MAC motor spining both forwards and backwards

[CrowCreature]

Hi yall, im newly registered, but definitely not new to this forum, i have been reading it since i got my ebike kit from em3ev 4 years ago and since then the motor has broken down a few times because of my own fault, which i wont go into too much, but the first repair was of changing a broken planetary gear, second time fixing a broken cable and a hall sensor.

So after fixing the broken cable and changing a hall sensor a month ago, i have a weird issue that if i demand too much power at once from the motor, like if i go full throttle from standing still, it will start spinning backwards (i presume), as i can hear the motor whiring and spining, but the wheel stands still. If i release the throttle, i can hear it winding down, followed sometimes by a clank. If i push the throttle ever so gently increasing the power, it will lock on and spin forwards and spin the wheel and if get to full speed, it will spin forwards without any interuptions, except if its not going full speed and up the hill, it will sometimes stop, make a clanking sound and start spining backwards. Usually if i go slowly and hear the initial magnet whiring sound without the wheel spining and increase the power gently from there, it starts slowly spining forwards and i can increase the throttle more rapidly to the max without problems.

Ive read that the motor can spin backwards when hall sensor/phase wires are mixed up for some reason, but i dont understand that it also spins normally with less load/torque. When i was testing it without any load, it was working completely fine.

It also could be that i messed up something when i was soldering in new cable or hall sensor, im pretty sure i soldered all the colors correctly. I also greased the clutch for better freewheeling, but somehow i dont think its the clutch slipping (or could it be?)

Tomorrow i will try to do some testing, but i really dont want to dissasemble the motor again, i tried to make it watertight with silicone and i would have to redo it and its just messy.

I will also upload some videos shortly.

Anyways, maybe someone has any ideas at least where should i start, what to check and what could cause this peculiar issue?

Many thanks

 

[CrowCreature]

Its not a great video and the end is interrupted by my courier app notification, but the main point was that it can spin forwards and if i stop and then full throttle quickly it produces the whirring motor spinning sound, but no torque whatsoever, then if i stop and push a bit more carefully again, it starts going again

 

[amberwolf]

If your controller is sensorless capable, then if there were sensor failures or problems, it may just ignore all the sensors and read phases for position data instead. Under high currents that data may be less reliable (more noisy).

Same thing can happen even in sensored mode if there's no shielding on the hall wires and they run inside the same cable as the phases.

Either one can cause the controller to not know where the rotor is in relation to the phases, and drive it the wrong direction.


You can verify sensor operation without opening the motor; there are various ways to test, such as the ones at ebikes.ca 's Learn - Troubleshooting section.


If cable damage included phase wire shorts to hall wires, it can also damage not just the halls, but the hall sensing electronics inside the controller, which can cause wierd or wrong behavior even when the sensors themselves have been replaced and verified working.

Controller FETs can also be damaged, so they don't all pass the same amount of current, but "work", and various weird behaviors can happen.

In either of those cases, it's easier to replace the controller than to fix it (if it's even possible to fix).



BTW, greasing the clutch, if it's the typical roller/ramp type, can cause slippage of the rollers especially if it's older / worn. This will make it harder for it to grab, but it's unlikely to actaully keep it from grabbing at all. The problem you're seeing is much more likely to be the motor spinning the wrong way inside the casing.

The only "easy" way to tell which is happening is to open the motor, but you'd have to do the testing with it mounted on the bike under the same conditions (with you on it, presumably) that you already know cause it to happen, to be sure. If it's a clutch issue, you may not see it happen unless there's enough load on it to cause a slip. If it's reversal, it'll happen anyway, most likely.

 

[CrowCreature]

If your controller is sensorless capable, then if there were sensor failures or problems, it may just ignore all the sensors and read phases for position data instead. Under high currents that data may be less reliable (more noisy).

Same thing can happen even in sensored mode if there's no shielding on the hall wires and they run inside the same cable as the phases.

Either one can cause the controller to not know where the rotor is in relation to the phases, and drive it the wrong direction.


You can verify sensor operation without opening the motor; there are various ways to test, such as the ones at ebikes.ca 's Learn - Troubleshooting section.


If cable damage included phase wire shorts to hall wires, it can also damage not just the halls, but the hall sensing electronics inside the controller, which can cause wierd or wrong behavior even when the sensors themselves have been replaced and verified working.

Controller FETs can also be damaged, so they don't all pass the same amount of current, but "work", and various weird behaviors can happen.

In either of those cases, it's easier to replace the controller than to fix it (if it's even possible to fix).



BTW, greasing the clutch, if it's the typical roller/ramp type, can cause slippage of the rollers especially if it's older / worn. This will make it harder for it to grab, but it's unlikely to actaully keep it from grabbing at all. The problem you're seeing is much more likely to be the motor spinning the wrong way inside the casing.

The only "easy" way to tell which is happening is to open the motor, but you'd have to do the testing with it mounted on the bike under the same conditions (with you on it, presumably) that you already know cause it to happen, to be sure. If it's a clutch issue, you may not see it happen unless there's enough load on it to cause a slip. If it's reversal, it'll happen anyway, most likely.

Thanks for sharing your thoughts on this. My controller seems not to be sensorless capable as the motor doesnt spin with hall cables disconnected.

Hall sensor wire did get shorted, yes, hence the new hall sensor and cable change. I didnt think about it, but it kinda makes sense that the controller insides also got shorted, FETs or something else, i dont really know how the controller really works, but it makes sense somehow and, like you have said, it produces these weird behaviours.

I will try to get a new controller, probably best would be to get the same one, but em3ev is not supplying them anymore, so i will have to figure out which would work for me as a replacement. My current controller is infineon 6fet 3077 36-50V.

Thanks again