RE: Simulation resolution weirdness

Mark,
You should take a close look at the IBIS data you were given. Check the
ramp data. I recently had a case of an oscillation like you describe that
went away after I got an updated and corrected IBIS file. Your supplier can
put numbers in the file that will cause it to pass the ibis checker, but not
the "spice" or human checker.

Aubrey Sparkman
Signal Integrity
Aubrey_Sparkman@Dell.com
(512) 723-3592

> -----Original Message-----
> From: mark.browne@nokia.com [mailto:mark.browne@nokia.com]
> Sent: Tuesday, September 19, 2000 8:51 AM
> To: ibis-users@eda.org
> Cc: Andrew.Ingraham@compaq.com
> Subject: RE: Simulation resolution weirdness
>
>
> I am using Interconnectix. Do the tools not generally work
> in the same way?
>
> Regards,
> Mark
>
> > -----Original Message-----
> > From: EXT Ingraham, Andrew [mailto:Andrew.Ingraham@compaq.com]
> > Sent: 19 September 2000 13:13
> > To: 'mark.browne@nokia.com'
> > Cc: 'ibis-users'
> > Subject: RE: Simulation resolution weirdness
> >
> >
> > There are limitations and imperfections in all circuit
> > simulators. (You
> > didn't mention which simulator program you are using, which
> might help
> > others to help you.) Some programs are better than others.
> > These problems
> > can be finicky and difficult to avoid.
> >
> > Yours sounds like a "numerical oscillation" problem, which
> > occasionally
> > happens in SPICE or SPICE-like simulators. In your case, it
> > apparently
> > happens to be made worse by certain time resolution settings.
> > There are
> > likely some controls (such as the Method=Gear vs. Trap
> > setting in SPICE)
> > that may help here.
> >
> > Regards,
> > Andy
> >
> >
> > > ----------
> > > From: mark.browne@nokia.com[SMTP:mark.browne@nokia.com]
> > > Sent: Tue, 19 September 2000, 05:25
> > > To: ibis-users@eda.org
> > > Subject: Simulation resolution weirdness
> > >
> > > Hi all,
> > >
> > > I am fairly new to using IBIS, so I am hoping that the
> > symptoms that I am
> > > getting will strike a chord with someone out there.
> > >
> > > I have run the same simulation several times, changing
> only the time
> > > resolution each time. When the resolution is set to 20ps,
> > the results
> > > look
> > > fine. However, at 15ps the resultant waveform is just an
> increasing
> > > oscillation, and at 10ps it look like a range of mountains.
> > >
> > > Obviously I am misusing the method somehow - all I need is
> > for someone to
> > > tell me how.
> > >
> > > Any ideas?
> > >
> > > Mark Browne
> > > Senior Engineer
> > > Nokia UK Ltd.
> > >
> >
>