Re: [SI-LIST] : Automatic/Semi-automatic design check of PCB layout database for crosstalk/SI violat

Kim Helliwell ()
Thu, 02 Sep 1999 16:34:29 -0700

See remarks interspersed:

Pat Zabinski wrote:
>
> Ron,
>
> Cadence Allegro already has an integrated tool called SigNoise.
> I'm not sure what it costs, but it's an option you can purchase
> to Allegro.
>
> I dabbled with it several years ago, and it contained several
> basics that I think would full under the category you're
> interested in.
>
> However (unless they've updated it since I've used it last),
> it will not take into account things like SSN, ground bounce,
> nonlinearities, etc. In other words, it should work
> great for the first-order effects, but I would be hesitant
> in counting on it for second-order effects.
>

Cadence has updated it considerably since a few years ago, and
it will do SSN, ground bounce, crosstalk, etc. I have been
using it and like it. There are some holes, but they are
not show-stoppers, and it's one of the products at Cadence that
is apparently still under development. I talk to an AE at
Cadence frequently about my wish list for improvements, and many
of the items are under development or about to be released.
So they are trying to meet the needs.

> In terms of your requirements list below, I would say it fits
> well with 1, 2, and 4. However, 3 (cost) might be high (again,
> it's been a while). One other issue not in your list is setup
> time.
>

There is no question that the cost is high. However, if
you're already using Allegro, the incremental cost is lower
than some of the competitive products in this market.

I disagree somewhat as to ease of learning the tool. It's a
complex tool used to solve a complex problem, and there is
a pretty steep learning curve, in my opinion. But that's a
one-time investment, and it should continue to pay dividends
down the road. The same can be said for any of the competitors
in this class.

> In addition to defining the parts like you've always been
> doing, you would now need to include additional parameters in
> your parts definitions (risetime, voltage swing, etc). If
> you could automate this somehow with scripts, it might
> not be a big deal, but it could take considerable
> time if you needed to do this pin-by-pin.
>

Getting and maintaining models is the biggest headache for any
of these tools. Cadence is neither better or worse in this
regard. The trouble is that parts vendors are slow to generate
models, yet that's the logical entity to provide such models.
Some vendors know that IBIS (or at least SPICE) models are
necessary, but don't always provide them timely. Other vendors
appear to be clueless as to the need for models at all. It is
to be hoped that this set will approach the null set in the next
few years! Anyway, this issue is neutral with regard to selection
of tool vendor.

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

**** To unsubscribe from si-list: send e-mail to majordomo@silab.eng.sun.com. In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at http://www.qsl.net/wb6tpu/si-list ****