Thermal resistance = 135 C/Watt

Power: 3W/mm

Wich Temperature model did you use ?, in that sense also Which mobility model ?

Do you see any difference in the lattice temperature?

We found out when simulating your voltage source as a transient PWL voltage source while having your thermal simulation active delivers more accurate results

 Hello Mr. @Hady Yacoub,

We used both Hydrodynamic and Thermodynamic models but we could not manage to see any current drop in IdVd graph at higher voltages. Another interesting result is that we find out, while making thermodynamic simulations the entire device becomes too hot, that is Tmin and Tmax are very close and very high even at the very begining of the simulation. By the way we use quasistationary solver so far, we haven't tried any transient simulation yet.

Regards, 

Talha Furkan Canan

I had similar problems as i started using TCAD, So it seems that your thermode boundary conditions are too extreme. Additionally, do not rely on quasi-static.. there is no real heat transfer  there being solved it only solves for a steady state. I cannot stress enough on transient simulation it gave us till now perfect results (we can really fit the simulated current to our measurement data). As for Canali model, we used this too to optimize our saturation velocity.

I hope this helps

Regards

Hady Yacoub

Thanks for all responses

For hydrodynamic simulation you should choose a proper mobility model, as the mobility in fact function of energy!

i am applying thermodynamic model, but it is showing lattice temperature variation in the gate section, and at the channel it is showing 300K?what can i do?