I'm working on getting a prototype running and have had three DRV8301 failures in the last week. The system is being powered from a 12v Duracell Ultra deep cycle, 35Ah battery, through an ideal diode. Two of the failures involved Gvdd_uv being reported by the chip, and the chip was very hot to the touch but only AFTER the failure occurred.
The third failure was not reported by the chip, but it manifested itself as GH_A and SH_A both floating up to about 10v if EN_GATE was asserted. In this case, the DRV8301 was not warm. I suspect the high side driver of phase A failed open.
I've gone through our schematic and compared it to the DRV8301-HC-EVM and couldn't find any significant differences except that we're only using a single current shunt monitor instead of both. As for driving it, the GUI that comes with the kit drives the motor using a flux control method whereas I'm attempting to use a BEMF method.
I also have three 20v TVS diodes (SMDJ20A) on the board connecting the power plane to the ground plane which I would expect to handle any issues with voltage spikes that might come from switching the motor (was well as 3 470uF capacitors).
What can cause the Gvdd_uv failures to occur? And what might cause the second failure mode that I saw?