MathGroup Archive 2004

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Filterquestion

  • To: mathgroup at
  • Subject: [mg47636] Re: Filterquestion
  • From: Paul Abbott <paul at>
  • Date: Mon, 19 Apr 2004 04:33:06 -0400 (EDT)
  • Organization: The University of Western Australia
  • References: <c5ob2l$1gn$>
  • Sender: owner-wri-mathgroup at

In article <c5ob2l$1gn$1 at>, "bamse" <bamse at kyllingen.dkkkk> 

> I am wondering if it is possible to design a filter with the impulseresponse
> h(t) such that the convolution of
> a signal p(t)=f(t)*cos(w*t)+g(t)*sin(w*t) with h(t) is equal to
> f(t)*sin(w*t)+g(t)*cos(w*t)???
> In other words, I am looking for a filter H(s) in the s-domain that has the
> following property:
> InverseLaplace { H(s)*P(s) } =  f(t)*sin(w*t)+g(t)*cos(w*t)
> where P(s)=Laplace{p(t)}
> Some more info:
> f(t) is a train of half-sine pulses with the period T
> g(t) is also a train of half-sine pulses with the period T, but
> g(t) is delayed 0.5T in relation to f(t)

For such f and g you should be able to compute P(s) and H(s) in 
closed-form. Extracting h(t) is likely to be problematic though. See


Paul Abbott                                   Phone: +61 8 9380 2734
School of Physics, M013                         Fax: +61 8 9380 1014
The University of Western Australia      (CRICOS Provider No 00126G)         
35 Stirling Highway
Crawley WA 6009                      mailto:paul at 

  • Prev by Date: Re: Alternative to defining 'operator' function?
  • Next by Date: Re: undocumented function: StringQ
  • Previous by thread: Filterquestion
  • Next by thread: Re: Saving as HTML -- Page width problems