[SI-LIST] Re: Frequency v/s Time Domain analysis
- From: "Andrew Ingraham" <a.ingraham@xxxxxxxx>
- To: <darshanmehta2k@xxxxxxxxx>, <si-list@xxxxxxxxxxxxx>
- Date: Thu, 4 Nov 2004 09:43:50 -0500
This is an interesting question. But it's important to distinguish between
*looking* at the data in the frequency or time domain, and *simulating* it
in the frequency or time domain.
There's little reason why you can't simulate in one domain and then look at
the results in the other, as long as you understand the limitations.
People's opinions may differ, but the way I look at it, the world is
fundamentally a time domain universe. Things happen in time. Frequency is
artificial. Our notion of frequency is an interesting and extremely useful
way of looking at some things, but even in tuned circuits, those electrons
are behaving on a moment-by-moment basis, they don't know about our man-made
concept we call frequency. It just happens that when you look at things in
the frequency domain, you may see some extremely useful things, things that
you might completely miss in a time domain view, which is why we do it.
Traditionally, going back some years, simulators had to simulate in the time
domain to handle anything nonlinear. Frequency domain analysis was OK only
for purely linear circuits. Now this distinction is blurred, because some
programs will handle nonlinearities while doing frequency domain analysis.
But not all programs. If you take a traditional SPICE based simulator and
do a frequency domain analysis, it's going to make everything linear when
doing the analysis, which is fine for small-signal RF analysis but not good
for most logic circuits or digital waveforms ... even though you have
given the simulator the full nonlinear device model parameters.
I am somewhat biased because I have used mostly time domain simulations,
unless I'm looking just at passive linear structures (transmission lines,
connector or package models, etc.).
One case where frequency domain simulations have traditionally been better,
is when some of the input data is better (or more easily) expressed in the
frequency domain, such as s-parameters, even skin effect loss. But there
again, some time domain simulators can now handle this kind of frequency
domain device data, to varying degrees.
So, the answer to your question depends in part on which simulator you are
using, and what kind of device models you have.
If you are interested in loss vs. frequency, you are probably looking at
just the analog portion of your whole circuit which is fairly linear.
Otherwise, loss vs. frequency doesn't have much meaning, does it? For
everything else, time domain analysis probably works best.
A DDS might have special considerations. For example, to look at spectral
purity, it might be that a time domain simulation doesn't easily provide
enough precision to simulate a really low noise floor. But I'm only
guessing here.
Regards,
Andy
------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field
or to administer your membership from a web page, go to:
//www.freelists.org/webpage/si-list
For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field
List FAQ wiki page is located at:
http://si-list.org/wiki/wiki.pl?Si-List_FAQ
List technical documents are available at:
http://www.si-list.org
List archives are viewable at:
//www.freelists.org/archives/si-list
or at our remote archives:
http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
http://www.qsl.net/wb6tpu
Other related posts:
