“You learn as long as long as you live” is I guess a bad translation from Swedish… Thanks to Aamir, I got a chance to look into a way to plot parameters as function of swept DC variables. I haven’t used this since back in the old HSPICE days. I remember that I once concluded it was not possible in Cadence spectre. But it is.
OK, so I want to plot the gm of the transistor as function of the swept voltage (or whatever). To do this in spectre you can add an additional model file to your Setup > Model Libraries in the ADE window. Store it whereever you like, but preferrably somewhere close to the test bench.
To this model file, add for example the following lines:
or whatever setup you have. Save, apply, change, etc., etc. Rerun the simulation. Then go to the results browser and pick the results from the dc-dc branch of the tree. The expression would typically look something like:
getData( "M0.m1:gm" ?result "dc-dc" )
Now one can start looking for those optimum points a little bit more practically than running parametric sweeps!
PS. Notice also that on the web you can find some additional tricks around this for spectre version 6.2 and later. For example one can do:
to save all stuff.
Thanks Aamir (if you read this — could you post a couple of links too?).