IIR implementation

[julianoys]

Hi Everyone...

I've implemented sucessfully, one FIR filter and now, I'm trying t
implement an IIR filter on a TMS320C5402 DSK board...
I have doubts about how to implement it. I thought about using direct for
I, but it is not efficient and I heard something about the quantizatio
problem on IIR filters... so, I implemented the direct form II, with hal
buffers than form I, but all I have in the output is noise...
I want to implement a cascade or parallel form to avoid the quantizatio
error, but first I wanted to try the form II... how do I quantize th
poles and zeros to a 16 bit value for a form II implementation? th
processor operates using 16 bit fixed point math....
for example these coefficients:

Zeros:
0.0018 -0.0061 0.0113 -0.0142 0.0151
-0.0142 0.0113 -0.0061 0.0018

Poles:
1.0000 -5.7444 14.6876 -21.7719 20.4258
-12.4016 4.7534 -1.0506 0.1024

-----------------------------
Coefficients got from Matlab
Chebyshev Tipo II - Lowpass
wp = 0.1250
ws = 0.1875
[n,wn]=cheb2ord(wp,ws,3,60)
n = 8
wn = 0.1875
[b,a]=cheby2(n,60,wn)
tf(b,a,1/16000)
-----------------------------

also for the cascade implementation... how do I find the coefficients fo
it?

Thanks a lot...!!

This message was sent using the Comp.DSP web interface o
www.DSPRelated.com