LDMOS RF Power Amplifier Drain Current Overshoot and Oscillation After Device Replacement

I am using an asymmetrical ldmos amplifier as final stage amplifier in a RF card. When I tested the card initially i got the gain and linearity as per the design requirements.But the thermal dissipation was not good, hence transistor got damaged when operated at high power. We took measures to improve thermal coupling and hence thermal dissipation by using extra screws between PCB and heatsink and also by suing thermal plate.

But when the damaged device was replaced with a new one, we observed some abnormalities while powering on.We switched on the gate bias as per the quiescent current requirement mentioned in test report and datasheet.
As soon as we switch on the main (carrier) device drain supply, the drain current is shooting up. At output we observed oscillations as well.
But since the same card gave stable output before, we could not find the exact reason for this instability?
Kindly share your suggestions and feedbacks

Your added thermal plate added stray capacitance and inductance between ldmos drain and ground.

Just some basic points. Don’t screw packages like TO220 to heatsinks. They need spring clips. I know they have screw holes -- don’t use them! Secondly, don’t solder parts to the pcb then spring clip the package to the heatsink. You need to spring clip the part to the heatsink then do the soldering, otherwise the spring clip is fighting the solder joints. It costs more to do it the way I have said, but it is worthwhile in terms of improved reliability.

RF circuits “like to oscillate” if they have any bandwidth. Just because one example device works does not mean the design is resilient against device-to-device variation.

A circuit diagram would help for us to see the issue. Maybe you need a small (22R) gate suppression resistor close to the output FET gate. Maybe it would benefit from a ferrite bead. Hard to see without a circuit and layout.