troubleshooting intermittent sensored gear hub motor

[RVD]

I have a bike with a 350 watt geared hub. I am running 12s lipo (44v) on it.

I do take it up a steep hill sometimes which causes some heating but pedal hard and go up in a zigzag pattern to try to alleviate some of the overheating. This is towards the end of my ride so after about 0.5 miles of riding flat to try to cool things down, I'm home and just put the bike away...motor and controller are usually still hot/warm when I put the bike away.

Recently the bike has been having intermittent periods of failure. When applying throttle it does nothing. Sometimes the motor seems to sputter and jerk and sometimes it does this when I'm not applying throttle. This usually happens in the morning when I leave for work. I have tried unplugging and replugging things in and sometimes that helps. Usually I'll just sort of pedal a lot when it doesn't work and will keep trying the throttle and maybe 1-2 miles in, things will start working and will keep working for the rest of my commute (roughly 8 miles each way).

Any ideas on how to troubleshoot?

Should I check my phase and hall wires and controller?

RVD.

 

[d8veh]

Which motor is it?

 

[dogman dan]

I've had controllers start acting that way, after quite some use. Not a wiring problem, though any intermittent fault is 90% likely to be a wiring issue.

My assumption is that something in the controller stops working when warm, but next day, works again when cold. Weird, because I'm more used to the more obvious kind of failure. Zap-pop, stinks, obvious what the problem is.

So check all that wiring, try a different throttle, etc. No joy? If a new controller fixes it, that was the problem.

 

[RVD]

I'll try a new controller as this one is pretty old (about 4-5 years old or so and it is a lyen 9 fet controller). the motor is a 350 watt geared hub motor from ebikekit from back in 2011 or so. i believe that it's a mxus but not sure. i have an extra 12 fet sensorless controller that i can try out for testing.

i have checked the connections and they seem fine but will try to unplug and replug everything in just to check.

the general behavior is:

works fine
go up big steep hill with lots of pedaling, some motor use, zig zag pattern
motor usually starts to bog as i approach the top of the hill
at top of the hill when road flattens, apply power to motor
motor works on/off but sputters a bit
pedal and periodically try applying power to motor
eventually works and go downhill for about 1/2 mile
get home...feel motor and controller...both are hot but not really hot (i can hold my hand there forever)
next morning when everything is cold, motor responds on/off...sputters...sometimes just plain doesn't work
keep trying to apply power every 20 seconds or so
about 1 mile or so into commute, things start working

i was afraid that maybe i fried something...phase wire, hall wires, etc. but if it's not so obvious, then maybe it's a controller or connection issue like y'all said. thanks!

 

[Ykick]

I don't believe a Lyen controller will need replacing. Maybe repair but not replacement due to age.

You may have iffy Hall Sensors and/or wiring? If lose timing cause the controller to go into overtemp protect mode. I've had the later happen on a Q128 motor and Lyen 9FET.

Wires, connectors and wires - careful checks of everything? Try a different controller to help isolate and if needed Edward will probably repair very reasonablly, if required?

You may have Hall sensors suffering from overheating? They often simply die when a hot motor cooks off but anything's possible at this stage of guesswork.

 

[littleskull99]

I had this happen to me very recently. Intermitent working/not, didnt seem to be any rhyme or reason to it. It would be randomly jerky/fine/nothing. It ended up being the throttle - the wires going to it. I stripped the old throttle and found the green wire only holding on by a single strand. It was the part of the wire that moved whilst steering. The wire had enough slack but over time it must have slowly broken strands. Process of elimination will find it, change out each part till it works again but id start with throttle as its cheap to replace.

 

[amberwolf]

Check connections first, see if pins are backing out, crimps are good, corrosion, etc.

Open controller and check capacitors--if they get hot enough, espeically as they age, they'll swell up and be obviously bad. See badcaps.net or wikipedia capacitor plague for pics (many also here on ES but harder to find).

 

[dogman dan]

Again, 90% likely it's the wiring, including the possibility of faulty throttle, or a short on the ebrakes. I was just saying it could be the controller. 10%.

 

[dnmun]

those tubular crimp bullet connectors can swell as they get hot from the current and that would cause them to lose the tension needed to keep what little contact they have.

solder everything.

 

[motomech]

A 9-FET Lyen and a 12-FET controller are both too big for that motor.
My experience with the MXUS is, 22 Amps is about it.
It might be a connector as I see the bike is a folder and places stress on the wires/connectors.
But, likely your next step will be to test the Hall sensors.
Will either of those controllers run sensorless?

 

[dnmun]

hall sensors would just die and not come back and you would have groaning from missed timing, so i agree that it has to be a connector heating up and losing contact surface pressure.

 

[RVD]

The 12 fet will run sensorless so after I pull everything out and re-connect everything and secure the connections with tape (not solder quite yet), if it still has problems I'll try using the sensorless controller.

The bike is dirty with lots of road dust and such because I usually ride along a gravel bike trail and with the lack of rain in Seattle this summer, it has really done a dust number on the bike.

 

[Ricky_nz]

I have had older controllers start acting up but still working sometimes. I traced it back the the main electrolytic capacitors in the controllers either failing or in one case completely exploded. This can happen and the controller can sort of function. The additional electrical noise can mess things up.

 

[dnmun]

The 12 fet will run sensorless so after I pull everything out and re-connect everything and secure the connections with tape (not solder quite yet), if it still has problems I'll try using the sensorless controller.

The bike is dirty with lots of road dust and such because I usually ride along a gravel bike trail and with the lack of rain in Seattle this summer, it has really done a dust number on the bike.

maybe i wasn't clear. your hall sensors did not fail. if the hall sensors failed it would not run. it would not run sometime and then run at other times. if your hall sensors had failed it would not run at all. i thought that was clear. electrical tape does not make a connection better than one without electrical tape.