Re: Temperature variations

Jerry,

You are correct. An IBIS model is extracted at a process corner with
a fixed temperature. It is possible to perform multiple extractions at
various temperatures for each process corner and create multiple
IBIS models to show this behavior ... although this is not normally done.

What is normally done is that a behavioral model is extracted for CMOS
under the following conditions:

fast-n, fast-p, low temperature
nominal-n, nominal-p, room temperature
slow-n, slow-p, high temperature

If there is some reason to have extractions at more than one temperature
for each process corner, then separate IBIS models could be extracted.
These models would then be valid at those new temperatures. Normally,
if device behavior is linear with respect to temperature, then it should
only be necessary to perform the extraction at the temperature extremes,
as is usually done.

regards,

scott

--
Scott McMorrow
Principal Engineer
SiQual, Signal Quality Engineering
18735 SW Boones Ferry Road
Tualatin, OR  97062-3090
(503) 885-1231
http://www.siqual.com
Hayes wrote:
> To my knowledge an IBIS model is unable to account for temperature
> variations at a fixed process corner.  My customer claims otherwise.
> Without additional information within a simple CMOS model, I do not see
> how this possible. Could someone clarify my understanding on this issue?
>
> Thanks,
> Jerry
>
>