ABES ENGINEERING COLLEGE, GHAZIABAD

DEPARTMENT OF ELECTRONICS &

COMMUNICATION ENGINEERING

COURSE MATERIAL
Subject Name: DIGITAL SIGNAL PROCESSING
Subject Code: KEC -503
Branch/Semester: ECE / 5th
Session: 2023-24
2022
2022-23(Odd-Semester)

Faculty Members
Dr. Devvrat Tyagi
Ms. Tania Gupta
Dr. Mangal Deep Gupta
 KEC503
DIGITAL SIGNAL PROCESSING

Contents as per syllabus:

UNIT-2 Syllabus

1. Infinite Impulse Response Digital (IIR) Filter Design:

1.1 Introduction to Filters,
1.2 Need of IIR Filters
1.3 Impulse Invariant Transformation,
1.4 Bi-Linear Transformation
1.5 Advantages and Disadvantages of IIR Filters
1.6 Industrial exposure of IIR Systems

2. All- Pole Analog Filters:

2.1 Butterworth Filter,
2.2 Chebyshev Filter,
2.3 Design of Digital Butterworth Filters,
2.4 Design of Digital Chebyshev Filters,

3. Frequency Transformations
1. Infinite Impulse Response Digital (IIR) Filter Design
1.1 Introduction to Filters

The analog filter design is well-developed and the techniques discussed in this
chapterare all based on taking an analog filter and converting it into a digital
filter. Thus, the design of an IIR filterinvolves design of a digital filter in the
analog domain and transforming the design into the digital domain.

The system function describing an analog filter may be written as

∑ 𝑏 𝑠
𝐻 (𝑠) =
∑ 𝑎 𝑠
Where{𝑎 } and {𝑏 } are the filter coefficients. The impulse response of these
filter coefficients isrelated to Ha(s) by the Laplace transform

𝐻 (𝑠) = ℎ(𝑡)𝑒 𝑑𝑡 (1)

The analog filter having the rational system function H (s) given in Eq.(1) can
also be describedby the linear constant-coefficient differential equation
𝑑 𝑦(𝑡) 𝑑 𝑥(𝑡)
𝑎 = 𝑏
𝑑𝑡 𝑑𝑡

The above three equivalent characterisation of an analog filter leads to three

alternative methods for transforming the filter into the digital domain.

1.2 Need of IIR Filters

The design techniques for IIR filters are presented with the restriction that the
filters be realisable and stable. Recall that an analog filter with system function
H(s) is stable if all its poles lie in the left-half of the s-plane.
As a result, if the conversion techniques are to be effective, the technique
should possess the following properties:
(i) The jΩ axis in the s-plane should map onto the unit circle in the z-
plane. This gives a direct relationship between the two frequency
variables in the two domains.
 (ii) The left-half plane of the s-plane should map into the inside of the unit
circle in the z-plane to convert a stable analog filter into a stable
digital filter.

Fig. 1Mapping from s-plane to z-plane.

1.3 IIR Filter Design by Impulse Invariant Method

In this technique, the desired impulse response of the digital filter is obtained by
uniformly samplingthe impulse response of the equivalent analog filter. That is,

ℎ(𝑛) = ℎ (𝑛𝑇)

Where𝑇 is the sampling interval. The transformation technique can be well

understood by first consideringa simple distinct pole case for the analog filter’s
system function, as shown below.
𝐴
𝐻 (𝑠) = (2)
𝑠−𝑝

The impulse response of the system specified by Eq. (2) can be obtained by
taking the inverseLaplace transform and it will be of the form

ℎ (𝑡) = 𝐴𝑒 𝑢 (𝑡)
Where𝑢 (𝑡) is the unit step function in continuous time. The impulse response
ℎ(𝑛) of the equivalentdigital filter is obtained by uniformly sampling ℎ (𝑡):

ℎ(𝑛) = ℎ (𝑛𝑇) = 𝐴𝑒 𝑢 (𝑛𝑇) (3)

The system response of the digital system of Eq. (3), can be obtained by taking
the 𝑧-transform,i.e.

𝐻(𝑧) = ℎ(𝑛)𝑧

𝐻(𝑧) = 𝐴𝑒 𝑢 (𝑛𝑇) 𝑧

𝐻(𝑧) = 𝐴𝑒 𝑢 (𝑛𝑇) 𝑧

𝐴
𝐻(𝑧) = (4)
1−𝑒 𝑧
Now, by comparing Eqs. (2) and (4), the mapping formula for the impulse
invariant transformationis given by

1 1
→ (5)
𝑠−𝑝 1−𝑒 𝑧

Equation (5) shows that the analog pole at 𝑠 = 𝑝 is mapped into a digital pole
at 𝑧 = 𝑒 . Therefore,the analog poles and the digital poles are related by the
relation
𝑧=𝑒
Ω
𝑟𝑒 =𝑒 𝑒

𝑟=𝑒

𝜔 = Ω𝑇

Consequently, 𝑠 < 0 implies that 0 < 𝑟 < 1 and 𝑠 > 0 implies that 𝑟 > 1.
When 𝑠 = 0, we have 𝑟 = 1.
Therefore, the left-half of s-plane is mapped inside the unit circle in the z-plane
and the right-half ofs-plane is mapped into points that fall outside the unit circle
in z. This is one of the desirable propertiesfor stability.

The 𝑗Ω-axis is mapped into the unit circle in z-plane. However, the mapping of
the 𝑗Ω-axisis not one-to-one. The mapping 𝑤 = Ω𝑇 implies that the interval -
𝜋/𝑇 ≤ Ω ≤ 𝜋/𝑇 maps into the correspondingvalues of − 𝜋 ≤ 𝑤 ≤ 𝜋.

Further, the frequency interval 𝜋/𝑇 ≤ Ω ≤ 3 𝜋/𝑇 also maps into theinterval
− 𝜋 ≤ 𝑤 ≤ 𝜋 and, in general, any frequency interval (2𝑘 − 1) 𝜋/𝑇 ≤ Ω ≤
(2𝑘 + 1) 𝜋/𝑇, where 𝑘 isan integer, will also map into the interval − 𝜋 ≤ 𝑤 ≤
𝜋 in the z-plane. Thus the mapping from the analogfrequency Ω to the
frequency variable 𝑤 in the digital domain is many-to-one, which simply
reflectsthe effects of aliasing due to sampling of the impulse response. Fig. 2
illustrates the mapping fromthe s-plane to the z-plane.

Fig. 2The Mapping of𝑧 = 𝑒

Some of the properties of the impulse invariant transformation are given below.

1 (−1) 𝑑 1
→ ;𝑠 → 𝑠
(𝑠 + 𝑠 ) (𝑚 − 1)! 𝑑𝑠 1−𝑒 𝑧

𝑠+𝑎 1−𝑒 (𝑐𝑜𝑠𝑏𝑇)𝑧

→
(𝑠 + 𝑎) + 𝑏 1 − 2𝑒 (𝑐𝑜𝑠𝑏𝑇)𝑧 + 𝑒 𝑧
 𝑏 𝑒 (𝑠𝑖𝑛𝑏𝑇)𝑧
→
(𝑠 + 𝑎) + 𝑏 1 − 2𝑒 (𝑐𝑜𝑠𝑏𝑇)𝑧 + 𝑒 𝑧

1.4 IIR FILTER DESIGN BY THE BILINEAR

TRANSFORMATION
The IIR filter design using (i) approximation of derivatives method and (ii) the
impulse invariant methodare appropriate for the design of low-pass filters and
bandpass filters whose resonant frequencies arelow.

These techniques are not suitable for high-pass or band-reject filters. This
limitation is overcomein the mapping technique called the bilinear
transformation. This transformation is a one-to-one mappingfrom the s-domain
to the 𝑧-domain. That is, the bilinear transformation is a conformal mappingthat
transforms the 𝑗Ω-axis into the unit circle in the 𝑧-plane only once, thus
avoiding aliasing of frequencycomponents.

Also, the transformation of a stable analog filter results in a stable digital filter
asall the poles in the left half of the 𝑠-plane are mapped onto points inside the
unit circle of the 𝑧-domain.

The bilinear transformation is obtained by using the trapezoidal formula for

numerical integration.

Letthe system function of the analog filter be

𝑏
𝐻(𝑠) = (6)
𝑠+𝑎

𝑌(𝑠) 𝑏
𝐻(𝑠) = =
𝑋(𝑠) 𝑠 + 𝑎

𝑠𝑌(𝑠) + 𝑎𝑌(𝑠) = 𝑏𝑋(𝑠)

Taking inverse Laplace transform,

𝑑𝑦(𝑡)
+ 𝑎𝑦(𝑡) = 𝑏𝑥(𝑡) (7)
𝑑𝑡

Equation (7) is integrated between the limits (𝑛𝑇 − 𝑇) and 𝑛𝑇

 𝑑𝑦(𝑡)
𝑑𝑡 + 𝑎 𝑦(𝑡)𝑑𝑡 = 𝑏 𝑥(𝑡)𝑑𝑡
𝑑𝑡

The trapezoidal rule for numeric integration is given by

𝑇
𝑎(𝑡)𝑑𝑡 = [𝑎(𝑛𝑇) + 𝑎(𝑛𝑇 − 𝑇)]
2

We get,
𝑎𝑇 𝑎𝑇
𝑦(𝑛𝑇) − 𝑦(𝑛𝑇 − 𝑇) + 𝑦(𝑛𝑇) + 𝑦(𝑛𝑇 − 𝑇)
2 2
𝑏𝑇 𝑏𝑇
= 𝑥(𝑛𝑇) + 𝑥(𝑛𝑇 − 𝑇)
2 2
Taking z-transform, the system function of the digital filter is
𝑌(𝑧) 𝑏
𝐻(𝑧) = = (8)
𝑋(𝑧) +𝑎
Comparing Eqs. (6) and (8), we get

2 1−𝑧 2 𝑧−1
𝑠= =
𝑇 1+𝑧 𝑇 𝑧+1

2 𝑧−1 2 𝑟𝑒 −1
𝑠= =
𝑇 𝑧+1 𝑇 𝑟𝑒 +1

Substituting 𝑒 = 𝑐𝑜𝑠 𝑤 − 𝑗 𝑠𝑖𝑛 𝑤 and simplifying, we get

2 𝑟 −1 2𝑟𝑠𝑖𝑛𝜔
𝑠= +𝑗
𝑇 1 + 𝑟 + 2𝑟𝑐𝑜𝑠𝜔 1 + 𝑟 + 2𝑟𝑐𝑜𝑠𝜔

2 𝑟 −1
𝜎= (9)
𝑇 1 + 𝑟 + 2𝑟𝑐𝑜𝑠𝜔
2 2𝑟𝑠𝑖𝑛𝜔
Ω= (10)
𝑇 1 + 𝑟 + 2𝑟𝑐𝑜𝑠𝜔
From Eq. (9), it can be noted that if 𝑟 < 1, then 𝑠 < 0, and if 𝑟 > 1, then
𝑠 > 0. Thus, the left-halfof the 𝑠-plane maps onto the points inside the unit
circle in the 𝑧-plane and the transformation results ina stable digital system.
Consider Eq. (10), for unity magnitude (𝑟 = 1), 𝑠 is zero. In this case
2 𝑠𝑖𝑛𝜔
=
𝑇 1 + 𝑐𝑜𝑠𝜔

2 2𝑠𝑖𝑛 cos
=
𝑇 𝑐𝑜𝑠 + 𝑠𝑖𝑛 + 𝑐𝑜𝑠 − 𝑠𝑖𝑛

2 𝜔 (11)
Ω= 𝑡𝑎𝑛
𝑇 2
Ω𝑇
𝜔 = 2𝑡𝑎𝑛 (12)
2
Pre-Warping: Eq. (12) gives the relationship between the frequencies in the
two domains andthis is shown in Fig. 3. It can be noted that the entire range in
Ω is mapped only once into the range− 𝜋 ≤ 𝑤 ≤ 𝜋. However, as seen in Fig. 3,
the mapping is non-linear and the lower frequencies in analogdomain
areexpanded in the digital domain, whereas the higher frequencies are
compressed. Thedistortion introduced in the frequency scale of the digital filter
to that of the analog filter due to the non-linearity of the arctangent function and
this effect of the bilinear transform is usually called frequencywarping the filter
design.

The analog filter is designed to compensate for the frequency warping by setting
Eqn. (11) forevery frequency specification so that the corner frequency or center
frequency is controlled. This iscalled pre-warping the filter design. When a
digital filter is designed as an approximation of an analogfilter, the frequency
response of the digital filter can be made to match the frequency response of
theanalog filter by considering the following:
 Fig.3 Relationship between 𝜔 and Ω

1.5 Advantages and Disadvantages of IIR Filters

Advantages-

 IIR filter is better than the FIR in that, it can produce the same response
using some fewer delay blocks.
 It is easy to implement.
 It is easy to design.
 This filter is useful only when some analog filter is bandlimited.
 They are more susceptible to the problem of line finite length arithmetic.
 An IIR filter has a lesser number of side lobes in the stopband than an
FIR filter.
 Implementation of IIR filter involves fewer parameters, less memory
requirement, and lower computational complexity.

Disadvantages-

 There are more susceptible to the problem of finite length arithmetic,

Such as the noise is generated by calculations, and limit cycle.
 They are harder to implement using fixed-point arithmetic.
 IIR filter becomes unstable.
 Analog frequency and digital frequency are linearly related.
  IIR filters usually have a non-linear response, while the FIR usually has a
linear phase.
 They don't offer the computational advantages of the FIR filter for multi-
rate applications.
 IIR filter has a feedback loop so they will accumulate rounding and noise
error.
 The realization of the IIR filter is not very easy as compared to FIR
filters.
 The implementation of an IIR filter involves fewer parameters needed.
 Less memory requirement.
 It is a recursive filter the number of the coefficient is very large and the
memory requirement is also high.
 It is hard to optimize than the FIR filter.


1.6 Industrial exposure of IIR Systems

Followings are the key domain where IIR systems are deployed

 Telecommunication
 Clock recovery in data communication
 Receiver anti-imaging filter
 Digital telephony called digital dual tone multi-frequency touch-tone
receiver
 Signal monitoring application

2. All- Pole Analog Filters

An all-pole filter has a frequency response function that goes infinite (poles) at
specific frequencies, but there are no frequencies where the response function is
zero. Basically the filter function (also called the transfer function) is a ratio
with a constant in the numerator and a polynomial in the denominator.

3.1 BUTTERWORTH FILTERS

The Butterworth low-pass filter has a magnitude response given by
𝐴
|𝐻(𝑗Ω)| = .
(13)
[1 + (Ω/Ω ) ]

Where 𝐴 is the filter gain and Ω is the 3dB cut-off frequency and 𝑁 is the order
of the filter. The magnituderesponse of the Butterworth filter is shown in Fig.
4(a). The magnitude response has a maximally flat passband and stopband. It
can be seen that by increasing the filter order 𝑁, the Butterworth
responseapproximates the ideal response. However, the phase response of the
Butterworth filter becomes morenon-linear with increasing.

Fig. 4(a)Magnitude Response of a ButterworthLow-pass Filter

Fig.4(b) Magnitude Response of ButterworthLPF with Design specifications.

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

𝛿 ≤ |𝐻(𝑗Ω)| ≤ 1, 0≤Ω≤Ω (14)

 |𝐻(𝑗Ω)| ≤ 𝛿 , Ω ≤Ω≤𝜋

𝛼 = −20 𝑙𝑜𝑔 𝛿

𝛼 = −20 𝑙𝑜𝑔 𝛿
1
𝛿 =
√1 + 𝜀
1
𝛿 =
√1 + 𝜆
WhereƐ and 𝛿 are the parameters specifying allowable passband, and 𝜆 and 𝛿
are the parametersspecifying allowable stopband.

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

process. Using Eq.(13) in Eq. (14) and if 𝐴 = 1, we get
1
𝛿 ≤ ≤ 12
1 + (Ω /Ω )
1
≤𝛿
1 + (Ω /Ω )
1
(Ω /Ω ) ≤ −1 (15A)
𝛿
1 (15B)
(Ω /Ω ) ≤ −1
𝛿
Equality is assumed in Eq. (15) in order to obtain the filter order 𝑁 and the 3dB
cut-off frequency Ω .Dividing Eq. (15A) by Eq. (15B)

(Ω /Ω ) = ((1/𝛿 ) − 1)/((1/𝛿 ) − 1)

(16)
1 𝑙𝑜𝑔 ( − 1)/( − 1)
𝑁=
2 log (Ω /Ω )
The value of 𝑁 is chosen to be next nearest integer to the value of 𝑁 as given by
Eq. (16.). UsingEq. (15), we get
Ω
Ω =
[ 1/𝛿 − 1] /
The values of Ω and Ω are obtained using the bilinear transformation or
impulse invariant transformationtechniques:Ω = 2/𝑇 𝑡𝑎𝑛(𝑤/2)for bilinear
transformation or Ω = 𝑤/𝑇for the impulse invarianttransformation.

The transfer function of the Butterworth filter is usually written in the factored
form as given below.
/
𝐵 Ω
𝐻(𝑠) = 𝑁 = 2, 4, 6, …. (17)
𝑠 +𝑏 Ω 𝑠+𝑐 Ω
Or
( )/
𝐵 Ω 𝐵 Ω
𝐻𝐻(𝑠) = 𝑁 = 3, 5, 7, …. (18)
𝑠+𝑐 Ω 𝑠 +𝑏 Ω 𝑠+𝑐 Ω

The coefficients 𝑏 and 𝑐 are given by

(2𝑘 − 1)𝜋
𝑏 = 2 sin and𝑐 = 1
2𝑁

The parameter 𝐵 can be obtained from

/

𝐴= 𝐵 , for even 𝑁

( )/

𝐴= 𝐵 , for odd 𝑁

The system function of the equivalent digital filter is obtained from 𝐻(𝑠) (Eq.
(17) or Eq. (18))using the specified transformation technique, viz. impulse
invariant technique or bilinear transformation.

For bilinear transformation,

2 1−𝑧 2 𝑧−1
𝑠= =
𝑇 1+𝑧 𝑇 𝑧+1

For impulse invariant transformation,

 1 1
→
𝑠−𝑝 1−𝑒 𝑧

1 (−1) 𝑑 1
→ ;𝑠 → 𝑠
(𝑠 + 𝑠 ) (𝑚 − 1)! 𝑑𝑠 1−𝑒 𝑧

𝑠+𝑎 1−𝑒 (𝑐𝑜𝑠𝑏𝑇)𝑧

→
(𝑠 + 𝑎) + 𝑏 1 − 2𝑒 (𝑐𝑜𝑠𝑏𝑇)𝑧 + 𝑒 𝑧

𝑏 𝑒 (𝑠𝑖𝑛𝑏𝑇)𝑧
→
(𝑠 + 𝑎) + 𝑏 1 − 2𝑒 (𝑐𝑜𝑠𝑏𝑇)𝑧 + 𝑒 𝑧

3.1 CHEBYSHEV FILTERS

The Chebyshev low-pass filter has a magnitude response given by
𝐴
|𝐻(𝑗Ω)| =
.
(19)
[1 + 𝜀 𝐶 (Ω/Ω )]

Where𝐴 is the filter gain, 𝑒 is a constant and Ω is the 3dB cut-off frequency.
The Chebyshev polynomialof the I kind of 𝑁 order, 𝐶 (𝑥) is given by

cos(𝑁𝑐𝑜𝑠 𝑥) , 𝑓𝑜𝑟 |𝑥| ≤ 1

𝐶 (𝑥) =
cos(𝑁𝑐𝑜𝑠 ℎ 𝑥) , 𝑓𝑜𝑟 |𝑥| ≥ 1

The magnitude response of the Chebyshev filter is shown in Fig. 5. The

magnitude response hasequiripplepassband and maximally flat stopband. It can
be seen that by increasing the filter order 𝑁,the Chebyshev response
approximates the ideal response. The phase response of the Chebyshev filteris
more non-linear than the Butterworth filter for a given filter length 𝑁.
 Fig.5 Magnitude Response of a Low-pass Chebyshev Filter.
The design parameters of the Chebyshev filter are obtained by considering the
low-pass filter withthe desired specifications as below.

𝛿 ≤ |𝐻(𝑒 )| ≤ 1, 0≤𝜔≤𝜔
(20)
𝐻 𝑒 ≤𝛿 , 𝜔 ≤𝜔≤𝜋

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

process. Using Eqs.(19) in (20) and if 𝐴 = 1, we get
1
𝛿 ≤ ≤1 (21A)
1 + 𝜀 𝐶 (Ω /Ω )
1 (21B)
≤𝛿
1 + 𝜀 𝐶 (Ω /Ω )
Assuming Ω = Ω , we will have 𝐶 (Ω /Ω ) = 𝐶 (1) = 1. Therefore, Eq.
(21A) can be written as
1
𝛿 ≤
1+𝜀
Assuming equality in the above equation, the expression for Ɛ is
.
1
𝜀= −1
𝛿
The order of the analog filter 𝑁 can be determined from Eq. (21B). Assuming
Ω = Ω ,
.
1 1
𝐶 (Ω /Ω ) ≥ −1
𝜀 𝛿
Since Ω > Ω ,
.
1 1
cosh [𝑁𝑐𝑜𝑠ℎ (Ω /Ω )] ≥ −1
𝜀 𝛿

Or
.
𝑐𝑜𝑠ℎ −1 (22)
𝑁≥
𝑐𝑜𝑠ℎ (Ω /Ω )

Choose 𝑁 to be next nearest integer to the value given by Eq. (22).

The values of Ω and Ω are obtainedusing the bilinear transformation or

impulse invariant transformation techniques; Ω = 2/𝑇 𝑡𝑎𝑛(𝑤/2)for bilinear
transformation or Ω = 𝑤/𝑇for the impulse invarianttransformation.

The transfer function of the Butterworth filter is usually written in the factored
form as given below.
/
𝐵 Ω
𝐻(𝑠) = 𝑁 = 2, 4, 6, …. (17)
𝑠 +𝑏 Ω 𝑠+𝑐 Ω
Or
( )/
𝐵 Ω 𝐵 Ω
𝐻𝐻(𝑠) = 𝑁 = 3, 5, 7, …. (18)
𝑠+𝑐 Ω 𝑠 +𝑏 Ω 𝑠+𝑐 Ω

The coefficients 𝑏 and 𝑐 are given by

(2𝑘 − 1)𝜋
𝑏 = 2 sin
2𝑁

(2𝑘 − 1)𝜋
𝑐 =𝑦 + 𝑐𝑜𝑠
2𝑁
 𝑐 =𝑦

The parameter 𝑦 is given by

. .
1 1 1 1 1
𝑦 = +1 + − +1 +
2 𝜀 𝜀 𝜀 𝜀

The parameter 𝐵 can be obtained from

/
𝐴 𝐵
.
= , for 𝑁 even
(1 + 𝜀 ) 𝑐

𝐵
𝐴= for 𝑁 𝑜𝑑𝑑
𝑐

The system function of the equivalent digital filter is obtained from 𝐻(𝑠) (Eq.
(17) or Eq. (18))using the specified transformation technique, viz. impulse
invariant technique or bilinear transformation.

For bilinear transformation,

2 1−𝑧 2 𝑧−1
𝑠= =
𝑇 1+𝑧 𝑇 𝑧+1

For impulse invariant transformation,

1 1
→
𝑠−𝑝 1−𝑒 𝑧

1 (−1) 𝑑 1
→ ;𝑠 → 𝑠
(𝑠 + 𝑠 ) (𝑚 − 1)! 𝑑𝑠 1−𝑒 𝑧

𝑠+𝑎 1−𝑒 (𝑐𝑜𝑠𝑏𝑇)𝑧

→
(𝑠 + 𝑎) + 𝑏 1 − 2𝑒 (𝑐𝑜𝑠𝑏𝑇)𝑧 + 𝑒 𝑧

𝑏 𝑒 (𝑠𝑖𝑛𝑏𝑇)𝑧
→
(𝑠 + 𝑎) + 𝑏 1 − 2𝑒 (𝑐𝑜𝑠𝑏𝑇)𝑧 + 𝑒 𝑧

3. Frequency Transformations
There are basically four types of frequency selective filters, viz. low-pass,
highpass, bandpass andbandstop. In the design techniques discussed so far we
have considered only lowpass filters. This lowpassfilter can be considered as a
prototype filter and its system function can be obtained. Then, if ahighpass or
bandpass or bandstop filter is to be designed, it can be easily obtained by using
frequencytransformation. Frequency transformation can be accomplished in two
ways. In the analog frequencytransformation, the analog system function 𝐻 (𝑠)
of the prototype filter is converted into another analogsystem function 𝐻(𝑠) of
the desired filter. Then using any of the mapping techniques, it is convertedinto
the digital filter having a system function 𝐻(𝑧). In the digital frequency
transformation, the analogprototype filter is first transformed to the digital
domain, to have a system function 𝐻 (𝑠). Then usingfrequency transformation,
it can be converted into the desired digital filter.

3.1 Analog Frequency Transformation

The frequency transformation formulae used to convert a prototype lowpass
filter into a lowpass (witha different cut-off frequency), highpass, bandpass or
bandstop are given below.

i. Low-pass with cut-off frequency Ω to low-pass with a new cut-off

frequency Ω ∗
Ω
𝑠→ ∗
Ω
Thus, if the system response of the prototype filter is 𝐻 (𝑠), the system
response of the newlow-pass filter will be
Ω
𝐻(𝑠) = 𝐻 𝑠
Ω ∗
ii. Low-pass with cut-off frequency Ω to high-pass with cut-off frequency
Ω ∗
Ω Ω ∗
𝑠→
𝑠
The system function of the high-pass filter is then,
Ω Ω ∗
𝐻(𝑠) = 𝐻
𝑠
iii. Low-pass with cut-off frequency Ω to band-pass with lower cut-off
frequency Ω and higher cut-off frequency Ω
 𝑠 +Ω Ω
𝑠→Ω
𝑠(Ω − Ω )
The system function of the high-pass filter is then
𝑠 +Ω Ω
𝐻(𝑠) = 𝐻 Ω
𝑠(Ω − Ω )
iv. Low-pass with cut-off frequency Ω to bandstop with lower cut-off
frequency Ω and higher cut-off frequency Ω
𝑠(Ω − Ω )
𝑠→Ω
𝑠 +Ω Ω
The system function of the bandstop filter is then,
𝑠(Ω − Ω )
𝐻(𝑠) = 𝐻 Ω
𝑠 +Ω Ω

Table: 1Analog frequency transformation.

Type Transformation
Low pass Ω
𝑠→
Ω ∗
∗
High pass Ω Ω
𝑠→
𝑠
Band pass 𝑠 +Ω Ω
𝑠→Ω
𝑠(Ω − Ω )
Band stop 𝑠(Ω − Ω )
𝑠→Ω
𝑠 +Ω Ω

University Questions Related to Unit-2

Two Mark Questions

Q-1. Differentiate Butterworth Low Pass Filter with Chebyshev LPF in terms
ofFilter Order. [AKTU 2021-22]

Q-2. What are difference between Inpulse invariant and bilinear transformation
method? [AKTU 2020-21]
Q-3. Write down the advantages & disadvantages of bilinear
transformation.[AKTU 2019-20]

Q-4. What is the warping effect? [AKTU 2018-19]

Five Marks Questions

Q-5. Design Digital Butterworth filter to satisfy the following constraints using
bilinear transformation method, the sampling Interval is 2 second: assume
missing data if required:[AKTU 2021-22]

Q-6. Design Chebyshev Digital LPF filter to satisfy the following constraints
using Impulse Invariant method.[AKTU 2021-22]

Q-7. Design Chebyshev Digital LPF filter to satisfy the following constraints
using Bilinear Transformation method, assume that the sampling time is
one second.[AKTU 2021-22]

Q-8. The system function of analog filter is given by

(𝑺 + 𝟎. 𝟏)
𝑯(𝒔) =
(𝒔 + 𝟎. 𝟏)𝟐 + 𝟏𝟔
Obtain the system function of digital filter by using impulse invariant
technique. Assume T-1sec.[AKTU 2019-20]

Q-9. Design a Butterworth low pass analog filter for the following specification:
[AKTU 2019-20]
i. Pass band gain required:0.9
ii. Frequency up to which pass band gain must remain more or less
steady:100 rad/sec
iii. Gain in attenuation band:0.4
iv. Frequency from which the attenuation must start: 200 rad/sec

Q-10. What is frequency warping effect? How this problem is overcome in

bilinear transform method of IIR filter design? Also write down the
advantages & disadvantages of bilinear transformation.[AKTU 2019-20]
Q-11. Find the order and cut off frequency of a digital filter with the following
specification-
0.89<=|H(ejω)|<=1, 0<=ω<=0.4π
|H (ejω)|<=0.18, 0.6π<=ω<=π
Use the impulse invariance method?[AKTU 2018-19]

Q-12. Using bilinear transformation, design a Butterworth filter which satisfies

the following condition:[AKTU 2018-19]
0.8<=|H(ejω)|<=1, 0<= ω<=0.2π
|H(ejω)|<=0.2, 0.6π<= ω<=π

Q-13. What is the difference between Butterworth and Chebyshev? Explain the
frequency transformation is done? [AKTU 2018-19]