Re: [SI-LIST] : Bad IBIS models!

About this list Date view Thread view Subject view Author view

From: Kim Helliwell (khelliwe@acuson.com)
Date: Thu Mar 16 2000 - 14:15:58 PST


Tadashi ARAI wrote:
>
  ... previous complaints about IBIS models deleted, then:

> I have many "abnormal" IBISs in my collection (in which there is few
> accurate model ;() and will show you an exmaple.
>
  ... stuff deleted, then:

> > [GND Clamp]
> > | Voltage I(typ) I(min) I(max)
> > |
> > -3.60000 -1.000e+25 -1.000e+25 -1.000e+25
> > -2.20000 -1.000e+25 -1.000e+25 -1.000e+25
> > -2.00000 -1.000e+25 -1.000e+25 -1.000e+25
> > -1.80000 -1.000e+25 -1.000e+25 -1.000e+25
> > -1.60000 -9.468e+22 -1.911e+22 -1.000e+25
> > -1.40000 -9.894e+19 -4.472e+19 -5.569e+20
> > -1.20000 -1.034e+17 -1.046e+17 -1.136e+17
> > -1.00000 -1.080e+14 -2.448e+14 -2.319e+13
> > -800.00000m -112.9147G -573.0318G -4.7331G
> > -600.00000m -117.9969M -1.3409G -965.8735k
> > -400.00000m -123.3078k -3.1377M -197.1048
> > -200.00000m -128.7234 -7.3251k -40.2150m
> > 0. -1.1095n -1.0994n -851.9528p
> > 200.00000m 134.5172m 17.1407 8.2060u
> > 400.00000m 134.6579m 17.1810 8.2077u
> > 600.00000m 134.6579m 17.1810 8.2078u
> > 800.00000m 134.6579m 17.1810 8.2078u
> > 1.00000 134.6579m 17.1810 8.1905u
>
> This is an audio device for PCI. WHAT A CRAZY CURRENT! -1e+25 Amps!?

This comes about because they used ideal diodes in their SPICE model
for the part. Of course, aside from the large currents that will
break your SI simulator, you also have to wonder just what other
garbage that model generated, since it's obviously not a very careful
SPICE model to start with. Use of ideal components in SPICE models
is well-known to cause problems, yet it's amazing how many vendors
persist in doing so. With unusable results, as you can see.

>
> I was crazed to complain to the vendor, but they wouldn't make corrected
> model.

The only response to this is: then I won't design your part in. I sure
hope you had options!

>
> So that I will add one more line, to the list.
>
> 19."Not guaranteed" model.
> Hey, if you couldn't guarantee your model, I cannot guarantee my
> simulation. Also I cannot guarantee to continue to use your devices.

I sympathize with this notion, but in today's litigous society, this is
probably unrealistic. I think the best we can hope for is a clear explanation
of how the model was generated and verified. There will always be disclaimers
as to the accuracy of the model or the fitness for a particular purpose. This
is to keep the vendor's lawyers happy. But there are other ways to tell whether
a vendor actually cares about the results, and you have to key off those.

****

-- 
Kim Helliwell
Senior CAE Engineer
Acuson Corporation
Phone: 650 694 5030  FAX: 650 943 7260

**** To unsubscribe from si-list or si-list-digest: send e-mail to majordomo@silab.eng.sun.com. 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 - 11:35:39 PDT