40 phase amps with the wheel off the ground

[avh]

I have my first ebike setup coming together. A KT 48SVPRL-GLT01 controller, KT-LCD3 display, 52V 16S LiFePO4 battery, and 9C 2705 direct-drive hub motor (I think I got the last one Grin had in stock). No cadence or torque sensor yet, just a throttle. I finally have it together enough to try running the motor (the battery isn't mounted yet, it's sitting on the floor with a short cable going to the controller), and I have a problem that I don't understand.

Holding the wheel off the ground and giving it throttle, the motor spins smoothly enough, but the display shows it as drawing many hundreds of watts. A clamp meter reveals that each phase is getting 40 amps, even though there's no load on the motor. It actually started out much worse than that, drawing 1200+ watts and being unable to get above 12 miles per hour indicated, so I tried all 6 permutations of hall effect wires to get it here.

I don't think it's mechanical drag in the motor; it spins with little enough resistance that I barely notice while pedaling.

 

[Chalo]

Sounds like a false phase/Halls arrangement to me. If you get a false positive combination, the motor will run too fast, with minimal torque and way too much unloaded current.

 

[avh]

I'm pretty sure there are only 6 combinations of hall sensor wiring, and I tried all of them.
Power measurements made at full throttle with the wheel off the ground.

Controller	Motor	Result
Y G B	Y G B	spins forwards slowly, stuttering, >1200W, >50A
Y G B	Y B G	doesn't spin, stutters / vibrates loudly
Y G B	G Y B	doesn't move, stutters / vibrates loudly
Y G B	G B Y	spins backwards. vibrates / hums. ~50W, ~10A. Cuts out after a few seconds.
Y G B	B Y G	spins forward, stutters / vibrates at low speed. ~600W, ~50A
Y G B	B G Y	doesn't move, stutters / vibrates loudly

None of these seem like correct wiring. The YGB -> GBY one is weird; I have no idea why it cuts out after a few seconds.

 

[avh]

Do I also need to try different permutations of motor phase wires? I know that on sensorless motors, swapping the phase wires around will change which way the motor spins. On sensored motors, will it change the phase relationship between the sensor and the coils?

 

[molybdenum]

Yes you do, here's the template