Subject: Re: [IBIS-Users] FW: Extreme currents????
From: AC (comeral@ix.netcom.com)
Date: Wed Jul 10 2002 - 18:28:51 PDT
Hello:
Although Spice may predict extremely high currents, I doubt whether they
are predicted accurately. I've seen IBIS models with 1.0E25 diode
currents,
from big manufacturers; a quick sanity check indicates that such a diode
would dissipate more power than the sun puts out! And it probably
doesn't
take much current to blow up a small diode, so blowing one up in the lab
doesn't really prove that the high currents were accurate -
I think a little more "physicality" in these models would be desireable
-
AC
Scott McMorrow wrote:
>
> Kim and all,
>
> In many cases these are not ideal diodes. The IBIS spec suggests that
> the input voltage for model creation be swept from -Vcc to 2Vcc. It
> turns out that for many real devices of small geometry that the Spice
> simulation accurately predicts extremely high currents. The real device
> would have blown up in a lab, which I know that Tom Dagastino of Mentor
> can attest to in creating models from measurement.
>
> Spice to ibis does not allow the user to change the input voltage sweep
> range. These excessive voltages, especially on devices with low core
> voltages, will cause all sorts of problems in I/V curve extraction. The
> least of which is excessive currents. I have had a wonderful time with
> convergence issues under curve extraction, too. The solution I found
> was to write a bit of script code to place a limiting amplifier in the
> path of the voltage sweep for spice to ibis. (This was much easier than
> going into the spice to ibis code, finding the many offending sections
> and assumptions, and fixing it.)
>
> Generally one can remove the offending high current table entries from
> the IBIS file and perform the limiting yourself with no ill effect in
> the actual active region of the device.
>
> Besides, in IBIS simulators that use a universal "perfect ground" (which
> doesn't exist in reality) reference, the reference for the clamp diode
> (and output I/V curves) on the silicon is all wrong anyway. Packages
> and boards are much more perverse than we like to imagine. And with
> high enough clamping currents, the simultaneous switching input (SSI)
> power/ground rail bounce can be quite significant.
>
> My best advise to all, after a couple years of experience, is to never
> count on device clamp diodes, unless you are willing to perform some
> substantial Spice simulation runs, including power and ground rail
> package modeling ... or are prepared for some interesting surprises in
> the actual system.
>
> Best regards,
>
> Scott
>
> --
> Scott McMorrow
> Teraspeed Consulting Group LLC
> 2926 SE Yamhill St.
> Portland, OR 97214
> (503) 239-5536
> http://www.teraspeed.com
>
> Kim Helliwell wrote:
>
> > Frankly, I wish that suppliers of models (whether IBIS
> > or SPICE) would stop using these "ideal" diodes that
> > allow such large currents. I don't understand how anyone
> > can look at such results and think that's OK to ship to
> > customers!
> >
> > Maybe they just don't look, but that's hardly an acceptable
> > excuse, either.
> >
> > There, I feel better now!
> >
> > Kim
> >
> >
> > On Wednesday, July 10, 2002, at 03:19 PM, Lynne Green wrote:
> >
> >> Hello, Chris,
> >>
> >> Different EDA tools have different upper limits on currents that can
> >> be handled
> >> by the simulator.
> >>
> >> The warning is needed because "Spice" models that use the default series
> >> resistance (0.0 ohms) for diodes and junction leakage currents result
> >> in large
> >> IBIS table currents (often above 1 MegAmp). Most EDA tools don't go
> >> quite
> >> _that_ high.
> >>
> >> Best regards,
> >> Lynne
> >>
> >>
> >>
> >> -----Original Message-----
> >> From: John Angulo [mailto:jangulo@innoveda.com]
> >> Sent: Tuesday, July 09, 2002 6:02 PM
> >> To: ibis-users@eda.org
> >> Subject: [IBIS-Users] FW: Extreme currents????
> >>
> >>
> >> The following is forwarded on behalf of Chris DeWitt at Motorola TCG.
> >> Please send all private replies to Chris.Dewitt@motorola.com.
> >>
> >> John Angulo
> >> IBIS Open Forum Postmaster
> >> -----------------------------------
> >>
> >> If I run IBISCHK3 on many models I download I get an "Extreme currents
> >> present" warning.
> >>
> >> What are the limits on currents in the VI tables? I can't find it
> >> documented anywhere.
> >>
> >> Thanks
> >>
> >> Chris DeWitt
> >> Principal Staff Engineer
> >> Motorola TCG
> >> |------------------------------------------------------------------
> >>
> >> |------------------------------------------------------------------
> >> |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
> >>
> >>
> > Kim Helliwell
> > Apple Computer
> > kimgh@apple.com
> > 408 974 9936
> >
> > |------------------------------------------------------------------
> > |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
> >
> >
>
> |------------------------------------------------------------------
> |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
|------------------------------------------------------------------
|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 : Wed Jul 10 2002 - 18:39:01 PDT