= H (z ) = 1
X ( z) 2 1 z
1
+a
T 1+z
Comparing this with the analog filter system function Ha(s) we get
1
2 1 z 2 z 1
s= 1
=
T 1+z T z +1
Rearranging, we can get
T
1+ s
z= 2
T
1 s
2
This is the relation between analog and digital poles in bilinear transformation. So to convert
an analog filter function into an equivalent digital filter function, just put
1
2 1 z
s= 1
in Ha(s)
T 1+z
The general characteristic of the mapping z = esT may be obtained by putting s = + j! and
expressing the complex variable z in the polar form as z = re j in the above equation for s.

2 z 1 2 re j 1
Thus, s= =
T z 1 T re j + 1

2 (re j 1) (re j
+ 1) 2 r2 1 2r sin
or s= = +j
T (re j + 1) (re j
+ 1) T 1 + r 2 + 2r cos 1 + r 2 + 2r cos
Since s = + j!, we get

2 r2 1
=
T 1 + r 2 + 2r cos

2 2r sin
and =
T 1 + r 2 + 2r cos

From the above equation for , we observe that if r < 1 then < 0 and if r > 1, then > 0,
and if r = 1, then = 0. Hence the left half of the s-plane maps into points inside the unit
 Infinite-duration Impulse Response (IIR) Filters ! 567

circle in the z-plane, the right half of the s-plane maps into points outside the unit circle in
the z-plane and the imaginary axis of s-plane maps into the unit circle in the z-plane. This
transformation results in a stable digital system.

5(.$&-+%* 6(&7((%* $%$.+'* $%"* "-'-&$.* ,/(94(%:-()

On the imaginary axis of s-plane = 0 and correspondingly in the z-plane r = 1.

2 2 sin 2 sin
# = =
T 1 + 1 + 2 cos T 1 + cos

2 sin cos
2 2 2 2
= 2
= tan
T 1 + 2cos /2 1 T 2

# The relation between analog and digital frequencies is:

2
= tan
T 2

T 1
or equivalently, we have = 2 tan .
2
The above relation between analog and digital frequencies shows that the entire range in !
is mapped only once into the range – % $% % $% . The entire negative imaginary axis in the
s-plane (from ! = – " to 0) is mapped into the lower half of the unit circle in z-plane (from
= – to 0) and the entire positive imaginary axis in the s-plane (from ! = " to 0) is
mapped into the upper half of unit circle in z-plane (from = 0 to + ).
But as seen in Figure 8.4, the mapping is non-linear and the lower frequencies in
analog domain are expanded in the digital domain, whereas the higher frequencies are

*+,-./& '(2 4%$$')-* 63.;33)* !* %)0* * ')* 6'/')3%&* .&%)(+"&#%.'")<

568 ! Digital Signal Processing

compressed. This is due to the nonlinearity of the arctangent function and usually known as
frequency warping.
The effect of warping on the magnitude response can be explained by considering an
analog filter with a number of passbands as shown in Figure 8.5(a). The corresponding
digital filter will have same number of passbands, but with disproportionate bandwidth, as
shown in Figure 8.5(a).
In designing digital filter using bilinear transformation, the effect of warping on
amplitude response can be eliminated by prewarping the analog filter. In this method, the
specified digital frequencies are converted to analog equivalent using the equation
2
= tan . This analog frequencies are called prewarp frequencies. Using the prewarp
T 2
frequencies, the analog filter transfer function is designed, and then it is transformed to
digital filter transfer function.
This effect of warping on the phase response can be explained by considering an
analog filter with linear phase response as shown in Figure 8.5(b). The phase response of
corresponding digital filter will be nonlinear.

*+,-./& '(3 A83* ;%&$')-* 3++39.* ")* =%>* #%-)'.?03* &3($")(3* %)0* =6>* $8%(3* &3($")(3<

From the earlier discussions, it can be stated that the bilinear transformation preserves
the magnitude response of an analog filter only if the specification requires piecewise
constant magnitude, but the phase response of the analog filter is not preserved. Therefore,
the bilinear transformation can be used only to design digital filters with prescribed
magnitude response with piecewise constant values. A linear phase analog filter cannot be
transformed into a linear phase digital filter using the bilinear transformation.

EXAMPLE 8.10 Convert the following analog filter with transfer function

s + 0.1
H a (s ) =
(s + 0.1)2 + 9
 Infinite-duration Impulse Response (IIR) Filters ! 569

into a digital IIR filter by using bilinear transformation. The digital IIR filter is having a
resonant frequency of r = /2.
Solution: From the transfer function, we observe that !c = 3. The sampling period T can
be determined using the equation:
2
c = tan r
T 2

2 2 /2
# T= tan r
= tan = 0.6666 s
c 2 3 2
Using the bilinear transformation, the digital filter system function is:

H ( z) = Ha (s) 2 1 z 1
= Ha ( s)
s
T 1 z 1 1 z 1
s 3
1 z 1

#
s + 0.1
H ( z) =
( s + 0.1)2 + 9 s 3
1 z 1
1 z 1

1
1 z
3 1
+ 0.1
= 1+z
1 2
1 z
3 1
+ 0.1 +9
1+z

3(1 z 1 ) + 0.1(1 + z 1 ) [1 + z 1 ]
= 2
3(1 z 1 ) + 0.1 (1 + z 1 ) + 9(1 + z 1 ) 2

3.1 + 0.2 z 1 2.9 z 2

=
18.61 + 0.02 z 1 + 17.41 z 2

s + 0.5
EXAMPLE 8.11 Convert the analog filter with system function H a (s) =
(s + 0.5) 2 + 16
into a digital IIR filter using the bilinear transformation. The digital filter should have a
resonant frequency of r = /2.

Solution: From the system function, we observe that !c = 4. The sampling period T can be
2
determined using the equation = tan .
T 2
570 ! Digital Signal Processing

# 2
c = tan r
T 2

2 2
i.e. T= tan r
= tan = 0.5 s
c 2 4 4
Using the bilinear transformation, the digital filter system function is:

H ( z ) = H ( s) = H (s )
2 1 z 1
s
T 1 z 1 1 z 1
s 4
1 z 1

s + 0.5
H (z ) =
(s + 0.5) 2 + 16 s 4
1 z 1
1 z 1

1
1 z
4 1
+ 0.5
1+z
= 2
1
1 z
4 1
+ 0.5 + 16
1+z

4(1 z 1 ) + 0.5(1 + z 1 ) [1 + z 1 ]
= 2
4(1 z 1 ) + 0.5(1 + z 1 ) 16[1 + z 1 ]2

4.5 + z 1 3.5 z 2
=
36.25 + 0.5 z 1 + 28.25 z 2

EXAMPLE 8.12 Apply the bilinear transformation to

4
H a (s) =
(s + 3) (s + 4)
with T = 0.5 s and find H(z).

4
Solution: Given that H a (s) = and T = 0.5 s.
(s + 3) (s + 4)
1
2 1 z
To obtain H(z) using the bilinear transformation, replace s by 1
in Ha(s)
T 1+z
 Infinite-duration Impulse Response (IIR) Filters ! 571

4 4
# H (z ) = =
( s + 3) (s + 4) s
2 1 z 1
( s + 3) (s + 4) s 4
1 z 1
T 1 z 1 1 z 1

4
=
1 1
1 z 1 z
4 1
+3 4 1
+4
1+z 1+z

4
= 1 1
4 4z + 3 + 3z 4 4z 1 + 4 + 4z 1

1 z 1 1+z 1

4(1 + z 1 ) 2
=
(7 z 1) 8
1 (1 + z 1 )2
=
2 (7 z 1 )

EXAMPLE 8.13 Obtain H(z) from Ha(s) when T = 1 s and

3s
H a (s ) = 2
s + 0.5 s + 2
using the bilinear transformation.

3s
Solution: Given H a (s) = and T = 1 s.
s 2 + 0.5 s + 2
1
2 1 z
To get H(z) using the bilinear transformation, put s = 1
in Ha(s).
T 1+z

3s
# H ( z ) = H a (s ) 1 = 2
2 1 z s + 0.5s + 2
s
T 1 z 1 2 1 z 1
s
T 1 z 1

1
1 z
3 2 1
1 z
= 2
1 1
1 z 1 z
2 1
+ 0.5 2 1
+2
1 z 1 z
572 ! Digital Signal Processing

1
1 z
6 1
1+z
=
4(1 z 1 )2 + (1 z 1 ) (1 + z 1 ) + 2(1 + z 1 )2
(1 + z 1 ) 2

6 (1 + z 1 )
= 1
4(1 2z + z 2 ) + (1 z 2 ) + 2(1 + 2z 1
+ z 2)
6 + 6z 1
=
7 4z 1 + 5z 2

EXAMPLE 8.14 Using the bilinear transformation, obtain H(z) from Ha(s) when T = 1s
s3
and H a (s) =
(s + 1)(s 2 + 2s + 2)

s3
Solution: Given that H a (s) = and T = 1 s.
(s + 1)(s 2 + 2s + 2)

1
2 1 z
To obtain H(z) using the bilinear transformation, put s = 1
in Ha(s).
T 1+z
Given T = 1 s,

s3
H ( z ) = H a (s) =
2 1 z 1 (s + 1)( s 2 + 2s + 2)
s=
T 1 z 1 1 z 1
s 2
1 z 1

3
(1 z 1 )
2
(1 + z 1 )
=
2
(1 z 1 ) (1 z 1 ) (1 z 1 )
2 +1 2 +2 2 +2
1+z 1 1+z 1 1+z 1

8(1 z 1 )3
=
2(1 z 1 ) + (1 + z 1 ) 4(1 z 1 )2 + 4(1 z 1 ) (1 + z 1 ) + 2(1 + z 1 ) 2

8(1 z 1 )3
=
(3 z 1 )[10 4 z 1 + 2z 2 ]
 Infinite-duration Impulse Response (IIR) Filters ! 573

4(1 z 1 )3
=
(3 z 1 ) (5 2z 1 + 2z 2 )

(1 3z 1 + 3z 2 z 3 )
=4
15 11 z 1 + 8 z 2 2 z 3

EXAMPLE 8.15 A digital filter with a 3 dB bandwidth of 0.4 is to be designed from the
analog filter whose system response is:

c
H (s) =
s +2 c

Use the bilinear transformation and obtain H(z).

2
Solution: We know that c = tan c
T 2
Here the 3 dB bandwidth c = 0.4

2 0.4 1.453
# c = tan =
T 2 T
The system response of the digital filter is given by

H ( z) = H a ( s)
2 1 z 1
s
T 1 z 1

1.453
= c
= T
1 1
2 1 z 2 1 z 1.453
1
+2 c 1
+2
T 1+z T 1+z T

1.453 (1 + z 1 )
=
2 (1 z 1 ) + 2(1 + z 1 ) 1.453
1+z 1
= 1
3.376 0.624 z

EXAMPLE 8.16 The normalized transfer function of an analog filter is given by

1
H (sn ) =
sn2 + 1.6 sn + 1
574 ! Digital Signal Processing

Convert the analog filter to a digital filter with a cutoff frequency of 0.6 , using the bilinear
transformation.
Solution: The prewarping of analog filter has to be performed to preserve the magnitude
response. For this the analog cutoff frequency is determined using the bilinear
transformation, and the analog transfer function is unnormalized using this analog cutoff
frequency. Then the analog transfer function is converted to digital transfer function using the
bilinear transformation.
Given that, digital cutoff frequency, c = 0.6 rad/s. Let T = 1s.
In the bilinear transformation,
2 0.6
Analog cutoff frequency c = tan c = 2 tan = 2.753 rad/s.
T 2 2
1
Normalized analog transfer function H a (sn ) =
sn2 + 1.6 sn + 1
The analog transfer function is unnormalized by replacing sn by s/!c.
Therefore, unnormalized analog filter transfer function is given by

1 1
H a (s) = 2
= 2
s s s s
+ 1.6 +1 + 1.6 +1
c c
2.753 2.753

2.7532 7.579
= =
s 2 + 1.6 2.753 s + 2.7532 s2 + 4.404 s + 7.579
1
2 1 z
The digital filter system function H(z) is obtained by substituting s = 1
in
T 1+z
Ha(s). Here T = 1. Therefore, the digital filter transfer function is:

7.579
H ( z) = 2
1 1
1 z 1 z
2 1
+ 4.404 2 1
+ 7.579
1 z 1+z

7.579(1 + z 1 ) 2
= 1
4(1 2 z + z 2 ) + 4.404(1 + z 1 ) 2(1 z 1 ) + 7.579(1 + z 1 )2

7.579[1 + 2 z 1 + z 2 ]
=
20.387 + 7.158 z 1 + 2.771 z 2

0.371 + 0.742 z 1 + 0.371 z 2

=
1 + 0.351 z 1 + 0.136 z 2
 Infinite-duration Impulse Response (IIR) Filters ! 575

!"? 09-+$2$+3&$(%01 (21 &>-1 6(@A93001 2$6&-'

The magnitude response of low-pass filter in terms of gain and attenuation are shown in
Figure 8.6.

*+,-./& '(4 4%-)'.?03* &3($")(3* "+* /";5$%((* +'/.3&* =%>* B%')* @(* * %)0* =6>* 1..3)?%.'")* @(* <

Let 1 = Passband frequency in rad/s.

2 = Stopband frequency in rad/s.
Let the gain at the passband frequency 1 be A1 and the gain at the stopband frequency
2 be A2, i.e.
A1 = H ( ) and A2 = H ( )
1 2

The filter may be expressed in terms of the gain or attenuation at the edge frequencies.
Let 1 be the attenuation at the passband edge frequency 1, and 2 be the attenuation at the
stopband edge frequency 2.

1 1 1 1
i.e. 1 = = and 2 = =
A1 H( ) A2 H( )
1 2

The maximum value of normalized gain is unity, so A1 and A2 are less than 1 and 1
and 2 are greater than 1. In Figure 8.6, A1 is assumed as 1/ 2 and A2 is assumed as 0.1.
Hence 1 = 2 = 1.414 and 2 = 1/0.1 = 10.
Another popular unit that is used for filter specification is dB. When the gain is
expressed in dB, it will be a negative dB. When the attenuation is expressed in dB, it will be
a positive dB.
Let k1 = Gain in dB at a passband frequency 1
k2 = Gain in dB at a stopband frequency 2
576 ! Digital Signal Processing

The gain can be converted into normal values as follows:

20 log A1 = k1 20 log A2 = k2
log A1 = k1/20 log A2 = k2/20
A1 = 10k1/20 A2 = 10k2/20
When expressed in dB, the gain and attenuation will have only change in sign because
log = log(1/A) = –log A. (Hence when dB is positive it is attenuation and when dB is
negative it is gain).
When A1 = 0.707, k1 = 20 log(0.707) = –3.0116 = –3 dB
When A2 = 0.1, k2 = 20 log(0.1) = –20 dB
The magnitude response of low-pass filter in terms of dB-attenuation is shown in Figure 8.7.

*+,-./& '(5 4%-)'.?03* &3($")(3* "+* /";5$%((* +'/.3&* =%>* 0C5B%')* @(* * %)0* =6>* 0C5%..3)?%.'")* @(* #

Sometimes the specifications are given in terms of passband ripple p and stopband
ripple s. In this case, the dB gain and attenuation can be estimated as follows:
k1 = 20 log (1 – p) 1 = –20 log (1 – p)
k2 = 20 log s 2 = –20 log s

If the ripples are specified in dB, then the minimum passband ripple is equal to k1 and
the negative of maximum passband attenuation is equal to k2.

!"B )-0$5%1 (21 6(@A93001 )$5$&361 7*&&-'@('&>1 2$6&-'

The popular methods of designing IIR digital filter involves the design of equivalent analog
filter and then converting the analog filter to digital filter. Hence to design a Butterworth IIR
digital filter, first an analog Butterworth filter transfer function is determined using the given
specifications. Then the analog filter transfer function is converted to a digital filter transfer
function using either impulse invariant transformation or bilinear transformation.
 Infinite-duration Impulse Response (IIR) Filters ! 577

!%$.+'* ;4&&(/7+/&<* ,-.&(/

The analog Butterworth filter is designed by approximating the ideal frequency response
using an error function. The error function is selected such that the magnitude is maximally
flat in the passband and monotonically decreasing in the stopband. (Strictly speaking the
magnitude is maximally flat at the origin, i.e., at ! = 0, and monotonically decreasing with
increasing !).
The magnitude response of low-pass filter obtained by this approximation is given by
2 1
Ha ( ) = 2N

1+
c

where !c is the 3 dB cutoff frequency and N is the order of the filter.

=/(94(%:>* /()3+%)(* +,* &<(* ;4&&(/7+/&<* ,-.&(/

The frequency response of Butterworth filter depends on the order N. The magnitude
response for different values of N are shown in Figure 8.8. From Figure 8.8, it can be
observed that the approximated magnitude response approaches the ideal response as the
value of N increases. However, the phase response of the Butterworth filter becomes more
nonlinear with increasing N.

*+,-./& '(' 4%-)'.?03* &3($")(3* "+* C?..3&;"&.8* /";5$%((* +'/.3&* +"&* @%&'"?(* @%/?3(* "+* $<

?/"(/* +,* &<(* ,-.&(/

Since the frequency response of the filter depends on its order N, the order N has to be
estimated to satisfy the given specifications.
Usually the specifications of the filter are given in terms of gain A or attenuation at
a passband or stopband frequency as given below:
A1 $ |H( )| $ 1, 0 $% % $ 1
|H( )| $ A2, 2 $% % $
578 ! Digital Signal Processing

The order of the filter is determined as given below.

Let !1 and !2 be the analog filter edge frequencies corresponding to digital frequencies
1 and 2. The values of !1 and !2 are obtained using the bilinear transformation or
impulse invariant transformation.
1
# A12 2N
1
1
1+
c

1
and 2N
A22
2
1+
c

These two equations can be written in the form

2N
1 1
1
c A12

2N
1
and 2
1
c A22

Assuming equality we can obtain the filter order N and the 3 dB cutoff frequency !c.
Dividing the first equation by the second, we have

1
2N 1
1 A12
=
1
2 1
A22
From this equation, the order of the filter N is obtained approximately as:

1 1
log 1 1
1 A22 A12
N=
2 2
log
1

If N is not an integer, the value of N is chosen to be the next nearest integer. Also we can get

1
c = 1/2N
1
1
A12

when parameters A1 and A2 are given in dB.

 Infinite-duration Impulse Response (IIR) Filters ! 579

A1 in dB is given by
A1 dB = –20 log A1

A1 dB
i.e. log A1 =
20
A1 dB
or A1 = 10 20

# 1 1
1= 1
A12 A1 dB
2

10 20

1
i.e. 1 = 10 0.1 A1 dB 1
A12

1
Similarly 1 = 10 0.1 A2 dB 1
A22

1 1 10 0.1 A2dB 1
log{[ 2 1] /[ 2 1]} log
1 A2 A1 1 10 0.1 A1dB 1
# N= =
2 2
log 2 log 2

1 1

and !c is given by

1 2
c = 0.1 A1dB 1/2N
or c = 0.1 A2 dB
(10 1) (10 1)1/2 N

1 1 2
In fact, c = 1/ 2 N
+ 1/2N
2 10 0.1 A1dB 1 100.1 A2 dB 1

;4&&(/7+/&<* .+7@3$))* ,-.&(/* &/$%),(/* ,4%:&-+%

The unnormalized transfer function of the Butterworth filter is usually written in factored
form as:
N/2 2
c
H a (s ) = (when N is even)
k 1 s 2 + bk cs +
2
c
580 ! Digital Signal Processing

N 1
2 2
c c
or H a (s) = (when N is odd)
s+ c k 1 s 2 + bk cs +
2
c

(2 k 1)
where bk = 2 sin
2N
If s/!c (where !c is the 3 dB cutoff frequency of the low-pass filter) is replaced by sn,
then the normalized Butterworth filter transfer function is given by
N/2
1
H a (s) = (when N is even)
k 1 sn2 + bk sn + 1

N 1
1 2 1
or H a (s) = (when N is odd)
sn + 1 k 1 sn2 + bk sn + 1

(2 k 1)
where bk = 2 sin
2N

0()-'%* 3/+:("4/(* ,+/* .+7@3$))* "-'-&$.* ;4&&(/7+/&<* 115* ,-.&(/

The low-pass digital Butterworth filter is designed as per the following steps:
Let A1 = Gain at a passband frequency 1
A2 = Gain at a stopband frequency 2
! !1 = Analog frequency corresponding to 1
! !2 = Analog frequency corresponding to 2

Step 1 Choose the type of transformation, i.e., either bilinear or impulse invariant transformation.
Step 2 Calculate the ratio of analog edge frequencies !2/!1.
For bilinear transformation
2 2 tan 2 /2
1 tan 1 , 2 = tan 2
= 2
=
T 2 T 2 1 tan 1 /2
For impulse invariant transformation,
1 2 2 2
1 = , 2 = =
T T 1 1

Step 3 Decide the order N of the filter. The order N should be such that

1 1
log 1 1
1 A22 A12
N
2 2
log
1

Choose N such that it is an integer just greater than or equal to the value obtained above.
 Infinite-duration Impulse Response (IIR) Filters ! 581

1
Step 4 Calculate the analog cutoff frequency c = 1/2 N
1
1
A12

2
tan 1 /2
For bilinear transformation = T
c 1/2N
1
1
A12

1 /T
For impulse invariant transformation c = 1/2 N
1
1
A12
Step 5 Determine the transfer function of the analog filter.
Let Ha(s) be the transfer function of the analog filter. When the order N is even,
for unity dc gain filter, Ha(s) is given by
N/2 2
c
H a ( s) =
k 1 s 2 + bk cs + 2
c

When the order N is odd, for unity dc gain filter, Ha(s) is given by
N 1
2 2
c c
H a (s ) =
s+ c k 1 s 2 + bk cs + 2
c

The coefficient bk is given by

(2 k 1)
bk = 2 sin
2N
For normalized case, !c = 1 rad/s
Step 6 Using the chosen transformation, transform the analog filter transfer function Ha(s)
to digital filter transfer function H(z).
Step 7 Realize the digital filter transfer function H(z) by a suitable structure.

A+.()* +,* &<(* %+/2$.-B("* ;4&&(/7+/&<* ,-.&(/

The Butterworth low-pass filter has a magnitude squared response given by

2 1
Ha ( ) = 2N

1+
c
582 ! Digital Signal Processing

We know that the frequency response Ha(!) of an analog filter is obtained by substituting
s = j! in the analog transfer function Ha(s). Hence the system transfer function is obtained
by replacing ! by (s/j) in the above equation.

1 1
# H a (s) H a ( s) = 2N
= 2N
s s2
1+ 1+
j c j2 2
c

In the above equation, when s/!c is replaced by sn (i.e. !c = 1 rad/s), the transfer function is
called normalized transfer function.
1
# H a ( sn ) H a ( sn ) =
1 + ( sn2 ) N
The transfer function of the above equation will have 2N poles which are given by the roots
of the denominator polynomial. It can be shown that the poles of the transfer function
symmetrically lie on a unit circle in s-plane with angular spacing of /N.
For a stable and causal filter the poles should lie on the left half of the s-plane. Hence
the desired filter transfer function is formed by choosing the N-number of left half poles.
When N is even, all the poles are complex and exist in conjugate pairs. When N is odd, one
of the pole is real and all other poles are complex and exist as conjugate pairs. Therefore,
the transfer function of Butterworth filters will be a product of second order factors.
The poles of the Butterworth polynomial lie on a circle, whose radius is c. To
determine the number of poles of the Butterworth filter and the angle between them we use
the following rules.
& Number of Butterworth poles = 2N
& Angle between any two poles = 360°/(2N)
If the order of the filter N is even, then the location of the first pole is at /2 w.r.t. the
positive real axis, with the angle measured in the counter-clockwise direction. The location
of the subsequent poles are respectively, at

+ , +2 , +3 , ..., 360
2 2 2 2
If the order of the filter N is odd, then the location of the first pole is on the X-axis. The
location of subsequent poles are at , 2 , ..., (360 – ) with the angle measured in the
counter-clockwise direction.
If is the angle of a valid pole w.r.t. the X-axis, then the pole and its conjugate are
located at [ c(cos ± j sin )].

A/+3(/&-()* +,* ;4&&(/7+/&<* ,-.&(/)

1. The Butterworth filters are all pole designs (i.e. the zeros of the filters exist at ").
2. The filter order N completely specifies the filter.
 Infinite-duration Impulse Response (IIR) Filters ! 583

3. The magnitude response approaches the ideal response as the value of N increases.
4. The magnitude is maximally flat at the origin.
5. The magnitude is monotonically decreasing function of !'
6. At the cutoff frequency !c, the magnitude of normalized Butterworth filter is 1/
2 . Hence the dB magnitude at the cutoff frequency will be 3 dB less than the
maximum value.

EXAMPLE 8.17 Design a Butterworth digital filter using the bilinear transformation. The
specifications of the desired low-pass filter are:

0.9 H( ) 1; 0
2
3
H( ) 0.2;
4
with T = 1 s
Solution: The Butterworth digital filter is designed as per the following steps.
From the given specification, we have

A1 = 0.9 and 1 =
2
3
A2 = 0.2 and 2 = and T = 1 s
4
Step 1 Choice of the type of transformation
Here the bilinear transformation is already specified.
Step 2 Determination of the ratio of the analog filter’s edge frequencies, !2/!1

2 2 (3 /4) 3
2 = tan 2 = tan = 2 tan = 4.828
T 2 1 2 8

2 2 ( /2)
1 = tan 1 = tan = 2 tan =2
T 2 1 2 4

4.828
# 2
= = 2.414
1 2

Step 3 Determination of the order of the filter N

1 1
log 1 1
1 A22 A12
N
2 2
log
1
584 ! Digital Signal Processing

1 1
log 2
1 1
1 (0.2) (0.9)2
2 log 1.207

1 log 24 0.2345
2.626
2 log 2.414
Since N ( 2.626, choose N = 3.
Step 4 Determination of the analog cutoff frequency !c (i.e., –3 dB frequency)

1 2
c = 1/2 N
= 1/2 3
= 2.5467
1 1
1 1
A12 0.92

Step 5 Determination of the transfer function of the analog Butterworth filter Ha(s)
N 1
2 2
c c
For odd N, we have H a (s) =
s+ c k 1 s2 + bk cs + 2
c

(2 k 1)
where bk = 2 sin
2N
For N = 3, we have
2
c c
H a (s) =
s+ c s 2 + b1 cs + 2
c

(2 1 1)
where b1 = 2 sin = 2sin =1
2 3 6

2.5467 (2.5467)2
Therefore, H a (s) =
s + 2.5467 s 2 + 1(2.5467) s + (2.5467)2
Step 6 Conversion of Ha(s) into H(z)
Since bilinear transformation is to be used, the digital filter transfer function is:
H (z) = H a (s) s
2 1 z 1 = H a ( s) s 2
1 z 1
T 1+ z 1 1 z 1

2.5467 (2.5467)2
H (z ) = 2
1
1 z 1 z 1
1 z 1
2 1
+ 2.5467 2 + 2.5467 2 + (2.5467)2
1+ z 1+z 1
1+z 1
 Infinite-duration Impulse Response (IIR) Filters ! 585

0.2332(1 + z 1 )3
=
1 + 0.4394 z 1 + 0.3845 z 2 + 0.0416 z 3

EXAMPLE 8.18 Design a digital Butterworth filter satisfying the following constraints:

0.8 H( ) 1; 0 0.2

H( ) 0.2; 0.32
with T = 1 s. Apply impulse invariant transformation.
Solution: From the given specifications, we have
A1 = 0.8 1 = 0.2
A2 = 0.2 2 = 0.32 and T = 1 s
The Butterworth IIR digital filer is designed as per the following steps.
Step 1 Choice of the type of transformation
Here, the impulse invariant transformation is already specified.
Step 2 Determination of the ratio of analog filter’s edge frequencies, !2/!1

2 0.32
2 = = = 0.32
T 1

1 0.2
1 = = = 0.2
T 1

2 0.32
= = 1.6
1 0.2

Step 3 Determination of the order of the filter N

1 1
log 1 1
1 A22 A12
N
2 2
log
1

1 1
log 1 1
1 0.22 0.82
2 log 1.6

1 log 24/0.5625
3.9931
2 log 1.6

So the order of the filter N ( 3.9931. Choose N = 4.

586 ! Digital Signal Processing

Step 4 Determination of the analog cutoff frequency !c (i.e., – 3 dB frequency)

1 0.2
c = 1/2N
= 1/ 2 4
= 0.675 rad/s
1 1
1 1
A12 0.82

Step 5 Determination of the transfer function of analog low-pass Butterworth filter. The
transfer function of the low-pass filter for even values of N is:
N/2 2
c
H a (s ) =
k 1 s 2 + bk cs + 2
c

(2k 1)
where bk = 2sin
2N
Here N = 4; # k = 1, 2
(2 1)
When k = 1, bk = b1 = 2sin = 0.765
2 4

(2 2 1)
When k = 2, bk = b2 = 2sin = 1.848
2 4
2 2
#% H a (s) = c c
s 2 + b1 cs + 2
c s 2 + b2 cs + 2
c

(0.675)2 (0.675) 2
=
s 2 + (0.765 0.675) s + 0.6752 s 2 + (1.848 0.675) s + (0.675) 2
0.2076
=
(s 2 + 0.516 s + 0.456) (s 2 + 1.247 s + 0.456)
Step 6 Determination of the digital filter transfer function H(z)
By partial fraction expansion, Ha(s) can be expressed as:
0.2076
H a (s) =
( s + 0.516 s + 0.456) ( s 2 + 1.247 s + 0.456)
2

As + B Cs + D
= 2
+ 2
s + 0.516 s 0.456 s + 1.247 s + 0.456
On cross multiplying the above equation and simplifying, we get
0.2076 = (A + C ) s3 + (1.247 A + B + 0.516 C + D) s 2 + (0.456 A + 1.247B + 0.456C
+ 0.516 D)s + (0.456 B + 0.456 D)
On solving, we get
A = – 0.622, B = – 0.321, C = 0.622 and D = 0.776
 Infinite-duration Impulse Response (IIR) Filters ! 587

Therefore, Ha(s) can be written as:

0.622s 0.321 0.622s + 0.776

H a (s ) = 2
+ 2
s + 0.516 s + 0.456 s + 1.247 s + 0.456

0.622 (s + 0.516)
= 2
( s2 + 2 0.258s + 0.2582 ) + 0.456 0.2582

0.622(s + 1.248)
+ 2
( s2 + 2 0.624s + 0.6242 ) + 0.456 0.6242

0.622 (s + 0.258 + 0.258) 0.622 (s + 0.624 + 0.624)

= +
(s + 0.258)2 + (0.624)2 (s + 0.624)2 + (0.258)2

s + 0.258 0.624
= 0.622 2 2
0.257
(s + 0.258) + (0.624) (s + 0.258)2 + (0.624) 2
s + 0.624 0.258
+ 0.622 2 2
+ 1.504
( s + 0.624) + (0.258) (s + 0.624)2 + (0.258)2
The analog transfer function of the above equation can be transformed to digital
transfer function using the following standard impulse invariant transformations.

s +a 1 e aT (cos bT ) z 1
(s + a) 2 + b2 1 2e aT (cos bT ) z 1 + e 2 aT z 2

b e aT (sin bT ) z 1
( s + a) 2 + b 2 1 2e aT (cos bT ) z 1 + e 2 aT
z 2

Taking T = 1 s, the above transformation can be applied to Ha(s) to get H(z).

1 e 0.258 (cos 0.624) z 1

H ( z) = 0.622
1 2e 0.258 (cos 0.624) z 1 + e 2 0.258 z 2

0.258
e (sin 0.624) z 1
0.257
1 2e 0.258 (cos 0.624) z 1 + e 2 0.258
z 2

1 e 0.624 (cos 0.258) z 1

+ 0.622
1 2e 0.624 (cos 0.258)z 1 + e 2 0.624 z 2

0.624
e (sin 0.258) z 1
+ 1.504
1 2e 0.624 (cos 0.258) z 1 + e 2 0.624
z 2
588 ! Digital Signal Processing

0.622 + 0.39 z 1 0.116 z 1

= +
1 1.254 z 1 + 0.597 z 2
1 1.254 z 1 + 0.597 z 2

0.622 0.322 z 1 0.206 z 1

+ +
1 1.036 z 1 + 0.287 z 2
1 1.036 z 1 + 0.287 z 2

1 1
0.622 + 0.274 z 0.622 0.116 z
= 1 2
+ 1 2
1 1.254 z + 0.597 z 1 1.036 z + 0.287 z
0.0224 z 1 0.0544 z 2 0.0094 z 3
=
1 2.29 z 1 + 2.1831 z 2 0.977 z 3 + 0.1713 z 4

EXAMPLE 8.19 Design a low-pass Butterworth digital filter to give response of 3 dB or

less for frequencies upto 2 kHz and an attenuation of 20 dB or more beyond 4 kHz. Use the
bilinear transformation technique and obtain H(z) of the desired filter.
Solution: The specifications of the desired filter are given in terms of dB attenuation and
frequency in Hz. First the gain is to be expressed as a numerical value and frequency in rad/s.
Here attenuation at passband frequency ( 1) = 3 dB
Therefore, gain at passband edge frequency ( 1) is k1 = –3 dB

# 1
A1 =10 k1/20 = 10 3/20
= 0.707 =
2
Attenuation at stopband frequency ( 2) = 20 dB
Therefore, gain at stopband edge frequency ( 2) is k2 = –20 dB

# A2 = 10 k2 /20 = 10 20/20
= 0.1
Passband edge frequency = 2 kHz,
Stopband edge frequency = 4 kHz,
The design is performed as given below.
Let the sampling frequency be 10000 Hz.
f1 2000
# Normalized 1 =2 =2 = 0.4
fs 10000

f2 4000
Normalized 2 =2 =2 = 0.8
fs 10000
Step 1 Bilinear transformation is chosen
Step 2 Ratio of analog filter edge frequencies !2/!1
2 2 0.4
1 = tan 1 = tan = 14530.8 rad/s
T 2 T 2
 Infinite-duration Impulse Response (IIR) Filters ! 589

2 2 0.8
2 = tan 2 = tan = 61553.6 rad/s
T 2 T 2

2
tan
# 2
= 2 = tan 0.4 = 4.236
tan 0.2
1 tan 1
2
Step 3 Order of the filter

1 1
log 1 1
1 A22 A12
N
2 2
log
1

1 1
log 1 1
1 (0.1)2 (1/ 2)2
2 log 4.236

1 log[99/1]
1.59
2 log 4.236
# N=2
Step 4 Analog cutoff frequency !c

1 1.4530
c = 1/2 N
= 1/ 2 2
= 1.4530
1 1
1 1
A12 (1/ 2)2

Unnormalized c = f s 1.4530 = 14530 rad/s

Step 5 Transfer function Ha(s)
2
For N = 2, H a (s ) = c
s 2 + b1 c s+ 2
c

(2 1 1)
where b1 = 2sin = 2sin = 1.414
2 2 4

(14530) 2
# H a (s) = 2
s + 1.414 14530 s + (14530) 2

2.1112 108
= 2
s + 20545.42 s + 2.1112 10 8
590 ! Digital Signal Processing

Step 6 Conversion of Ha(s) into H(z)

H ( z) = Ha (s) s
2 1 z 1 = Ha (s) s 20000
1 z 1
T 1 z 1 1 z 1

2.112 108
H (z ) = 2
1 1
1 z 1 z
20 10 3 1
+ 2.0545 10 4 20 103 1
+ 2.112 10 8
1+z 1+z

0.528
=
2.5552 0.946 z 1 + 0.5008 z 2

EXAMPLE 8.20 Design a low-pass Butterworth filter using the bilinear transformation
method for satisfying the following constraints:
Passband: 0–400 Hz Stopband: 2.1– 4 kHz
Passband ripple: 2 dB Stopband attenuation: 20 dB
Sampling frequency: 10 kHz
Solution: Given
1 = 2 dB, ! " k1 = –2 dB and 1 = 10 k1/20 = 10 2/20 = 0.794
2 = 20 dB, " k2 = –20 dB and 2 = 10 k2 /20 = 10 20/20 = 0.1

Step 1 Type of transformation

Bilinear transformation is already specified.
Step 2 Ratio of analog edge frequencies #2/#1.
Here fs = 10 kHz
Passband edge frequency f1 = 400 Hz
Stopband edge frequency f2 = 2.1 kHz
Normalizing the frequencies, we have
f1 400
1 =2 =2 = 0.25 rad
fs 10000

f2 2100
2 =2 =2 = 1.319 rad
fs 10000
Therefore, the analog filter edge frequencies are:
2 0.25
1 = tan 1 = 2 10000 tan = 2513.102 rad/s
T 2 2
2 1.319
and 2 = tan 2 = 2 10000 tan = 15,506.08 rad/s
T 2 2
 Infinite-duration Impulse Response (IIR) Filters ! 591

# 2 15506.08
= = 6.1703
1 2513.102
Step 3 Order of the filter N

1 1 10 0.1A2 dB 1
log 1 1 log
1 A22 A12 1 10 0.1A1 dB 1
N or N
2 2 2 log (6.1703)
log
1

1 1 10 0.1 20 dB 1
log 1 1 log
1 (0.1)2 (0.794)2 1 100.1 2 dB 1
i.e. N or N
2 log (6.1703) 2 log (6.1703)

i.e. N 1.409 2 or N 1.410 2

Step 4 The cutoff frequency !c

c = 2 or c = 2
0.1 A2 dB
1
1/2 N
[10 1]1/2 N
1
A22
i.e.
15506.08 15506.08
c = 1/2 2
= 4915.7 or c = = 4915.788 rad/s
1 [10 0.1 20 1]1/2 2
1
(0.1) 2
Step 5 The system function Ha(s)
2
c (2 1 1)
H a (s) = 2 2
where b1 = 2 sin = 1.414
s + b1 cs + c 2 2
(4915.788)2
=
s 2 + 1.414 4915.788 s + (4915.788)2
2.416 10 7
= 2
s + 6950.92 s + 2.416 10 7

Step 6 Digital transfer function H(z)

H (z ) = H a (s)
2 1 z 1
s
T 1 z 1
592 ! Digital Signal Processing

2.416 10 7
= 2
1 1
2 1 z 2 1 z
1
+ 6950.92 1
+ 2.416 10 7
T 1 z T 1+z

2.416 10 7
= 2
1 1
1 z 1 z
20000 1
+ 6950.92 20000 1
+ 2.416 10 7
1+z 1+z

0.042 + 0.085 z 1 + 0.042 z 2

=
1 1.335 z 1 + 0.506 z 2

j (2k + N + 1)
The poles are given by Pk = ( c) e , k = 0, 1 N
2N
3
j
# P0 = ( c)e
4
= 4.915.788( 0.707 + j 0.707) = 3475.6 + j 3475.46
5
j
4
P1 = ( c )e = 3475.6 j 3475.6

EXAMPLE 8.21 A digital low-pass filter is required to meet the following specifications.
Passband attenuation $ 1 dB Passband edge = 4 kHz
Stopband attenuation ( 40 dB Stopband edge = 8 kHz
Sampling rate = 24 kHz
The filter is to be designed by performing the bilinear transformation on an analog
system function. Design the Butterworth filter.

Solution: Given 1 = 1 dB, # k1 = –1 dB and A1 = 10 k1 /20 = 10 1/20

= 0.8912
k /20 40/20
2 = 40 dB, # k2 = – 40 dB and A2 = 10 2 = 10 = 0.01
Since fs = 24 kHz, normalized angular frequencies are:
2 f1 4000
f1 = 4 kHz, # 1 = =2 = 1.047 rad/s
fs 24000

2 f2 8000
f2 = 8 kHz, # 2 = =2 = 2.094 rad/s
fs 24000
The Butterworth filter is designed as follows:
Step 1 Type of transformation
Bilinear transformation is already specified.
 Infinite-duration Impulse Response (IIR) Filters ! 593

Step 2 Ratio of analog edge frequencies, !2/!1

2 1.047
1 = tan 1 = 2 24000 tan = 27706.49 rad/s
T 2 2
2 2.094
2 = tan 2 = 2 24000 tan = 83100.52 rad/s
T 2 2
83000.52
" 2
= = 2.9957
1 27706.49
Step 3 Order of the filter N

1 1 10 0.1 A2 dB 1
log 1 1 log
1 A22 A12 1 100.1 A1 dB 1
N or N
2 log ( 2/ 1)
2 log ( 2/ 1)

1 1 100.1 40 1
log 2
1 1 log
1 (0.01) (0.8912) 2 1 10 0.1 1 1
or N
2 log (2.9957) 2 log (2.9957)

1 log{9999/0.2590} 1 log (9999/0.2589)

or N
2 log (2.9957) 2 log (2.9957)

1 4.586 1 4.586
or N
2 0.476 2 0.476

4.8 5 or N 4.8 5
Step 4 The cutoff frequency !c

c = 1 or c = 1
1/2N 0.1 A1 dB
1 [10 1]1/2N
1
A12

27706.49 27706.49
= or c =
1/2 5 1/2 5
1 100.1 1 1
1
(0.8912)2

= 31,715 rad/s or c = 31,715 rad/s

Step 5 Analog filter transfer function Ha(s)
2 2
c c c
For N = 5, H a (s) =
s+ c s 2 + b1 cs + 2
c s 2 + b2 cs + 2
c
594 ! Digital Signal Processing

(2 1 1)
where b1 = 2sin = 2sin = 0.618
2N 10

(2 2 1) 3
b2 = 2sin = 2sin = 1.618
10 10

31708 (31708) 2
" H a (s) =
s + 31708 s 2 + 0.618 31708 s + (31708)2

(31708)2
s 2 + 1.618 31708 s + (31708)2

Step 6 Digital filter function H(z)

Using the bilinear transformation, we have

H ( z ) = H a (s ) = H a ( s)
2 1 z 1
s
T 1 z 1 1 z 1
s 48000
1 z 1

31708 (31708)2
=
s + 31708 s2 + 0.618 31708 s + (31708) 2

(31708) 2
s 2 + 1.618 31708 s + (31708)2 1 z 1
s 48000
1 z 1

EXAMPLE 8.22 Design a digital IIR low-pass filter with passband edge at 1000 Hz and
stopband edge at 1500 Hz for a sampling frequency of 5000 Hz. The filter is to have a
passband ripple of 0.5 dB and a stopband ripple below 30 dB. Design a Butterworth filter
using the bilinear transformation.
Solution: Given fs = 5000 Hz, the normalized frequencies are given as:
f1 1000
Passband edge f1 = 1000 Hz, " 1 =2 =2 = 0.4 rad/s
fs 5000

f2 1500
Stopband edge f2 = 1500 Hz, " 2 =2 =2 = 0.6 rad/s
fs 5000

1 = 0.5 dB, " k1 = –0.5 dB and A1 = 10 k1/20 = 10 0.5/20

= 0.9446
2 = 30 dB, # " k2 = – 30 dB and A2 = 10 k2 /20 = 10 30/20
= 0.0316
 Infinite-duration Impulse Response (IIR) Filters ! 595

The Butterworth filter is designed as follows:

Step 1 Type of transformation.
Bilinear transformation is to be used.
Step 2 Ratio of analog filter edge frequencies, !2/!1

2 0.4
1 = tan 1 = 2 5000 tan = 7265.425 rad/s
T 2 2
2 0.6
2 = tan 2 = 2 5000 tan = 13763.819 rad/s
T 2 2

2 13763.819
= = 1.8944
1 7265.425

Step 3 Order of the filter N

1 1
log 1 1
1 A22 A12
N
2 log ( 2/ 1)

1 1
log 2
1 1
1 (0.0316) (0.9446) 2
2 log (1.844)

1 log{1000.44/0.1207}
2 log(1.844)
$ 7.35 % 8
Step 4 The cutoff frequency !c

1 7265.425
c = 1/2 N
= 1/2 8
= 8292 rad/s
1 1
1 1
A12 0.94462

Step 5 The system function Ha(s)

N /2 2
c
H a (s) =
k 1 s 2 + bk cs + 2
c

(2 k 1)
where bk = 2sin
2N
596 ! Digital Signal Processing

3
" b1 = 2 sin = 0.390 b2 = 2sin = 1.111
16 16

5 7
b3 = 2 sin = 1.662 b4 = 2 sin = 1.961
16 16

(8292)2 (8292)2
H a (s) =
s 2 + 0.39 8292 s + (8292) 2 s 2 + 1.111 8292 s + (8292)2

(8292)2 (8292)2
s 2 + 1.662 8292 s + (8292)2 s 2 + 1.961 8292 s + (8292)2

Step 6 Digital filter function H(z)

Using the bilinear transformation, we have

H (z ) = H a (s) = H a (s)
2 1 z 1
s
T 1+ z 1 1 z 1
s 10000
1 z 1

(8292) 2 (8292)2
s 2 + 3233.8 s + (8292)2 s 2 + 9212.4 s + (8292) 2
H ( z) =
(8292) 2 (8292)2
s 2 + 13781.3 s + (8292)2 s 2 + 16260.6 s + (8292) 2 1 z 1
s 10000
1 z 1

EXAMPLE 8.23 Find the filter order for the following specifications:

0.5 H( ) 1 0 /2

H( ) 0.2 3 /4

with T = 1 s. Use the impulse invariant method.

Solution: Given A1 = 0.5, 1 =

2
3
A2 = 0.2, 2 =
4
T = 1 s.
The impulse invariant transformation is to be used.
 Infinite-duration Impulse Response (IIR) Filters ! 597

" 2 2 /T 2 3 /4
= = = = 1.5
1 1 /T 1 /2
Order of the low-pass Butterworth filter N

1 1
log 1 1
1 A22 A12
N
2 log ( 2/ 1)

1 1
log 1 1
1 (0.2)2 ( 0.5)2
2 log(1.5)

1 log (24)
3.919 4
2 log (1.5)

EXAMPLE 8.24 Determine the order and the poles of a low-pass Butterworth filter that
has a –3dB bandwidth of 500 Hz and an attenuation of 40 dB at 1000 Hz.
Solution: Given
Passband edge frequency f1 = 500 Hz, " 1 = 2 f1 = 1000
Gain at passband edge k1 = –3 dB, " A1 = 10 k1 /20 = 10 3/20
= 0.707
Stopband edge frequency f2 = 1000 Hz, " 2 = 2 f2 = 2000
Gain at stopband edge k2 = – 40 dB, " A2 = 10 k2 /20 = 10 40/20
= 0.01
Let the sampling frequency fs = 2000 Hz.
The normalized frequencies are:

f1 500
1 =2 =2 = 0.5
fs 2000
f2 1000
2 =2 =2 =
fs 2000
For impulse invariant transformation,

2 2
= =2
1 1
Therefore, order of the filter is:

1 1
log 1 1
1 A22 A12
N
2 2
log
1
598 ! Digital Signal Processing

1 1
log 1 1
1 (0.01)2 (0.707)2
2 log 2

1 log{999/1}
6.64 7
2 log (2)
The pole positions are:
j (2 k 1) /2N
2
sk = ce

j (2 k +1) /14
= 1000 e 2
, k = 0, 1, 2, 3, 4, 5, 6
where !c is 3 dB cutoff frequency.

EXAMPLE 8.25 Determine the order of a Butterworth low-pass filter satisfying the
following specifications:
fp = 0.10 Hz, p = 0.5 dB
fs = 0.15 Hz, s = 15 dB; f = 1 Hz

Solution: Given
fp = 0.10 Hz, "# p = 1 = 2 fp = 2 (0.1) = 0.2
fs = 0.15 Hz, "# s = 2 = 2 fs = 2 (0.15) = 0.30
p = 1 = 0.5 dB, "# k1 = –0.5 dB, so A1 = 10 k1/20 = 10 0.5/20
= 0.944

s = 2 = 15 dB, "# k2 = –15 dB, so A2 = 10 k2 /20 = 10 15/20

= 0.177

1 1
f = 1 Hz, T= = = 1s.
f 1
1. The type of transformation is not specified. Let us use bilinear transformation.
2 2 0.3
tan 2 tan
2 = 1 2 1.019
2.
2
= T = = 1.57
2 2 0.2 0.649
1 tan 1 tan
T 2 1 2

1 1
log 1 1
1 A22 A12
3. N
2 2
log
1
 Infinite-duration Impulse Response (IIR) Filters ! 599

1 1
log 1 1
1 0.1772 0.9442
2 log (1.57)
6.16 7
So the order of the low-pass Butterworth filter is N = 7.

!"! #$%&'() *+) ,*-./0%%) 12$34%2$5) +&,6$7

For designing a Chebyshev IIR digital filter, first an analog filter is designed using the given
specifications. Then the analog filter transfer function is transformed to digital filter transfer
function by using either impulse invariant transformation or bilinear transformation.
The analog Chebyshev filter is designed by approximating the ideal frequency response
using an error function. There are two types of Chebyshev approximations. In type-1
approximation, the error function is selected such that the magnitude response is equiripple
in the passband and monotonic in the stopband. In type-2 approximation, the error function
is selected such that the magnitude function is monotonic in the passband and equiripple in
the stopband. The type-2 magnitude response is also called inverse Chebyshev response. The
type-1 design is presented in this book.
The magnitude response of type-1 Chebyshev low-pass filter is given by

2 1
Ha ( ) =
2 2
1+ cN
c

1
1 2
where is attenuation constant given by = 1
A12

A1 is the gain at the passband edge frequency 1 and cN is the Chebyshev polynomial
c

of the first kind of degree N given by

cN (x ) = {cos(N cos 1 x ), for x 1

1
= {cos( N cosh x ), for x 1

and !c is the 3 dB cutoff frequency.

The frequency response of Chebyshev filter depends on order N. The approximated
response approaches the ideal response as the order N increases. The phase response of the
Chebyshev filter is more nonlinear than that of the Butterworth filter for a given filter length
N. The magnitude response of type-1 Chebyshev filter is shown in Figure 8.9.
600 ! Digital Signal Processing

!"#$%&'()* !"#$%&'()* +),-.$,)* ./* &0-)12* 34)50,4)6* /%7&)+8

The design parameters of the Chebyshev filter are obtained by considering the low-pass
filter with the desired specifications as given below.

A1 H( ) 1 0 1

H( ) A2 2

The corresponding analog magnitude response is to be obtained in the design process.

We have
1
A12 2 2
1
1+ cN ( 1/ 2)

1
2 2
A22
1+ cN ( 1/ 2)

Assuming !c = !1, we will have cN(!1/!c) = cN(1) = 1.

Therefore, from the above inequality involving A12 , we get

1
A12 2
1+
Assuming equality in the above equation, the expression for is
1
1 2
= 1
A12

The order of the analog filter, N can be determined from the inequality for A22 .
 Infinite-duration Impulse Response (IIR) Filters ! 601

Assuming !c = !1,
1
1 1 2
cN ( 2/ 1) 1
A22
Since !2 > !1,
1
1 1 2
cosh[ N cosh 1 ( 2/ 1 )] 1
A22

1 1 1 2
cosh 1
A22
or N
cosh 1 ( 2/ 1)

Choose N to be the next nearest integer to the value given above. The values of !2 and
!1 are determined from 1 and 2 using either impulse invariant transformation or bilinear
transformation.
The transfer function of Chebyshev filters are usually written in the factored form as
given below.
N
2 2
Bk c
When N is even, H a (s) =
k 1 s 2 + bk cs + ck 2
c

N 1
2 2
B0 c Bk c
When N is odd, H a (s) =
s+ c k 1
s 2 + bk cs + ck 2
c

where
(2k 1)
b k = 2 yN sin
2N

(2 k 1)
c k = yN2 + cos2
2N
c0 = yN
1 1
1 N 1 N
1 1 2 1 1 2 1
yN = 2
+1 + 2
+1 +
2
602 ! Digital Signal Processing

For even values of N and unity dc gain filter, the parameter Bk are evaluated using the
equation:

1
H a (s ) s 0 = 2 1/2
[1 + ]
For odd values of N and unity dc gain filter, the parameter Bk are evaluated using the
equation:
H a (s) s 0 =1

!"#$%& "'& (& )"*+(#,-$.& /0$12%0$3& ',#4$*

The transfer function of the analog system can be obtained from the equation for the
magnitude squared response as:
1
H a ( s) H a ( s) =
2 s/j
1+ cN2
c

For the normalized transfer function, let us replace s/!c by sn.

1
" H a (sn ) H a ( sn ) =
1 + 2 cN2 ( jsn )

The normalized poles in the s-domain can be obtained by equating the denominator of
the above equation to zero, i.e., 1 + 2 cN2 ( jsn ) to zero.
The solution to the above expression gives us the 2N poles of the filter given by
sn = – sin x sinh y + jcos x cosh y = n + j!n
where n = 1, 2, ..., (N+1)/2 for N odd
= 1, 2, ..., N/2 for N even

(2n 1)
and x= n = 1,2, ..., N
2N
1 1 1
y= sin h n = 1, 2, ..., N
N

The unnormalized poles, s&n can be obtained from the normalized poles as shown below.
s&n = sn!c
The normalized poles lie on an ellipse in s-plane. Since for a stable filter all the poles
should lie in the left half of s-plane, only the N poles on the ellipse which are in the left half
of s-plane are considered.
For N even, all the poles are complex and exist in conjugate pairs. For N odd, one pole
is real and all other poles are complex and occur in conjugate pairs.
 Infinite-duration Impulse Response (IIR) Filters ! 603

5$%,6)& 7*"8$.9*$& '"*& #":;7(%%& .,6,4(#& /0$12%0$3& <<=& ',#4$*

The low-pass Chebyshev IIR digital filter is designed following the steps given below.
Step 1 Choose the type of transformation.
(Bilinear or impulse invariant transformation)
Step 2 Calculate the attenuation constant .
1
1 2
= 1
A12

Step 3 Calculate the ratio of analog edge frequencies !2/!1.

For bilinear transformation,

2
tan 2 tan 2
2
= T 2 = 2
2
1 tan 1 tan 1
T 2 2

For impulse invariant transformation,

2 2 /T 2
= =
1 1 /T 1

Step 4 Decide the order of the filter N such that

1 1 1
cosh 1
A22
N
1 2
cosh
1

Step 5 Calculate the analog cutoff frequency !c.

For bilinear transformation,

2
tan 1
= 1
= T 2
c 1/2N 1/2N
1 1
1 1
A12 A12

For impulse invariant transformation

1 1 /T
c = 1/2N
= 1/2N
1 1
1 1
A12 A12
604 ! Digital Signal Processing

Step 6 Determine the analog transfer function Ha(s) of the filter.

When the order N is even, Ha(s) is given by
N/2 2
Bk c
H a (s) =
k 1 s 2 + bk cs + ck 2
c

When the order N is odd, Ha(s) is given by

N 1
2 2
B0 c Bk c
H a (s) =
s + c0 c k 1 s 2 + bk cs + ck 2
c

(2 k 1)
where bk = 2yN sin
2N

(2 k 1)
ck = y 2N + cos2
2N
c0 = yN
1 1
1 N 1 N
1 1 2 1 1 2 1
yN = 2
+1 + 2
+1 +
2

For even values of N and unity dc gain filter, find Bk!s such that
1
H a (0) = 2 1/2
(1 + )
For odd values of N and unity dc gain filter, find Bk!s such that
N 1
2 Bk
=1
k 0
ck

(It is normal practice to take B0 = B1 = B2 = ... = Bk)

Step 7 Using the chosen transformation, transform H a(s) to H(z), where H(z) is the
transfer function of the digital filter.
[The high-pass, band pass and band stop filters are obtained from low-pass filter design by
frequency transformation].

!*"7$*4,$%& "'& /0$12%0$3& ',#4$*%& !>27$& "#

1. The magnitude response is equiripple in the passband and monotonic in the
stopband.
 Infinite-duration Impulse Response (IIR) Filters ! 605

2. The chebyshev type-1 filters are all pole designs.

3. The normalized magnitude function has a value of 1/ 1 + 2 at the cutoff
frequency !c.
4. The magnitude response approaches the ideal response as the value of N increases.

EXAMPLE 8.26 Design a Chebyshev IIR digital low-pass filter to satisfy the constraints.

0.707 H( ) 1, 0 0.2
H( ) 0.1, 0.5

using bilinear transformation and assuming T = 1 s.

Solution: Given
A1 = 0.707, 1 = 0.2
A2 = 0.1, 2 = 0.5
T = 1 s and bilinear transformation is to be used. The low-pass Chebyshev IIR digital filter is
designed as follows:
Step 1 Type of transformation
Here bilinear transformation is to be used.
Step 2 Attenuation constant
1 1
1 2 1 2
= 1 = 1 =1
A12 0.7072

Step 3 Ratio of analog edge frequencies, !2/!1.

Since bilinear transformation is to be used,

2 0.5
tan 2 tan
2 = 2 2
2
= T = = 3.0779
2 0.2 0.6498
1 tan 1 tan
T 2 2

Step 4 Order of the filter N

1

1 1 1 2 0.5
cosh 1 1 1 1
A22 cosh 1
1 0.12
N 1
1.669 2.
1 2 cosh 3.0779
cosh
1
606 ! Digital Signal Processing

Step 5 Analog cutoff frequency !c

2
tan 1
1 T 2 0.6498
c = 1/2 N
= 1/ 2 N 1
= 0.6498
1 1 1 4
1 1 1
A12 A12 0.7077

Step 6 Analog filter transfer function Ha(s)

N/2 2 2
Bk c B1 c
H a (s ) = =
k 1 s 2 + bk cs + ck 2
c s 2 + b1 cs + c1 2
c

1 1
1 N 1 N
1 1 2 1 1 2 1
yN = 2
+1 + 2
+1 +
2

1 1
1 2 1 2
1 1 2 1 1 2 1
= +1 + +1 +
2 12 1 12 1

1 1
1
= 2.414 2 2.414 2 = 0.455
2

(2 k 1) (2 1 1)
b1 = 2yN sin =2 0.455 sin = 0.6435
2N 2 2

(2k 1) (2 1 1)
c1 = y2N + cos2 = (0.455) 2 + cos2 = 0.707
2N 2 2
For N even,
N
2 Bk A
= 2 0.5
= 0.707
k
c
1 k (1 + )

That is B1 = c1 ' 0.707 = 0.707 ' 0.707 = 0.5.

Therefore, the system function is:
0.5(0.6498)2
H a (s ) = 2
s + (0.6435)(0.6498) s + (0.707)(0.6498)2
 Infinite-duration Impulse Response (IIR) Filters ! 607

On simplifying, we get

0.2111
H a (s) = 2
s + 0.4181 s + 0.2985

Step 7 Digital filter transfer function H(z)

0.2111
H ( z ) = H a (s ) 21 z 1 2
s s + 0.4181 s + 0.2985 1 z 1
T 1 z 1 s 2
1 z 1

0.2111
= 2
1 1
1 z 1 z
2 1
+ 0.4181 2 1
+ 0.2985
1+z 1+z

0.2111(1 + z 1 ) 2
=
5.1347 7.403 z 1 + 3.463 z 2

0.0411(1 + z 1 )2
=
1 1.441 z 1 + 0.6744 z 2

EXAMPLE 8.27 Determine the system function H(z) of the lowest order Chebyshev IIR
digital filter with the following specifications:
3 dB ripple in passband 0 (# # ( 0.2
25 dB attenuation in stopband 0.45 (# # (
Solution: Given

1 = 3 dB, " k1 = 3dB and hence A1 = 10 k1/20 = 10 3/20

= 0.707

2 = 25 dB, " k2 = 25dB and hence A 2 = 10 k2 /20 = 10 25/20

= 0.0562
1 = 0.2 and 2 = 0.45
Let T = 1 and bilinear transformation is used
1
1 2 1
Attenuation constant = 1 = 1 =1
A12 0.7072

2 0.45
tan 2 tan
Ratio of analog frequencies, 2
= T 2 = 2 = 2.628
2 0.2
1 tan 1 tan
T 2 2
608 ! Digital Signal Processing

1 1 1 2
cosh 1
A22
Order of filter N
1 2
cosh
1

1
1 1 1 2
cosh 1
1 0.05622

cosh 1{2.628}

3.569
2.20 3
1.621

2
tan 1
Analog cutoff frequency = 1
= T 2 = 1.708
c 1/ 2 N 1/ 6
1 1
1 1
A12 0.7072

Analog filter transfer function for N = 3.

2
B0 c B1 c
H a (s ) =
s + c0 c s 2 + b1 cs + c1 2
c

1 1
1 N 1 N
1 1 2 1 1 2 1
yN = 2
+1 + 2
+1 +
2

1 1
1 3 1 3
1 1 2 1 1 2 1
" yN = +1 + +1 + = 0.5959
2 12 1 12 1

c0 = yN = 0.5959
(2 1 1)
b1 = 2yN sin =2 0.5959 sin = 0.5959
2N 6

(2 1 1)
c1 = y2N + cos2 = 0.59592 + cos2 = 1.105
2N 6
 Infinite-duration Impulse Response (IIR) Filters ! 609

( N 1)/2
Bk
For N odd =1
k 0
ck

" B0 = c0 = 0.5959, B1 = c1 = 1.105

0.5959 1.708 1.105(1.708)2

H a (s) =
s + 0.5959 1.708 s 2 + 0.5959 1.708 s + 1.105(1.708)2

1.01 3.223
=
s + 1.01 s 2 + 1.01 s + 3.223
Using bilinear transformation, H(z) is given by

1.01 3.223
H (z ) = Ha (s) 1 =
s
21 z s +1.01 s 2 + 0.1 s + 3.223
T 1 z 1 1 z 1
s 2
1 z 1

3.25
= 2
1 1 1
1 z 1 z 1 z
2 1
+ 1.01 2 1
+ 0.1 2 1
+ 3.223
1+z 1+z 1+z

(3.25) (1 + z 1 )3
=
7.423 1.554 z 1 + 7.023 z 2

EXAMPLE 8.28 The specification of the desired low-pass filter is:

0.9 H( ) 1.0; 0 0.3

H( ) 0.15; 0.5
Design a Chebyshev digital filter using the bilinear transformation.
Solution: Given
A1 = 0.9, 1 = 0.3
A2 = 0.15, 2 = 0.5
The Chebyshev filter is designed as per the following steps:
Step 1 The bilinear transformation is used.
Step 2 Attenuation constant
1/2 1/2
1 1
= 1 = 1 = 0.484
A12 (0.9)2
610 ! Digital Signal Processing

Step 3 Ratio of analog edge frequencies !2/!1

2
tan 2
2
= T 2 = tan 0.25 = 1.962
2 tan 0.15
1 tan 1
T 2
Step 4 Order of the filter N
1
1
1 1 1 2
1 1 2
cosh 1 cosh 1
1
A22 0.484 0.152
N
1 2 cosh 1 1.962
cosh
1

cosh 1 13.618
2.55 = 3
cosh 1 1.962
So order of the filter is N = 3. Let T = 1 s.
Step 5 Analog cutoff frequency !c

2
tan 1
2 = 1.019
c = 1
1
= T 1 1/6
= 1.13 rad/s
1
1 2N 1 2N 1
1 1 0.92
A12 A12
Step 6 Analog transfer function Ha(s)
2
B0 c B1 c
For N = 3, H a ( s) =
s + c0 c s 2 + b1 cs + c1 2
c

1 1
1 N 1 N
1 1 2 1 1 2 1
yN = 2
+1 + 2
+1 +
2

1 1
1 3 1 3
1 1 2 1 1 2 1
= +1 + +1 +
2 (0.484) 2 0.484 (0.484) 2 0.484

1
= {1.634 0.612} = 0.511
2
 Infinite-duration Impulse Response (IIR) Filters ! 611

" c0 = yN = 0.511
(2k 1)
ck = yN2 + cos2
2N

2 2
When k = 1, c1 = yN + cos = (0.511)2 + 0.75 = 1.011
6
(2k 1)
bk = 2yN sin
2N
1
When k = 1, b1 = yN + sin =2 0.511 = 0.511
6 2

B0 (1.13) B1 (1.13)2
" H a (s ) =
s + 0.511 1.13 s 2 + 0.511 1.13s + 1.011(1.13)2

B0 B1 (1.442)
When s = 0, H a (s) = Ha (0) = = 1.935 B0 B1
(0.511)(1.13)(1.011)(1.13)2

B0 (1.13) B1 (1.13)2
" H a (s ) =
s + 0.511 1.13 s 2 + 0.511 1.13s + 1.011(1.13)2

Let Ha(0) = 1, " 1.935 B0B1 = 1

1
Let B0 = B1, B02 = = 0.516 or B0 = 0.718
1.935
" B0 = B1 = 0.86
0.516 (1.442)
H a (s) =
(s + 0.577) ( s 2 + 0.577s + 1.29)
0.744
=
(s + 0.577) ( s 2 + 0.577s + 1.29)
Step 7 Digital transfer function

H (z ) = H a (s) 21 z 1
= H a ( s)
s=
T 1 z 1 1 z 1
s 2
1+ z 1

0.744
=
( s + 0.577) (s 2 + 0.577s + 1.29) s 2
1 z 1
1 z 1
612 ! Digital Signal Processing

0.744
= 2
1 1 1
1 z 1 z 1 z
2 1
+ 0.577 2 1
+ 0.577 2 1
+ 1.29
1+ z 1+z 1+z

0.744 (1 + z 1 )3
=
(2.577 1.423 z 1 ) (6.83 5.42 z 1
+ 3.75)

EXAMPLE 8.29 Determine the system function of the lowest order Chebyshev digital
filter that meets the following specifications.
2 dB ripple in the passband 0 (# # ( 0.25
Atleast 50 dB attenuation in stopband 0.4 (# # (
Solution: Given
Ripple in passband = 2 dB, i.e. k1 = –2 dB " A1 = 10 k1 /20 = 10 2/20
= 0.794

Attenuation in stopband = 50 dB, i.e. k2 = –50 dB " A2 = 10 k2 /20 = 10 50/20

= 0.0031

" A1 = 0.794, 1 = 0.25

A2 = 0.003, 2 = 0.4
The Chebyshev filter is designed as per the following steps:
Step 1 Type of transformation
Let us choose bilinear transformation.
Step 2 Attenuation constant
1/2 1/ 2
1 1
= 2 1 = 1 = 0.765
A1 0.794 2

Step 3 Ratio of analog edge frequencies, !2/!1

2
tan 2
2
= T 2 = tan 0.4 / 2 = 1.453 = 1.754
2 tan 0.25 / 2 0.828
1 tan 1
T 2
Step 4 Order of the filter N
1

1 1 1 2
cosh 1
A22
N
1 2
cosh
1
 Infinite-duration Impulse Response (IIR) Filters ! 613

1
1 1 1 2
cosh 1
0.765 (0.0031)2
1
cosh 1.754

6.718
5.786 6
1.161
" N=6
Step 5 Analog cutoff frequency !c

2
tan 1
1 T 2 0.828
c = 1
= 1
= 1/12
= 0.866 rad/s
1
1 2N 1 2N 1
1 1 0.7942
A12 A12

Step 6 Analog transfer function Ha(s)

2 2 2
B1 c B2 c B3 c
For N = 6, H a (s) =
s 2 + b1 cs + c1 2
c s 2 + b2 cs + c2 2
c s 2 + b3 cs + c3 2
c

1 1
1 N 1 N
1 1 2 1 1 2 1
yN = 2
+1 + 2
+1 +
2

1 1
1 6 1 6
1 1 2 1 1 2 1
= +1 + +1 +
2 (0.765) 2 0.765 (0.756) 2 0.765

1
= {1.197 0.83} = 0.183
2
" c0 = yN = 0.183
(2 k 1)
ck = yN2 + cos2
2N
(2 1 1)
c1 = yN2 + cos2 = (0.183) 2 + cos2 = 0.9664
2 6 12
614 ! Digital Signal Processing

(2 1 1)
b1 = 2yN sin =2 0.183 sin = 0.094
2 6 12

(2 2 1) 3
c2 = yN2 + cos2 = (0.183) 2 + cos2 = 0.5334
2 6 12

(2 2 1) 3
b2 = 2yN sin =2 0.183 sin = 0.258
2 6 12

(2 3 1)
c3 = yN2 + cos2 = 0.1
2 6

(2 3 1)
b3 = 2yN sin = 0.353
2 6

Let B1 = B2 = B3 and let Ha(0) = 1.

6
B1 B2 B3 c
" 6
=1
c1c2 c3 c

1 1
" B1 = B2 = B3 = (c1c2 c3 )3 = (0.964 0.533 0.1) 3 = 0.371

0.371 (0.866)2
" H a (s ) = 2
s + 0.094 0.866 s + 0.966 (0.866)2

0.371 (0.866) 2
s 2 + 0.258 0.866 s + 0.533 (0.866)2

0.371 (0.866)2
s 2 + 0.353 0.866 s + 0.1 (0.866)2

0.278 0.278 0.278

= 2 2 2
s + 0.018s + 0.724 s + 0.223s + 0.399 s + 0.305s + 0.074

Step 7 Digital filter transfer function H(z) taking T = 1s.

H ( z) = H a (s) = H a (s)
2 1 z 1
s=
T 1 z 1 1 z 1
s 2
1 z 1
 Infinite-duration Impulse Response (IIR) Filters ! 615

0.278
H (z ) =
2
1 1
1 z 1 z
2 1
+ 0.081 2 1
+ 0.724
1+z 1+z

0.278
2
1 1
1 z 1 z
2 1
+ 0.223 2 1
+ 0.399
1+z 1+z

0.278
2
1 1
1 z 1 z
2 1
+ 0.305 2 1
+ 0.074
1+z 1+z

0.278(1 + z 1 )2 0.278(1 + z 1 ) 2
=
4.886 6.552 z 1 + 4.562 z 2
4.845 7.202 z 1 + 3.953 z 2

0.278(1 + z 1 ) 2
4.684 7.852 z 1 + 3.464 z 2

EXAMPLE 8.30 Find the Chebyshev filter order for the following specifications:

0.6 H( ) 1; 0
2
3
H( ) 0.2 ;
2
with T = 1 s. Use the impulse invariant transformation.
Solution: Given

A1 = 0.6 = 0.774, 1 =
2
3
A2 = 0.25, 2 =
2
T = 1s and impulse invariant transformation is to be used.
616 ! Digital Signal Processing

The order of the filter is found as follows:

1 1
1 2 1 2
Attenuation constant = 1 = 1 = 0.818
A12 (0.774) 2

1
Analog passband edge frequency 1 = =
T 2

2 3
Analog stopband edge frequency 2 = =
T 2

2 3 /2
Ratio of edge frequencies = =3
1 /2

1 1 1 2
cosh 1
A22
Order of the filter N
1 2
cosh
1

1
1 1 1 2
cosh 1
0.818 0.252

cosh 1 (3)
1.268 2
So the order of the filter is N = 2.

EXAMPLE 8.31 Find the filter order for the following specifications:

0.5 H( ) 1; 0
2
3
H( ) 0.2;
4
with T = 1 s. Use the impulse invariant method.
Solution: Given
A1 = 0.5 = 0.707, 1 =
2
3
A2 = 0.2, 2 =
4
T = 1s and impulse invariant transformation is to be used.
 Infinite-duration Impulse Response (IIR) Filters ! 617

Since the type of filter is not specified, let us find the order of Chebyshev type-1 filter.
1 1
1 2 1 2
Attenuation constant = 1 = 1 =1
A12 (0.707)2

2 2 /T 3 /4
Ratio of analog edge frequencies = = = 1.5
1 1 /T /2

1
1
1 1 1 2
1 1 2
cosh 1 cosh 1
1
A22 1 0.22
Order of the filter N
1 2 cosh 1 (1.5)
cosh
1

2.271
2.36 3
0.962
The order of the filter N = 3.

EXAMPLE 8.32 Determine the lowest order of Chebyshev filter that meets the following
specifications:
(i) 1 dB ripple in the passband 0 0.3
(ii) Atleast 60 dB attenuation in the stopband 0.35
Use the bilinear transformation.
Solution: Given 1 = 0.3 , 2 = 0.35

1 dB ripple, so 1 = 1 dB or k1 = –1 dB " A1 = 10 k1 /20 = 10 1/20

= 0.891
k /20 60/20
60 dB attenuation, so 2 = 60 dB or k2 = – 60 dB " A2 = 10 2 = 10 = 0.001

Step 1 Bilinear transformation is to be used.

1 1
1 2 1 2
Step 2 Attenuation constant = 1 = 1 = 0.509
A12 (0.891) 2

2 0.35
tan 2 tan
Step 3 Ratio of analog edge frequencies 2
= T 2 = 2 = 1.2
2 0.3
1 tan 1 tan
T 2 2
618 ! Digital Signal Processing

1
1
1 1 1 2
1 1 2
cosh 1 cosh 1
1
A22 0.509 0.0012
Step 4 Order of the filter N
1 2 cosh 1 (1.2)
cosh
1

13.338 14
So the lowest order of the filter is N = 14.

EXAMPLE 8.33 Determine the lowest order of Chebyshev filter for the following specifications.
(i) Maximum passband ripple is 1 dB for !# ( 4 rad/s
(ii) Stopband attenuation is 40 dB for !# $ 4 rad/s
Solution: Using the impulse invariant transformation,

2 2 /T 2 4
= = = =1
1 1 /T 1 4

p = 1 dB, k1 = 1dB and A1 = 10 k1 /20 = 10 1/ 20

= 0.891

2 = 40 dB, k2 = 40 dB and A2 = 10 k2 /20 = 10 40 / 20

= 0.01
1 1
1 2 1 2
Attenuation constant = 1 = 1 = 0.509
A12 (0.891)2

1
1
1 1 1 2
1 1 2
cosh 1 cosh 1
1
A22 0.509 0.012
Order of the filter N
1 2 cosh 1 (1)
cosh
1

5.97
=
0
So the order of the filter required is N = )*

!"8 &(5$7%$) 12$34%2$5) +&,6$7%

Inverse Chebyshev filters are also called type-2 Chebyshev filters. A low-pass inverse
Chebyshev filter has a magnitude response given by
 Infinite-duration Impulse Response (IIR) Filters ! 619

cN ( 2/ )
H( ) = 1
2
[1 + c 2N ( 2/ )]2
where is a constant and !c is the 3 dB cutoff frequency. The Chebyshev polynomial cN(x)
is given by
cN ( x ) = cos( N cos 1 x ), for x 1
1
= cosh ( N cosh x ), for x 1
The magnitude response of the inverse Chebyshev filter is shown in Figure 8.10. The
magnitude response has maximally flat passband and equiripple stopband, just the opposite
of the Chebyshev filters response. That is why type-2 Chebyshev filters are called the inverse
Chebyshev filters.

!"#$%&' ()+, !"#$%&'()* +),-.$,)* ./* &4)* 7.91-",,* %$6)+,)* 34)50,4)6* /%7&)+8

The parameters of the inverse Chebyshev filter are obtained by considering the low-
pass filter with the desired specifications:
0.707 H( ) 1; 0 c

H( ) A2 ; 2
The attenuation constant is given by
A2
= 1
2 2
(1 A2 )
The order of the filter N is given as:
1

1 1 1 2
cosh 1 cosh 1
A22
N =
cosh 1 ( 2/ c) cosh 1 ( 2/ c)

The value of N is chosen to be the nearest integer greater than the value given above.
620 ! Digital Signal Processing

!"9: $,,&/6&1) +&,6$7%

The elliptic filter is sometimes called the Cauer filter. This filter has equiripple passband and
stopband. Among the filters discussed so far, for a given filter order, pass band and stop
band deviations, elliptic filters have the minimum transition bandwidth. The magnitude
response of an elliptic filter is given by

2 1
H( ) =
1 + 2 UN ( / c)

where UN(x) is the Jacobian elliptic function of order N and is a constant related to the
passband ripple.

!"99 +7$;<$(14) 670(%+*7=06&*(

Basically there are four types of frequency selective filters, viz. low-pass, high-pass, band pass
and band stop. In Figure 8.11, the frequency response of the ideal case is shown in solid
lines and practical case in dotted lines.

!"#$%&' ()++ :+);')$<0* +),-.$,)* ./* =">* ?.91-",,* /%7&)+@* =5>* A%#41-",,* /%7&)+@* =<>* B"$(* -",,* /%7&)+* "$(
=(>* B"$(* ,&.-* /%7&)+8
 Infinite-duration Impulse Response (IIR) Filters ! 621

In the design techniques discussed so far, we have considered only low-pass filters.
This low-pass filter can be considered as a prototype filter and its system function Hp(s) can
be determined. The high-pass or band pass or band stop filters are designed by designing a
low-pass filter and then transforming that low-pass transfer function into the required filter
function by frequency transformation. Frequency transformation can be accomplished in two
ways.
(1) Analog frequency transformation
(2) Digital frequency transformation

!"99"9 0>?@AB) +CDEFD>GH) 6C?>IJACK?LMA>

In the analog frequency transformation, the analog system function Hp(s) of the prototype
filter is converted into another analog system function H(s) of the desired filter (a low-pass
filter with another cutoff frequency or a high-pass filter or a band pass filter or a band stop
filter). Then using any of the mapping techniques (impulse invariant transformation or
bilinear transformation) this analog filter is converted into the digital filter with a system
function H(z).
The frequency transformation formulae used to convert a prototype low-pass filter into
a low-pass (with a different cutoff frequency), high-pass, band pass or band stop are given in
Table 8.2. Here !c is the cutoff frequency of the low-pass prototype filter. !c* cutoff
frequency of new low-pass filter or high-pass filter and !1 and !2 are the cutoff frequencies
of band pass or band stop filters.

-./01'()2 C$"7.#*:+);')$<0*D+"$,/.+E"&%.$

Type Transformation

s
Low-pass s c *
c

*
High-pass s c
c
s

Band pass s2 + 1 2
s c
s( 2 1)

Band stop s( 2 1)
s c
s2 + 1 2

!0 is the centre frequency !0 = 1 2

0
Quality factor Q =
2 1
622 ! Digital Signal Processing

1
EXAMPLE 8.34 A Prototype low-pass filter has the system function H p (s) = .
s 2 + 3s + 2
Obtain a band pass filter with !0 = 3 rad/s and Q = 12.

0
Solution: We know that the centre frequency 0 = 1 2 and quality factor Q = .
2 1
From Table 8.2, we have the low-pass to band pass transformation

s2 + 1 2 s 2 + 20
s c = c
s( 2 1) s( 0 /Q)

s 2 + 32 s2 + 9
s c =4 c
s(3/12) s

Therefore, the transfer function of band pass filter is:

H ( s ) = H p (s ) s2 +9
s 4 c
s

1
= 2
s2 + 9 s2 + 9
4 c +3 4 c +2
s s

1 s2
= 2
16 c s 4 + 0.75 c s 3 + (18 2
c + 0.125) s 2 + 6.75 cs + 81 2
c

EXAMPLE 8.35 Transform the prototype low-pass filter with system function

H ( s) = c
into a high-pass filter with a cutoff frequency !c*.
s+2 c

Solution: We know that the desired transformation from low-pass to high-pass is

*
c
s c .
s

c s
Thus, we have H hpf (s) = = *
c c 2s + c
+2 c
s
 Infinite-duration Impulse Response (IIR) Filters ! 623

!"99"N #MBML?@) +CDEFD>GH) 6C?>IJACK?LMA>

As in the analog domain, frequency transformation is possible in the digital domain also. The
frequency transformation is done in the digital domain by replacing the variable z–1 by a
function of z–1, i.e., f(z–1). This mapping must take into account the stability criterion. All the
poles lying within the unit circle must map onto itself and the unit circle must also map onto
itself. Table 8.3 gives the formulae for the transformation of the prototype low pass digital
filter into a digital low-pass, high-pass, band pass or band stop filters.

-./01'()3 F%#%&"7*:+);')$<0*D+"$,/.+E"&%.$

Type Transformation Design parameter

*
z 1 sin ( c c )/2
1
Low-pass z =
1 *
1 z sin ( c + c )/2

*
1 z 1+ cos ( c c )/2
High-pass z 1
=
1+ z cos ( c + *
c )/2

z 2
1z
1
+ 2 k
Band pass z 1 2
1 =
z 2 1
+1 (k + 1)
2 1z

(k 1)
2 =
(k + 1)

cos [( 2 + 1 )/2]
=
cos [( 2 1 )/2]

2 1 c
k = cot tan
2 2

z 2
z 1
+ 2
Band stop z 1 1 2 1 =
2 1
+1 (k + 1)
2z 1 z
(1 k )
2 =
(1 + k )

cos ( 2 + 1 )/2
=
cos ( 2 1 )/2

2 1 c
k = tan tan
2 2
624 ! Digital Signal Processing

The frequency transformation may be accomplished in any of the available two

techniques, however, caution must be taken to which technique to use. For example, the
impulse invariant transformation is not suitable for high-pass or bandpass filters whose
resonant frequencies are higher. In such a case, suppose a low-pass prototype filter is
converted into a high-pass filter using the analog frequency transformation and transformed
later to a digital filter using the impulse invariant technique. This will result in aliasing
problems. However, if the same prototype low-pass filter is first transformed into digital
filter using the impulse invariant technique and later converted into a high-pass filter using
the digital frequency transformation, then it will not have any aliasing problem. Whenever
the bilinear transformation is used, it is of no significance whether analog frequency
transformation is used or digital frequency transformation. In this case, both analog and
digital frequency transformation techniques will give the same result.

%2*76) ;<$%6&*(%) -&62) 0(%-$7%)

1. What are the basic types of digital filters and on what basis are they classified?
Ans. There are following two basic types of digital filters:
(i) Finite impulse response filters (FIR filters)
(ii) Infinite impulse response filters (IIR filters).
They are classified on the basis of the number of sample points used to determine
the unit sample (i.e. impulse) response of a LTI discrete-time system.
2. Define an IIR filter.
Ans. An IIR (Infinite-duration Impulse Response) filter is a digital filter designed
by considering all the infinite samples of impulse response.
3. Define an FIR filter.
Ans. An FIR (Finite-duration Impulse Response) filter is a digital filter designed
by considering only a finite number of samples of impulse response.
4. Distinguish between IIR and FIR filters.
Ans. The filter design starts from ideal frequency response. The desired impulse
response which consists of infinite number of samples is obtained by taking the
inverse Fourier transform of the ideal frequency response of the system. The digital
filters designed by using all the infinite samples of the impulse response are called
IIR filters and the digital filters designed by using only a finite number of samples
of the impulse response are called FIR filters.
 Infinite-duration Impulse Response (IIR) Filters ! 625

5. Compare IIR and FIR filters.

Ans. The IIR and FIR filters are compared as follows:

IIR filter FIR filter

(i) Design is based on all the infinite (i) Design is based on only a finite number
samples of the impulse response. of samples of impulse response.
(ii) The impulse response cannot be (ii) The impulse response can be directly
directly converted to digital filter converted to digital filter transfer function.
transfer function.
(iii) The digital filter cannot be directly (iii) The digital filter can be directly designed.
designed. First an analog filter is to
be designed and then it has to be
transformed to a digital filter.
(iv) The specifications include the desired (iv) The specifications include the desired
characteristics for magnitude response characteristics for both magnitude and
only. phase response.
(v) Linear phase characteristics cannot (v) Linear phase characteristics can be
be achieved. achieved.

6. Based on frequency response how are filters classified?

Ans. Based on frequency response, filters are classified as low-pass filters, high-
pass filters, bandpass filters and bandstop filters.
7. What are the requirements for an analog filter to be causal and stable?
Ans. For an analog filter to be causal and stable, the requirements are as follows
(i) The analog filter transfer function Ha(s) should be a rational function of s and
the coefficients of s should be real.
(ii) The poles should lie on the left half of s-plane.
(iii) The number of zeros should be less than or equal to the number of poles.
8. What are the requirements for a digital filter to be causal and stable?
Ans. The requirements for a digital filter to be causal and stable are as follows:
(i) The digital filter transfer function H(z) should be a rational function of z and
the coefficients of z should be real.
(ii) The poles should lie inside the unit circle in z-plane.
(iii) The number of zeros should be less than or equal to the number of poles.
9. How a digital filter is designed?
Ans. For designing a digital IIR filter, first an equivalent analog filter is designed
using any one of the approximation technique and the given specifications. The
result of the analog filter design will be an analog transfer function Ha(s). The
analog filter transfer function is transformed to digital filter transfer function H(z)
using either bilinear or impulse invariant transformation.
10. Mention any two techniques for digitizing the transfer function of an analog filter.
Ans. Two techniques for digitizing the transfer function of an analog filter are:
(i) Impulse invariant transformation, and (ii) Bilinear transformation.
626 ! Digital Signal Processing

11. Compare the analog and digital filters.

Ans. The analog and digital filters are compared as follows:

Analog filter Digital filter

(i) It operates on analog signals (i)
It operates on digital samples (or sampled
(or actual signals) version) of the signal.
(ii) It is governed by the linear differential (ii)
It is governed by the linear difference
equation. equation.
(iii) It consists of electrical components (iii)
It consists of adders, multipliers and
like resistors, inductors and capacitors.delay elements implemented in digital
logic (either in hardware or software
or both).
(iv) In analog filters, the approximation (iv) In digital filters, the filter coefficients
problem is solved to satisfy the are designed to satisfy the desired
desired frequency response. frequency response.

12. Mention the important features of IIR filters.

Ans. The important features of IIR filters are as follows:
(i) The physically realizable IIR filters do not have linear phase.
(ii) The IIR filter specifications include the desired characteristics for the
magnitude response only.
13. What is the impulse invariant transformation?
Ans. The transformation of analog filter to digital filter without modifying the
impulse response of the filter is called impulse invariant transformation (i.e. in this
transformation, the impulse response of the digital filter will be the sampled
version of the impulse response of the analog filter).
14. What is the main objective of impulse invariant transformation?
Ans. The main objective of impulse invariant transformation is to develop an IIR
filter transfer function whose impulse response is the sampled version of the impulse
response of the analog filter. Therefore, the frequency response characteristics of the
analog filter is preserved.
15. How analog poles are mapped to digital poles in the impulse invariant transfor-
mation (Bilinear transformation)?
Ans. In the impulse invariant transformation (or in bilinear transformation) the
mapping of analog poles to digital poles is as follows:
(i) The analog poles on the imaginary axis of s-plane are mapped onto the unit
circle in the z-plane.
(ii) The analog poles on the left half of s-plane are mapped into the interior of
unit circle in z-plane.
(iii) The analog poles on the right half of s-plane are mapped into the exterior of
unit circle in z-plane.
 Infinite-duration Impulse Response (IIR) Filters ! 627

16. What is the importance of poles in filter design?

Ans. The importance of poles in filter design is the stability of a filter is related
to the location of the poles. For a stable analog filter the poles should lie on the
left half of s-plane. For a stable digital filter the poles should lie inside the unit
circle in the z-plane.
17. What is aliasing?
Ans. The phenomena of high frequency sinusoidal components acquiring the
identity of low frequency sinusoidal components after sampling is called aliasing
(i.e., aliasing is higher frequencies impersonating lower frequencies). The aliasing
problem will arise if the sampling rate does not satisfy the Nyquist sampling
criteria.
18. What is bilinear transformation?
Ans. The bilinear transformation is a conformal mapping that transforms the
s-plane to z-plane. In this mapping, the imaginary axis of s-plane is mapped into
the unit circle in z-plane, the left half of s-plane is mapped into interior of unit
circle in z-plane, and the right half of the s-plane is mapped into exterior of unit
circle in z-plane. The bilinear mapping is a one-to-one mapping and it is
accomplished when
1
2 1 z
s= 1
T 1+z

19. What is frequency warping?

Ans. The distortion in frequency axis introduced when the s-plane is mapped into
z-plane using the bilinear transformation, due to the nonlinear relation between
analog and digital frequencies is called frequency warping.
20. What are the advantages and disadvantages of bilinear transformation?
Ans. The advantages and disadvantages of bilinear transformation are:
Advantages
(i) The bilinear transformation is one-to-one mapping.
(ii) There is no aliasing and so the analog filter need not have a band limited
frequency response.
(iii) The effect of warping on amplitude response can be eliminated by prewarping
the analog filter.
(iv) The bilinear transformation can be used to design digital filters with
prescribed magnitude response with piecewise constant values.
Disadvantages
(i) The nonlinear relationship between analog and digital frequencies introduces
frequency distortion which is called frequency warping.
(ii) Using the bilinear transformation, a linear phase analog filter cannot be
transformed to a linear phase digital filter.
628 ! Digital Signal Processing

21. What is prewarping?

Ans. In IIR filter design using bilinear transformation, the conversion of the
specified digital frequencies to analog equivalent frequencies is called prewarping.
The prewarping is necessary to eliminate the effect of warping on amplitude
response.
22. Explain the technique of warping.
Ans. In IIR filter design using bilinear transformation, the specified digital
frequencies are converted to analog equivalent frequencies, which are called
prewrap frequencies. Using the prewarp frequencies, the analog filter function is
designed and then it is transformed to digital filter transfer function.
23. How bilinear transformation is preformed?
1
2 1 z
Ans. The bilinear transformation is performed by substituting s = 1
in
T 1+z
the analog filter transfer function, i.e. H ( z) = [H a (s)] s 2 1 z 1 .
T 1 z 1

24. Compare the impulse invariant and bilinear transformation.

Ans. The impulse invariant and bilinear transformations are compared as follows:

Impulse invariant transformation Bilinear transformation

(i) It is many-to-one mapping. (i) It is one-to-one mapping.
(ii) The relation between analog and (ii) The relation between analog and digital
digital frequency is linear. frequency is nonlinear.
(iii) To prevent the problem of aliasing, (iii) There is no problem of aliasing and so
the analog filters should be band the analog filter need not be band limited.
limited.
(iv) The magnitude and phase responses (iv) Due to the effect of warping, the phase
of analog filter can be preserved by response of analog filter cannot be
choosing low sampling time or high preserved. But the magnitude response
sampling frequency. can be preserved by prewarping.

25. What is the relation between analog and digital frequencies in impulse invariant
transformation?
Ans. The relation between analog and digital frequencies in impulse invariant
transformation is given by
Digital frequency = Analog frequency ' Sampling time period
i.e. = !T
26. What is the relation between digital and analog frequency in the bilinear transformation?
Ans. In bilinear transformation, the digital frequency is given by
T 1
Digital frequency = 2 tan
2
where, ! = Analog frequency, and T = Sampling time period.
 Infinite-duration Impulse Response (IIR) Filters ! 629

27. Why impulse invariant transformation is not considered to be one-to-one?

Ans. In impulse invariant transformation any strip of width 2 /T in the s-plane for
values of s in the range (2k – 1) /T (# ! ( (2k + 1) /T (where k is an integer) is
mapped into the entire z-plane. The left half portion of each strip in s-plane maps
into the interior of the unit circle in z-plane, right half portion of each strip in
s-plane maps into the exterior of the unit circle in z-plane and the imaginary axis of
each strip in s-plane maps into the unit circle in z-plane. So the entire s-plane is
mapped infinite number of times on to the entire z-plane. Hence the impulse
invariant transformation is many-to-one and not one-to-one.
28. What is Butterworth approximation?
Ans. Butterworth approximation is one in which the error function is selected
such that the magnitude response is maximally flat at the origin (i.e., at ! = 0) and
monotonically decreasing with increasing !*
29. How the poles of Butterworth transfer function are located in s-plane?
Ans. The poles of the normalized Butterworth transfer function symmetrically lie
on an unit circle in s-plane with angular spacing of /N.
30. Write the magnitude function of low-pass Butterworth filter?
Ans. The magnitude function of low-pass Butterworth filter is given by

1
Ha ( ) =
2N

1+
c

where, !c = Cutoff frequency, N = Order of the filter

31. How the order of the filter affects the frequency response of Butterworth filter?
Ans. The magnitude response of the Butterworth filter approaches the ideal
response as the order of the filter is increased.
32. Write the transfer function of unnormalized Butterworth low-pass filter.
Ans. The transfer function of unnormalized Butterworth low-pass filter Ha(s) is
N/2 2
When N is even, H a (s ) = c

k 1 s 2 + bk cs + 2
c

( N 1)/2 2
c c
When N is odd, H a (s) =
s+ c k 1 s 2 + bk cs + 2
c

(2 k 1)
where, bk = 2sin
2N
N = Order of filter
! !c = Analog cutoff frequency
630 ! Digital Signal Processing

33. How will you choose the order N for a Butterworth filter?
Ans. The orders N for a Butterworth filter is chosen such that

1 1
log 1 1
1 A22 A12
N
2 2
log
1

34. Write the properties of Butterworth filter.

Ans. The properties of Butterworth filter are as follows:
(i) The Butterworth filters are all pole designs.
(ii) At the cutoff frequency !c, the magnitude of normalized Butterworth filter is
1/ 2 .
(iii) The filter order N, completely specifies the filter and as the value of N increases
the magnitude response approaches the ideal response.
(iv) The magnitude is maximally flat at the origin and monotonically decreasing
with increasing !.
35. What is Chebyshev approximation?
Ans. Chebyshev approximation is one in which the approximation function is
selected such that the error is minimized over a prescribed band of frequencies.
36. What is type-1 Chebyshev approximation?
Ans. Type-1 Chebyshev approximation is one in which the error function is
selected such that the magnitude response is equiripple in the passband and
monotonic in the stopband.
37. What is type-2 Chebyshev approximation?
Ans. Type-2 Chebyshev approximation is one in which the error function is
selected such that the magnitude response is monotonic in the passband and
equiripple in the stopband. The type-2 Chebyshev response is called inverse
Chebyshev response.
38. Write the expression for the magnitude response of Chebyshev low-pass filter.
Ans. The magnitude response of type-1 Chebyshev low-pass filter is given by

1
Ha ( ) =
2
1+ C N2
c

where = attenuation constant

CN = Chebyshev polynomial of the first kind of degree N.

c
 Infinite-duration Impulse Response (IIR) Filters ! 631

39. How the order of the filter affects the frequency response of Chebyshev filter?
Ans. The magnitude response of type-1 Chebyshev filter approaches the ideal
response as the order of the filter increases.
40. Write the transfer function of unnormalized Chebyshev low-pass filter?
Ans. The transfer function Ha(s) of unnormalized type-1 Chebyshev low-pass
filter is given as:
N
2 2
Bk c
When N is even, H a (s) =
k 1 s 2 + bk cs + ck 2
c

N 1
2
B0 c 2 Bk c
When N is odd, H a (s) =
s + c0 c k 1 s 2 + bk cs + ck 2
c

(2 k 1) (2 k 1)
where bk = 2 yN sin ; ck = yN2 + cos2 ; c0 = yN
2N 2N

1 1
1 N 1 N
1 1 2 1 1 2 1
yN = 2
+1 + 2
+1 +
2

41. How will you determine the order N of Chebyshev filter?

Ans. The order N of a Chebyshev filter is such that
1

1 1 1 2
cosh 1
A22
N
1 2
cosh
1

where = Attenuation constant

A2 = Gain at stopband edge frequency,
!2 and !1 = Analog stopband and passband edge frequencies.
42. How the poles of Chebyshev transfer function are located in s-plane?
Ans. The poles of the Chebyshev transfer function symmetrically lie on an ellipse
in s-plane.
43. Write the properties of Chebyshev type-1 filters?
Ans. The properties of Chebyshev type-1 filter are as follows:
(i) The magnitude response is equiripple in the passband and monotonic in the
stopband.
(ii) The type-1 Chebyshev filters are all pole designs.
632 ! Digital Signal Processing

(iii) The normalized magnitude function has a value of 1/ 1 + 2 at the cutoff

frequency !c.
(iv) The magnitude response approaches the ideal response as the value of N
increases.
44. Compare the Butterworth and Chebyshev type-1 filters.
Ans. The Butterworth and Chebyshev type-1 filters are compared as follows:

Butterworth Chebyshev type-1

(i) All pole design. (i) All pole design.
(ii) The poles lie on a circle in s-plane. (ii) The poles lie on an ellipse in s-plane.
(iii) The magnitude response is maximally (iii) The magnitude response is equiripple
flat at the origin and monotonically in passband and monotonically decreasing
decreasing function of !. in the stopband.
(iv) The normalized magnitude response (iv) The normalized magnitude response
has a value of 1/ 2 at the cutoff has a value of 1/ 1 + 2 at the cutoff
frequency !c. frequency !c.
(v) Only a few parameters have to be (v) A large number of parameters have to
calculated to determine the transfer be calculated to determine the transfer
function. function.

7$5&$-) ;<$%6&*(%)
1. Compare analog and digital filters. State the advantages of digital filters over
analog filters.
2. Define infinite impulse response and finite impulse response filters and compare.
3. Justify the statement IIR filter is less stable and give reason for it.
4. Describe digital IIR filter characterization in time domain.
5. Describe digital IIR filter characterization in z-domain.
6. Discuss the impulse invariant method.
7. What are the limitations of impulse invariant method?
8. Compare impulse invariant and bilinear transformation methods.
9. Discuss the magnitude and phase responses of digital filters.
10. Explain method of constructing Butterworth circle in the z-plane using the bilinear
transformation method.
11. Compare Butterworth and Chebyshev approximations.
12. Discuss the magnitude characteristics of an analog Butterworth filter and give its
pole locations. Discuss about the pole location for the digital Chebyshev filters.
13. What is frequency warping? How it will arise?
 Infinite-duration Impulse Response (IIR) Filters ! 633

14. What is warping effect? Discuss influence of warping effect on amplitude response
and phase response of a derived digital filter from a corresponding analog filter.
15. Discuss the concept of frequency transformation in analog domain.
16. Discuss the digital frequency transformation.
17. Obtain transformation for Butterworth filters between s and z using the bilinear
transformation.

+&,,) &() 62$) 3,0(O%)

1. Filters designed by considering samples of the impulse response are called
IIR filters.
2. The physically realizable IIR filters do not have phase.
3. The IIR filter specifications includes the desired characteristics for the
response only.
4. Filters designed by considering samples of the impulse response are called
FIR filters.
5. The impulse response is obtained by taking the inverse Fourier transform of the
.
6. The bandwidth of the discrete signal is limited by the .
7. The popular methods for design of IIR digital filters uses the technique of
an analog filter into an digital filter.
8. The bandwidth of a real discrete signal is the sampling frequency.
9. The three techniques used to transform an analog filter to digital filter are
, and .
10. The two properties which are to be preserved in analog to digital transformation are
and .
11. The tolerance in the passband and stopband are called .
12. In transformation the impulse response of digital filter is the sampled
version of the impulse response of analog filter.
13. In impulse invariant (bilinear) transformation, the poles of s-plane are
mapped into the interior of unit circle in z-plane.
14. In impulse invariant (bilinear) transformation, the right half poles of s-plane are
mapped into the of unit circle in z-plane.
15. In impulse invariant (bilinear) transformation, the poles on the imaginary axis of
s-plane are mapped into the in z-plane.
16. In impulse invariant transformation any strip of width in s-plane is
mapped into the entire z-plane.
17. The phenomenon of high frequency components acquiring the identity of low
frequency components is called .
634 ! Digital Signal Processing

18. is higher frequencies impersonating lower frequencies.

19. Aliasing occurs only in transformation.
20. The impulse invariant mapping is mapping, whereas bilinear mapping is a
mapping.
21. The due to nonlinear relationship between analog and digital frequencies
is called frequency warping.
22. In bilinear transformation, the effect of warping on can be eliminated by
the analog filter.
23. A linear phase analog filter cannot be transformed into a linear phase digital filter
using transformation.
24. The two popular techniques used to approximate the ideal frequency response are
and approximations.
25. In approximation, the magnitude response is maximally flat at the origin
and monotonically decreases with increasing frequency.
26. At the cutoff frequency, the magnitude of the Butterworth filter is times
the maximum value.
27. In approximation, the magnitude response is equiripple in the passband
and monotonic in the stopband.
28. In approximation, the magnitude response is monotonic in the passband
and equiripple in the stopband.
29. The type-2 Chebyshev response is also called response.
30. In Chebyshev approximation, the normalized magnitude response has a value of
at the cutoff frequency.

*3P$16&5$) 64/$) ;<$%6&*(%)

1. The condition for a digital filter to be causal and stable is

(a) h( n) = 0 for n 0 and h ( n)

n

(b) h( n) = 0 for n 0 and h( n)

n

(c) h( n) = 0 for n 0 and h ( n)

n

(d) h( n) = 0 for n 0 and h( n)

n
 Infinite-duration Impulse Response (IIR) Filters ! 635

2. IIR filters are

(a) recursive type (b) non-recursive type
(c) neither recursive nor non-recursive (d) none of the above
3. For same set of specifications
(a) IIR filter requires fewer filter coefficients than an FIR filter
(d) FIR filter requires fewer filter coefficients than IIR filter
(c) FIR and IIR filters require same number of filter coefficients
(d) none of the above
4. In the impulse invariant transformation, relationship between ! and is
(a) ! = T (b) ! = /T (c) + = !/T (d) = T/
5. In the impulse invariant transformation
1 1 1 1
(a) (b) pi T 1
s pi 1 e piT z 1 s pi 1 e z

1 1 1 1
(c) (d)
s pi 1 e piT z s pi 1 e pi T
z
6. Non-linearity in the relationship between ! and is known as
(a) aliasing (b) frequency warping
(c) unwarping (d) frequency mixing
7. In the bilinear transformation, the relationship between ! and is
2
(a) = 2 tan (b) = tan
2 T 2
1 T
(c) = tan (d) = tan
T 2 2
8. In the bilinear transformation, the relation between s and z is
1 1
2 1+z 1 1+z
(a) s = 1
(b) s = 1
T 1 z T 1 z

1 1
2 1 z 1 1 z
(c) s = (d) s = 1
T 1+z 1 T 1+z

9. Butterworth filters have

(a) wideband transition region (b) sharp transition region
(c) oscillation in the transition region (d) none of the above
10. Chebyshev filters have
(a) wideband transition region (b) sharp transition region
(c) oscillation in the transition region (d) none of the above
636 ! Digital Signal Processing

11. Type-1 Chebyshev filter contains

(a) oscillations in the passband (b) oscillations in the stopband
(c) oscillations in stop and pass bands (d) oscillations in the transition band
12. Type-2 Chebyshev filter is also called
(a) inverse Chebyshev filter (b) elliptic filter
(c) reverse Chebyshev filter (d) none of the above
13. The attenuation constant in the design of Chebyshev filter is given by
1 1
1 2
1 2
(a) = 1 (b) = 1
A12 A22

1 1
1 2N 1 2N
(c) = 1 (d) = 1
A12 A22

14. The cutoff frequency !c of a low-pass Butterworth filter is given by

1 2
(a) c = 1
(b) c = 1
1 2N 1 2N
1 1
A12 A12

1
(c) c = 1 (d) c = 1
1
1 2 1 2N
1 +1
A12 A12

15. For Butterworth filter, when A1 and A2 are in dB, filter order N is given by

100.1 A2 dB + 1 10 0.1 A2 dB 1
log log
1 10 0.1 A1 dB + 1 1 10 0.1 A1 dB 1
(a) N (b) N
2 log ( 2 / 1 ) 2 log ( 2 / 1 )

10 0.1 A2 dB + 1 100.1 A2 dB + 1
log log
1 10 0.1 A1 dB + 1 1 10 0.1 A1 dB + 1
(c) N (d) N
2 log ( 1 / 2 ) 2 log ( 1/ 2)

16. The magnitude response of a Butterworth filter is given by

2 1 2 1
(a) H a ( ) = 2N (b) H a ( ) = 2N
c
1+ 1+
1 c
 Infinite-duration Impulse Response (IIR) Filters ! 637

2 1 2 1
(c) H a ( ) = 1
(d) H a ( ) = 1
2N 2N
1+ c 1+
1 c

17. The magnitude response of type-1 Chebyshev filter is given by

2 1 2 1
(a) Ha ( ) = (b) H a ( ) = 1
2
1+ cN2 2
2N
c
1+ cN2
c

2 1 2 1
(c) H a ( ) = 2
(d) H a ( ) = 1
2N
1 + cN 1 + cN
c
c

/7*3,$=%)
1. Use the backward difference for the derivative to convert the analog low-pass filter
with system function given below to digital filter assuming T = 1 s.
1 1 1
(a) H ( s) = (b) H (s) = 2
(c) H ( s) =
s+4 s + 25 (s + 0.2)2 + 16
2. Covert the analog filter with system function given below into a digital filter using
impulse invariant transformation assuming T = 1s.
1 s + 0.2
(a) H ( s) = (b) H ( s) =
(s + 3) (s + 4) (s + 0.2)2 + 9
1
(c) H ( s) =
( s + 0.5) (s 2 + 0.5s + 2)

s + 0.3
3. Convert the analog filter with system function H ( s) = into a
(s + 0.3) 2 + 16
digital filter using the bilinear transformation. The digital filter should have
resonant frequency of r = /2.
4. Convert the analog filter with system function into a digital filter using bilinear
transformation. Take T = 1s.
4 2s
(a) H ( s) = (b) H (s) = 2
(s + 1) (s + 3) s + 3s + 4
638 ! Digital Signal Processing

5. A digital filter with a 3 dB bandwidth of 0.4 is to be designed from the analog

c
filter whose system response is H ( s) = .
s +3 c
Use the bilinear transformation and obtain H(z).
6. The specification of the desired low-pass filter is
1
H( ) 1.0; 0 < 0.2
2
H( ) 0.08; 0.4
Design a Butterworth digital filter using the bilinear transformation.
7. The specification of the desired low-pass digital filter is
0.9 H( ) 1.0; 0 0.25

H( ) 0.24; 0.5
Design a Chebyshev digital filter using the impulse invariant transformation.
8. Determine H(z) for a Butterworth filter satisfying the following constraints:

0.5 H( ) 1; 0 /2

H( ) 0.2; 3 /4

with T = 1 s. Apply the impulse invariant transformation.

9. Design (a) Butterworth low-pass digital filter, (b) Chebyshev low-pass digital filter
satisfying the following specifications:
Sampling time = 1 s
Passband frequency = 0.06 rad/s
Stopband frequency = 0.75 rad/s
Passband attenuation = 6 dB
Stop band attenuation = 20 dB
10. Design a Butterworth low-pass digital filter satisfying the following specifications:

0.89 H( ) 1.0; 0 0.2

H( ) 0.18; 0.3

1
11. A prototype low-pass filter has the system response H (s) = 2
. Obtain a
s + 2s + 4
bandpass filter with !0 = 4 rad/s and Q = 10.
c
12. Transform the prototype low-pass filter with system function H ( s) = into
s +3 c
a high-pass filter with cutoff frequency !c*.
 Infinite-duration Impulse Response (IIR) Filters ! 639

!"#$"%& '()*("!+

',-.,/0& 123
!" #$%&'(" )*" *&+,$-" .%&('" /&+&($0-" 1-0(%*)-20,&)(
!"!#$ !"%&'$ &""#$ !"()%$ &""#
*)+,---#$ .$ )&/0"123$ *'%45%2!6
*2+*)78#
*!+9--#$ .$ !5:(**$ *'%45%2!6
2+;#
<=>0>?@+A5::%'B2>*!7*2C#
A+?D0("6B=C#.$ =%'()
&+0("6B0C#.$ 0("%)
<E>(/@+*'%4=BA>&>;,8>*)C#
)5A0"(:B8>,>,C>0"(:B(/>8-D"(3,-B&A)BECCC#
F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG3&12$ 12$ IAKC
:1:"%BG/&321:5I%$ '%)0(2)%KC
)5A0"(:B8>,>8C>0"(:B(/>&23"%BECC#
F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG0E&)%$ 12$ '&I1&2)KC
:1:"%BG0E&)%$ '%)0(2)%KC

!"#$"#%

Magnitude response
100

0
Gain in db

–100

–200

–300
0 50 100 150 200 250 300 350 400 450 500
Normalized frequency
640 ! Digital Signal Processing
Phase response
4

2
Phase in radians

–2

–4
0 50 100 150 200 250 300 350 400 450 500

Normalized frequency

!"#$"%&' ()*
!" #$%%&'()'%*" +)(,-.//" 01+%&'
!"!#$ !"%&'$ &""#$ !"()%$ &""#
&"*+&)$ ,$ -.#$ /$ *&))$ 0&12$ &33%14&35(1$ 51$ 26
&"*+&*$ ,$ .78#$ /$ )3(*$ 0&12$ &33%14&35(1$ 51$ 26
9*&)),:...#$ /$ *&))$ 0&12$ 9'%;4%1!<$ 51$ =>
9)3(*,:8..#$ /$ )3(*$ 0&12$ 9'%;4%1!<$ 51$ =>
9)&?,8...#$ /$ )&?*"51@$ 9'%;4%1!<$ 51$ =>
A*,BC9*&))D9)&?#
A),BC9)3(*D9)&?#$ /$ *&))$ 0&12$ &12$ )3(*$ 0&12$ 9'%;4%1!5%)
E1FA1G$ ,$ 0433('2HA*FA)F&"*+&*F&"*+&)I#$ $ /$ ?515?&"$ ('2%'F$ +&"9J*(A%'$ 9'%;4%1!<
E0F&G$ ,$ 0433%'H1FA1I#$ /$ !(%995!5%13)$ (9$ 2%)5@1%2$ 95"3%'
E+FAG$ ,$ 9'%;>H0F&I#
)40*"(3HBF:F:I#*"(3HAD*5FB.C"(@:.H&0)H+III#
K"&0%"HLM('?&"5>%2$ N'%;4%1!<OI
<"&0%"HL@&51$ 51$ 20OI
353"%HL?&@15342%$ '%)*(1)%OI
)40*"(3HBF:FBI#*"(3HAD*5F&1@"%H+II#
K"&0%"HLM('?&"5>%2$ N'%;4%1!<OI
<"&0%"HL*+&)%$ 51$ '&25&1)OI
353"%HL*+&)%$ '%)*(1)%OI
1$ ,
$$$$$P
A1$ ,
$ $ $ $ .7QRQQ
 Infinite-duration Impulse Response (IIR) Filters ! 641

!"#$"#%
Magnitude response
200

0
Gain in db

–200

–400
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency

Phase response
4
Phase in radians

–2

–4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency

',-.,/0& 124
!" /.,,$-3)-,4" 4&'4560%%" *&+,$-
!"!#$ !"%&'$ &""#$ !"()%$ &""#
&"0E&)$ +$ ;-#$ .$ 0&))$ A&2I$ &::%25&:1(2$ 12$ IL
&"0E&0+$ ,#$ .$ ):(0$ A&2I$ &::%25&:1(2$ 12$ IL
*0+,-;-#$ .$ 0&))$ A&2I$ *'%45%2!6$ 12$ M=
*)+N--#$ .$ ):(0$ A&2I$ *'%45%2!6$ 12$ M=
*)&/+9;--#$ .$ )&/0"123$ *'%45%2!6$ 12$ M=
O0+8D*07*)&/#
O)+8D*)7*)&/#
<2>O2@$ +$ A5::('IBO0>O)>&"0E&0>&"0E&)C#$ .$ /121/&"$ ('I%'>$ E&"*P0(O%'$ *'%45%2!6
<A>&@$ +$ A5::%'B2>O2>KE13EKC#$ .$ !(%**1!1%2:)$ (*$ :E%$ I%)132%I$ *1":%'
<E>O@$ +$ *'%4=BA>&C#
)5A0"(:B8>,>,C>0"(:BO701>8-D"(3,-B&A)BECCC#
642 ! Digital Signal Processing

F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG3&12$ 12$ IAKC
:1:"%BG/&321:5I%$ '%)0(2)%KC
)5A0"(:B8>,>8C>0"(:BO701>&23"%BECC#
F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG0E&)%$ 12$ '&I1&2)KC
:1:"%BG0E&)%$ '%)0(2)%KC
2$ +
$$$$$Q

O2$ +
$ $ $ $ -R;NSN

!"#$"#%
Magnitude response
200

0
Gain in db

–200

–400
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency

Phase response
4

2
Phase in radians

–2

–4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency
 Infinite-duration Impulse Response (IIR) Filters ! 643

!"#$"%&' ()*
!" #$%%&'()'%*" +,-." /%)0" 123%&'
!"!#$ !"%&'$ &""#$ !"()%$ &""#
*)+,-./$ -.01#$ 2$ )3(4$ 5&67$ 8'%9:%6!;$ <6$ '&7<&6)
*4+,-.=$ -.>1#$ 2$ 4&))$ 5&67$ 8'%9:%6!;$ <6$ '&7<&6)
&"4?&4+-./#$ 2$ 4&))$ 5&67$ &33%6:&3<(6$ <6$ 7@
&"4?&)+A-#$ 2$ )3(4$ 5&67$ &33%6:&3<(6$ <6$ 7@
,6B*61$ +$ 5:33('7C*4B*)B&"4?&4B&"4?&)D#
,5B&1+5:33%'C6B*6BE)3(4ED#
,?B*1$ +$ 8'%9FC5B&D#
):54"(3CGBHBHD#4"(3C*I4<BG-J"(KH-C&5)C?DDD#
L"&5%"CMN('O&"<F%7$ P'%9:%6!;ED
;"&5%"CMK&<6$ <6$ 75ED
3<3"%CMO&K6<3:7%$ '%)4(6)%ED
):54"(3CGBHBGD#4"(3C*I4<B&6K"%C?DD#
L"&5%"CMN('O&"<F%7$ P'%9:%6!;ED
;"&5%"CM4?&)%$ <6$ '&7<&6)ED
3<3"%CM4?&)%$ '%)4(6)%ED
6$ +
$$$$$Q

*6$ +
$ $ $ $ -.=G/=$ $ $ $ -.0>A>

!"#$"#%
Magnitude response
100

0
Gain in db

–100

–200

–300
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency
644 ! Digital Signal Processing

Phase response
4
Phase in radians

–2

–4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency

!"#$"%&' ()+
!" 4*&+5/*&6" 123%&'" 3)(70,//" %50&78
!"!#$ !"%&'$ &""#$ !"()%$ &""#
&"4?&4+-.HA#$ 2$ 4&))$ 5&67$ &33%6:&3<(6$ <6$ 7@
&"4?&)+-.Q#$ 2$ )3(4$ 5&67$ &33%6:&3<(6$ <6$ 7@
*4+-.=J4<#$ 2$ 4&))$ 5&67$ 8'%9:%6!;$ <6$ '&7<&6)
*)+-.AJ4<#$ 2$ )3(4$ 5&67$ 8'%9:%6!;$ <6$ '&7<&6)
,6B*61+!?%5H('7C*4I4<B*)I4<B&"4?&4B&"4?&)D#
,5B&1$ +$ !?%5;HC6B&"4?&4B*6D#$ $ 2$ !(%88<!<%63)$ (8$ 7%)<K6%7$ 8<"3%'
,?B*1$ +$ 8'%9FC5B&D#
):54"(3CGBHBHD#4"(3C*I4<BG-J"(KH-C&5)C?DDD#
L"&5%"CMN('O&"<F%7$ P'%9:%6!;ED
;"&5%"CMK&<6$ <6$ 75ED
3<3"%CMO&K6<3:7%$ '%)4(6)%ED
):54"(3CGBHBGD#4"(3C*I4<B&6K"%C?DD#
L"&5%"CMN('O&"<F%7$ P'%9:%6!;ED
;"&5%"CM4?&)%$ <6$ '&7<&6)ED
3<3"%CM4?&)%$ '%)4(6)%ED
 Infinite-duration Impulse Response (IIR) Filters ! 645

!"#$"#%
Magnitude response
0

–50
Gain in db

–100

–150
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency
Phase response
0

–1
Phase in radians

–2

–3

–4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency

',-.,/0& 125
!" 74$89%4$:" *&+,$-" 4&'4560%%" ,96$5;
!"!#$ !"%&'$ &""#$ !"()%$ &""#
&"0E&0+,#$ .$ 0&))$ A&2I$ &::%25&:1(2$ 12$ IL
&"0E&)+,;#$ .$ ):(0$ A&2I$ &::%25&:1(2$ 12$ IL
O0+-R9D01#$ .$ 0&))$ A&2I$ *'%45%2!6$ 12$ '&I1&2)
O)+-R8D01#$ .$ ):(0$ A&2I$ *'%45%2!6$ 12$ '&I1&2)
<2>O2@+!E%A,('IBO0701>O)701>&"0E&0>&"0E&)C#
<A>&@$ +$ !E%A6,B2>&"0E&0>O2>KE13EKC#$ .$ !(%**1!1%2:)$ (*$ I%)132%I$ *1":%'
<E>O@$ +$ *'%4=BA>&C#
)5A0"(:B8>,>,C>0"(:BO701>8-D"(3,-B&A)BECCC#
F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG3&12$ 12$ IAKC
:1:"%BG/&321:5I%$ '%)0(2)%KC
646 ! Digital Signal Processing

)5A0"(:B8>,>8C>0"(:BO701>&23"%BECC#
F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG0E&)%$ 12$ '&I1&2)KC
:1:"%BG0E&)%$ '%)0(2)%KC

!"#$"#%
Magnitude response
0

–50
Gain in db

–100

–150

–200
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency
Phase response
4

2
Phase in radians

–2

–4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency

',-.,/0& 126
!" 74$89%4$:" 80(<" 60%%" *&+,$-" ,96$5;
!"!#$ !"%&'$ &""#$ !"()%$ &""#
T0$ +$ <N-$ 8--@ 7;--#
T)$ +$ <;-$ 8;-@ 7;--#
&"0E&0$ +$ 9#$ .$ 0&))$ A&2I$ &::%25&:1(2$ 12$ IL
&"0E&)$ +$ S-#$ .$ ):(0$ A&2I$ &::%25&:1(2$ 12$ IL
<2>T0@$ +$ !E%A,('IBT0>T)>&"0E&0>&"0E&)C#
<A>&@$ +$ !E%A6,B2>&"0E&0>T0C#
 Infinite-duration Impulse Response (IIR) Filters ! 647

<E>O@$ +$ *'%4=BA>&C#
)5A0"(:B8>,>,C#0"(:BO701>8-D"(3,-B&A)BECCC#
F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG3&12$ 12$ IAKC
:1:"%BG/&321:5I%$ '%)0(2)%KC
)5A0"(:B8>,>8C#0"(:BO701>&23"%BECC#
F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG0E&)%$ 12$ '&I1&2)KC
:1:"%BG0E&)%$ '%)0(2)%KC

!"#$"#%
Magnitude response
0

–200
Gain in db

–400

–600
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency
Phase response
4

2
Phase in radians

–2

–4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency

',-.,/0& 121
!" " 74$89%4$:" +)3560%%" *&+,$-" ,96$5=
!"!#$ !"%&'$ &""#$ !"()%$ &""#
T0$ +$ S-7;--#$ .$ 0&))$ A&2I$ *'%45%2!6$ 12$ '&I1&2)
T)$ +$ ,;-7;--#$ $ .$ ):(0$ A&2I$ *'%45%2!6$ 12$ '&I1&2)
648 ! Digital Signal Processing

&"0E&0$ +$ 9#$ .$ 0&))$ A&2I$ &::%25&:1(2$ 12$ IL

&"0E&)$ +$ N-#$ .$ ):(0$ A&2I$ &::%25&:1(2$ 12$ IL
<2>T)@$ +$ !E%A8('IBT0>T)>&"0E&0>&"0E&)C#
<A>&@$ +$ !E%A68B2>&"0E&)>T)C#
<E>O@+*'%4=BA>&C#
)5A0"(:B8>,>,C>0"(:BO701>8-D"(3,-B&A)BECCC#
F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG3&12$ 12$ IAKC
:1:"%BG/&321:5I%$ '%)0(2)%KC
)5A0"(:B8>,>8C>0"(:BO701>&23"%BECC#
F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG0E&)%$ 12$ '&I1&2)KC
:1:"%BG0E&)%$ '%)0(2)%KC

!"#$"#%

Magnitude response
50

0
Gain in db

–50

–100

–150
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency

Phase response
4

2
Phase in radians

–2

–4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency
 Infinite-duration Impulse Response (IIR) Filters ! 649

',-.,/0& 127
!" 74$89%4$:" 80(<" 60%%" *&+,$-" ,96$5=
!"!#$ !"%&'$ &""#$ !"()%$ &""#
T0$ +$ <N-$ 8--@7;--#$ .$ 0&))$ A&2I$ *'%45%2!6$ 12$ '&I1&2)
T)$ +$ <;-$ 8;-@7;--#$ .$ ):(0$ A&2I$ *'%45%2!6$ 12$ '&I1&2)
&"0E&0$ +$ 9#$ .$ 0&))$ A&2I$ &::%25&:1(2$ 12$ IL
&"0E&)$ +$ S-#$ .$ ):(0$ A&2I$ &::%25&:1(2$ 12$ IL
<2>T)@$ +$ !E%A8('IBT0>T)>&"0E&0>&"0E&)C#
<A>&@$ +$ !E%A68B2>&"0E&)>T)C#
<E>O@+*'%4=BA>&C#
)5A0"(:B8>,>,C#0"(:BO701>8-D"(3,-B&A)BECCC#
F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG3&12$ 12$ IAKC
:1:"%BG/&321:5I%$ '%)0(2)%KC
)5A0"(:B8>,>8C#0"(:BO701>&23"%BECC#
F"&A%"BGH('/&"1=%I$ J'%45%2!6KC
6"&A%"BG0E&)%$ 12$ '&I1&2)KC
:1:"%BG0E&)%$ '%)0(2)%KC

!"#$"#%
Magnitude response
100

0
Gain in db

–100

–300

–300
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency
Phase response
4
Phase in radians

–2

–4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Normalized frequency
650 ! Digital Signal Processing

Magnitude response
100

0
Gain in db

–100

–200

–300
0 50 100 150 200 250 300 350 400 450 500
Normalized frequency
Phase response
4

2
Phase in radians

–2

–4
0 50 100 150 200 250 300 350 400 450 500
Normalized frequency