Reply-To: "Kevin Aylward"
From: "Kevin Aylward"
Newsgroups: sci.electronics.design
References: <4c3c095a.0212231022.fdd6810@posting.google.com> <85fg0vc4obu9t3plhdilkbk0i564mafv4e@4ax.com>
Subject: Re: what is monte caro analysis ?
Organization: AnaSoft
X-Newsreader: Microsoft Outlook Express 6.00.2800.1106
X-Inktomi-Trace: public1-pete2-5-cust66.pete.broadband.ntl.com 1040770215 9623 80.0.216.66 (24 Dec 2002 22:50:15 GMT)
Message-ID:
Date: Tue, 24 Dec 2002 22:50:15 -0000
NNTP-Posting-Host: 80.1.224.4
NNTP-Posting-Date: Tue, 24 Dec 2002 22:50:16 GMT
John Woodgate wrote:
> I read in sci.electronics.design that Kevin Aylward
> wrote (in
> ) about 'what is
> monte caro analysis ?', on Tue, 24 Dec 2002:
>> I think so. This is called Worst Case (re-runs). Indeed, SuperSpice
>> does this with a very simple GUI setup
>> (http://www.anasoft.co.uk/screenshot.html:-)
>>
>> I personally, don't use MC at all. I think it has limited use because
>> you have to do a lot of runs to get reasonable results, and it might
>> not even use the max min values at all. I also think it can give one
>> the wrong mindset, e.g. too much trail and error.
>
> OTOH, worst case is very pessimistic, except for circuits with very
> low parts count.
I agree, that it can be pesemistic, but thats usually a good thing. Only
if you really having trouble in meeting spec, is it worth trying to fine
tune the sim a bit.
100 Monte Carlo runs give a good guide to the
> sensitivity of the circuit performance against component values.
But so does a WC, and in a lotv fewer runs.
If
> the values used for each run are logged, it's possible to see which
> parts are responsible for any suspiciously larger excursion of
> performance.
But this is still not usually practical for transient runs on anything
but small circuits.
Kevin Aylward
sales@anasoft.co.uk
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.