# Calculating power in FETs using the P() measurement function

For a simple DC circuit I would expect the avg(p(m1)) measurement to give me an expected power measurement for m1 but after doing some investigating I have no idea what avg(p(m1)) does.

Does any one know what the avg(p(m1)) function is measuring as shown in the image above:
(.meas p_m1 avg(p(m1)) ? I would think that maybe the gate power is incorporated in p(m1) but that should make it more not less than the power dissipated in the channel right? It’s not actually switching here anyway so what is going on?

P( ) calculates true power dissipation while ignoring displacement currrents. Here is a study of P( ) of NMOS in Qspice, and P( ) calculates not only drain but also gate power. But in your case, as gate is DC, gate power can normally be ignored.

Replicate your circuit, drain power is same as P(M1), as gate is DC and contribution from gate power can be ignored. But in Qspice, I cannot find a default model UF3C065080B7S, did you modify a sub-circuit?
Consider to upload your schematic file when you post a question as to get more valuable support and to save the review time.

Hello @KSKelvin , as always thank you for the detailed response! In response to the fist post talking about P() calculating the “true” power and ignoring displacement currents. This is a DC test so there should be no displacement current (only real power) also since there is no switching the gate power should be much much less than the channel power dissipation such that the channel power dissipation dominates any power measurement for m1.

For the second post, it’s interesting that you don’t have the model for UF3C065080B7S. It has been in all versions of Qspice since I started using it around January 2024. And to your point of uploading the spice model, you are right of course, I need to be better about that. The circuit below (and attached) hopefully shows in better detail what I am confused about when using the P() function:

Of note is that I put in some current limiting that will allow me to perform this exact test on the bench : )

I believe that in this DC circuit above avg(p(m1)) should be approximately equal to the power supplied to m1, but it is not and that is why I don’t understand what the p(m1) function is calculating here.

qspice power calc.qsch (6.5 KB)

Oh, I forgot about Symbol & IP Browser which uses level2010.txt library.
A quick review suggests Qspice may have problem in calculating P() for model of NMOS level=2010.
The example I shown in above is VDMOS from nmos.txt library.

Qspice may have problem in calculating P() for model of NMOS level=2010

hmm, well for me personally, I would rather have P() not work at all than return bogus values. Maybe Qspice should return at lest a warning when you run a simulation, especially for built in models! I guess it’s a word of caution in the tone of “Caveat emptor”