For the thermal models, is the turn-on and turn-off energy data accurate for soft switching, or is it double-pulse-test (DPT) data? The DPT assigns Eoss stored energy to turn-off rather than turn on, so the sum is correct (hard switching), but it is not accurate for ZVS turn-on. Do I need to manually correct for Eoss for a ZVS turn-on converter?
Best,
Noah
Hi Noah,
The answer depends on the thermal model you are using and how it is implemented. Some semiconductor vendors include an additional term +lookup('Eoss', v, T). Others do not specifically include the Eoss losses or fold them into the Eon/Eoff tables, but the distinction becomes critical at light loads or ZVS where the Eoss losses can dominate.
This also was the topic of a great talk at APEC this year by Wolfspeed, noting that other vendors (but not all) include this loss mechanism as well.
Regards,
Bryan