Re: Assymmetrical IIR bandpass filters doable?

On Jun 13, 9:11 pm, "gelliot" <gibsonell...@xxxxxxxxx> wrote:

What I was wondering if someone can tell me if it is possible to create a
single IIR filter not greater than 4th order, or close to it, that can do
the same thing? Can an IIR be made with an intentionally assymmetrical

now the way that these IIRs are symmetrical is in LOG frequency,
right? this is what you would get for an analog BPF or a digital BPF
where the peak frequency is much less than 1/2 Nyquist. for the
digital filter to have its peak at 1/2 Nyquist, then it is symmetrical
in linear frequency (and of course look asymmetrical in log
frequency). as the peak frequency gets closer to Nyquist, the
asymmetry gets more pronounced because the response has to be
reflected at Nyquist. if you have a zero at Nyquist (which you will
if you designed this using bilinear transform) then the response will
dive to -inf dB at Nyquist, no matter how close you push the peak
frequency to Nyquist.

profile in a more efficient manner than having to use a lowpass and a
highpass cascaded? If it's a coefficient trick, can someone explain to me
how this is done?

if you have MATLAB or some means of easily and repeatedly looking at
trial frequency responses, try cascading two IIRs. when the peak
frequencies lie on top of each other (and they're very low so the
thing behaves more like an analog filter), the result will be
symmetrical. then move one of the peaks to the right of the other
(and maybe reduce its gain or the Q) and you will get an asymmetrical
BPF filter. experimentation is, i think, the best way for you to find
the one that fits best.

another simlar thing to try is to move one of the zeros at Nyquist
(you'll have one for each 2nd-order BPF if you used bilinear to
prototype them) away from Nyquist but remaining on the real axis.
that will reduce the falling slope to some extent.

have fun.

r b-j

.

Relevant Pages

  • Re: filtering decimation of a signal
    ... The main disadvantage of IIR filters are ... A single pole IIR doesn't have a limit cycle oscillation unless you do ... Thus under what condition does a single pole IIR filter never have ...
    (sci.electronics.design)
  • Re: Phaeless filtering
    ... >> 1) Filter a block of data using an IIR filter. ... >> 2) Filter the resulting output in reverse direction through the same IIR ... >> Having made the decision when the output stops, you have defined a FIR ... if you reverse the FIR as a truncated IIR, ...
    (comp.dsp)
  • Re: Signal Spectrum
    ... I was thinking about your idea of using a set of bandpass filters. ... Does someone know a specific filter type, ... have been too quick to reject IIR approaches... ... the incoming frequencies down to DC, ...
    (microsoft.public.win32.programmer.mmedia)
  • Re: signal detection in doppler
    ... coefficients - it doesn't do the online resampling. ... need that many doppler replicas to find a peak. ... But now as I have changed my receiver architecture, ... requirements and the filter. ...
    (comp.dsp)
  • Re: trigonometric
    ... -i come out with a quite high error vector, if the timing error was ... The window filter lets this peak pass and attenuates the rest (actually is ... but significant reason for oversampling. ...
    (comp.dsp)
Loading