Subject: Re: ibischk3 V3.2.8 [Warning]
From: Takanobu Ura (t-ura@pb.jp.nec.com)
Date: Tue Feb 12 2002 - 19:13:50 PST
Dear all,
I appreciate for all advices.
But, How should I do after all? Should I describe the current of all
voltage to all model?
Or,Is the model out of which this warning does not come what does not
need to describe the current of all voltage?
How should any really right model be? Would you teach?
Regards,
T.Ura
Takanobu Ura wrote:
>
> Dear all:
>
> I have checked our model with ibisckk3.
>
> As for model of xxx,The following WARNING is reported.
>
> WARNING -
> Model xxx: POWER Clamp : Typical value never becomes zero
>
> [Power Clamp] <- model of xxx
> 0 4.957u 5.049u 4.825u
> -76m 5.07u 5.376u 4.918u
> -152m 5.348u 6.904u 5.053u
> -227m 7.611u 15.19u 6.135u
>
> But,As for model of yyy, The above WARNING is not reported.
>
> [Power Clamp] <- model of yyy
> 0 0.557u 0.949u 0.225u
> -76m 0.567u 0.976u 0.229u
> -152m 0.548u 0.984u 0.253u
> -227m 0.611u 1.190u 0.335u
>
> What is the reason of this difference?
> Therefore,I have any question.
> a/ How did you decide the limitation value as like 1uA?
> b/ I cannot understand what this WARNING warns to us.
> Can I ignore this WARNING?
> Because,I think that clamp current is not surely zero.
.....
> Dear T.Ura,
>
> The warning is meant to alert you that the data in your clamp table
> indicates
> that the clamp never turns off, regardless of voltage.
>
> I know that you're probably thinking that if the data in the table is
> extrapolated, then the current will stop flowing at some voltage outside of
> the table. That is a problem, because the simulator is forced to guess the
> behavior of the device outside of the boundaries of the table. The
> simulator might use linear extrapolation, cubic-spline extrapolation,
> quadratic, etc. In fact, the simulator may not extrapolate at all and may
> use the first and last currents in the table for all voltages beyond the
> table (since extrapolation can result in ridiculous currents like 1e+26A).
>
> The I-V tables are supposed to include data for all of the "interesting"
> voltages of the buffer. The voltage at which current stops flowing is
> certainly "interesting" and should be included in the table.
>
> Since there may exist some technology for which current never stops or to
> allow for measurement/calculation inaccuracies, IBISCHK3 will not report a
> warning if the current get very close to zero. Very close was arbitrarily
> chosen to be within 1uA. It seemed reasonable.
>
> Best regards,
> Matthew Flora
.
.
.
.
.
> Andy,
>
> You may be assuming that s2ibis2 users mostly know SPICE, and IBIS
> is a "second language" for them. But I am guessing that s2ibis is
> used more by people that know IBIS well, or may even know both
> quite well. Anyone who uses s2ibis2 and doesn't know IBIS fairly
> well is probably not turning out useful models. OK, maybe I just
> supported your premise :-)
>
> But I think the majority of s2ibis2 users are IBIS model end users,
> who have to fuss with it until they get a good IBIS model. That takes
> IBIS knowledge. Also, I find that to use s2ibis2 you have to know
> a 3rd language, "C".
>
> Mike LaBonte
>
> "Ingraham, Andrew" wrote:
> >
> > I'm not sure precisely what s2ibis* does (since I've not used it), but I
> > believe it does just the minimum required by the IBIS spec. Since most
> > s2ibis* users probably don't know much about IBIS, this is bound to
> > cause problems.
> >
> > Regards,
> > Andy
This archive was generated by hypermail 2b28 : Tue Feb 12 2002 - 19:38:43 PST