18BECC301J - DIGITAL SIGNAL

PROCESSING

UNIT II
DESIGN OF DIGITAL FILTERS
 IIR Filters
Design of IIR filters from analog filters (LPF, HPF)
Design of Butterworth Filter
Design of Chebyshev filter
Impulse invariance technique
Bilinear transformations
FIR Filters
Linear Phase FIR Filter characteristics
Design of Digital FIR Filter using windowing Techniques
Rectangular
Hamming
Hanning
Triangular
Frequency sampling method
2
 Introduction
Basically a digital filter is a linear time-invariant discrete
time system
Two types of filters
Infinite Impulse Response Filter (IIR Filter)
Finite Impulse Response Filter (FIR Filter)
FIR filters are non-recursive type  present output samples
depends on present and previous(past) input samples
IIR filters are recursive type  present output samples
depends on present input, previous(past) input and output
samples
3
 Design of IIR Filters
The impulse response h(n) for a realizable filter is
h(n) = 0 for n≤0
For the stability it must satisfy the condition

IIR digital filters have the transfer function of the form

Where, are filter coefficients

4
 Frequency Selective Filters
A filter is one, which rejects the unwanted frequencies from
the input signal and allow the desired frequencies.
Passband the range of frequencies of signal passed through
the filter
Stopband  the range of frequencies of signal blocked by the
filter
Types of filters
 Lowpass filter
 Highpass filter
 Bandpass filter
 Bandreject filter or bandstop filter
5
 Design Procedure for Analog Butterworth Filter
Step 1: from the given specification identify the values of
Where,
 - Passband attenuation in dB
 - Stopband attenuation in dB
 - Passband frequency in rad/sec
 - Stopband frequency in rad/sec

Step 2: find the value of ε and λ

or
or

6
 Design Procedure for Analog Butterworth Filter
Step 3: Determine the order of the filter N and roundoff it to the next higher
integer
or
Step 4: Find the transfer function H(s) for for the value of N

7
 Design Procedure for Analog Butterworth Filter

Step 5: Calculate the value of cutoff frequency

or

Step 6: Find the transfer function for the above value by

substituting

8
 Properties of Butterworth Filter
The Butterworth filters are all pole design
The filter order N completely specifies the filter
The magnitude is maximally flat at the origin
At the cut-off frequency , the magnitude of the normalized
Butterworth filter is or 0.707
The magnitude is monotonically decreasing the function of Ω
The magnitude response approaches the ideal response as the
value of N increases

9
Problem 2.1 Given the specification , . Determine the order of the filter.

Solution:

Order of the filter:

√ √ √
( 100.1 𝛼 − 1 )
𝑠 ( 100.1 𝑋 30 − 1 ) ( 10 3 − 1 )
log log log
( 100.1 𝛼 −1 )
𝑝 ( 10 0.1 𝑋 1 −1 ) ( 100.1 −1 )
𝑁≥ 𝑁≥ 𝑁≥
Ω𝑠 600
log log log 3
Ω𝑝 200
log 62.115
𝑁≥ N≥
log 3
3.758

Rounding off N to the next higher integer we get N = 4

10
Problem 2.2 determine the order of lowpass Butterworth filter that has
a 3dB attenuation at 500Hz and an attenuation of 40dB at 1000Hz.
Solution:

Order of the filter:

√
( 100.1 𝛼 − 1 )
√
( 100.1 𝑋 40 −1 )
√ ( 104 −1 )
𝑠

log log log

( 100.1 𝛼 −1 )
𝑝
( 10 0.1 𝑋 3 −1 ) ( 100.3 −1 )
𝑁≥ 𝑁≥ 𝑁≥
Ω𝑠 2000 π log 2
log log
Ω𝑝 1000 π
log 100.233
𝑁≥ N≥
log 2
6.647

Rounding off N to the next higher integer we get N = 7

11
Problem 2.3 Design an analog Butterworth filter that has a 2dB passband
attenuation at a frequency of 20 rad/sec and atleast 10dB stopband attenuation at
30rad/sec
Solution:

Order of the filter:

√
( 100.1 𝛼 − 1 )
√
( 100.1 𝑋 10 − 1 )
√ ( 10 − 1 )
𝑠

log log log

( 100.1 𝛼 −1 )
𝑝
( 10 0.1 𝑋 2 −1 ) ( 100.2 −1 )
𝑁≥ 𝑁≥ 𝑁≥
Ω𝑠 30 log1.5
log log
Ω𝑝 20
log 3.923
𝑁≥ N≥
log 1.5
3.371

Rounding off N to the next higher integer we get N = 4

12
 Problem 2.3 Continued
1
ormalized Lowpass Butterworth Filter for order N=4 𝐻 ( 𝑠 )=
( 𝑠 2+ 0.76537 𝑠+1 ) ( 𝑠 2 +1.8477 𝑠+1 )
Ω𝑝 20
The Cut-off frequency Ω 𝑐= Ω 𝑐= 1 /(2 𝑋 4) 21.3868 rad/sec
( 10 0.1 𝛼 𝑝
−1 )
1/ 2 𝑁
( 100.1 𝑋 2 −1 )

Substitute in H(s), here

1
𝐻 ( 𝑠 )=

(( ) ( ) )(( ) ( ) )
2 2
𝑠 𝑠 𝑠 𝑠
+0.76537 +1 +1.8477 +1
21.3868 21.3868 21.3868 21.3868

1
𝐻 ( 𝑠 )=

[( ) ][( ) ]
2 2
1 1
(𝑠 2
+0.76537 𝑋 21.3868 ( 𝑠 ) + ( 21.3868 )
2
) ( 𝑠 2 +1.8477 𝑋 21.3868 ( 𝑠 )+ ( 21.3868 )2 )
21.3868 21.3868

6
0.20921 𝑋 10
𝐻 ( 𝑠 )= 2
( 𝑠 + 16.3688 𝑠 +457.395 ) ( 𝑠2 +39.5164 𝑠 + 457.395 )
13
LOW PASS FILTER

1
≤| 𝐻 ( 𝑗 Ω )|≤ 1 𝑓𝑜𝑟 0 ≤ Ω ≤ Ω 𝑝
√ 1+ 𝜀 2

1
|𝐻 ( 𝑗 Ω )| ≤ 𝑓𝑜𝑟 Ω 𝑠 ≤ Ω ≤ 𝜋
√ 1+ 𝜆 2
Problem 2.4 for the given specification design an analog Butterworth filter.

Solution:

1
√ 1+ 𝜀 2
= 0.9 𝜀=
√( 1 2
0.9 )
−1=0.484

√( )
2
1 1
= 0.2 λ= −1=4. 8 98
√ 1+ 𝜆 2
0. 2
Order of the filter:

𝑁≥
𝜆
𝜀
log ( ) 𝑁≥
log (
4.898
0.484 ) N≥
Ω𝑠 0.4 𝜋
log log
0.2 𝜋 3.339
Ω𝑝

Rounding off N to the next higher integer we get N = 4

15
 Problem 2.4 Continued
1
ormalized Lowpass Butterworth Filter for order N=4 𝐻 ( 𝑠 )=
( 𝑠 2+ 0.76537 𝑠+1 ) ( 𝑠 2 +1.8477 𝑠+1 )
Ω𝑝 0.2 π
The Cut-off frequency Ω 𝑐= Ω 𝑐= rad/sec
ɛ
1/𝑁
( 0.484 )1 / 4

Substitute in H(s), here

1
𝐻 ( 𝑠 )=

(( ) ( ) )(( ) ( ) )
2 2
𝑠 𝑠 𝑠 𝑠
+ 0.76537 +1 +1.8477 +1
0.7533 0.7533 0.7533 0.7533

1
𝐻 ( 𝑠 )=

[( ) ][( ) ]
2 2
1 1
(𝑠 2
+0.76537 𝑋 0.7533 ( 𝑠 ) + ( 0.7533 )
2
) ( 𝑠2 +1.8477 𝑋 0.7533 ( 𝑠 ) + ( 0.7533 )2 )
0.7533 0.7533

0.322
𝐻 ( 𝑠 )=
( 𝑠 2+ 0.5766 𝑠+ 0.5675 ) ( 𝑠 2+1.3919 𝑠+ 0.5675 )
16
Homework 2.1 Design a lowpass Butterworth filter that has a 3dB cut-
off frequency of 1.5kHz and an attenuation of 40dB at 3kHz.

Homework 2.2 Design an analog Butterworth filter that has

Homework 2.1 Design a lowpass Butterworth filter that has a 3dB cut-
off frequency of 1.5kHz and an attenuation of 40dB at 3kHz.
Solution:

Order of the filter:

√
( 100.1 𝛼 − 1 )
√
( 100.1 𝑋 40 −1 )
√ ( 104 −1 )
𝑠

log log log

( 100.1 𝛼 −1 )
𝑝
( 10 0.1 𝑋 3 −1 ) ( 100.3 −1 )
𝑁≥ 𝑁≥ 𝑁≥
Ω𝑠 6000 π log 2
log log
Ω𝑝 3000 π
log 100.233
𝑁≥ N≥
log 2
6.647

Rounding off N to the next higher integer we get N = 7

18
 Homework 2.1 Continued
ormalized Lowpass Butterworth Filter for order N=7
1
𝐻 ( 𝑠 )=
(𝑠 +1) ( 𝑠 +1.80194 𝑠+1 ) ( 𝑠 +1.247 𝑠+1 ) ( 𝑠 + 0.445 𝑠+1 )
2 2 2

Ω𝑝 3000 π
The Cut-off frequency Ω 𝑐= 1/ 2 𝑁
Ω 𝑐= 1 /(2 𝑋 7 ) 3001 rad/sec
( 10 0.1 𝛼 𝑝
−1 ) ( 10 0.1 𝑋 3
−1 )
Substitute in H(s), here
1
𝐻 ( 𝑠 )=
( )(( ) ( ) )(( ) ( ) )(( ) ( ) )
2 2 2
𝑠 𝑠 𝑠 𝑠 𝑠 𝑠 𝑠
+1 +1.80194 +1 +1.247 +1 + 0.445 +1
3001 𝜋 3001 𝜋 3001 𝜋 3001 𝜋 3001 𝜋 3001 𝜋 3001 𝜋

1
𝐻 ( 𝑠 )=

[( ) ] [( )
1 2 2
][ ( ) ( ][( ) ]
2
1 1 1 2 2
(𝑠+3001 𝜋) ( 𝑠 +1.80194 𝑋 3001𝜋 ( 𝑠 ) +( 3001 𝜋 )2 ) 2
𝑠 +1.247 𝑋 3001 𝜋 ( 𝑠 ) + ( 3001 𝜋 )
2
) ( 𝑠 +0.445 𝑋 3001𝜋 ( 𝑠 )+ ( 3001𝜋 )2 )
3001 𝜋 3001 𝜋 3001𝜋 3001 𝜋

27
6.605 𝑋 10
𝐻 ( 𝑠 )=
[ ( 𝑠+3001 𝜋 ) ] [ ( 𝑠 2 +5407.62 𝑠 + ( 3001 𝜋 ) ) ][ ( 𝑠 2 +3742.25 𝑠 + ( 3001 𝜋 ) ) ][ ( 𝑠 2 +1335.45 𝑠+ ( 3001 𝜋 ) ) ]
2 2 2

19
Homework 2.2 Design an analog Butterworth filter that has

Solution:

Order of the filter:

√ √ √
( 100.1 𝛼 − 1 )
𝑠 ( 100.1 𝑋 22 −1 ) ( 102.2 −1 )
log log log
( 100.1 𝛼 −1 )
𝑝 ( 100.1 𝑋 0.5 −1 ) ( 100.05 −1 )
𝑁≥ 𝑁≥ 𝑁≥
Ω𝑠 50000 π
log log log 2 .5
Ω𝑝 20000 π
log 35.926
𝑁≥ N≥
log 2.5
3.908

Rounding off N to the next higher integer we get N = 4

20
 Homework 2.2 Continued
1
ormalized Lowpass Butterworth Filter for order N=4 𝐻 ( 𝑠 )=
( 𝑠 2+ 0.76537 𝑠+1 ) ( 𝑠 2 +1.8477 𝑠+1 )
Ω𝑝 20000 𝜋
The Cut-off frequency Ω 𝑐= Ω 𝑐= 1/( 2 𝑋 4 ) 26015 rad/sec
( 10 0.1 𝛼 𝑝
−1 )
1/ 2 𝑁
( 100.1 𝑋 0.5 −1 )

Substitute in H(s), here

1
𝐻 ( 𝑠 )=

(( ) ( ) )(( ) ( ) )
2 2
𝑠 𝑠 𝑠 𝑠
+ 0.76537 +1 +1.8477 +1
26015 𝜋 26015 𝜋 26015 𝜋 26015 𝜋

1
𝐻 ( 𝑠 )=

[( ) ][ ( ) ]
2 2
1 1
(𝑠
2
+0.76537 𝑋 26015 𝜋 𝑠+ ( 26015 𝜋 )
2
) ( 𝑠 2 +1.8477 𝑋 26015 𝜋 𝑠 + ( 26015 𝜋 )2 )
26015 𝜋 26015 𝜋

19
4.4616 𝑋 10
𝐻 ( 𝑠 )= 2
( 𝑠 + 62552.56 𝑠+ ( 26015 𝜋 )2 ) ( 𝑠 2+ 151009.8 𝑠 + ( 26015 𝜋 )2 )
21
 Design Procedure for Analog Lowpass Chebyshev Filters

Step 1: From the given specification identify the values of

Where,
 - Passband attenuation in dB
 - Stopband attenuation in dB
 - Passband frequency in rad/sec
 - Stopband frequency in rad/sec

Step 2: Find the value of ε and λ

or
or

22
 Design Procedure for Analog Lowpass Chebyshev Filters
Step 3: Determine the order of the filter N and roundoff it to the next
higher integer
or
Step 4: Using the following formulas find the values of a and b, which are
minor and major axis of the ellipse respectively

[𝜇 1/𝑁
−𝜇 −1 / 𝑁
] [ 𝜇 1/ 𝑁 +𝜇− 1/ 𝑁 ]
𝑎=Ω 𝑝 𝑏=Ω 𝑝
2 2
Where

𝜇=𝜀 + √ 𝜀 + 1
−1 −2

𝜀=√ 10 0.1 𝛼 𝑝
−1
For Normalized Chebyshev filter 23
 Design Procedure for Analog Lowpass Chebyshev Filters
Step 5: Calculate the poles of Chebyshev filter which lie on an ellipse by
using the formula
𝑠𝑘 =𝑎cos 𝜙 𝑘 + 𝑗 𝑏 sin 𝜙𝑘 , 𝑘=1 , 2 ,3 … 𝑁
𝜋 2 𝑘− 1
𝑤h𝑒𝑟𝑒 𝜙𝑘 = +
2 2𝑁( )
𝜋 , 𝑘=1 , 2 , 3 … 𝑁
Step 6: Find the denominator polynomial of the transfer function using the
above poles
Step 7: The numerator of the transfer function depends on the value of N
(a) if N is odd, substitute s=0 in the denominator polynomial and find the
value. This value is equal to the numerator of the transfer function. [for N
is odd, the magnitude response |H(jΩ)| starts at 1]
(b) If N is even substitute s=0 in the denominator polynomial and divide the
results by . This value is equal to the numerator. [for N is even, the
oscillatory curve starts from ]
24
 Comparison between Butterworth filter and Chebyshev
Filters
The magnitude response of Butterworth filter decreases
monotonically as the frequency Ω increases from 0 to ∞, whereas the
magnitude response of the Chebyshev filter exhibits ripples in the
passband or stopband according to the type
The transition band is more in Butterworth filter when compared to
Chebyshev filter
The poles of Butterworth filter lie on a circle, whereas the poles of
the Chebyshev filter lie on an ellipse
For the same specifications, the number of poles in Butterworth are
more when compared to the Chebyshev filter i.e the order of the
Chebyshev filter is less than that of Butterworth filter. This is a great
advantage because less number of discrete components will be
necessary to construct the filter
25
Problem 2.5 Given the specifications . Determine the order of the
filter using Chebyshev approximation. Find H(s).
Solution:

Step 1: Order of0.1the filter:-

√ ( 10 −1 )
√ ( 10 0.1 𝑋 16 −1 )
√ ( 10 1. 6 −1 )
𝛼 𝑠
−1 −1 −1
𝑐𝑜𝑠h 𝑐𝑜𝑠h 𝑐𝑜𝑠h
( 10 0.1 𝛼 − 1 )
𝑝
( 100.1 𝑋 3 −1 ) ( 100. 3 −1 )
𝑁≥ 𝑁≥ 𝑁≥
Ω𝑠 −1 4000 𝜋 𝑐𝑜𝑠h−1 2
𝑐𝑜𝑠h
−1
𝑐𝑜𝑠h
Ω𝑝 2000 𝜋

−1
𝑐𝑜𝑠h 6.245
𝑁≥ −1
𝑐𝑜𝑠h 2 N≥
1.912

Rounding off N to the next higher integer we get N = 2 26

 Problem 2.5 Continued
ep 2: The values of minor axis and major axis can be found as below

𝜀=√ 10 0.1 𝛼 𝑝
−1 𝜀=√ 10 − 1=0.9976 ≈ 1
0.3

𝜇=𝜀 + √ 𝜀 + 1
−1 −2
𝜇=( 1 ) −1
+ √1 −2
+1 𝜇=2.414

[ 𝜇1 / 𝑁 −𝜇 −1 / 𝑁 ] [ ( 2.414 ) 1/ 2
− ( 2.414 )− 1/ 2
]
𝑎=Ω 𝑝 𝑎= 2000 𝜋 𝑋
2 2
𝑎=2859 ( 𝑜𝑟 ) 910 𝜋

𝑏=Ω 𝑝
[ 𝜇 1/ 𝑁 +𝜇− 1/ 𝑁 ]
𝑏=2000 𝜋 𝑋
[ ( 2.414 )1 /2 + ( 2.414 )−1 / 2 ]
2 2
𝑏=6903 ( 𝑜𝑟 ) 2197 𝜋
27
Problem 2.5 Continued
tep 3: The Pole values are given by
𝑠𝑘 =𝑎 cos 𝜙 𝑘 + 𝑗 𝑏 sin 𝜙𝑘 , 𝑘=1 , 2
𝜙𝑘 = (𝜋
2
+
2𝑁 )
2 𝑘− 1
𝜋 , 𝑘=1 , 2

𝜙 = +( )
𝜋 2(1) −1 𝜋 𝜋 𝜊
1 𝜋 𝜙1 = + =135
2 2𝑋2 2 4

𝜙 = +( )
𝜋 2( 2) −1 𝜋 3𝜋 𝜊
2 𝜋 𝜙2 = + =225
2 2𝑋2 2 4
𝑜 𝑜
𝑠1=𝑎cos 𝜙 1+ 𝑗 𝑏 sin 𝜙 1 𝑠1=2859 cos 135 + 𝑗 6903 sin 135
𝑠1=−2021.62 + 𝑗 4881.16
𝑜 𝑜
𝑠2 =𝑎cos 𝜙 2 + 𝑗 𝑏 sin 𝜙2 𝑠2 =2859 cos 225 + 𝑗 6903 sin 225
𝑠2 =−2021.62 − 𝑗 4881.16
28
 Problem 2.5 Continued
Step 4: Finding Denominator of H(s)
The denominator of
𝐻 ( 𝑠 )=(𝑠 −(−2021.62 + 𝑗 4881.16))(𝑠 −(− 2021.62 − 𝑗 4881.16))
𝐻 ( 𝑠 )=(𝑠+2021.62− 𝑗 4881.16)(𝑠 +2021.62 + 𝑗 4881.16 )¿
𝐻 ( 𝑠 )= ( 𝑠 +2021.62 )2 + ( 4881.16 )2

Step 5: Finding numerator of H(s)

= 19737238.5

Step 6: The Transfer Function H(s)

19737238.5 19737238.5
𝐻 ( 𝑠 )= 𝐻 ( 𝑠 )=
( 𝑠+2021.62 )2+ ( 4881.16 )2 𝑠2 + 4043.24 𝑠+ 27912670.37

29
Problem 2.6 Obtain an analog Chebyshev filter transfer function that satisfies
the constraints
Solution:

√ 1+ 𝜀
1
2
= 0.707 𝜀=
√( 1 2
0.707 )
−1=1

√ 1+ 𝜆
1
2
= 0.1 λ=
√( 0.1)
1 2
− 1=9.95

Step 1: Order of the filter:-

−1 𝜆 9.95
𝑐𝑜𝑠h 𝑐𝑜𝑠h
−1
𝜀 1
𝑁≥
−1 Ω𝑠 𝑁≥
−1 4
N≥
𝑐𝑜𝑠h 𝑐𝑜𝑠h
Ω𝑝 2 2.269

Rounding off N to the next higher integer we get N = 3 30

 Problem 2.6 Continued
ep 2: The values of minor axis and major axis can be found as below

𝜀=1

𝜇=𝜀 + √ 𝜀 + 1
−1 −2
𝜇=( 1 ) −1
+ √1 −2
+1 𝜇=2.414

[ 𝜇1 / 𝑁 −𝜇 −1 / 𝑁 ] [ ( 2.414 ) 1 /3
− ( 2.414 ) −1 / 3
]
𝑎=Ω 𝑝 𝑎= 2 𝑋
2 2
𝑎=0.596

𝑏=Ω 𝑝
[ 𝜇 1/ 𝑁 +𝜇− 1/ 𝑁 ]
𝑏=2 𝑋
[ ( 2.414 )1/ 3 + ( 2.414 )−1 /3 ]
2 2
𝑏=2.087
31
Problem 2.6 Continued
tep 3: The Pole values are given by
𝑠𝑘 =𝑎cos 𝜙 𝑘 + 𝑗 𝑏 sin 𝜙𝑘 , 𝑘=1 , 2 ,3

𝜙𝑘 =
𝜋
2
+ (
2 𝑘− 1
2𝑁 )
𝜋 , 𝑘=1 , 2 , 3

𝜋
𝜙1 = +
2 (
2(1) −1
2 𝑋3
𝜋 ) 𝜙1=
𝜋 𝜋
2
+
6
=120
𝜊
𝜙2 =
𝜋
2
+(2( 2) −1
2𝑋3 )
𝜋 𝜙2 =
𝜋
2
+
𝜋
2
=180
𝜊

𝜙3 =
𝜋
2
+(2 (3)− 1
2𝑋3
𝜋) 𝜙3 =
𝜋 5𝜋
2
+
6
=240 𝜊
𝑜 𝑜
𝑠1=𝑎cos 𝜙 1+ 𝑗 𝑏 sin 𝜙 1 𝑠1=0.596 cos 120 + 𝑗 2.087 sin 120
𝑠1=−0.298 + 𝑗 1.807
𝑜 𝑜
𝑠2 =𝑎cos 𝜙 2 + 𝑗 𝑏 sin 𝜙2 𝑠2 =0.596 cos 180 + 𝑗 2.087 sin 180
𝑠2 =−0.596
𝑜 𝑜
𝑠3 =𝑎 cos 𝜙 3 + 𝑗 𝑏sin 𝜙 3 𝑠3 =0.596 cos 240 + 𝑗 2.087 sin 180
𝑠3 =−0.298 − 𝑗 1.807 32
 Problem 2.6 Continued
Step 4: Finding Denominator of H(s)
The denominator of
𝐻 ( 𝑠 )= ( 𝑠 − ( − 0.298+ 𝑗 1.807 ) ) ( 𝑠 −(− 0.596) ) (𝑠 −(− 0.298 − 𝑗 1.807))
𝐻 ( 𝑠 )= ( 𝑠 +0.298 − 𝑗 1.807 ) ( 𝑠 +0.596 ) (𝑠 +0.298 + 𝑗 1.807)
(

Step 5: Finding numerator of H(s)

𝐻 ( 𝑠 )=2

Step 6: The Transfer Function H(s)

2
𝐻 ( 𝑠 )= 2
¿
( 𝑠+ 0.596 ) ( 𝑠 +0.596 𝑠 +3.354 ¿
33
Problem 2.7 Design a Chebyshev filter with a maximum passband
attenuation of 2.5dB at and the stopband attenuation of 30dB at .
Solution:

Step 1: Order of0.1the filter:-

√ ( 10 −1 )
√ ( 10 0.1 𝑋 30 −1 )
√ ( 103 −1 )
𝛼 𝑠
−1 −1 −1
𝑐𝑜𝑠h 𝑐𝑜𝑠h 𝑐𝑜𝑠h
( 10 0.1 𝛼 − 1 )
𝑝
( 10 0.1 𝑋 2.5 − 1 ) ( 10 0.25 − 1 )
𝑁≥ 𝑁≥ 𝑁≥
Ω𝑠 −1 50 𝑐𝑜𝑠h−1 2.5
𝑐𝑜𝑠h
−1
𝑐𝑜𝑠h
Ω𝑝 20

−1
𝑐𝑜𝑠h 35.83
𝑁≥ −1
𝑐𝑜𝑠h 2.5 N≥
2.726

Rounding off N to the next higher integer we get N = 3 34

 Problem 2.7 Continued
ep 2: The values of minor axis and major axis can be found as below

𝜀=√ 10 0.1 𝛼 𝑝
−1 𝜀=√ 10 0.1 𝑋 2.5
−1=0.882

𝜇=𝜀 + √ 𝜀 + 1 𝜇=( 0.882 ) + √ ( 0.882 ) +1 𝜇=2.65

−1 −2 −1 −2

[ 𝜇1 / 𝑁 −𝜇 −1 / 𝑁 ] [ ( 2.65 ) 1 /3
− ( 2.65 ) −1 /3
]
𝑎=Ω 𝑝 𝑎= 20 𝑋
2 2
𝑎=6.61

𝑏=Ω 𝑝
[ 𝜇 1/ 𝑁 +𝜇− 1/ 𝑁 ]
𝑏=20 𝑋
[ ( 2.65 )1/ 3 + ( 2.65 )− 1/ 3 ]
2 2
𝑏=21.06
35
Problem 2.7 Continued
tep 3: The Pole values are given by
𝑠𝑘 =𝑎cos 𝜙 𝑘 + 𝑗 𝑏 sin 𝜙𝑘 , 𝑘=1 , 2 ,3

𝜙𝑘 =
𝜋
2
+ (
2 𝑘− 1
2𝑁 )
𝜋 , 𝑘=1 , 2 , 3

𝜋
𝜙1 = +
2 (
2(1) −1
2 𝑋3
𝜋 ) 𝜙1=
𝜋 𝜋
2
+
6
=120
𝜊
𝜙2 =
𝜋
2
+(2( 2) −1
2𝑋3 )𝜋 𝜙2 =
𝜋
2
+
𝜋
2
=180
𝜊

𝜙3 =
𝜋
2
+(2 (3)− 1
2𝑋3
𝜋) 𝜙3 =
𝜋 5𝜋
2
+
6
=240 𝜊
𝑜 𝑜
𝑠1=𝑎cos 𝜙 1+ 𝑗 𝑏 sin 𝜙 1 𝑠1=6.61 cos 120 + 𝑗 21.06 sin 120
𝑠1=−3.305 + 𝑗 18.24
𝑜 𝑜
𝑠2 =𝑎cos 𝜙 2 + 𝑗 𝑏 sin 𝜙2 𝑠2 =6.61 cos 180 + 𝑗 21.06 sin 180
𝑠2 =−6.61
𝑜 𝑜
𝑠3 =𝑎 cos 𝜙 3 + 𝑗 𝑏sin 𝜙 3 𝑠3 =6.61 cos 240 + 𝑗 21.06 sin 180
𝑠3 =−3.305 − 𝑗 18.24 36
 Problem 2.7 Continued
Step 4: Finding Denominator of H(s)
The denominator of
𝐻 ( 𝑠 )= ( 𝑠 − ( −3.305 + 𝑗 18.24 ) ) ( 𝑠 −(− 6.61) ) (𝑠 −(− 3.305 − 𝑗 18.24))
𝐻 ( 𝑠 )= ( 𝑠 +3.305 − 𝑗 18.24 ) ( 𝑠 +6.61 ) (𝑠+ 3.305+ 𝑗 18.24)
(

Step 5: Finding numerator of H(s)

𝐻 ( 𝑠 )=2271.33

Step 6: The Transfer Function H(s)

2271.33
𝐻 ( 𝑠 )= 2
¿
( 𝑠+6.61 ) ( 𝑠 + 6.61 𝑠+ 343.62¿
37
Homework 2.3 Design a lowpass Chebyshev filter with a pass band
attenuation of 3dB at the frequency of 1.5kHz and the stopband
attenuation of 40dB at 3kHz.

Homework 2.4 Design an analog lowpass Chebyshev filter that has the
following constraints
 Design of Digital IIR Filters from Analog Filters
Several methods used to design digital filters, but the conversion
technique is to be effective and it should possess the following
desirable properties
(a) The jΩ-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
(b) The left-half plane of the s-plane should map into the inside of the unit
circle in the z-plane. Thus a stable analog filter will be converted to a
stable digital filter
Most widely used techniques
1. Approximation of derivatives
2. The impulse invariant transformation
3. The bilinear transformation
39
4. The matched z-transformation technique
 Design of Digital IIR Filter – Impulse Invariant
transformation
Step 1: From the given specifications find the transfer function of an
analog filter
Step 2: Select the sampling rate of the digital filter, T seconds/sample
Step 3: Express the analog filter transfer function as the sum of single
pole filters

Step 4: Compute the z-transform of the digital filter by using the formula

For high sampling rates use

40
Problem 2.8 For the analog transfer function, , determine H(z) using
impulse invariant transformation if (a) T= 1 sec and (b) T=0.1 sec
Solution:
2 2
𝐻 ( 𝑠 )= 𝐻 ( 𝑠 )=
𝑠2 +3 𝑠+ 2 ( 𝑠+1 ) ( 𝑠 +2)
By partial fraction expansion
technique 𝐴 𝐵
𝐻 ( 𝑠 )= + 𝐴=2, 𝐵=−2
( 𝑠+1 ) ( 𝑠 +2 )
2 2 2 2
𝐻 ( 𝑠 )= − 𝐻 ( 𝑠 )= −
𝑠+1 𝑠+ 2 𝑠 −(−1) 𝑠 −( −2)
Using impulse invariance technique
𝑐1 𝑐2 𝑐1 𝑐2
𝐻 ( 𝑠 )= + 𝐻 ( 𝑧)= 𝑝1𝑇 −1
+ 𝑝2𝑇 −1
𝑠 −𝑝 1 𝑠 −𝑝 2 1 −𝑒 𝑧 1− 𝑒 𝑧
2 2
𝐻 ( 𝑧)= (− 1)𝑇 −1
− (−2 )𝑇 −1
1 −𝑒 𝑧 1− 𝑒 𝑧
(a) T= 1 sec
2 2 2 2
𝐻 ( 𝑧)= − 𝐻 ( 𝑧)= −
1 −𝑒 −1 𝑧 − 1 1 − 𝑒− 2 𝑧 − 1 1 − 0.3678 𝑧
−1
1 −0.1353 𝑧
−1

−1 −1
0.465 𝑧 0.1722 𝑧
𝐻 ( 𝑧)= −1 −2 ( 𝑏 ) 𝑇 =0.1 𝑠𝑒𝑐 , 𝐻 ( 𝑧 )= −1 41
−2
1 − 0.503 𝑧 + 0.04976 𝑧 1− 1.7235 𝑧 + 0.7408 𝑧
Problem 2.9 Convert the analog filter with system transfer function, ,
into a digital IIR filter by means of the impulse invariant method
Solution:
𝑠+ 0.1 𝑠+ 0.1 𝑠 +0.1
𝐻 ( 𝑠 )= 𝐻 ( 𝑠 )= 𝐻 ( 𝑠 )=
( 𝑠+ 0.1 )2 +9 𝑠2 +2 𝑥 0.1 𝑥𝑠+ 0.12 +9 2
𝑠 + 0.2 𝑠+ 9.01

𝑠+ 0.1
𝐻 ( 𝑠 )=
(𝑠 +0.1 − 𝑗 3)( 𝑠 +0.1+ 𝑗 3)

By partial fraction expansion

technique

42
 Problem 2.9 Continued
Using impulse invariance technique
𝑐1 𝑐2 𝑐1 𝑐2
𝐻 ( 𝑠 )= + 𝐻 ( 𝑧)= 𝑝1𝑇 −1
+ 𝑝2𝑇 −1
𝑠 −𝑝 1 𝑠 −𝑝 2 1 −𝑒 𝑧 1− 𝑒 𝑧

43
Problem 2.10 Design a Butterworth filter using the impulse invariance method for the following
specifications

Solution:
𝜔 𝑝 =Ω𝑝 𝑇 =0.2 π
𝜔 𝑠 =Ω𝑠 𝑇 =0.6 π
1
√ 1+ 𝜀 2
= 0.8 𝜀=
√( )
1 2
0.8
−1=0.75

√( )
2
1 1
= 0.2 λ= −1=4. 8 99
√ 1+ 𝜆 2
0. 2
Order of the filter:

𝑁≥
𝜆
𝜀
log ( ) 𝑁≥
log (
4.899
0.75 ) N ≥1.71
Ω𝑠 0.6 𝜋
log log
Ω𝑝 0.2 𝜋

Rounding off N to the next higher integer we get N = 2

44
 Problem 2.10 Continued
1
ormalized Lowpass Butterworth Filter for order N=2 𝐻 ( 𝑠 )=
( 𝑠 2+ √ 2 𝑠+ 1 )
Ω𝑝 0.2 π
The Cut-off frequency Ω 𝑐= Ω 𝑐= rad/sec
ɛ
1/𝑁
( 0.75 )1 /2

Substitute in H(s), here

1
𝐻 ( 𝑠 )=

(( ) ( ) )
2
𝑠 𝑠
+ √2 +1
0.231 𝜋 0.231 𝜋

1
𝐻 ( 𝑠 )=

[( ) ]
2
1
( 𝑠 2 +0.231 𝜋 𝑋 √ 2 𝑋 𝑠 + ( 0.231 𝜋 ) 2)
0.231 𝜋

0.5266
𝐻 ( 𝑠 )=
( 𝑠 2+ 1.03 𝑠 +0.5266 )
45
 Problem 2.10 Continued

𝐻 ( 𝑠 )=
0.5266 𝐴 𝐵
( 𝑠 2+ 1.03 𝑠 +0.5266 ) 𝐻 ( 𝑠 )= +
𝑠+ 0.51+ 𝑗 0.51 𝑠+ 0.51 − 𝑗 0.51
𝐴= 𝑗 0.516 , 𝐵=− 𝑗 0.516
0.516 𝑗 0.516 𝑗
𝐻 ( 𝑠 )= −
𝑠+ 0.51+ 𝑗 0.51 𝑠+ 0.51− 𝑗 0.51
0.516 𝑗 0.516 𝑗
𝐻 ( 𝑠 )= −
𝑠 −(−0.51 − 𝑗 0.51) 𝑠 −(− 0.5 1+ 𝑗 0.51)

Using impulse invariance technique

𝑐1 𝑐2 𝑐1 𝑐2
𝐻 ( 𝑠 )= + 𝐻 ( 𝑧)= 𝑝1𝑇 −1
+ 𝑝2𝑇 −1
𝑠 −𝑝 1 𝑠 −𝑝 2 1 −𝑒 𝑧 1− 𝑒 𝑧

0.516 𝑗 0.516 𝑗
𝐻 ( 𝑧)= − 0.51𝑇 − 𝑗 0.51 𝑇 −1
− −0.51 𝑇 𝑗 0.51 𝑇 −1
1 −𝑒 𝑒 𝑧 1 −𝑒 𝑒 𝑧
Substitute T=1 sec

0.516 𝑗 0.516 𝑗
𝐻 ( 𝑧)= − 𝑗 0.51 − 1
− 𝑗 0.51 −1
1 − 0 .6 𝑒 𝑧 1 − 0.6 𝑒 𝑧 46
Problem 2.10 Continued
0.516 𝑗 0.516 𝑗
𝐻 ( 𝑧)= − 𝑗 0.51 − 1
− 𝑗 0.51 −1
1 − 0 .6 𝑒 𝑧 1 − 0.6 𝑒 𝑧
0.516 𝑗 ( 1 −0.6 𝑒 𝑧 ) −0.516 𝑗 ( 1 −0.6 𝑒 𝑧 )
𝑗 0.51 −1 − 𝑗 0.51 −1
𝐻 ( 𝑧)= − 𝑗 0.51 −1 𝑗 0.51 −1
(1 −0 .6 𝑒 𝑧 )(1 −0.6 𝑒 𝑧 )
𝑗 0.51 −1 − 𝑗 0.51 −1
0.516 𝑗 −0. 3096 𝑒 𝑧 − 0.516 𝑗+ 0.3096 𝑒 𝑧
𝐻 ( 𝑧)= 𝑗 0.51 −1 − 𝑗 0.51 − 1 −2
1− 0.6 𝑒 𝑧 −0.6 𝑒 𝑧 + 0.36 𝑧

− 0.3096 (𝑒 𝑗 0.51 −𝑒 − 𝑗 0.51 ) 𝑧 −1

𝐻 ( 𝑧)=
1 − 0.6(𝑒 𝑗 0.51 +𝑒 − 𝑗 0.51) 𝑧 −1 + 0.36 𝑧 − 2
−1
− 0.3096 (2 𝑗 sin 0.51) 𝑧
𝐻 ( 𝑧)=
1 − 0.6(2 cos 0.51) 𝑧 −1 +0.36 𝑧 −2
−1
0.3022 𝑧
𝐻 ( 𝑧)=
1 −1.0473 𝑧 −1 +0.36 𝑧 −2

47
Problem 2.11 Design a Chebyshev filter using the impulse invariance method for the following
specifications

Solution:
𝜔 𝑝 =Ω𝑝 𝑇 =0.2 π
𝜔 𝑠 =Ω𝑠 𝑇 =0.6 π
1
√ 1+ 𝜀 2
= 0.8 𝜀=
√( 1 2
0.8 )−1=0.75

√( )
2
1 1
= 0.2 λ= −1=4. 8 99
√ 1+ 𝜆 2
0. 2
Step 1: Order of the filter:-

𝑁≥
𝜆
𝑐𝑜𝑠h−1
𝜀 ( ) 𝑁≥
𝑐𝑜𝑠h−1 (
4.899
0.75 ) N ≥1.45
−1 Ω 𝑠 0.6 𝜋
𝑐𝑜𝑠h 𝑐𝑜𝑠h− 1
Ω𝑝 0.2 𝜋

Rounding off N to the next higher integer we get N = 2

48
 Problem 2.11 Continued
ep 2: The values of minor axis and major axis can be found as below
𝜀=0.75

𝜇=𝜀 + √ 𝜀 + 1 𝜇=( 0.75 ) + √ ( 0.75 ) +1 𝜇=3

−1 −2 −1 −2

[ 𝜇1 / 𝑁 −𝜇 −1 / 𝑁 ] [ ( 3 ) 1/ 2
− ( 3 ) −1 / 2
]
𝑎=Ω 𝑝 𝑎= 0.2 π 𝑋
2 2
𝑎=0.3627

𝑏=Ω 𝑝
[ 𝜇 1/ 𝑁 +𝜇− 1/ 𝑁 ]
𝑏=0.2 π 𝑋
[ ( 3 )1 /2 + ( 3 )−1 /2 ]
2 2
𝑏=0.7255
49
Problem 2.11 Continued
tep 3: The Pole values are given by
𝑠𝑘 =𝑎 cos 𝜙 𝑘 + 𝑗 𝑏 sin 𝜙𝑘 , 𝑘=1 , 2
𝜙𝑘 = (𝜋
2
+
2𝑁 )
2 𝑘− 1
𝜋 , 𝑘=1 , 2

𝜙 = +( )
𝜋 2(1) −1 𝜋 𝜋 𝜊
1 𝜋 𝜙1 = + =135
2 2𝑋2 2 4

𝜙 = +( )
𝜋 2( 2) −1 𝜋 3𝜋 𝜊
2 𝜋 𝜙2 = + =225
2 2𝑋2 2 4
𝑜 𝑜
𝑠1=𝑎cos 𝜙 1+ 𝑗 𝑏 sin 𝜙 1 𝑠1=0.3627 cos 135 + 𝑗 0.7255 sin 135
0.513
𝑜 𝑜
𝑠2 =𝑎cos 𝜙 2 + 𝑗 𝑏 sin 𝜙2 𝑠2 =0.3627 cos 225 + 𝑗 0.7255 sin 225
𝑠2 =−0.2564 − 𝑗 0.513
50
 Problem 2.11 Continued
Step 4: Finding Denominator of H(s)
The denominator of
𝐻 ( 𝑠 )=(𝑠 −(−0.2564 + 𝑗 0.513))(𝑠 −(−0.2564 + 𝑗 0.513))
𝐻 ( 𝑠 )=(𝑠+0.2564 − 𝑗 0.513)(𝑠 +0.2564 + 𝑗 0.513)¿
𝐻 ( 𝑠 )= ( 𝑠 +0.2564 )2 + ( 0.513 )2

Step 5: Finding numerator of H(s)

= 0.264

Step 6: The Transfer Function H(s)

0.2 64 0.264
𝐻 ( 𝑠 )= 𝐻 ( 𝑠 )=
( 𝑠+ 0.2564 )2 + ( 0.513 )2 2
𝑠 + 0.513 𝑠 +0.33

51
 Problem 2.11 Continued Step 7:

𝐻 ( 𝑠 )=
0.264 𝐴 𝐵
2 𝐻 ( 𝑠 )= +
𝑠 + 0.513 𝑠 +0.33 𝑠+0.256 − 𝑗 0.513 𝑠+ 0.256+ 𝑗 0.513
𝐴= 𝑗 0.257 , 𝐵=− 𝑗 0.257
𝑗 0.257 − 𝑗 0.257
𝐻 ( 𝑠 )= +
𝑠+0.256 − 𝑗 0.513 𝑠+ 0.256+ 𝑗 0.513
𝑗 0.257 − 𝑗 0.257
𝐻 ( 𝑠 )= +
𝑠 −(−0.256 + 𝑗 0.513) 𝑠 −(− 0.256 − 𝑗 0.513)
Using impulse invariance technique
𝑐1 𝑐2 𝑐1 𝑐2
𝐻 ( 𝑠 )= + 𝐻 ( 𝑧)= 𝑝1𝑇 −1
+ 𝑝2𝑇 −1
𝑠 −𝑝 1 𝑠 −𝑝 2 1 −𝑒 𝑧 1− 𝑒 𝑧

0.257 𝑗 0. 257 𝑗
𝐻 ( 𝑧)= − 0.256 𝑇 − 𝑗 0.51 3 𝑇 −1
− −0. 256 𝑇 𝑗 0.513 𝑇 −1
1 −𝑒 𝑒 𝑧 1 −𝑒 𝑒 𝑧
Substitute T=1 sec
0.257 𝑗 0.257 𝑗
𝐻 ( 𝑧)= − 𝑗 0.51 3 −1
− 𝑗 0.513 −1
1 − 0.774 𝑒 𝑧 1− 0.774 𝑒 𝑧
52
Problem 2.11 Continued

0.257 𝑗 0.257 𝑗
𝐻 ( 𝑧)= − 𝑗 0.51 3 −1
− 𝑗 0.513 −1
1 − 0.774 𝑒 𝑧 1− 0.774 𝑒 𝑧

0.257 𝑗 ( 1 −0.774 𝑒 𝑗 0.513 𝑧 −1 ) − 0.257 𝑗 ( 1 −0.774 𝑒− 𝑗 0.513 𝑧 −1 )

𝐻 ( 𝑧)=
( 1− 0.774 𝑒 − 𝑗 0.51 3 𝑧 −1 ) ( 1 − 0.774 𝑒 𝑗 0.513 𝑧 −1 )
𝑗 0.513 −1 − 𝑗 0.513 −1
0.257 𝑗 − 𝑗 0.199 𝑒 𝑧 − 0.257 𝑗+ 𝑗 0.199 𝑒 𝑧
𝐻 ( 𝑧)= 𝑗 0.513 −1 − 𝑗 0.513 − 1 −2
1 − 0.774 𝑒 𝑧 −0.774 𝑒 𝑧 + 0.599 𝑧

− 𝑗 0.199 ( 𝑒 𝑗 0.513 − 𝑒 − 𝑗 0.513 ) 𝑧 − 1

𝐻 ( 𝑧)=
1 − 0.774 ¿ ¿ ¿
−1
0.1953 𝑧
𝐻 ( 𝑧)= −1 −2
1 −1.3487 𝑧 +0.599 𝑧

53
 Problem 2.12 Design a third order Butterworth digital filter using impulse invariant
technique. Assume sampling period T=1 sec.
Solution:
he transfer function of the third order normalised Butterworth filter is
1 1
𝐻 ( 𝑠 )= 𝐻 ( 𝑠 ) =
( 𝑆 +1 ) ( 𝑠2 + 𝑠+1 ) (𝑠 +1)(𝑠 +0.5 + 𝑗 0.866)(𝑠 +0.5 − 𝑗 0.866)
+

𝐴=1

1 1
𝐵= 𝐵=
(− 0.5− 𝑗 0.866 +1)(−0.5 − 𝑗 0.866+ 0.5 − 𝑗 0.866) (0.5 − 𝑗 0.866)(− 𝑗 1.732)
1 X − 1.5+ 𝑗 0.866
𝐵= 𝐵=
− 1.5 − 𝑗 0.866 3

𝐵=−0.5 + j 0.288 𝐶=−0.5 − j 0.288 54

 Problem 2.12 Continued

Using impulse invariance technique

𝑐1 𝑐2 𝑐3
𝐻 ( 𝑠 )= + + +
𝑠 −𝑝 1 𝑠 −𝑝 2 𝑠 −𝑝 3
Substitute T=1 sec
1 −0.5 + 𝑗 0.288 −0. 5 − 𝑗 0.288
𝐻 ( 𝑧)= −1 −1
+ − 0.5 − 𝑗 0. 866 − 1
+ −0.5 𝑗 0.866 −1
1 −𝑒 𝑧 1− 𝑒 𝑒 𝑧 1− 𝑒 𝑒 𝑧
1 − 0.5+ 𝑗 0.288 −0. 5 − 𝑗 0.288
𝐻 ( 𝑧)= −1
+ − 𝑗 0. 866 − 1
+ 𝑗 0.866 −1
1 − 0.3678 𝑧 1 − 0.6 1𝑒 𝑧 1− 0.61 𝑒 𝑧

𝐻 ( 𝑧)=
1
+
[ ( −0.5 + 𝑗 0.288 ) ( 1 − 0.61𝑒 𝑗 0.866 𝑧 −1 ) ] [ ( − 0.5 − 𝑗 0.288 ) ( 1− 0.6 1 𝑒− 𝑗 0.866 𝑧 − 1 )
1 − 0.3678 𝑧 −1
( 1 − 0.6 1 𝑒− 𝑗 0. 866 𝑧 − 1) ( 1− 0.61 𝑒 𝑗 0.866 𝑧 −1 )
1 [− 0.5+ 0.305 𝑒 𝑗 0.866 −1
𝑧 + 𝑗 0.288 − 𝑗 0.176 𝑒
𝑗 0.866
𝑧
−1
−0.5+ 0.305 𝑒
− 𝑗 0.866 −1
𝑧 − 𝑗 0.288+ 𝑗 0.176 𝑒
− 𝑗 0.866
𝑧
−1
]
𝐻 ( 𝑧)= −1
+ 𝑗 0.866 −1 − 𝑗 0.866 −1 −2
1 − 0.3678 𝑧 1 −0. 61 𝑒 𝑧 −− 0.61 𝑒 𝑧 + 0.3721 𝑧
55
Problem 2.12 Continued

1 [− 0.5+ 0.305 𝑒 𝑗 0.866 −1

𝑧 + 𝑗 0.288 − 𝑗 0.176 𝑒
𝑗 0.866
𝑧
−1
−0.5+ 0.305 𝑒
− 𝑗 0.866 −1
𝑧 − 𝑗 0.288+ 𝑗 0.176 𝑒
− 𝑗 0.866
𝑧
−1
]
𝐻 ( 𝑧)= −1
+ 𝑗 0.866 −1 − 𝑗 0.866 −1 −2
1 − 0.3678 𝑧 1 − 0.61 𝑒 𝑧 − 0.61𝑒 𝑧 + 0.3721 𝑧

− 1+0.305 [ 𝑒 ] 𝑧 − 𝑗 0.176 [ 𝑒 ] 𝑧 −1
𝑗 0.866 − 𝑗 0.866 −1 𝑗 0.866 − 𝑗 0.866
1 +𝑒 −𝑒
( )
𝐻 𝑧= +
1 − 0.61 [ 𝑒 𝑗 0.866 + 𝑒− 𝑗 0.866 ] 𝑧 − 1+ 0.3721 𝑧 −2
−1
1 − 0.3678 𝑧

1 − 1+0.305 [ 2 cos 0.866 ] 𝑧 − 1 − 𝑗 0.176 [ 2 𝑗 sin 0.866 ] 𝑧 −1

𝐻 ( 𝑧)= +
1 − 0.3678 𝑧 −1
1− 0. 61 [ 2 cos 0.866 ] 𝑧 − 1 +0.3721 𝑧 − 2

−1 −1
1 − 1+0.395 𝑧 − 0.268 𝑧
𝐻 ( 𝑧)= −1
+ −1 −2
1 − 0.3678 𝑧 1− 0.79 𝑧 + 0.3721 𝑧

−1
1 − 1+0.663 𝑧
𝐻 ( 𝑧)= −1
+ −1 −2
1 − 0.3678 𝑧 1 − 0.79 𝑧 +0.3721 𝑧

56
Problem 2.13 Design a Chebyshev lowpass filter with the specifications
ripple in the passband ripple in the stopband using impulse invariance
method.
Solution:

The analog frequency

ratio Ω𝑠 𝜔 𝑠 0.3 𝜋
= = =1.5
Ω𝑝 𝜔𝑝 0.2 𝜋
Step 1: Order of the filter:-

√ ( 100.1 𝛼 −1 )
√ ( 10 0.1 𝑋 15 −1 )
𝑠

𝑐𝑜𝑠h−1
−1
𝑐𝑜𝑠h
( 10 0.1 𝛼 − 1 )
𝑝
( 100.1 𝑋 1 − 1 )
𝑁≥ 𝑁≥
𝑐𝑜𝑠h
−1 Ω𝑠 𝑐𝑜𝑠h−1 1.5 N≥
Ω𝑝 3.197

Rounding off N to the next higher integer we get N = 4

57
 Problem 2.13 Continued
ep 2: The values of minor axis and major axis can be found as below

𝜀=√ 10 0.1 𝛼 𝑝
−1 𝜀=√ 10 0.1 𝑋 1
−1=0.508

𝜇=𝜀 + √ 𝜀 + 1 𝜇=( 0.508 ) + √ ( 0.508 ) +1 𝜇=4.17

−1 −2 −1 −2

[ 𝜇1 / 𝑁 −𝜇 −1 / 𝑁 ] [ ( 4.17 ) 1/4
− ( 4.17 ) −1 / 4
]
𝑎=Ω 𝑝 𝑎= 0.2 π 𝑋
2 2
𝑎=0.229

𝑏=Ω 𝑝
[ 𝜇 1/ 𝑁 +𝜇− 1/ 𝑁 ]
𝑏=0.2 π 𝑋
[ ( 4.17 )1/ 4 − ( 4.17 )−1/ 4 ]
2 2
𝑏=0.67
58
Problem 2.13 Continued
tep 3: The Pole values are given by
𝑠𝑘 =𝑎cos 𝜙 𝑘 + 𝑗 𝑏 sin 𝜙𝑘 , 𝑘=1 , 2 ,3

𝜙𝑘 =
𝜋
2
+ (
2 𝑘− 1
2𝑁 )
𝜋 , 𝑘=1 , 2 , 3

𝜋
𝜙1 = +
2 (
2(1) −1
2𝑋4
𝜋 ) 𝜙1= 112.5
𝜊
𝜙2 = ( 𝜋
2
+
2𝑋4 )
2( 2) −1
𝜋 𝜙2 =157.5
𝜊

𝜋
𝜙3 = +
2 (
2 (3)− 1
2𝑋4
𝜋 ) 𝜙3 =202.5
𝜊
𝜙 = +(
4
𝜋
2 2𝑋4 )
2( 4) −1
𝜋 𝜙 4= 247.5𝜊

𝑠1=𝑎cos 𝜙 1+ 𝑗 𝑏 sin 𝜙 1 𝑜
𝑠1=0.229 cos 112.5 + 𝑗 0.67 sin 112.5
𝑜
𝑠1=−0.0876 + 𝑗 0.619
𝑠2 =𝑎cos 𝜙 2 + 𝑗 𝑏 sin 𝜙2 𝑜
𝑠2 =0.229 cos 157.5 + 𝑗 0.67 sin 157.5
𝑜
𝑠2 =−0.2115 + 𝑗 0.2564
𝑠3 =𝑎 cos 𝜙 3 + 𝑗 𝑏sin 𝜙 3 𝑜
𝑠3 =0.229 cos 202.5 + 𝑗 0.67 sin 202.5
𝑜
𝑠3 =−0.2115 − 𝑗 0.2564
𝑠4 =𝑎cos 𝜙 4 + 𝑗 𝑏 sin 𝜙 4 𝑜
𝑠4 =0.229 cos 247.5 + 𝑗 0.67 sin 247.5
𝑜
𝑠4 =− 0.0876 − 𝑗 0.619
59
 Problem 2.13 Continued
Step 4: Finding Denominator of H(s)
The denominator of
𝐻 ( 𝑠 ) = ( 𝑠 − ( − 0.0876 + 𝑗 0.619 ) ) ( 𝑠 − (− 0.2115 + 𝑗 0.2564 ) ) ¿
𝐻 ( 𝑠 )= ( 𝑠 +0.0876 − 𝑗 0.619 ) ( 𝑠 +0.0876 + 𝑗 0.619 ) (𝑠 +0.2115− 𝑗 0.2564)(𝑠 +0.2115+ 𝑗 0.2564 )
)] )]
(

Step 5: Finding numerator of H(s)

𝐻 ( 𝑠 )=0.03834

Step 6: The Transfer Function H(s)

0.03834 2
𝐻 ( 𝑠 )= 2
( 𝑠 + 0.423 𝑠+0.11 ¿ ¿
( 𝑠 + 0.175 𝑠 +0.391 ¿
60
Problem 2.13 Continued