Your are welcome Paul ! I just looked your schematic rev 0.1 of the AC controller, and I see that you use a 30r Rgate, for switch ON and switch OFF of mosfets. I am almost sure you will have the autocommutation shoot throught condition. When we do not pay attention to this phenomenon, we don't see it because with the high mosfet derating mosfets will accept the peak current during the shoot through condition.
When you will do test of your AC controller, put the scope (channel 2) on Vgs of a low side mosfet for example, and trig on rising edge of channel 1, connected on the corresponding motor phase output. With 30r as Rgate and 144V bus (and even smaller bus voltage), I am almost sure you will see the Vgs signal rise up to the gate threshold voltage.
For comparison, last weeks I worked on a "small" BLDC controller, 72V/100A, based on 12 IRFB4110 mosfets (100V/120A), thus 2 mosfets in parallel on each half-leg. On the first version of the schematic my Rgates was based on 82r for switch ON and schottky + 10r for switch OFF.
Despite this I still was in limit of autocommutation under 84V (pack fully charged), because Vgs raised up to 3V when opposite mosfet switched ON. Thus to be sure, I passed the switch OFF Rgate to.....0r !!
Thus I let you imagine with 144V DC bus and 30r switch OFF Rgate !!
Regards
Cyril