RE: PCI spec question

Bob,

I think this is the problem that BIRD 66 starts to cover by putting Vref in
the [model spec]. My feeling is that IBIS needs to be a bit more flexible
and the IC vendors or commitees that specifiy test loads need to try really
hard to use loads that IBIS can describe. I think that we should extend
BIRD 66 to include Rref and Cref as well.

Regards,
Weston

 + -----Original Message-----
 + From: Haller, Robert [mailto:rhaller@cereva.com]
 + Sent: Friday, September 15, 2000 8:44 AM
 + To: 'ibis-users@eda.org'
 + Subject: PCI spec question
 +
 +
 +
 + First I would like to thank Bob Ross and Ed Sayre for a
 + great job on the
 + IBIS summit meeting !
 +
 + My question is how the PCI is specified and how IBIS is specified.
 + The 3 volts PCI spec has you drive 25 ohms// 10 Pf to gnd
 + (rising ) and
 + measure at .94 volts
 + Additionally it has 25 Ohms//10pf to 3.3 volts and measure
 + at 2.0 volts.
 +
 + The IBIS specification really doesn't appear to have a
 + mechanism to deal
 + with this (PLease tell me I am wrong).
 +
 + Do I have to do one set of simulations for rising, change
 + the model then do
 + another set for falling.
 + This seem tedious. PLease tell me there is another solution ??
 +
 + Thanks in advance,
 + bob
 +
 + Cereva Networks
 + 3 network drive
 + Marlboro MA. 01752
 + Phone: 508-486-9660 X 365
 + FAX: 508-486-9661
 + Email: rhaller@cereva.com
 +
 +

 
Received on Fri Sep 15 10:50:50 2000