Hello, I made my own 4.12 VESC where I basically just changed 50V to 100V ceramics, a ferrite bead, and a TVS so the VESC could work more reliably with a 12S pack. Also did my own layout so all transistors are on the same side so I can put a heatsink on the bottom of the board. Of course, it doesn't work.
FW 3.34, Hw 410, UUID 52 00 4E 00 18 51 36 34 32 35 34 36. Motor is a 5065 270Kv.
In VESC Tool, if I command 10% duty cycle the motor cogs around loudly and slowly. It does not enter controlled commutation mode (where it commutates from phase sense), and does not throw any codes. If I use any other mode, such as I or w, the motor jerks then stops when the 10A power supply current limit kicks in and it throws an undervoltage code. I interpret this as a symptom of the motor consuming lots of current from being driven incorrectly, ie without the supply limit, the driver would smoke the board and the motor.
Something I noticed is that phase B low side FET never seems to go inactive. Here are the phases as it is operating in open loop commutation mode (Phase A red, B Yellow, C green):
The commutation seems like its missing two steps: 'A+ C- Bsense' and 'C+ A- Bsense'. I assume this is some kind of startup technique? The waveform looks very wrong... Here is an image of B low drive always being on (Phase A red, B yellow, B low drive green):
Here is the current capture from the sample tool and oscilloscope of the phases when I run the motor in current mode set at 3A:
I suspect the problem has something to do with the un-driven phase stuck at 0 instead of floating so the vesc can sense it. This looks like a software problem but it is more likely a symptom of something I messed up. What might cause this? Why doesn't it try and commutate with all 6 steps? What other data should I investigate?