HellO
We busted EPC2034, MP6908 along with TPS724xx device in an implementation using UCC28782 controller.
It gives output as desired: but very fast , all secondary ICs popped!
Pl see the schematic screenshot attached.
The reason for the convoluted biasing of MP6908 is the datasheet does not specifically give any table for VG output values!
So we suspected that with HVC at any unknown voltage, VDD max could be 9V which is clear from datasheet. But nowhere any mention is made for VG value. Since they only show applications with MOSFET, most of which will take 12V, we assumed VG will be 9V- not acceptable for EPC GaN.
So we drop Vout from 20V to 5V & set up MP6908 with 5V.
Clearly, TPS724xx LDO cannot sustain charging current to VDD.
We won’t have to use this scheme if VG pin value under all VDD conditions is specified to be 5V.
Welcome to the MPS Technical Forum.
We had a couple of questions to understand your circuit. What is the switching frequency and output load current range?
Apologize for the delay- I did not notice your questions until today…
Freq or voltage or current has no relevance to the datasheet inquiry.
What I was asking is related strictly to ds information on Vg output voltage on MP6908.
It would be important feature to be very specific about Vg. #1 MOSFETs tically require full 12V Vg; MP6908 produces 9v so a short discussion on why all MOSFETs will be fully ON is mandatory . 99% engineers would not know channel behavior at 9V #2 GaN requires mostly 5V or 6v. This is not even presented in the datasheet. This will lead to misuse of the controller. #3 Since MPS6908 controlls by emulating dynamically diode rectification, I think it is confusing how it can do so when either device operates in the third quadrant.
Never seen any curves of MOSFET -Vds behavior. Gan with -Vds is problematic.
How does MPS6908 control under these unknown characteristics?
