EM_Linear Vs Non_EM_Linear RF PA Stability Simulation
I have designed an amplifier and stabilised the FET unconditionally and checked the K factor across wide band using AWR microwave office and PA was stable across the band. However, when I extracted the EM structures of the matching networks and checked the K factor again, I found out that K factor is below zero at some frequency points this time. Now question is which one should i consider more accurate and rely on. non EM or EM extracted networks.
What do you think is close to real time? It's also not enough just by looking at K factors, because they are just telling you about small signal stability.
EM simulations, particularly MoM(for PAs) are of higher order approximations. You can go as close as to measurement results depending on how you tell your EM simulator to act.
Ok, one thing is clear that what i was thinking is correct, and proven by realised hardware as well.
second question: now that i have made the board and based on just the non-EM simulated circuit and finding that FET gets blown out straight away i.e. its GaN FET @ 28V. when i am biasing is i am limiting the power supply to 1A but FET has 80uF of capacitance in its Drain, FET switches ON and then in few seconds GATE gives in i,e, FET starts drawing current from GATE. I have not seen something like this in my experience, which is not a lot any way. If FET is unstable then it should oscillate and with power supply limited to 1A should not let it blow up, unless transient current from drain capacitors is enough to kill it but again why is GATE giving up, usually GATE gives up under RF drive.
Confused! but million dollar question is that how can i rescue the design without re-spinning the PCB.