Smart Pie needs repair and troubleshooting

[Zeaeroman]

Last week while riding my Smart Pie bike to work it suddenly got all jumpy during starts and accelerations.

Tried reloading software but it has the same problem.  Monitored the voltage from my 48V 10Ah battery with a meter and seen 56V constant during the jumpy operations. 

The rear wheel was purchased direct from GM China using PayPal a few months ago and so far I have sent 4 emails to sales and others listed on this forum with no response yet.

The wheel was quoted with 1 year warranty and I was wondering what would be the best way to proceed.  Like to get some tech advice.

I have meters and very good electronic background.  Suspect a new internal controller under warranty would fix this problem.

[DeVi-Comfort]

Have you driven your bike through the rain right before the problem started?
Then replacing the controller shoudl help.

Good luck!

[Zeaeroman]

The weather was dry, not a fan of riding when its wet.

Got to reading all the helpful posts in the forum and put together a game plan.  Since unable to contact GM China for repair advice (although they did respond very quick when we did the paypal purchase 1/21/2013).  I guess it is about time to just crack open the rear wheel and try to get it running again. 

From what I learned here so far:
1.) I did not hear the two beeps for sensorless mode which usually means the hall's are ok, but will order quality hall replacements for install once inside the wheel.  Reading about the led sequence and voltage pulse information here makes it easy to monitor halls before and after replacements.

2.) Since the bike runs a while OK before getting jumpy as heck, it sounds like a bad factory connection within controller, hall or phase wiring so I will check solder and spade connections carefully and look for any moisture or condensation once in the wheel.  Best to check now before controller blows a mosfet from the jerky currents.

3.) Testing so far my battery seems OK, it was holding above 50V at wheel harness on my fluke taped to handle bars when running and then started jerking.

4.)  Plan to wire up a switch for throttle jumper and see if I can get it to start jerking on a test ride before working inside the wheel.

5.) I plan to put the old back wheel on the bike to support it then move the e wheel and cables to my bench to allow running when all is checked out.  Will order halls and post my results once I get inside the wheel.

If I am missing something or maybe another GM contact is available please advise.  Thanks, Dan


 

[DeVi-Comfort]

In the Smart Pie controller there is no horn, so you can never have heard an error message.

You can check if the hall sensors are the problem by throttling and turning the wheel by hand at the same time.
If the wheel turns, there is a big chance the hall sensors are the problem.
If nothing happens, the controller is probably broken.

[Zeaeroman]

Got the bike running again.

Here is more detail if your interested.  Took ewheel off and put on old wheel to support frame.  Turned off battery kill switch and removed the key.

Put ewheel up on table and removed the outside cover and released circlip for cable.  Took off controller cover and seen some white powder gook caked and loose all around the controller and some fine condensation on the back plate.

Unhooked all controller connections (blue, green and yellow phases and hall 6 pin plug)  and cleaned everything up with alcohol.  Cleaned the tiny pins for hall connector and squished the spade connectors a little to make a tight fit when replugged.  Did not power on.

Like a terrarium, moisture is moving around inside controller and motor from heat cycle. 

Once controller cavity was all cleaned up and rust removed I dabbed some bike lube on the rusty bits and a put layer on shaft.  Butyl o-ring will probably fail from lube but the seal is not working anyway.  The cable opening to the motor in the back was not sealed completely, so added some sealant and let cure overnight.  Put it all back together and bike is running again, will give it a 15 mile test on a ride to work tonight.

Some other things of note.  The flange on the motor side was so tight I did not pursue getting it off at this time.  May have to release spoke pressure and heat case and cool inside to get it off/release without wrecking it.  Also felt with the white gunk removed in controller area might fix the problem.

Next time this thing craps out, will get new oil proof o-ring to replace butyl, also will plan to get into wheel for new halls and general rust cleanup, and also thinking about pressurizing the wheel with a small dry air source to slowly force out moisture while riding.  Thinking about running a small 1/16" tube in with the motor wire and use dry aerosol air can or small pump that cycles on/off from 48V battery.

I think they built this wheel without deciding if it should be sealed or vented, so they ended up with an in between situation where moisture can enter and then is trapped for heat cycle condensation.  Like running a toaster through a car wash.  My guess is there is a small vacuum setup when wheel is spinning that is drawing moist air into the shaft of the wheel past the oring and positive dry pressure would help.

Still no warranty response from golden motor.

[Zeaeroman]

Rode the bike to work and back OK.  All is good again.