Does it make sense to parallel different GaN FETs for a DC-DC buck converter?
My idea is to use one faster switch with lower Qg and higher Rds_on (e.g. EPC2207), and another slower switch with lower Rds_on (e.g. EPC2305).
The gate drive signal of the slow switch will turn it off a bit earlier. The fast switch will take all the switching loss and the slow switch will take most of the conduction loss.
This way we could optimize loss for a specific load condition.
Fast FET means its Qoss is small. But if paralleling a fast FET and a slow FET, their Qoss should add together. This will not help to reduce any loss.