The Sampling theorem

And it is important to disambiguate between the theoretical
analysis undertaken in order to determine the minimum
frequency of sampling, from what is actually happening in
sampling circuits to get our samples. To treat the latter
as an extension of the former is to make a category error.

"glen herrmannsfeldt" <gah@ugcs.caltech.edu> wrote in message
news:cnb8e3$udg$1@gnus01.u.washington.edu...
> As mentioned somewhere else in this newsgroup, Nyquist was actually
> working on how fast telegraph signals could be sent through a band
> limited system and be reliably separated at the other end.

"Airy R. Bean" <me@privacy.net> wrote in message
news:2vr8ufF2ook8qU2@uni-berlin.de...
> Vague approximations are the stuff of technicians and not of engineers.
>

>
> How do I connect a spectrum analyser to a fictitious sampler? Hardly
> a respectable activity for a practising engineer!
>
>

You get an analogue multiplier and use a pulse generator. Multiply the
required signal by the impulse train and take the spectrum of the output-
easy. I saw it done over 20 years ago. If you are worried about the width of
the pulses you need not be concerned as you can very the width and see that
it satisfies the modified sampling theory where the pulses have finite
width.Of course if you know better then publish it or be dammed!

Tom

"Airy R. Bean" <me@privacy.net> wrote in message
news:2vttalF2p3f09U1@uni-berlin.de...
> And it is important to disambiguate between the theoretical
> analysis undertaken in order to determine the minimum
> frequency of sampling, from what is actually happening in
> sampling circuits to get our samples. To treat the latter
> as an extension of the former is to make a category error.

>
Fine but we are still waiting your mathematical explanation!

Tom

You're not describing a fictitious sampler which was the question put.

"Number 6" <No6@distant.island.nz> wrote in message
news:1100677357.781045@ftpsrv1...
> "Airy R. Bean" <me@privacy.net> wrote in message
> news:2vr8ufF2ook8qU2@uni-berlin.de...
> > Vague approximations are the stuff of technicians and not of engineers.
> > How do I connect a spectrum analyser to a fictitious sampler? Hardly
> > a respectable activity for a practising engineer!
> You get an analogue multiplier and use a pulse generator. Multiply the
> required signal by the impulse train and take the spectrum of the output-
> easy. I saw it done over 20 years ago. If you are worried about the width
of
> the pulses you need not be concerned as you can very the width and see
that
> it satisfies the modified sampling theory where the pulses have finite
> width.Of course if you know better then publish it or be dammed!

See lesson 3.

"Number 6" <No6@distant.island.nz> wrote in message
news:1100677435.865716@ftpsrv1...
> "Airy R. Bean" <me@privacy.net> wrote in message
> news:2vttalF2p3f09U1@uni-berlin.de...
> > And it is important to disambiguate between the theoretical
> > analysis undertaken in order to determine the minimum
> > frequency of sampling, from what is actually happening in
> > sampling circuits to get our samples. To treat the latter
> > as an extension of the former is to make a category error.
> Fine but we are still waiting your mathematical explanation!

Vladimir A. Kotelnikow died February 11, 2005...  8-(


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