Cytronex kit works for two revs, then cuts out for ten seconds

[Slimjim]

I've fitted a preloved Cytronex kit to my mountain bike. (180w motor)

I ordered a new crank sensor and magnet ring, but can't fit the ring yet as it won't work with Shimano MTB bottom bracket. So I'm swipinga single magnet past the pedal sensor at a fast rate.

When I do this, the motor kicks in for approx two revs (clunking a little on/off quickly) then cuts out for approx ten secs, then cuts in for two revs etc etc etc.

Any suggestions? It does this whether I've connected the old or new crank sensor.

The brake level sensor seems to work fine - I.e. if I place a magnet against it the motor never cuts in, take the magnet away and it acts as described above.

Any advice would be greatly appreciated.

Btw, I've had to refit the three hub position magnets as they fit knocked off, and I tell polarity. The motor position sensor is temporarily mounted, but approx 1mm from the passing magnets.

Thanks, Adam

 

[Slimjim]

Forgot to say, bike is mounted in a repair stand, so front wheel is in the air and this is happening with effectively zero load OK the front wheel. I'll try measuring battery voltage when the motor kicks in tomorrow night, in case the cells are weak and voltage is dropping below a minimum limit. Cheers, Adam

 

[Alan Quay]

A bad phase connection will cause this problem, as will a bad hall sensor/wiring.

 

[Slimjim]

Thanks Alan, by hall effect do you mean the red crank sensor, or a hall effect in the motor? I'm guessing the motor. I just couldn't work out why the symptoms are so repeatable, unless the error resets after ten seconds, then trips again.

 

[Alan Quay]

Thanks Alan, by hall effect do you mean the red crank sensor, or a hall effect in the motor? I'm guessing the motor. I just couldn't work out why the symptoms are so repeatable, unless the error resets after ten seconds, then trips again.

A hub motor is a complex, 3 phase device. It will only work when each phase is fired in the correct sequence.

This is performed by a micro controller.

In order to fire the in correct sequence it needs to know the position of the motor. That's what the hall sensors are there for.

Some controllers/motors can operate without the sensors, picking up information from the coils, and making some assumptions about position.

The upshot is that when you gave a bad connection from hall/phase wire a motor often 'kicks' (perhaps 1/3rd turn) then stops.

 

[Nealh]

Disconnect the 3 phase wires check all the wiring and connectors thoroughly for any signs of damage. Sometimes the motor will not work even if all coours are matched there are some 36 wirng combos with the 3 phase wires some will work some won't. Sticking, juddering or wrong way rotation are a sign of incorrect phasing, out of the 36 combos only 3 or 4 will work properly.
Sometimes the clutch mech in the hub sticks so you may have to give the tyre/wheel a clout clockwise and this may free it up if this is the issue.

3 phase wiring combo chart for trouble shooting.

 

[Deleted member 4366]

Hold on. Is your motor sensored or sensorless?

 

[Slimjim]

Thanks Alan and Neal for the advice.

I don't have any technical specs for the motor Dave, however there are only three pins in the motor connector therefore if the controller is in charge of phase voltage control, then it must use the signal from three external magnets glued to the outside of the hub end-face, and detected by the external black sensor - which I assumed originally contained a hall effect as it is three wire.

When i was first given the kit, I misunderstood that this black sensor was for a 'speedo' computer, hence ignored it and didn't notice that the three magnets had been knocked off the hub during my unsuccesful test spins in a vice. I later found two of the magnets and placed them temporarily in their original positions (I.e. within the glue 'wells'), but the third was lost hence for now I have taped on a small magnet that has approximately the same dimensions and strength.

I have no info on the motor or controller so am making guesses about how the system works.

Photo oh hub markings attached.

Cheers

 

[Slimjim]

Disconnect the 3 phase wires check all the wiring and connectors thoroughly for any signs of damage. Sometimes the motor will not work even if all coours are matched there are some 36 wirng combos with the 3 phase wires some will work some won't. Sticking, juddering or wrong way rotation are a sign of incorrect phasing, out of the 36 combos only 3 or 4 will work properly.
Sometimes the clutch mech in the hub sticks so you may have to give the tyre/wheel a clout clockwise and this may free it up if this is the issue.

3 phase wiring combo chart for trouble shooting.
View attachment 15315

Very useful, cheers Neal

 

[Slimjim]

A hub motor is a complex, 3 phase device. It will only work when each phase is fired in the correct sequence.

This is performed by a micro controller.

In order to fire the in correct sequence it needs to know the position of the motor. That's what the hall sensors are there for.

Some controllers/motors can operate without the sensors, picking up information from the coils, and making some assumptions about position.

The upshot is that when you gave a bad connection from hall/phase wire a motor often 'kicks' (perhaps 1/3rd turn) then stops.

Understood Alan. Is the micro processor typically within the hub motor, or in the separate motor controller?

 

[Alan Quay]

Understood Alan. Is the micro processor typically within the hub motor, or in the separate motor controller?

It's all done in the controller, which is usually a separate box.

 

[Slimjim]

It's all done in the controller, which is usually a separate box.

That's what I originally assumed, I'm not sure whether this motor is sensorless, or whether the three magnets on the hub give basic position signal, however the controller can't only the relative angle between the hub, and where the sensor is positioned on the frame so these must be used to give an indication of speed rather than true angular position?

And also the shaft/stator could be at any angular position since the dropout slot angle will vary between forks.

 

[Nealh]

A sensored hub would have 8 or 9 wires.

 

[Slimjim]

A sensored hub would have 8 or 9 wires.

I thought so, based on some wiring diagrams I've seen online. Any ideas what the signal from the three external magnets are used for Neal?

 

[Deleted member 4366]

I would say that you have the orientation of one or more of those magnets incorrect. As there are three magnets (odd number) we can assume that they should all have the same orientation. Check them with another magnet.

 

[Slimjim]

Tried all variations of polarity on both the crank magnets, and the hub magnets today and did various tests of voltage etc, I'll upload a short video if I can. I've stripped the controller down to measure DC voltage at PCB level etc. Basically symptoms still the same - I think at least that either the new hall sensor I bought and mounted next to the hub is faulty, or the circuitry in the controller.

Any idea how to upload/share a short video?

 

[Deleted member 4366]

Upload your video to Youtube, then paste a link.

 

[Slimjim]

Just done on a new thread D8ve, thanks