Subject: Re: [IBIS-Users] Waveform mismatch question
From: Peter LaFlamme (plaflamm@amcc.com)
Date: Tue Jan 28 2003 - 07:20:46 PST
Hi John,
These types of warnings are very common. It appears to me that you are either using an old IBISCHK or you are creating a Version 2.1 model because the latest IBISCHK will give errors for any mismatch greater than 1% or so. If your I/O is single-ended then
you can determine where the problem is pretty easily. You need to determine what your output load is. In this case 50 ohms. Then plot your I-V curves and V-T curves. On the I-V curves you need to draw a 50-ohm load line. The load line intersection with the
I-V curve should match the end point of the V-T curve (for rising waveform use the pull-up curve, Falling waveform use pull-down).
Is this a differential output? If so what technology is it? Creating an IBIS model of a differential output can be very tricky. IBIS actually models the differential output as a single-ended entity which it is definitely not. My experience with LVDS, PECL
and CML models has been that P and N output voltage levels are dependent on each other. For example on one of our LVDS I/O cells the sum of the P and N voltages will be 2.4V. So if the P side is 1.0V the N side will try to be 1.4V. To create a model as
such you need to set up your simulation such that the N side load is a voltage source that is dependent on the P output.
I have read the papers on the IBIS website and they are somewhat helpful, however the cases presented are applicable to certain cases only. There needs to be an IBIS cookbook that explains both the theory behind creating the models and the how to.
Well enough about Differential.
I hope that this helps.
Peter
Duane Mattheisen wrote:
>
> John
> I was writing for a bench measurement not a simulation. The 50 ohms
> is a common standard for the AC measurements. And if your on a AC bench
> measuring of actual devices, a capacitor helps when put it on the power
> supply side of the R_fixture not on the device side, what that does is
> take any voltage drop out of the wires going from the power supply to
> the R_fixture. For the DC measurement again on the bench the resistance
> to make as low as possible is the fixture resistance which is between
> the Voltage/Current measurement tool and the device.
>
> On another note I down loaded s2ibis2_fix_tar.Z and s2ibis2_tar.Z but
> did not know how to work with them. Is there any advice you can give me
> to start?
>
> Thanks Duane Matheisen
> John Phillippe wrote:
> >
> > Duane Mattheisen wrote:
> > > John Take out as much fixture resistance as possible in the DC
> > > measurement.
> > > Also for the AC measurement put a big capacitor (0.1uf and a >2.2uF
> > > tantalum cap) to ground on the Rfixture having the other side of the
> > > Rfixture that connects to the output as close as possible to the
> > > output.
> > > Sincerely Duane Mattheisen
> > > "Angulo, John" wrote:
> > >
> > >>The following is posted on behalf of John Phillippe at Motorola. Please send any private replies to john.phillippe@motorola.com only.
> > >>
> > >>--------------------------------------------------
> > >>
> > >>Ibis Guru's,
> > >> I get quite a few waveform mismatches between the rising/falling
> > >>edges and the I/V curves like the one shown below.
> > >>
> > >>WARNING - Model pad_26v_fast: The [Falling Waveform]
> > >> with [R_fixture]=50 Ohms and [V_fixture_max]=0V
> > >> has MAX column DC endpoints of 0.05V and 1.75v, but
> > >> an equivalent load applied to the model's I-V tables yields
> > >> different voltages ( 0.16V and 1.91V),
> > >> a difference of 6.66% and 8.94%, respectively.
> > >>
> > >>I'm curious if this is a common problem, or if there are some simple
> > >>solutions to get everything to match up. I check the waveforms and set
> > >>the sim time long enough to make sure the waveforms are stable. Any
> > >>help/suggestions would be greatly appreciated.
> >
> > Duane,
> > Since I'm using s2ibis2, I should make rload as small as possible?
> > On the order of 1ohm? I read in this group before to use an rload of
> > 50ohms since that is what most people are using, and that it is a bit of
> > a "standard." I've also been told to have NO capacitance in the AC
> > measurements, and again to only use a 50ohm resistor.
> > I did try a run with a 0.1uF cap on the pad for the AC measurements,
> > but that just slows the edge down.
> >
> > --
> > John Phillippe
> > SPS, 32 Bit Embedded Controller Division, IC Creation
> > Motorola - 512-895-1835
> > Austin, TX - john.phillippe@motorola.com
> |------------------------------------------------------------------
> |For help or to subscribe/unsubscribe, email majordomo@eda.org
> |with just the appropriate command message(s) in the body:
> |
> | help
> | subscribe ibis <optional e-mail address, if different>
> | subscribe ibis-users <optional e-mail address, if different>
> | unsubscribe ibis <optional e-mail address, if different>
> | unsubscribe ibis-users <optional e-mail address, if different>
> |
> |or email a written request to ibis-request@eda.org.
> |
> |IBIS reflector archives exist under:
> |
> | http://www.eda.org/pub/ibis/email_archive/ Recent
> | http://www.eda.org/pub/ibis/users_archive/ Recent
> | http://www.eda.org/pub/ibis/email/ E-mail since 1993
-- Peter LaFlammeApplied Micro Circuits Corp. System Applications Engineer 200 Minuteman Rd, 3rd Floor Andover, MA 01810
978-247-8470 phone 978-623-0055 Fax |------------------------------------------------------------------ |For help or to subscribe/unsubscribe, email majordomo@eda.org |with just the appropriate command message(s) in the body: | | help | subscribe ibis <optional e-mail address, if different> | subscribe ibis-users <optional e-mail address, if different> | unsubscribe ibis <optional e-mail address, if different> | unsubscribe ibis-users <optional e-mail address, if different> | |or email a written request to ibis-request@eda.org. | |IBIS reflector archives exist under: | | http://www.eda.org/pub/ibis/email_archive/ Recent | http://www.eda.org/pub/ibis/users_archive/ Recent | http://www.eda.org/pub/ibis/email/ E-mail since 1993
This archive was generated by hypermail 2b28 : Tue Jan 28 2003 - 07:51:52 PST