Smart Pie problem

[dBlues]

Story in 5 days:
Slight problems on friday, it was -4C. I came home after week of driving with no problems, recharged (as LiPoFE4 should endure -10C charging afaik) and connected.
After about 500m, the system was completely dead again. Pedalled on about 2km, stopped and went to shop. After I got back, motor was running for 300m and then no power again.
Took the battery indoors for the weekend, opened both sides and dried it. Today evening I tried it again (it's -1C) and everything worked smooth - for a short round of maybe 300m.

During a trip yesterday, again same problem. Maybe got a bit further, motor stopped working shortly after I used regen brake the first time.

Skipped re-charging the battery that evening, decided to try again tomorrow.

Next day: bike worked for maybe 10 meters! Sound of the motor was different, like jittering.

I left battery home, and went without. It's pretty easy to pedal the SmartPie actually, almost didn't notice the difference.
Until motor started to make weird noise and slow me down for few seconds. No battery, but motor is braking every now and then. Cable is tightly squeezed against axle with the metallic strap, so that was not making the sound.

EDIT: Today the motor was slowing me down continuously, if I went more than 9km/h it was really hard to go faster.

Testing
Took the bike indoors for some testing: voltages from battery were 53.3V.
Battery voltage stays constant while twisting the throttle, so I think it can't be the battery this time (if it ever was, need to improve my diagnostics).
Regen brakes don't work at all.
Blow dried connectors for some time, no effect.

Controller software responds ok by USB. Might still be a suspect.

I can't repeat the motor test mentioned in some other thread here, connecting the pins at 12'o clock and next cw, doesn't do anything. Is the pin configuration the same as in Magic Pie III?

Summary
At the moment I have my bike indoors in 20C,
- no controls work or give any indication,
- battery voltage is stable 53,3V when trying to activate controls (not fresh after recharge, but after about 1km ride)
- with or without battery, motor brakes continuously
- no PAS connected
- tested disconnecting all but throttle
- ambient temps +1C today
- battery kept indoors between rides
- bike is stored in dry place when not ridden

Edit: Based on similar symptoms here-> http://goldenmotor.com/SMF/index.php?topic=4604.0 , is it safe to say it's the controller?

I thought today that maybe the connectors are flooded within, need to open them also. If water is activating the regen brake, then the throttle wouldn't work and also no power would be taken from battery.

[Bikemad]

Hi Mikko,

If you turn the wheel by hand can you feel the braking resistance, and if so, is it jerky or smooth?

It might be worth disconnecting the front harness (and therefore all the controls) from the motor and see if the braking effect is still obvious, as this should rule out any water problems in the front harness connections or faulty brake switches etc.

If the strong braking effect is still evident with the main harness disconnected, then the controller will be the most likely cause, and you will need to contact your supplying dealer and explain the problem to them.

- no controls work or give any indication,
- battery voltage is stable 53,3V when trying to activate controls (not fresh after recharge, but after about 1km ride)
- with or without battery, motor brakes continuously

Are your battery gauge LEDs still lighting up? If not, it would indicate a problem with either the battery, it's BMS, the key switch, power connectors or even a poor connection somewhere in the motor harness.

Where exactly are you measuring the battery voltage, from the charging socket or the main output socket?

Are you able to measure the voltage directly from the cable side of the power connectors underneath the rubber sheath on the power cable?

Alan
 

[dBlues]

Thanks for the reply, Bikemad!

I was measuring from main output. Leds show full battery.
Turning the wheel there is braking resistance, jerky like magnets.

I opened the controller cover and discovered that there were couple of teaspoons of water rolling around.

The connector that connects power cable and controller was flooded. The no wonder it dried while stopped, took a few turns and got wet again.

With some luck I'm able to get the wheel back from the beyond, seems that it takes these situations quite well. Controller still responds to usb connection at least.
But I'm even more worried about the water that got into the motor itself, via the cables. It's probably got salt in it too. Probably why it seems so stiff to rotate it.

What is your advice, is it enough to open it up clean it and re-apply some bearing vaseline? I will check what happened.
Any suggestions how to re-seal the power/controller cable entrance?

Braking effect is there without battery. This started when I connected the usb last time, hmm...

[Bikemad]

Mikko, I would expect a small amount of resistance from the magnets, but if you use a 10mm spanner on the flats of the axle to exert some leverage, the axle should rotate relatively easily.

If it is difficult to turn with the spanner, make a note of which way around the three phase wires are connected and then unplug them from the controller.
If it the axle then becomes easy to turn, the problem will definitely be a failed controller.

no wonder it dried while stopped, took a few turns and got wet again.

It's a nice theory, apart from the fact that the controller and the housing does not rotate.

I'm concerned that any water on the surface of the circuit board may have formed an electrical pathway, allowing high voltage to flow between the insulated tracks, subsequently causing damage to some of the more sensitive components.

Any suggestions how to re-seal the power/controller cable entrance?

I would suggest sealing around the cable with some silicone sealant, but this won't be necessary if the controller has already failed.

Alan
 

[dBlues]

 Yes, I have to confess I didn't think about the rotation part (how did I possibly didn't think of this, how stupid!) but still, water might have well gone into connector(s) when accelerating and decelerating.

Mikko, I would expect a small amount of resistance from the magnets, but if you use a 10mm spanner on the flats of the axle to exert some leverage, the axle should rotate relatively easily.

If it is difficult to turn with the spanner, make a note of which way around the three phase wires are connected and then unplug them from the controller.
If it the axle then becomes easy to turn, the problem will definitely be a failed controller.

Thanks Alan again, seems that by disconnecting the first two connectors (connectors above the axle, mine are blue, green, yellow - note to self) the wheel rotates easily by twisting the axle with fingers only. When I plug the wires back, braking effect is there again. No battery connected.

Even though the CPU in controller is alive and USB connection works, the controller seems faulty.

In any case, I'm really glad we had this sorted. Hopefully I can get a replacement from China.

Now I need to seal the wire entrances properly so that no water will get in anymore. Thanks for the tip also for this.

I knew there would be some problems building my fist e-bike, but seems the support staff knows their business here. Can't wait to drive my bike again with Smart Pie
Will send you photos of the beauty too

BR,
Mikko


P.S. I know I messed up with the USB settings, that might be the cause for water leaking to get too much current. Seal was ok before that.
But there were problems before that also.
- Do you think my warranty is still ok or not? Didn't get any reply yet from China.
One thing is that the controller casing is screwed to the motor with 3 screws, and there is some non-sealed wire entry point behind it. Maybe there could be an O-ring to protect this place if water should enter the controller casing.

P.P.S. there is constant beeping sound from controller when power is on, I can only hear it when controller casing is open and I listen closely

[dBlues]

... the "beep" is not the controller's regular beeping when e.g. programming it but a more quiet, high frequency sound

[Bikemad]

As far as I'm aware, there is no audible beeper device fitted on the the Smart Pie's controller
The beeps you hear while programming come from the computer's speaker, not the controller.

The high pitch noise you are hearing is most likely being emitted from one of the solid state voltage regulators which constantly switch the power on and off at a very high frequency to effectively regulate the output voltage for the 5V supply which supplies the hall sensors, throttle and PAS, and a 12V output that the controller uses for switching the MOSFETs.

The USB settings will not have affected the sealing of the cables where they enter the controller.

The hole at the rear of the controller housing (where the phase and hall sensor wires enter from the motor) should not need to be sealed, but if you wish to put some silicone sealant around the wires it shouldn't do any harm.

The water level in the controller housing would need to be pretty high before it would reach the level of the hole and seep into the motor, unless the bike was leaned right over on its side. But don't forget, water should not be inside the controller housing in the first place.

Mikko, it is much better to make a new post rather than add additional comments to an existing post. A new post will cause an automated notification to be emailed to anyone who is subscribed to the original thread, but editing an existing post can easily go unnoticed.

Regarding the warranty, you should send an email to David (wyh@goldenmotor.com) and explain that the controller has stopped working and the wheel is difficult to turn even with the power disconnected, and see what he says.

Alan
 

[dBlues]

At the moment, I don't use a 'leg' in my bike. So it either leans to something or then ground. But it is quite a special case...
So controller is not beeping, was hard to believe but it's true. Laptop was just sitting too close to really be able to tell without paying enough attention

I'll check with David.

Cheers!

 -Mikko

[dBlues]

Solved!

Knock knock, I had no problems today riding to work.

I replaced the controller and made a way for condensed water to exit from the bottom of the controller box.
Put some vaseline into the o-ring and the face of the controller and axle, and the cable entry too, to avoid water getting in and dripping onto the controller itself.

Let's see what happens.

Difficult customer but very helpful staff here, thanks!


Best regards,
dBlues

[Bikemad]

It's good to hear you've got it going again Mikko, let's hope the new controller lasts a lot longer than the first one.



Alan
 

[dBlues]

Well, my Smart Pie is now water proof (absolutely no marks of water going in the controller) but still today:
I have a broken controller again.

It was almost the symptoms same as before, controller works ok for a while, then it needs some battery switching on/off after 300m of driving.
Difference was that due to my very careful precautions done, it really worked past weeks. Lately it always did stop once after specific distance (within 50m) from start, and required to unplug the battery completely and wait for short while to reconnect. After that, it worked all the way.

I did drive for 34km once in -15C.

You can now blame me for driving in minus degrees as that's what I've clearly been doing, but if you want to continue work on that controller, I can go on and test it more. Maybe you have a new version available.

CONCLUSION
Measurements, notes:

- coil side +/- = 4.65V
- 5 V inputs, hall sensors:
   - black/1: 4.3V - 0.01V
   - black/2: 4.7V - 0.01V
   - black/3: 4.7V - 0.01V
- axle is stiff to rotate, as if braking (familiar from last time, was fixed by replacing controller)
- all capacitors get slightly warm
- no water (no rust, dry everywhere)
- broken after power switch on, one hard stutter at switching and then nothing

[Lollandster]

I have put my replacement controller in a dark draw and installed a 3rd party controller on my bike. I believe this is the best way to protect the GM controller. I try to avoid opening the draw and look at the GM controller in fear that if you look at it too much it will break.

I have to give GM some credit, their service is actually very good. It can be hard to get in dialog with them, but when you do they are very helping and sent me a new controller complete cables, screws and heatsink. All plug and play. I also got very good help diagnosing on my first break down.

[dBlues]

Ok, thanks for the tip... I think you might be referring to Lyen controller?

[Lollandster]

Ok, thanks for the tip... I think you might be referring to Lyen controller?

I think most external controllers are going to be more reliable than the internal one, but that is just my opinion. I bought the Lyen controller. The controller from em3ev.com is exactly the same. Kelly controllers are also know to be very good. I have heard good things about the external GM controllers too, but only on this site.

Modifying the motor to accept external controller isn't very hard if you know how to use a soldering iron and/or a crimping tool. The biggest disadvantage with an external controller is that you need to find a place to mount it and the amount of cables more or less doubles.

You can check for shorts on the mosfets as that could explain a couple of your symptoms.

EDIT: I hade to strike through one of my sentences after seeing this post

[dBlues]

Roger mate... yup, soldering is not hard, just got to remember not to put your sandwich anywhere near your soldering station...

Thanks, I think I'll get a Lyen one too.