How to check if the magnets have come unstuck?

Joined
Feb 23, 2025
Messages
2
Location
London
On an electric bike when the motor is loaded, a creaking/grinding sound is heard and the traction is lost. If you lift the wheel off the ground and press the throttle, it spins normally, but if, for example, you hold the bike with your hands or sit on it, then this creaking and grinding appears and the traction drops almost to zero. If you accelerate on the pedals to 15 Km/h and switch to the motor, then this creaking does not exist, the motor works, pulls.

I suspect that either the magnets have come unglued or the freewheel has broken.

Can anyone please tell me, if it is possible to somehow check and make sure that the magnets have come unglued or they have not come unglued and everything is fine with them?

As far as I know, the freewheel cannot be checked just like that easily, so I think that by excluding the magnets from the list of possible breakdowns, I would know that only the freewheel remains.
 
What variety of hub motor do you have?
Unless you abuse the motor I doubt the problem is due to loose magnets. As your pedals work, the problem doesn't appear to be your freewheel. For a geared hub motor, it could be reduction gears or clutch. It could be a problem with the motor's hall sensors.
I had something similar, and that was a problem with the motor's hall sensor connector. So unplug and replug your motor's electrical connector and see if that effects a change.
 
What type of motor is it?

If it's a geared hub, it has an internal clutch (freewheel) that can slip. that doesn't usually make much of a sound, but depending on why it's slipping, it could.

It's also possible there are gears with broken teeth, and when the breaks all line up the motor just slips inside, with the broken tips of the gears grinding against each other, but if you get it spinning so the gears can sometimes line up with their unbroken teeth, it "works".

If it's a DD direct drive hub, there's no gearing or clutch inside, so the most likely mechanical issue is unglued magnets. It usually requires quite a bit of overheating or significant corrosion inside to cause glue failure, but there are some motors that don't appear to *have* glue on their magnets when this kind of failure is discovered.


Opening the motor to see what's happening is the only certain way to test which one of those is the case.


Another issue that can make some pretty awful racket is a sensor failure, causing the controller to be unable to tell exactly where the rotor is, and so it sends the wrong pulses or timing to the phases. Under loaded conditions that can cause quite a noise, but without a load or when already spinning it may coincidentaly send the right pulse timing and the motor begins to drive normally even though the sensors aren't working correclty. Testing for this can be done with one of those "ebike testers" that has various lights on it to show you what's happening; there are threads about them with more info:
You can also manually test with a 5v power supply (old usb charger) and a voltmeter:


Since yours does work if you're already moving, I'd expect the last problem to be the one you have.

If it is the sensor, they usualy fail either from extreme overheating, severe corrosion, or wire damage (usually at the axle exit and usually caused by a crash or the bike falling over, or axle spinout from incorrect or insufficient axle-retention hardware (torque arms, etc))
 
Last edited:
What type of motor is it?

If it's a geared hub, it has an internal clutch (freewheel) that can slip. that doesn't usually make much of a sound, but depending on why it's slipping, it could.

It's also possible there are gears with broken teeth, and when the breaks all line up the motor just slips inside, with the broken tips of the gears grinding against each other, but if you get it spinning so the gears can sometimes line up with their unbroken teeth, it "works".

If it's a DD direct drive hub, there's no gearing or clutch inside, so the most likely mechanical issue is unglued magnets. It usually requires quite a bit of overheating or significant corrosion inside to cause glue failure, but there are some motors that don't appear to *have* glue on their magnets when this kind of failure is discovered.


Opening the motor to see what's happening is the only certain way to test which one of those is the case.


Another issue that can make some pretty awful racket is a sensor failure, causing the controller to be unable to tell exactly where the rotor is, and so it sends the wrong pulses or timing to the phases. Under loaded conditions that can cause quite a noise, but without a load or when already spinning it may coincidentaly send the right pulse timing and the motor begins to drive normally even though the sensors aren't working correclty. Testing for this can be done with one of those "ebike testers" that has various lights on it to show you what's happening; there are threads about them with more info:
You can also manually test with a 5v power supply (old usb charger) and a voltmeter:


Since yours does work if you're already moving, I'd expect the last problem to be the one you have.

If it is the sensor, they usualy fail either from extreme overheating, severe corrosion, or wire damage (usually at the axle exit and usually caused by a crash or the bike falling over, or axle spinout from incorrect or insufficient axle-retention hardware (torque arms, etc))
This bike does not belong to me but to my colleague from work. Since I'm the most avid e-biker at work, I was asked to repair this bike (even though I don't understand much).
Look at the picture. I opened the wheel and it is in very good condition. It's definitely not the gears that broke.
At the moment I can hardly imagine what is the problem could be. I probably won't be able to repair this motor.
20250224_190016.jpg20250224_190220.jpg
 
since you have it open you can put it back in teh dropouts and fasten it down and connect it back up. then power it up ands pin it and see if you can stop the rotor byhand. if it's a magnet problem then you'll be able to sotp it and see the magnets moving around with the field. if it's a clutch problem you'd have to load down the planets so the clutch is loaded too and see if it makes the noise.
 
Back
Top