Changing the FET model parameters dynamically during Transient Simulation

This is just a global parameter, so you should be able to change it via the dynamic parameters in transient as before. If it's only this parameter you want to change, you can check the "dynamic parameters" checkbox on the transient form, specify mainfet_rdsw as the Parameter name, and then enter the parameter vector such that this is 5000 at time 0 and 7000 at some later time. You only need the paramset if you want to vary several parameters during the transient.

I just checked this with a similar setup using gpdk090 models and it worked fine.

Regards,

Andrew.

Andrew thanks for the answer, but also, let's say I have schematic with two devices using the same model with that parameter. And during the simulation I want this parameter to be changed only for one instance. If I do it your way, then it will change that parameter (since its global) for two of them.

How can I change so that only one instance at say 100ms has that parameter changed to 7000 but another instance continues using that model with old parameter 5000, any idea? Do I have to do some nesting syntax in that parameter box or any other method? I probably should have described that scenario better in my first post.

This is a very odd thing to do (even if you wanted to change all of them) because of course you can't really change the model parameters during real life (unless maybe you are trying to emulate aging effects, in which case there's RelXpert for doing that).

Anyway, the simple answer is that you can't do this (without some work on the model files). The only way I can think of would be to duplicate the model in the model file (or in an additional model file) with a different name and have that model parameter in the differently named  controlled by a different global parameter. Then you have the one instance you wish to change use the modified copy of the model, and you can then dynamically alter the new global parameter without affecting any other instance.

Regards,

Andrew.

I am trying to see behavior of the system when this parameter varies differently for different devices. Do you think even a better idea then would be just to run monte carlo with needed tolerance on that parameter during simulation? (reason why i wanted to change model is emulate the variation)

Yes. If you run with monte carlo mismatch, each instance has a unique model (assuming the models are parameterised) and so that allows it to assign different random values of the parameters to each instance.

Andrew.

well, I am looking at model description file, and there are several corner subsections with that parameter (mainfet_rdsw) in each of those subsections. Also, it has same value within each subsection.

I selected the MC simulation, then I selected the FET device using that model in the "Specify Instances for Mismatch" dialog window.

Then I ran the simulation with 10 points, however; right now I cannot see if the mainfet_rdsw was really varied during sim, and also I want to know which parameters were varied specifically, how can I do that?

Thanks!

Nevermind.., it is the statistics keyword in the model libraries *.scs file description which tells what and how will be varied.

You can also switch to Detail view after the simulation has finished, and do Right Mouse->Print Statistical Parameters over one of the monte carlo points - then you can see the values of all of the parameters.

Regards,

Andrew.