RE: Can I-V data be unsorted?

Andrew,

Did you experiment with any real .ibs files by re-arranging the V-I table
rows? I found that ibischk3 flags non-monotonic data when the lines are
jumbled and it also had the repercussion of causing a bogus warning about
the buffer being unable to drive through the Vmeas level.

I think this alone is a good enough indication that the data should be
ordered (sorted). However, I don't believe there is any requirement to order
in increasing or decreasing order. It seems most models are ordered from -ve
to +ve (e.g. -3.3V to +6.6V) but I have seen a few that go the other way and
they don't give any problems (except to the human eye parser that kind-of
expects them to go the other way).

On the lighter side, the state of many models out there are bad enough
without giving their creators an idea that they could jumble the data up as
they please!!

Was your question based on generalities or clarification of the spec., or do
you have a specific .ibs file that's an issue to you?

Regards,
Tony Dunbar
Mentor Graphics

-----Original Message-----
From: Ingraham, Andrew [mailto:Andrew.Ingraham@compaq.com]
Sent: Friday, June 09, 2000 1:21 PM
To: 'ibis-users'
Subject: Can I-V data be unsorted?

Is there any requirement to list V/I data in order of increasing table
voltage values?

From what I can tell, there is no need to even sort the data, let alone have
it in ascending order. The only mention I saw in the spec that seems
related, is that "numeric values (in the typical current column) MUST be
specified for the first and last voltage points on any I-V table." This
probably should have been worded, "lowest and highest voltage points" rather
than "first and last." Or does it really mean that the first and last lines
in each table always represent the voltage endpoints?

Regards,
Andy
Received on Mon Jun 12 16:42:41 2000