Subject: Re: [IBIS-Users] On IBIS Model Generation
From: Ege Engin (engin@izm.fhg.de)
Date: Wed Sep 18 2002 - 01:38:05 PDT
Andrew,
I-V and V-T curves are both obtained when the (CMOS) driver is connected
to an ideal supply voltage. The resultant driver model is a 2-pole
model, such that the signal return current is assumed to flow back on
the AC short-circuited power and/or ground. Do you think, it is possible
to use/extend IBIS (similar to the 4-terminal device model in the
japanese IMIC) to include the power supply noise? You mentioned that the
V-T curves at various loads is required to enhance accuracy. Maybe, a
possible way to include the power-supply noise is to have all these V-T
(and V-I) curves at various power-supply levels as well. I think the
simulator needs to be "smarter" in this case to figure out what to do.
Regards,
Ege
"Ingraham, Andrew" schrieb:
>
> Jairam,
>
> The I-V curves are essentially static or DC, so even if they did include the package, only the package resistance would affect them. L and C would not.
>
> They are also load-independent. Their purpose is to "map out" or describe the characteristics of the IO buffer alone. You may think of it as connecting a constant voltage source (or perhaps a current source) to the IO pin. At that voltage, you would measure some value of current, which depends on the strength of the IO buffer alone. Repeat at other voltage values to create a curve.
>
> I think it is extremely rare for a buffer's drive strength to depend on the load. (Actually, there are some technologies that are capable of adjusting themselves against their loads; but once so adjusted, the drive strength remains constant until the next re-cal.) So except for rare cases, while the output current does depend on the load as you say, the drive strength, i.e., the shape and magnitude of the I-V curve, does not.
>
> On the other hand, the V-T curves do greatly depend on load; and for that reason, it is highly desirable for the model creator to have at least one V-T curve whose test load fixture is similar to the actual load that the customer will use. The simulator tool will need to extrapolate or interpolate around V-T curves in order to derive the waveform when a somewhat different load is used. Accuracy is theoretically enhanced when the actual load is very similar to the test load. An exact termination is not required, but the simulator needs to be "smart" enough to figure out what to do.
>
> Regards,
> Andy
>
> |------------------------------------------------------------------
> |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 Sep 18 2002 - 01:45:33 PDT