GMAC 10T - Phaserunner_L10 - V3 Cycle Analyst; Sudden error Phaserunner (blinking), bike no longer works without any obvious cause

[Pandro]

Hello all,

If anyone could help; Please, I would much appreciate any feedback and/or assistance.

I have been riding around on a bike with a GMAC 10T motor in combination with a Phaserunner and V3 Cycle analyst since november 2019.
Last friday all was working well. Nothing out of the ordinary happened. Rode home from work. Parked my bike, but, saturday, it would not start again.

The phaserunner was and is still blinking. 1-3. (Restarted it, same result). Connected the phaserunner to my pc. 3 faults show up in the software.

Cycle analyst still powers up. Throttle nor pas work. All hall sensors do work (nicely light up in sequence when spinning the wheel).
Wheel does not block. Cables are not damaged. No damage to any of the cabling. Again, nothing special happened. Just rode home, 28km from my place of work. No issues during ride. 12 hours later, it would not engage any more.

So, the bike brought me home without issue. Just would not start again when leaving on saturday morning.
Screenshots of Phaserunner software, taken last saturday;

3 error codes:


The dev screen shows large amperages at yellow and green wires.



I checked, the cable is not bent, knicked, or damaged on the exterior.
(Assuming those ~90 amp readouts on the green and yellow in the dev screen of the app might point to a short circuit?! Whut?!)

Please help.

Dutch person here, quite dependent on bicycle for home-work transport. Any help, much appreciated.

Thank you for any assistance or ideas.

Br,

Jasper

 

[amberwolf]

Since the GMAC doesn't have a freewheel/clutch *and* it's geared, if the controller had shorted FETs or there was a direct phase-to-phase short in the motor or phase wires, it would be noticeably harder to turn than usual (moreso than a DD hub in the same situation). If it's not (at least when disconnected from the controller), then most likely the motor and all the wiring up to the controller/motor connector is fine.

If it is the same whether connected or not to the controller, then the controller probably doesn't have a phase short either.

If the high current is from a short, it's probably a soft-short, like damaged winding insulation allowing a short only when higher voltage is applied to or generated by the motor. This you can test for by disconnecting the controller from the motor, and pedalling as fast as you can down the road. The motor will naturally be harder to turn the faster it is spun even when working normally, but if there is such a high-voltage issue, it should somewhat suddenly become a lot harder to turn when you reach a high enough speed.

If it's a controller problem, such as inability to drive the FETs correctly, or one side of a phase bridge damaged, you might have to get hold of Grin Tech to get it repaired or replaced. You can test the FETs for hard shorts with the info at the grin tech ebikes.ca site in the learn - troubleshooting section, but if they are "soft shorts" that only happen as the voltage is applied to them, you'd need a "high voltage" insulation-test type of meter to directly test for that (though other characteristics might show it).

 

[Pandro]

Thank you very much for your response. I will carefully perform the tests you suggested.

First impression just now, might indeed be that the wheel feels harder to spin by hand. Was at first thinking this was because virtual freewheeling was now no longer working. Will try without the controller connected.

 

[Pandro]

I disconnected the controller from battery and motor, got on the bike, and rode it down the street.
It does get very hard to ride when speeding up. As if regenerative braking is active.

Keeping in mind english is not my native language; Could you please entertain me and reiterate;
Did you say that this could be the result of either or both of the copper windings inside the motor and/or the cable being broken? If only one of the two, which one of two cases would you suspect?

Thanks!


EDIT: Post scriptum; It doesn't matter, whichever one it is, I'll have to open the whole thing up again aparently. Thanks.

 

[amberwolf]

It is normal for a motor like the GMAC to be harder to turn the faster you go, and harder to turn than the equivalent DD (direct drive, non-geared) hubmotor, because the force required to turn it is multiplied by the gearing (the same way the force it outputs is multiplied to drive the wheel).

If it's really hard to turn by hand, regardless of speed, that's typically a phase-to-phase short, either in the windings or the motor-to-controller cable.

But if the motor gets hard to turn "suddenly", where it's easy enough up to some point then rapidly becomes hard to turn at some higher speed, then it could be the windings of the motor having damaged insulation (the clear coating on the copper).

Usually that damage is at the point it winds around the corners of the stator teeth (the metal layered laminations), on the inside parts of the coils. That kind of damage is usually started at the factory, with a nick in the insulation, then made worse over time from vibration or water intrusion/corrosion.

You can't see this kind of damage in a typical visual examination, unfortunately. Opening up the motor to check this way probably won't show you anything.

The "best" way to confirm this kind of damage is with an isolation tester, like this:

https://www.amazon.com/UNI-T-Insulation-Resistance-Tester/dp/B0BBG592YG

(I think that's the same as the one I have here, but there's plenty of them out there, some for less cost than this one)


A broken cable is usually open-circuit, which is the opposite of the problem you probably have.


It's possible, though unlikely, that the cable from the motor to the controller has this type of damage between the phase wires inside the cable. If this is the case, it is almost always at the exit of the motor axle, or where it changes direction near there to be routed along the frame to reach the controller, and often this has visible external damage to the cable from an impact or other event.


If it's the cable, you can fix that by replacing the cable, or at least repair the damaged section.

If it's the windings...it depends on the specific failure (which you can't really know). If there is sufficient gap between the damaged winding and the stator (or other winding) that is being shorted to, you can drizzle CoronaDope down into all of the windings and do your best to make sure it gets between all of them and the stator laminations, and let it cure for a few days or more. (CD is used to insulate high voltage wiring/coils/etc and even a very thin layer would fix this kind of problem--I used it to fix a motor with this problem once; it still works though I am not presently using it).


I'd also still recommend checking with ebikes.ca for recommendations, and doing the FET test of the controller, just in case.

 

[Pandro]

Just a follow-up;

@amberwolf; Thank you very much for your time spent on my question.

In the end, it appears my Phaserunner controller indeed had failed mosfets.

Do not know why, or what happened, but, given the amount of distance I have been able to ride it, I do not feel too bad purchasing a new one.

Thanks again for your assistance.

 

[E-HP]

Just a follow-up;

@amberwolf; Thank you very much for your time spent on my question.

In the end, it appears my Phaserunner controller indeed had failed mosfets.

Do not know why, or what happened, but, given the amount of distance I have been able to ride it, I do not feel too bad purchasing a new one.

Thanks again for your assistance.

Doing a little additional investigation might be wise before installing a replacement. If the controller failed due to a phase to phase or phase to ground short in the motor, installing a replacement controller could just lead to a repeat performance.

 

[amberwolf]

Yes; if there is a motor problem, or motor wiring / cable problem, including damage that is not obvious on the outside, it can blow up FETs like this. A poor connection between controller and motors can do it too, from arcing at the bad connection.

The only thing that *usually* causes FET failures (for unmodified controllers running well within their original specifications) is something wrong with the motor or it's cabling to the controller.

Excessive loading for unprotected controllers can also do this, overheating things, but the PR has built in protection to rollback power when it overheats, etc.

For tunable controllers like the PR, you can have failures from incorrect tuning, but for anything that had been working fine and no settings were changed, this is unlikely.

Sometimes FETs just fail....or electrolytic capacitors can fail, so they aren't filtering spikes and noise anymore, and FETs can then be damaged by that. (capacitors age over time, more the hotter they get and the longer they stay that way, so eventually they all die...it can take many years for good ones to go, though).

 

[Pandro]

Thank you. I appreciate all your feedback, I will replace the motor, too, to prevent blowing up my new controller.