If I have a diode connected NFET being biased by an ideal current source. As the temperature goes up, based square law, the Vds now equal to Vgs should go up too. But my Vds versus temp simulation result is actually opposite. Does anyone know why?
Also, in this configuration, how does the Vdsat change over temperature?
Why would you think vgs will increase with temperature based on square law?
vgs = vth + vov
Threshold has a negative TC typically around -2mV/K. As for overdrive, it also reduces with temperature due to increased mobility.