it's very strange, because it works with W and L, ...
# xschem
j
it's very strange, because it works with W and L, but with m and nf it doesn't work, it never iterates
s
Replace parameter M with MM. M is a special reserved multiplier. I see 3 different currents in I(VDD4) this means that ngspice is correctly iterating the MM parameter. Transistors have attribute
mult={MM}
j
I don't understand why it doesn't work for me, I changed to MM and I still don't see the iterations
fo4.sch
s
Here the working schematic. You see the 3 values for [m] in the graph.
j
can you look at the mean commands in the simulator? do they change with each iteration? is that if you look I have a line that saves the variations of that with respect to the change of M, that file does not change with multiplicity or fingers but with W and L, yes
s
Of course measured delays do not change. If you multiply the whole circuit you will have the same timings. Add some fixed capacitance on the inverter outputs and you will see a difference
j
Do you know if this technology is 130, is it overloaded with a very large W? I wanted to characterize that