Subject: RE: [IBIS] IBIS models for Intel LXT971A and LXT9785 Ethernet PHYs
From: Muranyi, Arpad (arpad.muranyi@intel.com)
Date: Tue Feb 11 2003 - 09:57:39 PST
You are correct. And anticipating your next question,
those can be reverse engineered from the data in the
Vt curves, with more or less accuracy.
Arpad
======================================================
-----Original Message-----
From: Brad Crowell [mailto:brad.crowell@amirix.com]
Sent: Tuesday, February 11, 2003 9:56 AM
To: Muranyi, Arpad; ibis@eda.org
Subject: RE: [IBIS] IBIS models for Intel LXT971A and LXT9785 Ethernet
PHYs
Arpad, Tom,
Thanks for your comments. If I understand you correctly, the IV curves in an IBIS file correspond to a particular "bias" of the IO transistors, which is fully on or fully off. Then to model a "real" analog IO using traditional IBIS with the assumptions you refered to would require a multitude of IV curves, a set for each bias point being modelled. Is that correct?
Brad
> -----Original Message-----
> From: owner-ibis@eda.org [mailto:owner-ibis@eda.org]On Behalf Of
> Muranyi, Arpad
> Sent: February 11, 2003 1:31 PM
> To: ibis@eda.org
> Subject: RE: [IBIS] IBIS models for Intel LXT971A and LXT9785 Ethernet
> PHYs
>
>
> Brad,
>
> The difference between IBIS analog, and analog-analog is that
> in IBIS the buffer is either in the high state, low state or
> high impedance state. This means that the pullup, or pulldown
> are fully on or off (except during transitions). In real
> analog, you have a transistor that is biased to a certain
> operating point, and then the signal sits on top of that bias
> level. The signal can't be larger than the bias, otherwise
> it will get clipped. Think of digital buffers as if they
> were analog amplifiers which get turned on/off by the signal,
> or which have signals that is the bias itself.
>
> The reason IBIS can't do it (yet) is because we made several
> assumptions for how the buffer is constructed so that we can
> have simple IV and Vt curves in the IBIS file, and this
> limitation was part of one of the assumptions. In the days
> when we started IBIS there was no real need for covering real
> analog signals with IBIS because there were practically no
> boards that had true analog signals on them. This is changing
> and that's where the recently approved BIRD75 will help.
>
> And you are right, IBIS simulations are actually analog, not
> digital, but as I stated above, what makes these buffers "digital"
> is how they are operated. You could probably make an open loop
> amplifier out of a 74244 buffer if you put a small sine wave
> on its input with a DC bias of Vcc/2. In theory, (if you are
> lucky), the output should give you an amplified sine wave around
> Vcc/2. I never tried this, but I can imagine that it could work.
> Can anyone on the list share some "experience" in this area?
>
> Have fun...
>
> Arpad
> ==============================================================
>
> -----Original Message-----
> From: Tom Dagostino [mailto:tom@teraspeed.com]
> Sent: Tuesday, February 11, 2003 8:46 AM
> To: 'Brad Crowell'; 'Hassan Ali'; ibis@eda.org
> Subject: RE: [IBIS] IBIS models for Intel LXT971A and LXT9785 Ethernet
> PHYs
>
>
> Brad
>
> In this case the driver is more analog than most. I believe it
> has 3 levels.
>
> IBIS 4.0 and prior was set up for full swing signals and cannot
> model intermediate levels. With the inclusion of BIRD 75 you
> should be able to model and simulate these signals.
>
>
> Tom Dagostino
> Teraspeed Consulting Group LLC Teraspeed Consulting Group LLC
> 2926 SE Yamhill St. Device Modeling Division
> Portland, OR 97214 13610 SW Harness Lane
> Beaverton, OR 97008
> http://www.teraspeed.com 503-430-1065
> tom@teraspeed.com
>
>
> -----Original Message-----
> From: Brad Crowell [mailto:brad.crowell@amirix.com]
> Sent: Tuesday, February 11, 2003 5:14 AM
> To: tom@teraspeed.com; 'Hassan Ali'; ibis@eda.org
> Subject: RE: [IBIS] IBIS models for Intel LXT971A and LXT9785 Ethernet
> PHYs
>
>
> Your reply brought a couple of questions to mind. First, with a
> little tongue in cheek, aren't all IO analog in nature, which is
> the point of IBIS? Second, more seriously, why can't IBIS be used
> to model these types of IO? I have come across this probelm
> several times myself. With only an IBIS simulator at my disposal
> for these types of problems, I don't have HSpice and if I did
> it's doubtful that I could get the needed models, this leaves me
> no means of simulating these cases.
>
> Brad
> ****************************
> Brad Crowell, P. Eng.
> Senior Hardware Designer
> AMIRIX Systems
> Phone: 902-450-1700 Ext 287
> Fax: 902-450-1704
> www.amirix.com
> ****************************
>
> -----Original Message-----
> From: owner-ibis@eda.org [mailto:owner-ibis@eda.org]
> Sent: February 10, 2003 5:29 PM
> To: 'Hassan Ali'; ibis@eda.org
> Subject: RE: [IBIS] IBIS models for Intel LXT971A and LXT9785
> Ethernet PHYs
>
>
> I believe you will find that these pins are really analog in
> nature not digital. Thus no IBIS models for them.
>
> Tom Dagostino
> Teraspeed Consulting Group LLC Teraspeed Consulting Group LLC
> 2926 SE Yamhill St. Device Modeling Division
> Portland, OR 97214 13610 SW Harness Lane
> Beaverton, OR 97008
> http://www.teraspeed.com 503-430-1065
> tom@teraspeed.com
> -----Original Message-----
> From: owner-ibis@server.eda.org
[mailto:owner-ibis@server.eda.org]On Behalf Of Hassan Ali
Sent: Monday, February 10, 2003 12:50 PM
To: ibis@server.eda.org
Subject: [IBIS] IBIS models for Intel LXT971A and LXT9785 Ethernet PHYs
I've obtained IBIS models "lxt971a_q_mii_3_31.ibs" and "lxt9785_q_rmii_3_3.ibs" of the above-mentioned devices through the following link:
http://www.intel.com/design/network/products/ethernet/linecard_ept.htm
Unfortunately the IBIS model for LXT9785 device does not show which I/O models should be used for the 100Mbps pins: TPFOP, TPFON, TPFIP, TPFIN. The IBIS model for LXT971A does not show the I/O models for the TPFOP, TPFON pins.
I need the models for those pins to be able to perform signal integrity analysis of a magneticless link through a backplane.
Does anyone have info on the missing models? Are those the right IBIS model files for my analysis?
Thanks.
Hassan.
-----------------------------------------------------------------
|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
-----------------------------------------------------------------
|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
This archive was generated by hypermail 2b28 : Tue Feb 11 2003 - 10:10:15 PST