Crystalyte X5304 Hall Sensors Shorting?

[TheThirdWheel]

Hello all!

I have relied on the endless-sphere forums for many years now every time something goes wrong with one of my electric bikes or trikes (usually BMS problems), but now I seem to have a problem I cannot find a definitive answer to.

Last Thursday night my Crystalyte X5304 on my tricycle-sidecar frame just stopped working when we were ~28Km from home. It wasn't much fun getting home on pedal power but we made it in the end. Anyway, a few weeks prior a similar thing happened where the hall sensor on the blue wire failed and left us stranded 20Km from home.
While fiddling with the bare hall wires to twist the blue wire connections together (in case the fault was in the plug), I got a fairly nasty shock while touching the hall wire and the motor casing at the same time. I'm guessing it was ~80V pack voltage!
After testing the hall signals I opened up the motor to replace the sensor and WOW, rust galore! 2 days later after rust cleaning I replaced the faulty sensor, sealed it in, painted over the entire stator and rotor with rust "proofing" non-metallic paint, and put it all back together.
After that the motor ran just fine, no cogging, no sign of overheating, and we took it on a 24Km test run at mostly full throttle just to see if it would fail again, but it passed with flying colours.

Of course as Murphy's Law would have it, the motor stopped again last Thursday. Initial testing suggested the recently replaced sensor was the only one that survived. I opened it up yesterday and there were scorch marks around the hall sensors and wires, the negative lead to the yellow wire sensor had blown apart completely, and the green wire had been partially exposed. Upon reconnecting the black wire the sensor returns a funny value on the multimeter. Instead of being 0V or 5V, it sits around 2.4V, the green wire sensor is toast since it stays on 5V continuously.

I suspect somehow battery voltage is reaching the hall sensors, but have no idea how. Continuity testing between the hall wires and phase wires (while the motor is disconnected from the controller and even removed from the frame) shows no connection between phase and hall sensors in any configuration.
All the phase wires show continuity to each other but I'm assuming that's normal since they all connect together in the wye.

I could just replace the other 2 sensors, but now I've asked myself why they are getting burned out and would rather solve the problem than have it happen again.

All suggestions are welcome!

Cheers!

Kev

 

[amberwolf]

A short between the phase and halls within the cabling, most likely inside the bundle exiting the motor (or inside the motor along the axle), is probably causing it. Similar things have happened before.


(P.S.: if you have posted here before, you should re-use your existing membership instead of creating a new login. If you can't access it for some reason we can fix that for you, and reassign this thread to it, and remove the duplicate membership, since multiple logins/accounts for one person aren't allowed.)

 

[John in CR]

Hall problems are commonly caused by problems in the wiring harness rather than simply failed sensors, and quite often people go to the trouble of changing halls only to have it happen again. The only time I had a hall problem it was intermittent and progressively worse, and it got worse as the motor warmed up. Inspection of the hall wires showed nothing, and a continuity check showed no problem, but that was always when it was cool. It turned out the hall wire insulation had become slightly deformed and fragile from heat, and they would start to short somewhere along the harness once things heated up.

Hall wires are often overlooked when we're putting premium wiring in our motors, but despite not carrying high current they see about the same heat that phase wires do, especially as they pass through the axle or bearing. Now any time I do a rework of a motor, the hall wires get swapped out too as I upgrade the phase wires.

Your shock is a significant additional concern, because you have a short somewhere. I have extra parts, so I'd change out all wiring including the controller unless I could find a visible short such as battery positive to the frame. All of my bikes run voltages that can be lethal, so it's not like a hard to track down problem in the 12V wiring of a car whose only risk is blowing a fuse or a light not working. Some might say that my bikes go like lightning, but I definitely don't want to "ride the lightning".

 

[TheThirdWheel]

Thanks for the input guys!

A short between the phase and halls within the cabling, most likely inside the bundle exiting the motor (or inside the motor along the axle), is probably causing it. Similar things have happened before.

(P.S.: if you have posted here before, you should re-use your existing membership instead of creating a new login. If you can't access it for some reason we can fix that for you, and reassign this thread to it, and remove the duplicate membership, since multiple logins/accounts for one person aren't allowed.)

That was the first thing I suspected since I've read all about similar happenings here on the forums, but there seems to be no continuity between hall and phase according to the multimeter. I doubt the problem lies within the controller since it has always been run well within specifications and looks pretty good inside. Looks can be deceiving though.

Oh when I say I've relied on these forums for years, I mean only as a visitor. Up until now I've always managed to find the answers I need without having to create an account and present the problem myself e.g. your documentation of that "CammyCC" eBay duct tape battery repair guided me to dismantle my own battery, locate, and replace the offending cells. Thanks for that btw

Hall problems are commonly caused by problems in the wiring harness rather than simply failed sensors, and quite often people go to the trouble of changing halls only to have it happen again. The only time I had a hall problem it was intermittent and progressively worse, and it got worse as the motor warmed up. Inspection of the hall wires showed nothing, and a continuity check showed no problem, but that was always when it was cool. It turned out the hall wire insulation had become slightly deformed and fragile from heat, and they would start to short somewhere along the harness once things heated up.

Hall wires are often overlooked when we're putting premium wiring in our motors, but despite not carrying high current they see about the same heat that phase wires do, especially as they pass through the axle or bearing. Now any time I do a rework of a motor, the hall wires get swapped out too as I upgrade the phase wires.

Your shock is a significant additional concern, because you have a short somewhere. I have extra parts, so I'd change out all wiring including the controller unless I could find a visible short such as battery positive to the frame. All of my bikes run voltages that can be lethal, so it's not like a hard to track down problem in the 12V wiring of a car whose only risk is blowing a fuse or a light not working. Some might say that my bikes go like lightning, but I definitely don't want to "ride the lightning".

Now that has got me thinking. Perhaps I have the same problem you once did? The problem does seem to occur after ~1 hour of cruising at 60% throttle. I shall strip off all the heatshrink around the phase and hall wires, and go about replacing those hall wires, and perhaps even the phase wires while I'm at it (assuming I can find where they connect under the coils).
The only potential problem I see is getting thicker hall wires to fit under the coils. Yes, the hall wires go UNDERNEATH the coils to the sensors (unlike the 9C motors). I'm guessing this is due to the clearance between stator and rotor being about 1mm in this motor.

As for the shock, anyone who got shocked only experienced it while touching the bare hall wire belonging to the dead hall sensor, while touching the frame. The battery isn't earthed to the frame, however the controller is, therefore the MOSFET heatsink and thus the source pin of the MOSFET's.

Should I be getting continuity from the phase wires to the stator/axle though? It's not great continuity, about 10MOhms, but when I compare it to the hub motor on another bicycle, I'm assuming it should be complete open circuit.

Again, thanks for the input!

 

[TheThirdWheel]

Fairly sure I found the problem!
It seems that the white goo Crystalyte uses to hold the phase wires and hall wires inside the axle had let go. The hall wires sprung upwards, making contact with the motor side cover, and of course the side cover spinning with the motor slowly wore away through the green and black hall wire insulation and eventually caused a short between them.
There is evidence of this by a black scorch ring with a clean line right down the middle, and some melted aluminium splatter in line with the ring. This sits 72mm from the centre of the bearing. I measured 72mm from the axle towards the hall sensors on the stator and it lines up with the break in the wires.
I'll replace the hall sensors and make sure to tie down those hall wires so they go nowhere near the side cover!