[SI-LIST] : 回覆: [SI-LIST] : board-level simulation or differential signal

About this list Date view Thread view Subject view Author view

From: Daniel Wei ([email protected])
Date: Thu Apr 13 2000 - 19:24:16 PDT


Marc,

Many thanks for your reply.
If you use MDC file, you can use a separate pin model, also.
Could you just add a MDC file to run simulation again ?
And the same topology file, model file, coupling file and transmission line
parameter file are used.
Compare the two results.
Are they the same ?

Daniel Wei

-----原始郵件-----
寄件者: [email protected] [mailto:[email protected].
sun.com]代表 Marc Humphreys
寄件日期: 2000年4月13日 PM 11:32
收件者: [email protected]
主旨: RE: [SI-LIST] : board-level simulation for differential signals

Daniel,

I use (Innoveda/Viewlogic/QUAD)'s XTK plus SPICE to generate the models.
I use the Differential Driver/Receiver model approach because this lets me
use a separate pin model for both the _P and _N outputs of the driver which
was important
because measured data and simulating with HSPICE models show some amount of
asymmetry between
the two pins, and this can effectively be duplicated in the XTK simulation
using this approach.
Most of the differential signals I used this on are LVDS type I/O's. I get a
reasonable
correlation with SPICE and measured data for my purposes.

The main interest in running the diff pair simulations is as a post route
check to insure the routed
traces are equal in length. The board extractions capabilites of the XTK
tools
do a great job of insuring that duplicate etch segments don't get simulated
and essentially double
counted. Its a lot more difficult insuring trace lengths purely on the bases
of routed trace length
reports for the CAD system, unless you can convince yourself there's no
overlapping segments
included in the etch length reports. This is not typically a big issue on
slow single-ended signal
because these descrepencies are usually in the noise, but when trying to
minimize skew on signals
with 300ps differential edges its a great way to validate the routing on the
high speed differential nets
is balanced.

Marc

> -----Original Message-----
> From: Daniel Wei [SMTP:[email protected]]
> Sent: Thursday, April 13, 2000 7:45 AM
> To: '[email protected]'
> Subject: [SI-LIST] : board-level simulation for differential signals
>
> Hi,
>
> Does anyone have the experience in simulating differential signals by
> QUAD,
> HSPICE, or IS ?
> I appreciate any suggestion or experience.
>
> However, I had simulated differential signals by QUAD, but there are two
> ways to do it.
> One method is to simulate by defining Differential Driver and Receiver
> PINs,
> another method is by defining Multi-Driver Control file.
> And I found that the results of the above two methods are not the same
> (waveform and timing).
> Note that the driver model had been adjusted to make both pull-up and
> pull-down symmetric.
> What make them different ?
>
> If you had the HSPICE simulation experiences, which model did you use ?
> Or do you have the IS experience ?
>
> Thanks a million.
>
> Regards,
>
> Daniel Wei
>
>
>
> **** To unsubscribe from si-list or si-list-digest: send e-mail to
> [email protected]. In the BODY of message put: UNSUBSCRIBE
> si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
> si-list archives are accessible at http://www.qsl.net/wb6tpu
> ****

**** To unsubscribe from si-list or si-list-digest: send e-mail to
[email protected]. In the BODY of message put: UNSUBSCRIBE si-list
or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
****

**** To unsubscribe from si-list or si-list-digest: send e-mail to [email protected]. In the BODY of message put: UNSUBSCRIBE si-list or UNSUBSCRIBE si-list-digest, for more help, put HELP.
si-list archives are accessible at http://www.qsl.net/wb6tpu
****


About this list Date view Thread view Subject view Author view

This archive was generated by hypermail 2b29 : Thu Apr 20 2000 - 11:36:10 PDT