Overdrive Voltage of a Transistor in Cadence Virtuoso/Spectre

It depends on the model type as to what the equation for vdsat is. For bsim4 in SPECTRE17.1 this is described in <SPECTREinstDir>/doc/spectremod/spectremod.pdf in equations 2-37 or 2-38 in the bsim4 chapter. As for plotting vgs-vth I would expect you'd subtract the vgs and vth operating point parameters.

Andrew.

Hello,

Thank you for your response.

So the overdrive voltage for bsim4 is then vgs-vth and not vdsat ? 

Thank you.

The overdrive voltage is (by definition) Vgs-Vth. In older, more simplistic models that aren't really valid for small geometry models, Vdsat could be modelled as the same as Vgs-Vth, but in the bsim4 model which is more realistic, the Vdsat equation is more realistic. So it was (I think this is a reasonable way of looking at it) that Vdsat=Vov in older (simpler) models, rather than the other way around. Vov is always Vgs-Vth, but Vdsat may need to be described by a more complex equation in more modern models which are a better fit for modern devices.

Put another way, Vov tells you how far above the threshold voltage you're driving the gate, whereas Vdsat is the voltage such that if Vds is greater than Vdsat, the drain current is essentially constant (the "essentially" is because it's usually not really constant because of channel length modulation and other effects). 

Regards,

Andrew.

Hello,

Thank you for your response.

So, for design purposes, should I take the Vdsat from the simulator as the actually "simulated" value of the overdrive voltage for that particular transistor for BSIM4 models ?

Thank you.

No. That's not what I said. Why don't you use vgs-vth since that actually is the overdrive voltage rather than something that isn't?

Andrew.