RE: [IBIS] RE: [IBIS-Users] hspice vs IBIS model correlation at different capacitive loads

From: Tom Dagostino <tom_at_.....>
Date: Tue Jan 03 2006 - 22:40:33 PST
I've been given many models to "fix" that had a C_fixture in them.
Universally, these models showed poor transient response into resistive
loads and had even worse problems trying to drive a reactive load.  I would
strongly recommend not to use a C_fixture in any model.

If your transmission line's length is on the order of Tr/Tf then you are
really into the range where you will see transmission line effects.  If this
line is not terminated then I would expect to see a lot of ringing in the
simulations.

You need to understand why you are seeing the ringing and proceed from
there.

Tom Dagostino
Teraspeed(R) Labs
13610 SW Harness Lane
Beaverton, OR 97008
503-430-1065
tom@teraspeed.com
www.teraspeed.com

Teraspeed Consulting Group LLC
121 North River Drive
Narragansett, RI 02882
401-284-1827

-----Original Message-----
From: owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org]On
Behalf Of Vinayak Ghatawade
Sent: Tuesday, January 03, 2006 10:07 PM
To: ibis-users@eda.org; ibis@eda.org
Subject: Re: [IBIS] RE: [IBIS-Users] hspice vs IBIS model correlation at
different capacitive loads


Hi all,

I had this query where in if the buffer is going to be
used for shorter TL (transmission line) then do we need
to generate the V-T data with C-load rather than R-load since the
buffer rise time is comparable to TL delay.

We have this case where in the *buffer has feedback* and with R-load we see
V-T table with lot of ringing(oscillations) when pad has reached steady
state.
As this is no real situation for us as TL delay that buffer would see in
real world is too small
comparable to buffer tr/tf.

How do we generate the V-T data in this case. Should we use C-load
instead of R-load
or we release the V-T data with some oscillations when pad reaches
steady state. Does
simulator really have a problem seeing such V-T data.

Thanks
vinayak

Ummalaneni, Venu Babu (Venu) wrote:

>Itzik Peleg and all,
>
>
>
>Thank you very much for your help. Please see my answers in line with your
comments.
>
>
>
>1. This buffer operation is intended for a load resistance of more than 5k
ohm.
>
>2. Yes, you are right. I should allow more transient simulation time for
min corner.
>
>However, reported behavior (correlation of hspice vs IBIS) is for typ
corner.
>
>
>
>Even I have tried with 50, 100, 500 & 5k ohm terminations for capturing V-t
waveforms. But I could not see any improvement in the correlation. I am
still investigating this issue.
>
>
>
>
>
>                 Thanks & Best Regards,
>
>Venu
>
> -----Original Message-----
> *From:* Itzik Peleg [mailto:itzikpe@marvell.com]
> *Sent:* Thursday, December 15, 2005 7:09 PM
> *To:* Ummalaneni, Venu Babu (Venu); ibis-users@eda.org; ibis@eda.org
> *Subject:* RE: [IBIS] RE: [IBIS-Users] hspice vs IBIS model
> correlation at different capacitive loads
>
> Hello Venu
>
> Here are my comments reviewing your model:
>
> 1. The loading condition is 1k ohm !? isn’t this buffer should be used
> in 50 ohm environment? If the answer is yes then you should use 50 ohm.
>
> 2. It seems like the slow didn’t finish the complete transition. You
> should allow more time for the V(t) simulation in order that the
> response get to steady state (like the maximum case in red). This also
> causing the warning that you have when using IBIS check:
>
> **************************************************************
>
> WARNING - Model ATA4N1K000: The [Rising Waveform]
>
> with [R_fixture]=1000 Ohms and [V_fixture_min]=0V
>
> has MIN column DC endpoints of 0.00V and 2.78v, but
>
> an equivalent load applied to the model's I-V tables yields
>
> different voltages (-0.00V and 2.92V),
>
> a difference of 0.00% and 4.69%, respectively.
>
> WARNING - Model ATA4N1K000: The [Falling Waveform]
>
> with [R_fixture]=1000 Ohms and [V_fixture_min]=3V
>
> has MIN column DC endpoints of 0.21V and 3.00v, but
>
> an equivalent load applied to the model's I-V tables yields
>
> different voltages ( 0.05V and 3.00V),
>
> a difference of 5.54% and 0.00%, respectively.
>
> ****************************************************************
>
> 3. Based on the 2 previous points I can think of a buffer for slow
> interface that should be used in 50 ohm environment. If this is the
> case you should allow the transient simulation more time on 50 ohm.
>
> Hope this will halp.
>
>--
>
>Regards
>
>
>
>Itzik Peleg
>
>Board Technology Group
>
>
>
>Marvell Semiconductor Israel Ltd
>
>6 Hamada Street
>
>
>
>
>Mordot HaCarmel Industrial Park
>
>
>
>Yokneam 20692, ISRAEL
>
>
>
>Email - itzikpe@marvell.com <mailto:itzikpe@marvell.com>
>
>Tel   - +972 4  9091192
>
>Cell  - +972 54 4452482
>
>Fax   - +972 4  9091501
>
>WWW Page: http://www.marvell.com <BLOCKED::http://www.marvell.com>
>
>========================================================================
>
>This message may contain confidential, proprietary or legally privileged
>
>information. The information is intended only for the use of the
>
>individual or entity named above. If the reader of this message is not
>
>the intended recipient, you are hereby notified that any dissemination,
>
>distribution or copying of this communication is strictly prohibited.
>
>If you have received this communication in error, please notify us
>
>immediately by telephone, or by e-mail and delete the message from
>
>your computer.
>
>
>
>Thank you!
>
>========================================================================
>
> ------------------------------------------------------------------------
>
> *From:* Ummalaneni, Venu Babu (Venu) [mailto:venubabu@agere.com]
> *Sent:* Thursday, December 15, 2005 1:07 PM
> *To:* Itzik Peleg; Todd Westerhoff (twesterh); ibis-users@eda.org;
> ibis@eda.org
> *Subject:* RE: [IBIS] RE: [IBIS-Users] hspice vs IBIS model
> correlation at different capacitive loads
>
>Itzik Peleg,
>
>
>
>Thanks for your concern. I think there is no such feedback.
>
>
>
>Here, I am attaching the IBIS model.
>
>
>
>Thanks & Regards,
>
>Venu
>
> -----Original Message-----
> *From:* Itzik Peleg [mailto:itzikpe@marvell.com]
> *Sent:* Wednesday, December 14, 2005 12:50 PM
> *To:* Ummalaneni, Venu Babu (Venu); Todd Westerhoff (twesterh);
> ibis-users@eda.org; ibis@eda.org
> *Subject:* RE: [IBIS] RE: [IBIS-Users] hspice vs IBIS model
> correlation at different capacitive loads
>
> Hello Venu
>
> Just to add one more possibility doe’s your buffer has a feedback?
>
> Could you attached the IBIS model, or show the I(V) and V(t) curves?
>
>--
>
>Regards
>
>
>
>Itzik Peleg
>
>Board Technology Group
>
>
>
>Marvell Semiconductor Israel Ltd
>
>6 Hamada Street
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>Mordot HaCarmel Industrial Park
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>Yokneam 20692, ISRAEL
>
>
>
>Email - itzikpe@marvell.com <mailto:itzikpe@marvell.com>
>
>Tel   - +972 4  9091192
>
>Cell  - +972 54 4452482
>
>Fax   - +972 4  9091501
>
>WWW Page: http://www.marvell.com <BLOCKED::http://www.marvell.com>
>
>========================================================================
>
>This message may contain confidential, proprietary or legally privileged
>
>information. The information is intended only for the use of the
>
>individual or entity named above. If the reader of this message is not
>
>the intended recipient, you are hereby notified that any dissemination,
>
>distribution or copying of this communication is strictly prohibited.
>
>If you have received this communication in error, please notify us
>
>immediately by telephone, or by e-mail and delete the message from
>
>your computer.
>
>
>
>Thank you!
>
>========================================================================
>
> ------------------------------------------------------------------------
>
> *From:* owner-ibis@eda.org [mailto:owner-ibis@eda.org] *On Behalf Of
> *Ummalaneni, Venu Babu (Venu)
> *Sent:* Tuesday, December 13, 2005 2:49 PM
> *To:* Todd Westerhoff (twesterh); ibis-users@eda.org; ibis@eda.org
> *Subject:* [IBIS] RE: [IBIS-Users] hspice vs IBIS model correlation at
> different capacitive loads
>
> Todd,
>
> Thanks for your inputs.
>
> I have verified the programming options of driver and it’s fine. Here,
> I am attaching the correlation waveforms of the simulations by using a
> 0.1 ohm resistor instead of transmission line in both load (5pF &
> 30pF) cases. Still, I am seeing a bad correlation in 30pF case.
>
> Please ignore my previously attached *.PDF.
>
> Once again thanks for your response.
>
> Best Reagrds,
>
> Venu
>
> -----Original Message-----
> *From:* Todd Westerhoff (twesterh) [mailto:twesterh@cisco.com]
> *Sent:* Tuesday, December 13, 2005 5:43 PM
> *To:* Ummalaneni, Venu Babu (Venu); ibis-users@eda.org; ibis@eda.org
> *Subject:* RE: [IBIS-Users] hspice vs IBIS model correlation at
> different capacitive loads
>
> Venu,
>
> How long is the transmission line you're using ... about 1ns?
>
> Your IBIS driver in the 30pf case seems to be taking several
> round-trips to charge the load where the HSpice driver doesn't - if
> this is a programmable driver, is it possible the driver settings for
> the two plots are different? In the top plot for the 30pf case, the
> IBIS driver looks like a much weaker output with a faster slew rate to
> me ...
>
> Todd.
>
>
> Todd Westerhoff
> High Speed Design Group Manager
> Cisco Systems
> 1414 Massachusetts Ave
> Boxboro, MA 01719
> email:twesterh@cisco.com
> ph: 978-936-2149
> ____________________________________________
> "Always do right.
> This will gratify some people and astonish the rest."
>
> - Mark Twain
>
> ------------------------------------------------------------------------
>
> *From:* owner-ibis-users@eda.org [mailto:owner-ibis-users@eda.org] *On
> Behalf Of *Ummalaneni, Venu Babu (Venu)
> *Sent:* Tuesday, December 13, 2005 5:38 AM
> *To:* ibis-users@eda.org; ibis@eda.org
> *Subject:* [IBIS-Users] hspice vs IBIS model correlation at different
> capacitive loads
>
> Hi Gurus,
>
> I have an ATA* IBIS model and it is correlating well with hspice model
> when I am using a 5pF load followed by a transmission line. If I
> replace the 5pF load with 30pF correlation is bad (particularly in
> 2^nd page waveform of attached PDF).
>
> Could some one comment on this, where it is going wrong?
>
> Your comments/inputs are highly appreciated.
>
> Attachments: ata4nik_5pf.pdf – hspice and IBIS model waveforms
> captured using 5pF load
>
> ata4nik_30pf.pdf – hspice and IBIS model waveforms captured using 30pF
> load
>
> Thanks & Best Regards,
>
> Venu
>
> <<ata4n1k_30pf.pdf>> <<ata4n1k_5pf.pdf>>
>

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


-----------------------------------------------------------------
|For help or to subscribe/unsubscribe, email majordomo@eda.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 email a request to ibis-request@eda.org.
|
|IBIS reflector archives exist under:
|
|  http://www.eda.org/pub/ibis/email_archive/  Recent
|  http://www.eda.org/pub/ibis/users_archive/  Recent
|  http://www.eda.org/pub/ibis/email/          E-mail since 1993
Received on Tue Jan 3 22:40:50 2006

This archive was generated by hypermail 2.1.8 : Tue Jan 03 2006 - 22:41:57 PST