awful grinding noise from hub motor

[cgLarry]

Hello everyone,

I have a a 250W hub motor and a 350W controller on my bike, both noname. Everything was going great until i decided to do the shunt mod (sometimes you gotta do what you gotta do). After that while the motor was under load (accelerating on a straight road) it would make that grindy/resonant noise which would result in:

Nasty resonant noise,
Loss of power,
vibrations and shakiness.

Since then i have to give it very little throttle while accelerating to avoid that.

And I also ran the motor in sensorless mode for a bit after the shunt mod (connected the 'learning' wire on the controller to make it 'learn' after removing the hall connector)
Then i plugged the hall connector back in and done the 'learning' step again, no difference.

Before the mod, while accelerating, it would do the resonant noise but it was brief and didnt result in power loss or a little amount (maybe?).

I always help my bike with pedals while accelerating from a dead stop and on inclines.
I run my dumb (non-smart) controller in sensored mode.

You can hear the godawful sound in the video. Even though Im helping it with the pedals. (Note: The hub motor is on the rear wheel, I could only put the camera in the basket in the front[if thats what its called]). I also suspect that a magnet might have fallen out of its slot and may be getting jammed somewhere, but its unlikely i suppose.
Any help would be appreciated, thanks.

https://imgur.com/j2QEKfE

 

[amberwolf]

If the problem started the very first time you used the systme after the shunt mod, the problem is likely the controller pouring so much current into it that it is shaking the windings, stator, etc., mechanically making the noise, or creating so much noise in the hall sensors from high phase currents that it causes timing problems so the phase currents aren't synced with the rotor like they should be.


If the problem started only after the motor was used this way for a bit, and especially if the motor heats up, then since it's probably a geared hubmotor, the gears may be damaged by the heat and deform.

Or, if it got the magnets hot enough, and the glue used on them isn't very good, it could be pulling some of them off the rotor and causing them to strike the stator while spinning.



Alternately, since the controller now has no way to protect itself against damage from overcurrent, becuase it doesn't have any idea how much current is actually flowing, the controller itself could be damaged, and unable to correclty drive the motor.

Or it could have insufficient capacitance in the vairous places tha'ts required, to deal with the extra electrical noise from high phase currents.

Etc.

 

[cgLarry]

If the problem started the very first time you used the systme after the shunt mod, the problem is likely the controller pouring so much current into it that it is shaking the windings, stator, etc., mechanically making the noise, or creating so much noise in the hall sensors from high phase currents that it causes timing problems so the phase currents aren't synced with the rotor like they should be.


If the problem started only after the motor was used this way for a bit, and especially if the motor heats up, then since it's probably a geared hubmotor, the gears may be damaged by the heat and deform.

Or, if it got the magnets hot enough, and the glue used on them isn't very good, it could be pulling some of them off the rotor and causing them to strike the stator while spinning.



Alternately, since the controller now has no way to protect itself against damage from overcurrent, becuase it doesn't have any idea how much current is actually flowing, the controller itself could be damaged, and unable to correclty drive the motor.

Or it could have insufficient capacitance in the vairous places tha'ts required, to deal with the extra electrical noise from high phase currents.

Etc.

Thanks for the answer, there are a lot of things to be looking out for, yeah?

Firstly, I decided to crack the motor open to take a peek inside. This required me to remove the hall connector's pins from its plastic housing. And I found this:



As I said before, I disconnected the hall connector just after doing the shunt mod, and went to ride it like that. It was making that noise and then I reconnected it, still the same. Maybe it was running in sensorless mode after I reconnected it due to the bent pin or it was running sensored but unable to read input from one which had its connector bent. Still the same noise as the sensorless if thats the case tho.

I think that i might have bent the pin at that time.

Does sensorless operation cause that much noise and vibration?