1
$\begingroup$

I am just starting to learn discrete filters and I could use some help. I understand continuous signals and filters.

I am trying to understand the math behind discrete filtering. For example in the s domain, a simple low pass filter can be recognized as $$\dfrac{1}{\tau s+1}$$ How does sampling rate affect a discrete low pass filter?

Improve this question
$\endgroup$
2
  • 1
    $\begingroup$ The cut-off frequency of a discrete-time low pass filter is specified relative to the sampling frequency. So by changing the sampling frequency, you also change its cut-off frequency. Is that what you're asking? $\endgroup$
    – Matt L.
    May 22 '15 at 8:36
  • $\begingroup$ I guess what I am trying to understand where if the equation above was brought to the z-domain, where would the sampling time have a role? $\endgroup$
    – Adam
    May 22 '15 at 12:59
0
$\begingroup$

IF we take the sampling period as T.First I will convert the filter to time domain impulse response as $$ h(s)=\frac{1}{\tau s+1}\\h(s)=\frac{1/\tau}{s+\frac{1}{\tau}} $$ this in time domain is $$ h(t)=\frac{e^{-t/\tau}}{\tau} $$ IF the above signal is sampled at time T the it can be written as $$ h(Tn)=\frac{e^{\frac{-Tn}{\tau}}}{\tau}\\h(n)=\frac{e^{\frac{-Tn}{\tau}}}{\tau} $$ if Z transform taken for the impulse response.Then the equation is $$ H(z)=\frac{1/\tau}{1-e^{\frac{T}{\tau}}z^{-1}} $$ Here we can see the sampling time coming in denominator which could put the pole inside or outside the unit circle

Improve this answer
$\endgroup$
3
  • $\begingroup$ I understand the first part, up to h(n). Could you explain how you got H(z) from h(n)? $\endgroup$
    – Adam
    May 22 '15 at 17:43
  • $\begingroup$ is T/\tau in H(z) supposed to be negative? $\endgroup$
    – Adam
    May 22 '15 at 18:14
  • $\begingroup$ since h(n) is a geometric sequence we can find standard H(z) for h(n) as $$ \frac{1}{1-a^{-1}z^{-1}}$$ here a is $$ e^{T/\tau}$$ and T\/tau should be negative $\endgroup$
    – Vinith
    May 23 '15 at 2:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.