Hi Nishan, thanks for the info - really helps visualize the issue you are describing.
Here are some questions:
- [So I want to show this on PLECS. … idea to implement this.] - in your application are conditions known for when ZVS is achieved in your converter or is this something that needs to be determined via simulation? If you know the conditions for ZVS operation then potentially you could use a variable that indicates whether ZVS is achieved or not based on the operating condition and utilize this with your switching loss lookup take (see Switching losses in partially soft-switching inverter - #2 by Oliver_Schwartz)
- If however, you don’t know prior to running your simulation which operating conditions will achieve ZVS, then I can’t think of a way how PLECS would be able to predict this for you without including the node capacitor of the switches.
- [I modeled the thermal … according to the look-up tables.] - this is because the PLECS thermal domain will use the blocking voltage right before turn on, the conducting current right after turn on and then use the lookup table to determine the losses.
When I probe some waveforms in your topology I see that when the gate turns on the block voltage is already zero (likely because the body diode is conducting prior to turn on). Thus the lookup table at the turn-on will always show 0J of switching energy at turn on as the blocking voltage is 0V as the gate turns on (see below).
