Quote:1. when monte-carlo varies the process stat parameter, say gauss distribution, it seems for one particular set of parameter applies to all devices of same type. which means process stat does not cause DC offset? is the conclusion valid? it looks so from a diff pair test circuit.
Yes, you are correct. Process variation means all the devices change in the same direction, not mismatch.
Quote:2.tsmc model provide regular device with 2 sets of symbols, one for regular purpose and one for mismatch model. and only limit sets of devices 0.13u pdk have mismatch model available. to run DC offset, i have to build same schematic but using mismatch symbol. it is not practical to maintain 2 sets of schematic to check DC offset of the circuit.
I have not any T 0.13um pdk on the hand. So I am not sure if each device has mismatch model. But you need not maintain two sets of schematic using T 0.13um pdk. I recall that there is a menu in CIW for replacing all symbols with their mismatch symbols and vice versa. If the pdk was installed following the related readme file, you would see the menu definitely.
Best Regards
Yawei
Home
Pages: 1
monte carlo simulation basic questions (Read 2150 times)