Troubleshooting help needed

[chrisvw]

After >4000 relatively troublefree miles, I'm having a bit of trouble with my bike and thought I'd ask for help before I start replacing parts. Here's what I have and the symptoms:
Crystalyte 5303 with Crystalyte 4840 controller (48V, 40A). 48V LIPO pack.
Problem: Intermittenly, the motor will stop working. I can get about 1 second of thrust, but the motor "growls" and vibrates. It sounds like the normal cogging noise the motor would make but about 10 times louder. The vibration is very noticeable. If I keep on the throttle, it will turn at 4 mph (no load speed) and continue "growling". It does this in any throttle position. The growling noise and vibration goes away the instant I release the throttle.
The intermittent is not "normal". Sometimes I can ride 10 miles with no problem, then the next 10 miles (e.g. on return) it will have the problem. It coincidentally likes to leave me as far from home as possible! Occasionally it will alternate between working and not working, but it usually either just works or doesn't.
The battery voltage is good (~48V), it pulls >5 amps for the 1 second when I blip the throttle and get thrust, then drops to less than 1/2 amp.
I've checked wires, switches and connectors including the hall sensor wires and don't see anything obvious.
I'm suspecting the controller or possibly an issue with the hall sensors but really can't think why either would cause this problem.

My wife's ebike is a 9C with 20A controller. I'd do a controller swap, but the motor connectors are different and there is no hall sensor wire on her controller. I know the 5303 didn't operate when I disconnected the hall sensor wires, but assume that may just be that the controller won't operate without hall sensor input.

Any suggestions are appreciated.
Chris

 

[amberwolf]

My best guess based on my recent problems I caused myself are that you have a bad hall connection on the motor, so it is not always able to read one of the halls, and so isnt' able to time things correctly when this happens, causing the growl.

First just try unplugging and replugging your hall connector a few times, then turn the system on and ride it for a bit on a bumpy road and see what happens. If there is corrosion on a pin it might require several reocnnectons to clean it.

It could be a phase connection too, but I doubt it.

If there is a broken wire or bad solder joint on the hall inside the motor, you'd have to open it up to see that.

BTW, the problem with mine was that I had accidentally swapped two of the hall wires inside the motor (blue and green) but it caused a growl during operation and it ran slower than it should have. While playing around with things troulbeshooting, I found that if I accidentally left a pin loose in the plug (I had popped all the pins out of the socket so I could more easily swap things around) then I also got a growly noise and high current.

 

[chrisvw]

Thanks Amberwolf! The hall wires connector has always seemed to be a weak point of my system. That sounds like a good place to start. I'll let you know what I find.

 

[dogman dan]

Having similar problems lately, I found a bike rack had nicked my wires on a front hub. So my problem was a cut halls wire. But while it was still cut but touching, it did some wierd intermittent cutouts.

 

[chrisvw]

Thanks again for the pointers. I'm pretty sure it's the hall wires, but am having a heck of a time determining where. There's no obvious damage anywhere and no sign of corrosion. I think it's time to pull the wheel and cover to look for problems. I'm not finding anything in the wiring or connector.

 

[amberwolf]

Do you have three LEDs and three 10K resistors? If so you can make a hall tester that will test them at the connector, before pulling the wheel.

On mine I was able to do it without even soldering, by taking each LED's cathode leg and stuffing it down into a hall signal pin on the connector, thru the back of the connector where the wires come out. I left the connectors plugged together, since I need the hall signal from the motor and the power for hte LEDs form the controller.

The anodes of each LED got one end of a resistor lead twisted onto them, then all three resistors' other ends were poked into the 5V pin on the hall connector.

Now with power on but no throttle, spinning the motor by hand will light each LED, sometimes in pairs and sometimes alone. As long as they all blink on and off, they are all working as they should. Sequence doesn't matter since you haven't changed any wire orders from when it was working.

If one doesn't blink, then you have a problem somewhere between the plug and the hall, possibly inside the motor, possibly a bad hall itself.

If none blink maybe the 5V isn't being supplied. That can be tested by putting one of the LEDs' cathodes to the ground wire instead of a hall signal, with the resistor still on it and 5V. If it doesnt' light up, the 5V may not be working, which you can test with a multimeter.

 

[chrisvw]

I have neither LEDs or resistors, but those are easy to get. So the 5V lead is providing the power and the hall leads are providing the sink (low end) that will turn the LED's on. I assume I just check each of the 5 pins with a voltmeter until I figure out which one is the 5V pin. I was thinking I'd just pull the motor apart and start looking around, but I like the idea of actually measuring what's going on to verify I'm looking in the right place.
I'm "grounded" at the moment. I'm bummed - I'm driving in to work today (Oct 25th) for only the second time this year. Way too busy at the moment - gotta find some time to work on this.

 

[amberwolf]

You can use the multimeter to measure the voltages instead of the LEDs; but it is prettier to watch teh blinkies.

If your halls are a 5 or 6 pin JST-SM connector (I think this is common on Crystalyte stuff?) then if they are like the Crystalyte controllers I have here, the one at one edge is 5V, and the one immediately inward from it is ground. The next three are the hall signals. On a couple of infineon controllers with the same connectors, I have found they're red=5v, yellow=ground, green/gray/brown are hall signals.

So measure from either 5V or ground for the black meter wire, and one hall signal with teh red meter wire, and rotate the wheel slowly. You'll see the signal change from close to 0V up to close to 5V and back and forth. Move the red meter lead to teh next hall wire and repeat; then agian ofor the last hall.

 

[chrisvw]

Thanks again Amberwolf. This uses a small round, 5 pin connector, not the connector you specified (I had to search for JST SM connector). Once I find the 5V pin, it should be easy.

 

[el_walto]

I would also guess you have a bad sensor, or sensor wire. One other thing it could be is corrosion built up in your motor, this happened to my X5. See example here: Hall Problem due to Wetness
http://ebikes.ca/troubleshooting.shtml