The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Other CAD Tools >> Physical Verification, Extraction and Analysis >> Cadence bug with via/contact resistance extraction
https://designers-guide.org/forum/YaBB.pl?num=1112039441

Message started by stephen on Jul 20^th, 2004, 10:08pm

Title: Cadence bug with via/contact resistance extraction
Post by stephen on Jul 20^th, 2004, 10:08pm

I found sometimes the via/contact resistance is not extracted when I do parasitic RC extraction from the layout. It comes out for some kind of layout but fails to show up for others.

As far as I can guess, the contact/via resistance may be lost during the so-called network reduction. For example, if I connect one short piece of metal1 and one long piece of metal2 with a m1-m2 via, the via resistance won't be extracted if the piece of metal1 is too short. Only if both piece are long, it will be extracted.

I can give other examples if you need.
I wonder whether it's a known and resolved bug.

Thanks.

Title: Re: Cadence bug with via/contact resistance extrac
Post by Andrew Beckett on Jul 21^st, 2004, 12:38am

Which extraction tool are you talking about? Diva or Assura?

Also, this may be dependent on the extraction rules.

If Assura, do you have the "Add Explicit Vias" option turned on? If
not, the vias are probably being reduced into the resistance for
the tracks to which they are connected.

Regards,

Andrew.

Title: Re: Cadence bug with via/contact resistance extrac
Post by stephen on Jul 21^st, 2004, 7:23am

I use Diva. The typical via resistance in the technology I use is 4-8ohm. When the resistance is not extracted, I only see values like 0.7ohm in the extracted view. Those 0.7ohm are the extracted resistors for the wire segments.

Title: Re: Cadence bug with via/contact resistance extrac
Post by Andrew Beckett on Jul 27^th, 2004, 2:56pm

Stephen,

Diva normally extracts contacts rather excessively, (many resistors per contact) unless you do things in the rules file to minimise this (usually tricks like oversizing and undersizing the contacts to merge them all together). Perhaps this is happening, and then the resulting via array is ending up a smaller resistance than the minimum specified in the rule file, and hence they are being filtered?

It's bound to be dependent on the specifics of the rule file rather than a generic tool issue (at least that's what I expect).

Regards,

Andrew.