RE: [IBIS-Users] IBIS question on C_comp

From: <April.Hachenburg_at_.....>
Date: Mon Aug 27 2007 - 08:12:45 PDT
Michael, 
Your response is helpful.  I too am doing my first differential buffer so 
on a slightly different note ....

Does the s2ibis3 script handle generation of differential buffers?  If 
not, is there a similar script available?

- April

---------------------------------------------------
April Hachenburg
Core Technology Design Engineer
SMSC - Phoenix




"Mirmak, Michael" <michael.mirmak@intel.com> 
Sent by: owner-ibis-users@server.eda.org
08/24/2007 05:10 PM

To
"Chitiz Mathema" <cit@cypress.com>, <ibis-users@server.eda.org>
cc

Subject
RE: [IBIS-Users] IBIS question on C_comp






Chitiz,

Thanks for your questions on IBIS.  Here are some hopefully useful
answers...

1,2) The purpose of C_comp is to describe the capacitance of the buffer
at the pad, as it appears "looking in."  You are correct that the V-t
tables and ramp information already take C_comp into account, for
drivers.  However, C_comp is still needed in order to describe the
capacitive load the buffer presents to waveforms (reflections) coming
back into the driver.

In general, tools ensure that C_comp as a discrete number and its
effects in the table and ramp data are not double-counted.  This may be
implemented by simply "backing out" or derating the V-t tables for the
effects of C_comp, then adding an actual capacitor to the driver circuit
in simulation.  In that way, the derated-table-plus-capacitor waveforms
will match the original IBIS driver data, but the system will "see" the
capacitance of the buffer as a load.  Specific tools may have different
ways to approach the problem.

3) C_comp should exclude the effects of the package, but include all the
buffer's on-die parasitic capacitances -- pad, die routing, etc.

4) Information on extraction methods for C_comp, as well as differential
treatments, can be found in the IBIS Cookbook:

http://www.eda.org/ibis/cookbook/cookbook-v4.pdf

In addition, several presentations on C_comp issues are available in the
IBIS Summit archive:

http://www.eda.org/ibis/summits/index-bytitle.htm

To address your concern, I would not say that IBIS per se is not
properly addressing high-frequency effects.    However, C_comp in
traditional table-driven IBIS is a single value.  Capacitance for a
buffer is frequency- and voltage-dependent.  Therefore, the method of
extraction and the frequency at which the extracted value applies are
critical when correlating the resulting IBIS model.  In the case you
describe, I would check to make sure that the C_comp value was not
extracted at a frequency below 400 MHz.

A more comprehensive treatment of buffer capacitance is possible using
advanced IBIS techniques, including multi-lingual methods (IBIS 4.1,
4.2).

I hope this helps!

- Michael Mirmak
  Intel Corp.
  Chair, EIA IBIS Open Forum


-----Original Message-----
From: owner-ibis-users@server.eda.org
[mailto:owner-ibis-users@server.eda.org] On Behalf Of Chitiz Mathema
Sent: Friday, August 24, 2007 4:31 PM
To: ibis-users@server.eda.org
Subject: [IBIS-Users] IBIS question on C_comp

Hello,

I have some questions about C_Comp parameter in IBIS models. I am trying

to resolve some issues for IBIS models of high frequency differential 
output buffers.
I am trying to compare my SPICE model transient simulations with the 
IBIS model I generated for a differential output buffer. At low 
frequencies (eg ~ 100MHz), the SPICE and IBIS model output simulation 
waveforms correlate closely. But at higher frequencies (>400MHz), the 
SPICE and IBIS model simulations don't match as well. I am trying to see

if the mismatch is due to the C_comp parameter, or if this is just a 
limitation of IBIS models not being able to capture high frequency
effects.

1. How IBIS simulators use the C_comp parameter during simulations?
2. Don't the V-T tables and ramp information already take into account 
the effect of C_comp?
3. What does the C_comp parameter exactly represent?
-should it include die routes, pad capacitance, etc?
4. How should it be measured?
5. Should C_comp be measured in a special manner for differential 
buffers like LVDS buffers?

I would appreciate your response.

Thank you,
-Chitiz Mathema
Cypress Semiconductor




-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

--------------------------------------------------------------------
|For help or to subscribe/unsubscribe, e-mail majordomo@eda-stds.org
|with 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 e-mail a request to ibis-request@eda-stds.org.
|
|IBIS reflector archives exist under:
|
|  http://www.eda-stds.org/pub/ibis/email_archive/ Recent
|  http://www.eda-stds.org/pub/ibis/users_archive/ Recent
|  http://www.eda-stds.org/pub/ibis/email/         E-mail since 1993

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.


--------------------------------------------------------------------
|For help or to subscribe/unsubscribe, e-mail majordomo@eda-stds.org
|with 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 e-mail a request to ibis-request@eda-stds.org.
|
|IBIS reflector archives exist under:
|
|  http://www.eda-stds.org/pub/ibis/email_archive/ Recent
|  http://www.eda-stds.org/pub/ibis/users_archive/ Recent
|  http://www.eda-stds.org/pub/ibis/email/         E-mail since 1993


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.


--------------------------------------------------------------------
|For help or to subscribe/unsubscribe, e-mail majordomo@eda-stds.org
|with 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 e-mail a request to ibis-request@eda-stds.org.
|
|IBIS reflector archives exist under:
|
|  http://www.eda-stds.org/pub/ibis/email_archive/ Recent
|  http://www.eda-stds.org/pub/ibis/users_archive/ Recent
|  http://www.eda-stds.org/pub/ibis/email/         E-mail since 1993
Received on Mon Aug 27 08:13:38 2007

This archive was generated by hypermail 2.1.8 : Mon Aug 27 2007 - 08:14:12 PDT