Clutch slipping in geared hub

shs

10 µW
Joined
Jan 19, 2018
Messages
6
Hello forum!

My geared hub motor, front wheel 500W, bought from Ali about a year ago has developed a problem! When I apply the throttle it has a rattling sound and barely moves. Sometimes it comes around after a while and then works for the rest of the trip.

I'm pretty sure it's the freewheeling clutch that is the problem, altough I can't see anything wrong with it?! Everything inside this motor looks like new, which isn't surprising because it has only gone 800 km.

The clutch works as it should when tested by hand but of course it has to handle much more torque when used with the motor...I even took the clutch apart and checked/cleaned everything, lubed with oil and welded back together. Only difference now is the rattling sound is not as harsh, slips just as much! :x

Well, any suggestions on this? Is it a common problem? Should the clutch be un-lubed to work as intended?

20180115_173215.jpg
20180117_165308.jpg
 
Two things;
1)If a Hall sensor fails, the motor tend not to "sync.-up" and sounds and acts like it has a mechanical problem. If I were you, I would test the Halls to make sure one or more is not the problem.
2) Are the rollers held against the ramps by sm. springs? If so, you might want to inspect those springs carefully to make sure one or more has not been damaged.
 
Hey, thanks for the tips!

Just assumed it was something mechanical that was failing, never thought a sensor issue would express itself like this! I will have to read up on how to test those Hall sensors.

Yes, there are springs in there but they all look fine to me.
 
A small update to my thread...

Tested the Hall sensors like this: https://www.ebikes.ca/documents/HallSensorTestingFinal.pdf and they all gave the 5V pulses which should mean they are ok?!

However I don't think it's the clutch either, everything looks good and is cleaned 100% with brake cleaner. It also works perfect by hand.

One new observation: I can make it work by heating the motor up with a hot air gun!! To me that suggests some connection is loose, and the heat makes it connect. Or can the Hall sensor(s) be bad after all, and start working with heat? :?
 
Heat also makes metal expand (different metals at different rates), which means if the clutch is heating enough in the right way it could expand and grab. (or expand in a different way and loosen).
 
shs said:
A small update to my thread...

Tested the Hall sensors like this: https://www.ebikes.ca/documents/HallSensorTestingFinal.pdf and they all gave the 5V pulses which should mean they are ok?!

However I don't think it's the clutch either, everything looks good and is cleaned 100% with brake cleaner. It also works perfect by hand.

One new observation: I can make it work by heating the motor up with a hot air gun!! To me that suggests some connection is loose, and the heat makes it connect. Or can the Hall sensor(s) be bad after all, and start working with heat? :?

If the halls tested OK when it was cold, then it seems like they are not the problem.
It may be possible something else is loose, like the sun gear, the ring gear or even the motor magnets. One test would be to mark the gears with black marker so you can see if there is a shift in position, then make it slip and check the marks again to see if one is no longer lined up.
 
shs said:
A small update to my thread...

Tested the Hall sensors like this: https://www.ebikes.ca/documents/HallSensorTestingFinal.pdf and they all gave the 5V pulses which should mean they are ok?!

However I don't think it's the clutch either, everything looks good and is cleaned 100% with brake cleaner. It also works perfect by hand.

One new observation: I can make it work by heating the motor up with a hot air gun!! To me that suggests some connection is loose, and the heat makes it connect. Or can the Hall sensor(s) be bad after all, and start working with heat? :?

If the wiring uses those white multi-pin, block connectors, sometimes one of the pins backs out when the two halves are pushed together.
 
Ok, some more testing has been done :!:

I ran the motor without any wheel/clutch/gears and the problem is still there, so those can definitely be ruled out now.

Video of how it behaves if you want to check it out: https://youtu.be/164MotbeMUo
I use the hot air gun and get it going at some point in the video.

Had the flywheel off too but can't see anything loose inside:
motor1.jpg

This looks kind of rough but must be from the factory, it hasn't touched the magnets:
motor2.jpg

Pain in the ass to get the flywheel back on :evil:
 
Think I finally located the issue... Where the cable enters the motor axle if I twist enough it gets worse and the motor won't go at all. THEN one of the hall sensors measurement shows strange values. So apparently it was bent too much or the chinese cable wasn't too good to begin with.
 
Good work. Those kind of problems can be very hard to track down. It's kind of a pain to rewire the hall cable, but not too bad. Where the cable bends to pass through the axle is always a difficult spot. If the cable was pulled hard at some point, it will tend to break at the bend. If you have enough cable length, it might be possible to pull some of the hall cable into the motor so you can cut out the bad spot and splice the wires. I've found that rubbing alcohol works well as a lubricant to help slide the wires through tight spots. Later it just dries out and leaves no residue.
 
Gah, still doesn't run :lol:

I cut out a good piece of cable + a connector (because there wasn't enough cable to leave it there). Rerouted the cable through the axle and soldered everything back.

One of the Hall sensors now show 3.3V regardless of the wheel spinning or not. This is the same result as earlier (when I twisted the cable). Can it be anything else than a faulty Hall sensor? I don't really know how they behave when they go bad...
 
About the only other thing would be if you still had a short between wires in the cable somewhere.
With no power, you could try to measure ohms between the bad signal wire and the red wire. It should not be near zero.

If the wire shorted, it could have blown the hall sensor. They are fairly inexpensive if you need to replace it.
 
Back
Top