Modeling method and Intel non-monotonic model

From: Kellee Crisafulli <71436.1314@CompuServe.COM>
Date: Tue Apr 05 1994 - 20:52:30 PDT

From: Kellee Crisafulli, HyperLynx
Re: Which modeling method does HyperLynx use, and Does the Intel
non-monotonic model work.

  How do you use the POWER/GND clamp and pullup/down curves in your simulations?

   a) summing method, or
   b) exclusive method

1) HyperLynx uses the summing method (per your definition).

  a) By the summing method I mean that the VI curve in the clamp section is added
     to the VI curve in the pullup/down section for the ON condition. For the OFF
     (3-stated) condition, only the clamp curves are used.

   b) By the exclusive method I mean that the clamp curves are strictly used for
      the 3-state condition, and the pullup/down curves are strictly used for the ON
      condition without adding in the currents from the clamp sections.

2) I have tested the Intel model using a falling edge driving a long transmission
   line resulting in a very large reflection below ground (> -1V). I examined
   numerous examples using the most current version of the LineSim Pro simulator
   version 3.2. I found that the model at appeared to work fine.

I am still against allowing non-monotonic data in IBIS models, however our simulator was
modified several months ago to allow non-monotonic data.

In conclusion we see no problem with the Intel model that would prevent it from being
released. On that other hand we also support requiring that all V/I tables be monotonic.

Have a great week... Kellee
Received on Tue Apr 5 20:59:51 1994

This archive was generated by hypermail 2.1.8 : Fri Jun 03 2011 - 09:52:28 PDT