[SI-LIST] Re: IBIS model problems


John,
I'm assuming that you are running your IBIS files in a version of HSPICE.
The rwf_ramp and fwf_ramp keywords specify to use either Ramp or V-T data.
A 0 will use the
Ramp data, a 1 will use the first rising/falling waveform set, and a 2 will
use the first and second rising/falling waveform sets.  One issue with
using IBIS files in HSPICE is that
the rising and falling waveform data has to have an adequate "time window"
or else it will not work in HSPICE.  The best explanation I've received is
below:

 All tools should be able to handle "overclocked" situations
where the time window in the model might be longer than the pulse width
used in simulation.  For example, your model would work with a 3 ns
period - or 333 MHz, whether or not you truncate constant trailing
waveform voltage information.  However, some EDA tool algorithms may
still have problems if the Waveform time window is longer than the
pulse width.  So truncating the data is just a suggestion to make the
model more portable.

I have not run into this problem using Mentor Graphics (formerly Innoveda)
tool XTK.  This is because the algorithm in XTK actually truncates the V-T
data.

As an IBIS model developer, I agree with you that more documentation is
needed with models.  One challenge is that every IBIS simulator uses it's
own algorithm.  The time window
issue is an example of how one IBIS file might work in one simulator, but
not in another one.

Regards,
Tim




|---------+---------------------------->
|         |           jellis@xxxxxxxxx |
|         |                            |
|         |           09/19/2002 04:40 |
|         |           PM               |
|         |                            |
|---------+---------------------------->
  
>------------------------------------------------------------------------------------------------------------------------------|
  |                                                                             
                                                 |
  |       To:       si-list@xxxxxxxxxxxxx@Internet                              
                                                 |
  |       cc:       (bcc: Timothy Coyle/Americas/NSC)                           
                                                 |
  |       Subject:  [SI-LIST] IBIS model problems                               
                                                 |
  
>------------------------------------------------------------------------------------------------------------------------------|





I too have seen the problems generated by many IBIS models.  Recently I've
found that about half the IBIS models I try to run cannot support the
application when rwf_ramp and fwf_ramp are set to 1 or 2.

In these cases I have to fall back on the 0 setting.  I think I was told
that it has to do with the frequency at which the IBIS waveforms are
derived.  If someone has a more complete explanation, please share it.

While I'm on the subject,  In IBIS there are multiple optional settings to
choose from which impact the nature of the waveforms.  Typically when I use
an IBIS model, its one I've downloaded from a component vendor's website
because an HSPICE model wasn't available.  There is usually ZERO
documentation for this model regarding what settings (rwf_ramp etc.) to use.
Does anyone who routinely builds IBIS models have any suggestions on how to
proceed in implmenting an undocumented model within a netlist so as not to
go too far afield?

Thanks,

John Ellis

TriCN, Inc.


------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List archives are viewable at:
            http://www.freelists.org/archives/si-list
or at our remote archives:
            http://groups.yahoo.com/group/si-list/messages
Old (prior to June 6, 2001) list archives are viewable at:
            http://www.qsl.net/wb6tpu








------------------------------------------------------------------
To unsubscribe from si-list:
si-list-request@xxxxxxxxxxxxx with 'unsubscribe' in the Subject field

or to administer your membership from a web page, go to:
http://www.freelists.org/webpage/si-list

For help:
si-list-request@xxxxxxxxxxxxx with 'help' in the Subject field

List archives are viewable at:     
                http://www.freelists.org/archives/si-list
or at our remote archives:
                http://groups.yahoo.com/group/si-list/messages 
Old (prior to June 6, 2001) list archives are viewable at:
                http://www.qsl.net/wb6tpu

Other related posts: