I try to characterize nfet_g5v0d10v5 with W/L = 7/...
# sky130f
Filippo
12/16/2022, 10:23 AMI try to characterize nfet_g5v0d10v5 with W/L = 7/8. In the online documentation Vth is presented as 0.811V. I found a threshold voltage about 1.2V. Can be correct?
a
Arman Avetisyan
12/16/2022, 12:42 PMhow did you measure it?
Arman Avetisyan
12/16/2022, 12:42 PMProbably incorrect
Arman Avetisyan
12/16/2022, 12:43 PMthe threshold is etested afaik
f
Filippo
12/16/2022, 1:05 PMI found the line equation of Ids vs Vgs of nmos
Filippo
12/16/2022, 1:05 PMThe Vth is the intercept of this line with 0 axis.
Filippo
12/16/2022, 1:05 PMCorrect?
Filippo
12/16/2022, 1:12 PMImmagine di iOS.jpg
a
Arman Avetisyan
12/16/2022, 1:25 PMnope. It even says "linear interpolation"
Arman Avetisyan
12/16/2022, 1:25 PMthe Vth is defined as the point where electrons == holes underneath the gate
s
Stefan Schippers
12/16/2022, 3:00 PMI tried to simulate this transistor: this is what i get with Vd=0.01V and tt corner:
Stefan Schippers
12/16/2022, 3:11 PMIf you are more interested in strong inversion this is probably where d(Id)/d(Vg) is maximum:
f
Filippo
12/16/2022, 3:13 PMOkay thank you. I use Vds = 10V. So, Vds must be in weak inversion, right?
s
Stefan Schippers
12/16/2022, 3:16 PMVds=10V is definitely wrong fo rmeasuring Vth. Usually a low drain voltage is used such that the transistor operates in the linear region, usually 10mV or 100mV.
👍 1
f
Filippo
12/16/2022, 3:17 PMOkay thank you very much
s
Stefan Schippers
12/16/2022, 3:23 PMHere the schematic file I used:
f
Filippo
12/16/2022, 3:35 PMOk, I use also mismatch models in my simulation
s
Stefan Schippers
12/16/2022, 3:38 PMmismatch (tt_mm) models will give you a slightly diferent value on every run.
f
Filippo
12/16/2022, 4:34 PMPerfect