Dear Todd,
The netlist
I used (..sp file) is a
transistor-level netlist which was extracted from the layout.
Therefore, the
netlist I used to create the IBIS buffer should already have parasitic
effects
catered inside.
My question
was more for the .s2i file. Inside
this file, there’s a parameter [C_comp] that we have to include. If my
netlist already has the parasitic effects catered, do I still need to
input the
C_comp value explicitly in the .s2i file?
Thank you.
Best regards,
Mei
Yee
MCPD,
Avago Technologies (Malaysia)
Tel no.:
+604-6108515
Mei-Yee,
Can you expand on the phrase "the netlist I used for the pad is an
extracted netlist"?
Do you mean the transistor-level netlist used to create the IBIS buffer
model
in the first place ... that is, a transistor level netlist used to
perform
simulation under controlled conditions, the results of which were
post-processed to create the IBIS buffer model?
If that's what you meant, and the
transistor level netlist you started with
included the output buffer's capacitance, then the conversion program
(s2ibis3
or other) you used should have calculated and included C_Comp
automatically in
your buffer model. Arpad is correct that the V-T curves in the IBIS
model should include the effects of C_Comp, and the IBIS simulator
needs to
combine the V-T data and C_Comp during simulation without
double-counting the
effect of C_Comp.
... but I'm trying to understand whether your question was whether or
not
C_Comp was to be added explicitly to the original transistor level
netlist. The usual answer to that question is - no.
Todd.
Todd Westerhoff
VP, Software Products
SiSoft
6 Clock Tower Place, Suite 250
Maynard, MA 01754
(978) 461-0449 x24
twesterh@sisoft.com
www.sisoft.com
NG,MEI-YEE wrote:
Hi
Arpad,
Just one
quick question, if the netlist I
used for the pad is an extracted netlist, does that mean the C_comp
value is already
incorporated inside (parasitics and all)? In that case, if I still put
in an
additional C_comp value, it would mean that I’m double-counting the
C_comp?
Thank you.
Best regards,
Mei
Yee
MCPD,
Avago Technologies (Malaysia)
Tel no.:
+604-6108515
Timmy,
You can do it
either way, it
depends on your needs.
If you want to
correlate to
your SPICE model, match
it to that, if
you want to
correlate to real life,
match it to your
measurements.
Arpad
====================================================
From: owner-ibis@server.eda.org
[mailto:owner-ibis@server.eda.org]
On Behalf Of Timmy
Leon
Sent: Thursday,
September 25, 2008
8:46 PM
To: ibis@server.eda.org
Cc: ibis-users@server.eda.org
Subject: [IBIS] C_comp
value
I have a question regarding
the c_comp value that we should put in IBIS model. Should the c_comp
value be based on measurement, or can we tune value so that IBIS model
correlate with its spice model or measurements? Thank you.
|
--
This message has been scanned for viruses and
dangerous content by MailScanner,
and is
believed to be clean.
--
This message has been scanned for viruses and
dangerous content by MailScanner,
and is
believed to be clean.
--
This message has been scanned for viruses and
dangerous content by MailScanner,
and is
believed to be clean.