What are the disadvantages of Impulse Invariant technique used for /JR filt

2 er des·
• • ----- • tgr
Convert the analog filter with transfer function H(s) = (s + 1) (s + 2) into a digital IIR .
"
• ,requency
• Impulse Invariant Technique. Use sampling as SHz. [Dec. 2 fllte
usmg
01 6, 6 f.
Solution : Following are disadvantages of Impulse Invariance Techniques :
(1) The many to one mapping of s-domain into z-domain is main disadvanta
creates aliasing problem. 9es. Th

(2) Highpass filter cannot be designed by Impulse Invariance Method.

(3) The aliasing cannot be controlled by sampling period.
Solution of problem :

Step 1: 2 C2
H(s) = (s + 1) (s + 2) +-
s+2
where, c1 = (s + 1) H(s)ls = -1 = 2

C2 := (s + 2) H(s)ls = -2 = -2
So, 2 2 .
H(s) = s + 1 - s + 2
Step 2: h(t) = 2e-t - 2e-2t
Step 3: h(n) =· h(nn = 2e-nT - 2e-2nT
Since sampling frequency is S Hz, therefore,
1
T = 5 = 0:2 sec

So, h(n) = 2e- 0.2 n _ 2e- o.4 n

Step 4: H(z) = z [2e- 0.2 n _ 2e-<>.4 "]
2 2
or,
H(z) = (1 - e--0.2 rl) (1 - e-<>.4 r 1)

or, H(z) . = 0.296 r1

(1 - 0.818 r 1) (1 - 0.67 rl)
Unit IV 14.10
~ ,7.
ixomple • •
convert the l 1.·1 "th 1. • 5 + 0.l
ana og ,, ter wt system ,unctt~ n H(s) = (s + 0.l)2 + 9
, into /IR filter using
T. The digital filter should have a resonant frequency of m,. = JT/4. [~ay 2017, 10 MJ
B.L • . .
5ofution:
. The analog resonant frequency is 3.
Step 1 • •
!2r = 3 and ror = rt/4 •
Unit IV I 4.21
 AU

_D_IG_IT_A.....:L;_;:S:..::..IG:::..:N..:..:A..!!L~P..!.:P~O~C~ES~S'.!!:IN~G~(~TE~E~&~T_sC)L___ _~__________ _IIR_F_IL.....:T..:::..:.~

2 (J)
• n T
= tan 2
2 1t 2 1t
3 = T tan 8 or T = tan 8 = 0.2761
or, 3

2 (1 - z-1)
St ep 2: 5
= T (1 + r 1)

= 0.2~61 ({ : ~: ) = 7•243 t:: ~:)

St ep -3 : Putting value of s, we ge t :

7.243 G:~~) + 0.1

H(z) = [7.243 (; : ~~) + o.1J + 9

2)
0.116 (1 + 0.027 r 1 - 0.92 r
or, H(z) = 1 - 1.69 r 1 + 0.83 r 2
 r·~· ..., -.v,lii, "
(,....,(,.-..(~e-..... : - ~....., ·~''<-~✓--'I,'(., ~

. ital filter has spedfication as : : {61

l, Ad19
passband frequency rop = 0.4 1t,
Stopband frequcney ros = 0.6 1t
What are the corresponding spedfications for passband and stopband frequencies
analog domain if.
ti) lmpu\se \~variance Techn~que is used for ~esigning.
tii) Bilinear Transformation Method is used for designing ...
~ns.: Passband frequency cop = 0.4 1t •
Stopband frequency COs = 0.6 1t •

Unit IV \ 4.103
 A Q!J

DIGITAL SIGNAL PROCESSING (TE E&TC

(i) Imp ulse Inva rian t Met hod

Analog passband frequ~ncy
Op = cop =0.4 1t
Analog stopband frequency
ns = cos =0.6 1t
(ii) Bilinear Transformation
(Assume T = 1)
Analog· passband frequency
2 ~ 0.41 t
flp = T tan
2 = 2 tan 2 .
flp =. 2 tan 0.2 1t.
Ana log stopband frequency,
2 ~ 0.61 t
ns = • T tan = 2 tan
2 2
ns = 2 tan 0.3 1t
rl
IMal201sl
... . ~-
l! oes1g. n a linear phase FIR band pass filter using hamming window with cut off
~ encies 0.2 rad/sec and 0.3 rad/sec. M = 7. (9)
1 frequ ·

'l
Af1S,:
J H (w)

l
• .
I
.
-0.3 -0.2
1

0.2 0.3
.
I

1t
-
(0

Fig. 5.47

The impulse response h(n) of the filter is calculated by taking inverse Fourier transform
,f H{ro) and is given as :

-02 0.3 ]
h(n) = -f; [
Jel'°" dro Jel'°" dro
-0.3
+
0.2

= 1_ [e-J0.2n - e -,0.3n + eJ03n - eio.2n1

2 7tJn .
1 . .
h{n) = 1tn [sin 0.3 n - sin 0.2 n]

0.3 sin 0.3 n 0.2 sin 0.2 n

h(O) = 7 0.3 n - 7 0.2 n

= Ql_.Ql = Qd = 00318
1t 1t 1t •

h{l) = h(-1) =
'
l1t [sin 0.3 - sin 0.2]
.

Unit VI 5.75
DIGITAL SIGNAL PROCESSING E E&TC
[0.3 - 0.2] 0.1
= - 1t = 1t = 0.0308

h(2) = h(-2) =
2
! 7t [sin 0.6 - sin 0.4]

= 0.0279

h(3) = h(-3) = ;7t [sin 0.9 - sin 0.6i

= 0.0232

Hamming Window W(o) = 1

. 21tn
W(n) = 0.54 + 0.46 cos M - l M =7

-3 -2 -1 · 0 1 2 3 n

Fig. 5.48
21tn
W(n). = 0.54 + o.~6 cos 3

W(l) = 0.54 + 0.46 cos 120 = 0.31

.
W(2) = 0.54 + 0.46 x COS 240 = 0.31

W(3) ,;, 0.54 + 0.46 x cos 360 = 1.0.

Table 1

±0 0.0318 1.0 - 0.0318

±1 0.0308 0.31 9.548 X 10-3
±2 0.0279 0.31 8.649 X 10-3
±3 0.0232 1.0 0.0232