Help! Replaced the controller and now the hub motor is making an odd noise

[oscahie]

Years ago I bought a second hand e-bike in pretty bad mechanical condition for dirt cheap. I did all the repairs and it's been running mostly fine all this time, except for intermittent waterproofing issues, specially during the winter. Unfortunately a few weeks ago the LED display finally died due to corrosion from rain water and it was impossible to find the same model anymore (brand-less, old electronics) so I decided to purchase a new controller and LED display. I found on Amazon a cheapish 250W 24V LED display + controller pack similar to the old one, but allegedly waterproof.

For context it was my first time doing this sort of thing but the task seemed easy. Connected everything together, battery, PAS sensor, brake sensors, phase lines and hall sensors (only had to cut and solder a few connectors that wouldn't match, like the one for the hall sensors) and that was it. With the bike still upside down everything appeared to be working properly. However... once tested on the ground it's evident that things weren't gonna be so easy.

The bike 'works', but every time you start pedaling from a standstill the hub motor makes a loud sort of squealing noise for a few seconds until it gains some speed, then it operates normally. If I apply very little power to the pedals to force the hub motor to do all the work, or start the pedaling on a slope, sometimes the hub motor seems to stall entirely with an even louder but shorter tone, and then once the bike has some speed it operates again. A video is probably better than a thousand words, so here you have two:

(on this one you can see the stalling at the very beginning)

I'm not an expert, hence why I'd like to ask for your opinions, but my theory is that the hub motor is not getting enough current to function normally. In fact this issue doesn't show up on the assistance level 1 (lowest) but it does on level 2 and specially so on level 3, i.e. when more power is demanded from the hub motor.

What I've done so far:

• Checked all the wires that supply electricity, from the battery to the controller and the 3 phase lines, but I couldn't find any evidence of a bad connection. Before I did the controller switch some of those connectors were slightly dirty and/or rusty and I used a fine file on them to get them to shine again.
• I've tried the self-calibration wire several times but the outcomes seem limited to 2 states: if I plug them to each other once the hub motor makes a noise as if it was spinning but the wheel doesn't move, and when I unplug and replug it again the wheel spins correctly. That's it.
• I've tried running with the hall sensors unplugged and the issue persists. I'd say it's somewhat worse even (the stalling happens more often) in sensorless mode.
• Contacted the Chinese vendor over Amazon to ask for help but besides offering a partial refund they have not been any helpful.

Could this be a case of a bogus controller, or simply one that is incompatible with my bike's hub motor? I'm really running out of ideas of what to try, besides returning the thing to Amazon (but I did modify a few plugs so...) and buying a different one.

For additional context, here's a pic of the old and the new controller.

 

[vfr400]

Firstly, be very careful about running your motor when it makes noise like that. Only give it minimum power otherwise there's a good chance that you'll blow the mosfets in the controller.

Firstly, if you have the 9 pin connector near your motor, make sure its in all the way to the line, not just in tight.

You mention the self-learning wires, but I didn't see any in the link you gave us, so I'm going to assume that it has them. You need to do the self-learning procedure. With everything switched off, connect the two wires. Lift the wheel off the ground and switch on. After a few seconds the motor should start to turn slowly. Switch off, disconnect the wiresand leave them disconnected. If the motor whirrs but doesn't turn the wheel, it's going backwards, so you switch off and repeat the prodedure.

If those two measures don't fix it, look carefully at the motor phase wire bullet connectors. Maybe they're not in properly or they're not gripping tight.

It runs without the hall sensors, so you have a dual mode controller. I'd say that it could be running sensorless whether you have the hall sensors connected or not, so there could be a problem with your hall sensors or their connection. You check the hall sensors with everything switched on. Measure the voltage between the black and each of the green, blur and yellow hall wires, while you rotate the wheel slowly backwards by hand. You should see 5v switching on and off several times for each rotation.

 

[oscahie]

On the self-learning, the controller has the wires named as "intelligent identification" on one of the pictures in Amazon. Anyway, I've done that procedure, though always with the power already on before connecting or disconnecting the wires, I don't know if that makes any difference.

I think I'm gonna give it a go at replacing the bullet connectors with brand new ones, to be safe. Perhaps even the hall sensor connector too, since I re-used the one from the old controller and it is definitely not in the best shape. I'll give it a go with the multimeter too, though I'd expect the hall sensors to be working, since the hub motor was running fine with the prev controller.

 

[vfr400]

On the self-learning, the controller has the wires named as "intelligent identification" on one of the pictures in Amazon. Anyway, I've done that procedure, though always with the power already on before connecting or disconnecting the wires, I don't know if that makes any difference.

I think I'm gonna give it a go at replacing the bullet connectors with brand new ones, to be safe. Perhaps even the hall sensor connector too, since I re-used the one from the old controller and it is definitely not in the best shape. I'll give it a go with the multimeter too, though I'd expect the hall sensors to be working, since the hub motor was running fine with the prev controller.

You have to do it right, not how you want.

 

[Nealh]

As vfr has already said. Connect learn wires, then turn on. Disconnect them when it runs correctly after turning off first.

 

[oscahie]

Ok, I've tried to perform the procedure the way you guys say but the outcome doesn't seem to what you'd expect.

With everything off:
1) Connect the self-learning wire
2) Switch power on
- the wheel spins correctly after a few secs
3) Switch power off and disconnect the self-learning wire
4) Switch on again and test by pressing the 6 km/h button
- the wheel does NOT spin, the hub motor is now going backwards.

No matter how many times I repeat those steps, the result is always the same time and time again. The only way to break this sort of loop is by disconnecting and reconnecting the self-learning wire with the power on, i.e.:

With everything off:
1) Connect the self-learning wire
2) Switch power on
- the wheel spins correctly after a few secs
3) disconnect and reconnect the self-learning wire
- the hub motor is now going backwards
4) Switch power off and disconnect the self-learning wire
5) Switch on again and test by pressing the 6 km/h button
- the wheel spins correctly now

Isn't this like the exact opposite of how it's supposed to work?

However if I perform the self-learning procedure the way I did it before posting here, i.e. without switching the power off, then the outcome makes more sense:

With everything off:
1) Switch power on
2) Connect the self-learning wire
- the hub motor is now going backwards
3) disconnect and reconnect the self-learning wire
- the wheel spins correctly after a few secs
4) disconnect the self-learning wire
5) test by pressing the 6 km/h button
- the wheel spins correctly and it keeps doing so after power cycles.

What do you make of this?

 

[oscahie]

Well well well... the hub motor is now working properly! The 'problem' is I fixed two things before giving it a quick test outside, so I don't yet know which of the two did the trick:

1) Replaced the old hall sensor plug with a big chunky electrical terminal block as a temporary workaround (I'll be buying a proper connector soon obviously). This also allowed me to easily test the hall sensors with the multimeter, and they all seem to work OK.
2) Noticed that the bullet connector of one of the phase wires was not a very tight fit. I could easily spin the male side with my fingers while plugged in, which wasn't the case with the other 2. So as yet another quick hack I wrapped a bit of aluminum foil on the head and plugged it in again, now much tighter.

Then I went outside (keep in mind it's over 1am here...) and bingo, the odd hub motor noise sure was gone! I'll have to do proper testing tomorrow and also some troubleshooting to figure out which of the above was the actual fix, plus hopefully find some spare time to buy and install proper connectors

Thanks for the help! I'm still curious to hear your thoughts about the odd self-learning results from my previous post though.

 

[Nealh]

Likely the phase bullet with intermittent sparking/contact, replace bullets. solder directly or try a MT60 tri-connector.

 

[oscahie]

Turns out the culprit was actually the connector of the hall sensors! First thing I did is remove the aluminum foil and the bike was still working fine. But the story didn't end there... I replaced the bullet connectors for the phase lines and then replaced the temporary terminal block I used the previous night for the hall wires with a PCB-style 6 pin plug with soldered wires (because I couldn't find anything better in the shop, so this will have to do until I can get something better). Tested the bike again and the problem was back

Long story short, one of the wires of that new plug had been crimped at the factory directly over the plastic sleeve, and so there was no continuity on the pin for the blue signal wire. The rest were working OK but it took a while to find out what the problem was, you just don't expect that sort of thing... Anyway, I fixed that and now we're back in business

Seeing is believing

 

[Mtrinh1]

Im having that same whirring noise right now after i replaced the controller, can it be as simple as the hall wires? Or the self learning wire?

 

[vfr400]

Whirring or grinding/rasping noise?

Whirring is because the motor is spinning backwards (self learning sorts that), because the gears are stripped (needs a new clutch assy) or because the clutch is stuck (needs a new clutch).

 

[dhav]

hi there i am getting the same sound i have done the self learning which has not worked and i have checked all the wires my hub motor has no hall sensor but the new controller has
i get the sound on high pas and when going up hills and slops

 

[vfr400]

hi there i am getting the same sound i have done the self learning which has not worked and i have checked all the wires my hub motor has no hall sensor but the new controller has
i get the sound on high pas and when going up hills and slops

Start from the beginning: tell us everything and show us what you have.

 

[dhav]

hi so i am getting the same sound as above i have done the self learning
by old ebike had a crack which was a e+plus
so took the hub motor and battery of it
bought a controller kit from amazon and bought a folding bike
put altogether
now i am getting the sound above even worse in high assistance there is no hall sensor on the front wheel hub motor

 

[vfr400]

We have the background now. Forget about anything that has been said before and tell us what your problem is. Pleas give plenty of details about it. Describe the sound, how it happens and when it doesn't happen.

Which LCD do you have?

 

[dhav]

its hard to say what sound it is but is the same sound as the post video
makes the sound as soon as the power motor kicks in worse on high assistance and going up hills
sending photo of the display is jp-lcd1

 

[matze]

Hi guys,

having the same problem here and in German ebike forum they can not help.

After replacing broken controller/display of a Qwic 20" 2010 foldable rear hub motor to a brainpower together with S866 Display, I get the same problem.

I did not connect hall sensors, because cables were cut and motor was running without hall sensors before.

After learning process wheel was turning in right direction. But as others mention, when trying to start from zeroat a hill or trying to start with heavy load ( motor used in cargo bike) on max. PAS level, motor stalls, makes squeezing sound and does not have any torque unless I switch to lower PAS level.

I did not have that problem before. Qwic controller could start from zero on highest PAS with full torque. Now I can only ride very slowly up on a hill.

Until now I tried nothing else but connecting hall wires but no change. I did not do the learning process again, I will do later.

Questions: When I hear the sound and torque is lost, what is happening inside the motor? Is it possible to destroy anything?

I hope you can help Thanks a lot.

 

[vfr400]

Hi guys,

having the same problem here and in German ebike forum they can not help.

After replacing broken controller/display of a Qwic 20" 2010 foldable rear hub motor to a brainpower together with S866 Display, I get the same problem.

I did not connect hall sensors, because cables were cut and motor was running without hall sensors before.

After learning process wheel was turning in right direction. But as others mention, when trying to start from zeroat a hill or trying to start with heavy load ( motor used in cargo bike) on max. PAS level, motor stalls, makes squeezing sound and does not have any torque unless I switch to lower PAS level.

I did not have that problem before. Qwic controller could start from zero on highest PAS with full torque. Now I can only ride very slowly up on a hill.

Until now I tried nothing else but connecting hall wires but no change. I did not do the learning process again, I will do later.

Questions: When I hear the sound and torque is lost, what is happening inside the motor? Is it possible to destroy anything?

I hope you can help Thanks a lot.

You won't damage the motor, but you can damage the controller because the current jumps up.

The problem you're experiencing is when the commutation pulses have incorrect timing. Sensorless controllers have to determine the timing from the back emf. Some controllers are better at it than others. The KU63 and KU65 from BMSbattery have always been good, but they don't have LCDs.

 

[matze]

You won't damage the motor, but you can damage the controller because the current jumps up.

The problem you're experiencing is when the commutation pulses have incorrect timing. Sensorless controllers have to determine the timing from the back emf. Some controllers are better at it than others. The KU63 and KU65 from BMSbattery have always been good, but they don't have LCDs.

Thanks. So I just had luck that old qwic controller worked well without hall sensors connected ?

I just tried to run the motor with hall sensors connected. No change. It still has loss of torque on high PAS and start from zero and making that odd noise.

Is there any chance to resolve that problem by checking hall sensors or using display programming options?

 

[vfr400]

Thanks. So I just had luck that old qwic controller worked well without hall sensors connected ?

I just tried to run the motor with hall sensors connected. No change. It still has loss of torque on high PAS and start from zero and making that odd noise.

Is there any chance to resolve that problem by checking hall sensors or using display programming options?

Before using the hall sensors, you have to do two things: 1. Test that they work; 2. Redo the self learning procedure.