Smart pie controller broken (mosfets blown). Suspect water damage

[thomab]

Hello,

I earlier made a post about my internal controller dying (see http://goldenmotor.com/SMF/index.php?topic=4604.0).
GM China sent me a replacement controller free of charge, which was great. My bike ran for a couple of weeks, before it died again, with
almost the exact same thing happening. Having experienced this twice, I now have a theory of what caused this, but I need some help.
I do not want to replace another internal controller, just to have it blow up again

So the first time this happens:
1. Bike is left outside in light rain
2. Bike does not respond to throttle
3. After a few days of being left in the garage, the bike works again
4. Next day while riding, one of the mosfets blows up.

The second time this happens:
1. Bike is left outside in light rain.
2. Bike gives throttle on its own for a few seconds, then stops responding to throttle at all.
3. I this time do the "throttle test" (shorting the throttle pins into the controller with a paper clip), this works and the motor is running.
4. I attempt to water proof the connector socket with grease on the connector. I might (not sure) have gotten some grease on a pin of the connector.
5. When I turn on the battery this time, one of the mosfets blows up.

Is it possible to blow up the mosfets if I get some grease inside the connector that goes into the controller?
My gut feeling is that this is unlikely, and I´m not sure if I got grease in there at all. Seeing as my first controller also had similar issues, and also a blown
mosfet a few days later, I suspect that water into the controller might be the real issue (or maybe something with the motor itself, though that seems unlikely).

My friend had a spare external controller, so I am strongly considering wiring that up instead. But I would really like some input on what is causing the problem
before I spend time fixing it again

[Bikemad]

Hi Thomas,

Unless you happen to have used Copaslip or an electrically conductive grease like electricians use which contains Silver or Graphite, I very much doubt that the grease will have been responsible for the failure of the MOSFET.

I think water in the controller could cause a MOSFET to fail if there was enough to make a path between a high voltage track and one of the MOSFETs gate connections, as this could easily result in both the high and low MOSFETs on one of the phases being switched at the same time, resulting in a direct short circuit through the two MOSFETs.

Having said that, I think it could just as easily have been caused by a malfunctioning electrical component on the controller's PCB.

Was there any sign of moisture inside your controller housing or on the surface of the controller's PCB when you removed it?

Alan
 

[thomab]

Hi Alan,

I used regular vaseline, which I believe is not conductive.

When I took apart the first controller, there was signs of water damage. There was visible corrosion on parts of the PCB.

Attaching the replacement controller, I took great care in attempting to make more water resistant. I put vaseline between the shaft and the controller
housing, and also attached the cable with a loop so that water could not slide down the cable and into the controller. I will try to open the replacement controller
this weekend, and see if I see any signs of water damage. If so, I think the best idea for me is to use an external controller.

Thanks for your help!

[thomab]

Hi,

I opened the controller, and Alan you are correct. It was clear that water had come into the controller housing again, with visible corrosion and rust.

On the positive side I have received reply from GM China, and they seem to be very service minded again and it looks like they will ship me a new external controller. It will be some rewiring from my side, but that is definitely worth it.

[GM Canada]

Are you sure the o-ring is positioned on the axle correctly before putting the controller in?

Gary

[dBlues]

Could it be condensed water?
I also had a dead controller and water in controller housing.
I am now intending to drill some small hole(s) to the controller housing to allow condense water to get out of the casing.

Was the power cable hole well enough sealed? Mine seems again a bit light in sealing agent.
Silicone is 'fractured', meaning some water could run through along cable if it reached that fracture and capillar effect would pull it through.
Any drip of water will go straight to the connector socket for the I/O. Mine was soaked, also couple of teaspoons of water visible in the casing

I just read the o-ring comment, my axle is rusty around the o-ring. From the wheel side too.

[GM Canada]

Condensation is definitely a possibility in many cases. I know it is for me. Currently it is around -2c when I go to work and +4c on the return trip. Gotta love living in Canada! I can just imagine the water build up I am creating. Not sure if its an issue in this case though.

Gary

[thomab]

I tried to be careful when attaching the controller for the second time. The controller wire seemed properly sealed, I put some vaseline on the o-ring (which seemed to be properly in place) and axle. Never the less, from looking at the controller, it seems that water has found its way through both by the axle, and the controller wire. You can have a look at the pictures yourselves, and see if you agree.

I'm not sure if it could be condensation.

Anyway, I have decided to move to an external controller now.