Problem with hub motor / hall / controller

Hi!

I recently bought this kit from aliexpress. Everything was running fine for about 200 km until today. Motor still runs fine with no load. But when I ride it and throttle up the motor makes weird noises like something is spinning inside. The speed on LCD also goes up although the actual road speed doesn't because the motor makes only little torque. Did I cook the hall sensors? I was really pushing it the last ride and the motor got quite hot. Could overheating hall sensors cause permanent damage to them? What are symptoms of weakened magnets in hub motors because of overheating?

Thanks a lot.

Here are few pictures.




The motor has two pairs of hall sensors. Each has its own 6 pin connector. The motor also has a 10k thermistor. But it's only connected to one hall connector. The other one is just has a wire and connector terminal connected to nothing...

I don't know what to guess here really. The noise and weak motor don't fit the normal faults. Especially the noise.. Could you post a video of this?

-If you damaged the sensors then motor wouldn't be going ok on no-load. Do you get same behaviour on both hall sets?

-If you are using temp limiting then it could be the controller is reducing power if temp sensor is measuring incorrectly after being overheated (not so likely)

-If the magnets are cooked then you'll have a changed kV, this means more speed and less torque. You can measure what noload rpm you get to verify this. Has it increased?

My best guess: the phase connectors or wires might be damaged, causing partial short circuit at higher load. Phase leads look pretty cooked here:

Check isolation with an isolation tester or DMM, between the phases and between phases and stator.

I've also had poor connectors giving weird motor running after they melted a bit (check them also)

How hot did your motor get? Does it smell foul?

Noise is like the controller getting out of sync. The controller is changing commutation steps by itself and not following hall sensors at higher loads.

- I get same behavior on both hall sets. I replaced one set with Honeywell sensors and nothing changed.
- It's not because I just noticed the controller doesn't even have a connection for temp sensor. Just 5 wires on hall sonnector.
- Noload rpm stays the same. Still at about 120 km/h.

So you think at higher load the insulation fails and shorts the motor?

Motor smells chinese.

Resistance between all three phases is ~55 mOhm. I was measuring with 4 wires so it should be quite accurate. Resistance between stator and phases is more than 10 MOhm.

The speed on LCD also goes up although the actual road speed doesn't because the motor makes only little torque...

..Could overheating hall sensors cause permanent damage to them?

Click to expand...

So it's not magnets, not temp sensor..

Controller will get speed from one of the halls, so if the shown speed wanders then something is certainly off. Overheating can kill the halls but if you have two sets working at noload and have replaced one of them then that's not it..

Might be controller itself, do you have another one to test with?
Can you unplug the display and still run the kit? If so, then try it, or just unplug and replug the display and try.
Do you use the brake switches? If so, then try to disconnect them.

I think I misinterpreted your pic, is the blackened wire a dead end or going to the phase wires? I believe it's not the phase wires but instead a dead end?
In that case it means you have a star connected motor and that it's the star connection that's so burned. If so: can you cut the cable tie and carefully bend the wires to see if they are burned to a crust or if it's still got some flexibility underneath?

The star point is always connected, doesn't matter if it's burnt but it indicates really high temps there. Are there other parts of the wires nearby that are burnt as well?

I'm currently making my own controller. https://vesc-project.com/node/655
Once I receive correct caps I'll be able to test the motor with it. The controller offers powerful diagnostics tools so it should be easy to see what's going on.

I also think it's the chinese controller. Seems like something happened inside and messed with commutation logic. It is only stable at low load. I'll try disconnecting the screen when I get home on friday but I don't think it will help.

It does seem like a star connection indeed. I assume wires are connected with a solder joint which has higher resistance than the winding. This explains why just this part of current path is burned and everything else seems fine.

Loose magnet ring? I know it's unlikely, but it fits the noise and apparent increase in speed.

when you have a root cause then please post it. It can help others :thumb:

Hooked up the motor to another controller and worked fine. The test was done up to 120A phase current which is much way more than the chinese controller capability.

My guess is the logic in chinese controller fried just a little to keep everything working under light loads but not under heavy load. If anyone has similar issues I'd recommend replacing the controller.

After some riding the weird sound is back. I replaced the controller with my custom VESC running in FOC mode. Under current lower than ~80A there's no sound after that it sounds like in the video below. It doesn't matter at what speed I'm riding. I also tried switching my controller to sensorless FOC mode to eliminate any potential hall problems. The same rattling behavior was observed.

In the video I'm holding rear brake and applying throttle until rattling begins.
[youtube]apzKfVEC0Ug[/youtube]

Could the noise be the stator vibrating for some reason? Or maybe it's just loose from the axle.

I made the same test again after the motor has cooled down. No weird sounds with 200A phase current... It seems like its only making noises when it gets above certain temperature. There's about 30% difference in thermal expansion between steel (axle) and aluminum (stator). Maybe at higher temperatures the stator expands and rigid mechanical connection to axle becomes loose which causes rattling sound. I'm just guessing here

That sound.. I'd guess something is actually rotating. But unlikely.. Only rotor could, otherwise your phases would be torn out. Could kdog be correct, magnet or backiron ring rotating inside hub shell?

it must get extra hot somewhere, where?

Magnet ring... yes that could be the explanation. The Chinese controller I had before didn't have temperature monitoring. Motor possibly reached extreme temperatures making the magnet ring loose. I also noticed on the day this noise started showing up (when the motor was overheated) the spokes weren't as tight as they normally are when the motor is cool indicating that thermal expansion of the motor housing is not insignificant. I'll try to disassemble the motor and see if I can glue the magnet ring or in some other way fix it to the rotor.

Thermal expansion of aluminum (motor housing)


Thermal expansion of iron/steel (magnet ring)


Difference between materials is ~0.2 mm which is a lot if the magnet ring is pressed in the motor housing. It could also be glued in place but the glue was destroyed during the extreme temperature incident.

I opened the motor and poked a magnet with a screwdriver and this happened:



All the magnets were loose. Looks like the manufacturer didn't use enough or correct glue.

I than took all the magnets out and cleaned them with acetone. Cleaned insides of the motor too. Then I used epoxy to glue magnets back in place.



Should be ok now. Thanks to everyone who helped.

Congratulations! There's a first for every issue, i guess you were first on this one
Poor quality.. you should send a notice to the seller. Hope it stays together now so you can ride! 8)
Cheers / larsb