The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Analog Verification >> Analog Performance Verification >> Verification on PVT corners and Monte-Carlo
https://designers-guide.org/forum/YaBB.pl?num=1334929887

Message started by weber8722 on Apr 20^th, 2012, 6:51am

Title: Verification on PVT corners and Monte-Carlo
Post by weber8722 on Apr 20^th, 2012, 6:51am

Hi,

to check a circuit I usually test it in sweeps upfront, then for PVT corners, and for mismatch I run a MC analysis. Usually mismatch data is only extracted at typical process corner, so I end up with e.g. at least 4*3*3=36 corner runs and 50 mismatch MC runs to get a pretty complete overview & verification. Of course for more deeper verification the number of process corners would increase e.g. to include corners like slowNslowP_fastRfastC :)

However, in tools like VSDE or ADEXL I can put in one active setup and datasheet only EITHER PVT results OR MC results. So to combine corner & MC results I would need to run MC at all corners, ending up e.g. with 36*50=1800 simulations - although for me it would be enough to do 36+50 MC runs at TT with mismatch=86 runs. In the run of 1800 simulations there is far too much redundancy, thus waste of time.

I wonder if there is a GUI-based automated way? Currently I create e.g. 2 datasheets from ADEXL, one for PVT only, one for MC only and combine them in Word or Excel.
Another solution I found was creating statistical corners, but that needs also some manual steps.

Are there yet only script-based methods (like Ocean) to do that?

Bye Stephan