Forums

bartlet

Started by umurakbas December 19, 2007
hi folks, i have a term project about high-pass bessel filter with fstop =
200Hz and fPass = 400Hz, there's no big deal in "iir" design, i could
overcome it, but at the "fir" design i was asked to perform this by a
"bartlett window" and my brain is almost to explode, i can't find any
articles, any techniques and unfortunately there's no assistants to help
me (and as we all know "if you don't know well enough sth., you'd better
don't know":) )... 

i'm looking for your answers... 
thank you all..


umurakbas wrote:
> ... but at the "fir" design i was asked to perform this by a > "bartlett window" and my brain is almost to explode, i can't find any > articles, any techniques ...
Can't find any articles? Google -> fir filter design -> I'm feeling lucky!
Hi,

  Bartlet window and triangular window are one and the same.
  
  step1 : Find the order of the filter using equation
          
          N = 3.05*Fs/abs(fs - fp)
          
          where Fs = sampling freq, fp = passband freq, fs = stop band   
          freq.

  step 2 : design a ideal hpf of same order using equation

          n  = -(N-1)/2 : (N-1)/2;
          hn = -2*fc*sin(n*2*pi*fc/Fs)/(2*pi*fc*n/Fs);
          fc = (fp+fs)/2;

  step 3 : design a traingular window using the function

          wn = 2*n/(M-1); 0<=n<=(M-1)/2
               2-2*n/(M-1); M-1/2 <= n <= M-1
               0; otherwise.

          this can also be obtained from matlab using 
          wn = bartlett(N+1);

  step 4 :
         coeffs = hn .* wn;

         element wise multiply ideally generated transfer function 
         with the generated window function.

Hope this helps.

Best Regards
Bharat Pathak

Founder and CEO
Arithos Designs
www.arithos.com ~dsp Simplified.







>hi folks, i have a term project about high-pass bessel filter with fstop
=
>200Hz and fPass = 400Hz, there's no big deal in "iir" design, i could >overcome it, but at the "fir" design i was asked to perform this by a >"bartlett window" and my brain is almost to explode, i can't find any >articles, any techniques and unfortunately there's no assistants to help >me (and as we all know "if you don't know well enough sth., you'd better >don't know":) )... > >i'm looking for your answers... >thank you all.. > > >