The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Modeling >> Semiconductor Devices >> model for DC mismatch simulation
https://designers-guide.org/forum/YaBB.pl?num=1202281040

Message started by Rad_31 on Feb 5^th, 2008, 10:57pm

Title: model for DC mismatch simulation
Post by Rad_31 on Feb 5^th, 2008, 10:57pm

I have used the nmos2v_mis model and section 'tt' in cadence spectre for performing dc mismatch simulation of a simple current mirror , however i am getting 0 % variance for all parameters.how do i include the dc mismatch parameters to get the correct results.

Title: Re: model for DC mismatch simulation
Post by Marq Kole on Mar 19^th, 2008, 1:41pm

Hi Rad,

It is difficult to put together an answer for the amount of missing information. I assume that this model is taken from a foundry PDK like TSMC or UMC. In that case you should also look for a section named "stat" or "mc" and include that as well. Most of these PDKs have a stack of interdependent sections hat you need to include to get all the information in your simulator needed to run monte carlo or mismatch simulations.

In the end it all depends on the company that put together the PDK.

Cheers,
Marq

Title: Re: model for DC mismatch simulation
Post by K.Riad on Sep 20^th, 2008, 4:43pm

Dear Marq and Rad,

dcmismatch simulation parameters have nothing to do with the mismatch data meant for MC statistical analysis. These are 2 things different altogether. There is a thread in the Cadence's sourcelink where the difference between DCmismatch and MC statistical mismatch is detailed. Can't remember the SR number off the to of my head. A little search would spot it out.

The dcmismatch data is often available within the TYP corners.
For BSIM3V3 MOSFETS, Spectre does support the following parameters for dcmismatch : (spectre -h bsim3v3)

;
DC-mismatch dependent parameters:
279 mvtwl=0.0 v*m Threshold mismatch area dependence.
280 mvtwl2=0.0 v*m^1.5 Threshold mismatch area square dependence.
281 mvt0=0.0 V Threshold mismatch intercept.
282 mbewl=0.0 m Beta mismatch area dependence.
283 mbe0=0.0 Beta mismatch intercept.
284 mismatchmod=0 select Mismatch mode.
285 mismatchdist=0 m Mismatch Distance.
;

One can then grep one of the above parameters in the model file to check whether the data is there or not.

There are ways of extracting these parameters otherwise, a bit tricky but still possible in some cases ...

Cheers ;)
Riad.