Designer’s Guide Community

‹ Previous topic | Next topic ›

Pages: 1

Monte Carlo: Mismatch and area (Read 2205 times)

idesign

New Member

Offline

Posts: 2

Monte Carlo: Mismatch and area
Jan 27^th, 2009, 12:53pm

in Monte Carlo analysis, the variation between two devices on the same die due to random mismatch is inversely proportional to the square root of the device area: sigma ~ 1/sqrt(Area).
If the transistor is made of multifingers, should I consider the total area or one finger area? An answer and a reference will be much appreciated.
Thanks

IP Logged

Tlaloc Community Member Offline Posts: 81 Dallas, TX	Re: Monte Carlo: Mismatch and area Reply #1 - Jan 27^th, 2009, 6:46pm It is due to the total area including all fingers and multiples. This is due to averaging effects since with larger areas, you are more likely to get areas of the device that are shifted in opposite directions that then tend to cancel.
Back to top	IP Logged

wave Senior Member Offline Posts: 117 Silicon Valley	Re: Monte Carlo: Mismatch and area Reply #2 - Jan 28^th, 2009, 2:33am Agreed - consider total area. Layout stripes so you don't introduce something systematic, which is not modeled. (That's another can of worms - Ref: Bob Pease) M. Pelgrom, JSSC, wrote the classic matching paper with the eqn you cite.
Back to top	IP Logged

Pages: 1

‹ Previous topic | Next topic ›