Sp4 acting up again, buzzing and vibrating

[Thuktun]

I was riding to work this morning, about halfway here when I stopped at a red light. When I went to go again something seemed off right away, I realize now that was the throttle not doing anything at first, I started moving after a second but with a lot of noise and vibration.

Walked it to work from there, only just now had a chance to go out in the parking lot and investigate further.

The throttle does nothing from a dead stop. No response at all. If I pedal up to about 7km/hr it will cut in but with the aforementioned noise and vibration. It's light at first but the faster I go the harder it shakes and louder it buzzes. It also buzzes and shakes when the brakes are applied. No problem when freewheeling or pedaling. I thought maybe the brakes were causing it since they were the last thing I touched before it went wonky so I disconnected them and the cruise control, same problem.

So what's up? Another bad controller? How can just stopping at a red light mess this thing up?

[Thuktun]

I rode it home because I didn't have much choice.

There are 3 wires from the controller to the motor, I assume this means it gets 3 pulses per revolution. I think what mine is doing is only pulsing once or twice per revolution. The rate of vibration is directly proportional to the speed of rotation and the strength of the vibration directly to the amount of work the motor is doing. If it only pulses on one side it would vibrate and hammer on one side, right?

Same question though, why, and how to fix?

[Bikemad]

It sounds like you may have one of the following:

• a failed Hall Sensor
• a poor phase wire connection
• a poor Hall sensor connection
• a faulty controller

You will need to remove the controller to check that the three phase wires are still properly connected:



Also check that the 6 pin Hall sensor connector is properly fitted and that none of the crimped sockets have been dislodged from the connector housing during assembly:


(NOTE: this is an MPIII connector shown here)

Make sure that the five crimped sockets on the hall sensor leads are fully inserted into the connector block as shown in the above picture, and then make sure the connector is clicked properly into the socket on the controller.

If nothing is visually obvious, you will need to check the operation of the Hall sensors using a voltmeter to check for pulses   of 0~5V on each of the Hall sensor signal wires (Yellow, Green and Blue) as the motor/axle is slowly turned by hand.

If all the connections are good and the sensor voltage readings are also correct, I suspect that the controller itself may be causing the problem.

Incidentally, you should have around 120 pulses per revolution (not 3) as I seem to recall there are approximately forty magnets that will all switch each of the three Hall sensors every revolution.

Alan
 

[Thuktun]

Looks like a hall sensor. All the connections are good. I'm reading 4.25v on the red wire and 3.22v on the blue and green depending where the wheel is. The yellow only shows 0.01v.

So where do I get suitable hall sensors and how do I replace them? If there is a more robust part then the one it comes with I'd prefer to go with that and swap out the 2 good ones if I can.

[Thuktun]

Never mind. Did a search then ordered 3 from that ES link. They don't look difficult to replace either.

[Thuktun]

Everyone is saying Honeywell SS41 in the forums, and I have them ordered but a friend told me he might have some hall sensors. If they aren't SS41s but operate in the same voltage and current range could I use them? I had a look at a few data sheets for generic Chinese hall sensors like you see on eBay, most are the same 4.5-24v and are 25mA max output current. How important are the other parameters like the switch on and off times? I would think if it mattered I would need 3 matching sensors and I doubt a motor made in China has Honeywell parts, it would have generic Chinese sensors so the SS41 is probably not a match for the other 2 and it is recommended. So, besides them being cheap Chinese parts (like those in the motor now lol) is there any reason not to use them if the electrical characteristics are the same as the SS41's? Assuming he can find them....

[Bikemad]

At maximum rpm, each of the hall sensor will be switching high and low ~120 times a second, so the switching on and off time might make a slight difference, but I don't suppose you would notice it.

Alan
 

[Thuktun]

My buddy found a bag of OH137s. Electrical characteristics are the same as the SS41, it's just the switching times are a bit faster. Shag it, I'm putting one in. Halfway there. I'm doing this at work so it will probably take all day, I can only work on it when I'm not busy. From 6 to 9 is dead here but now my phone starts ringing lol.

Had to take out the stator to get at it. Going to ask for help putting it back in, don't want to chip a magnet. Had to chisel the old sensor out with a small screwdriver, was glued in there good. I'm not going to replace all 3 due to the effort, it was tough not seperating the plates for the stator electromagnets while chipping away. I'll take my chances with the timing being off a fraction of a microsecond.

Oh, and there are 40 magnets for anyone who was wondering.

[Thuktun]

Got the new one in, the one with the green heat shrink. Seems to work fine. One pole of a magnet turns it on, the other turns it off. Now to glue it in, tie down the wires, clean all the old glue/silicone off everything, resilicone it and finally reassemble. Hours of work left. Ugh.

[Thuktun]

Got it all back together. Once I finished I thought I maybe should have checked everything one last time before siliconing around the controller and screwing it all together.

Now that I tested it I really wish I had!!! Doesn't work, same problem. Still runs, but not until I'm doing about 7km/hr and it vibrates and reduced power.

When I put the stator in it slammed in spite of my being careful. It went from no force to tug of war in an instant. I'm hoping I just scrunched a wire or something when I did. I gave it a quick visual inspection but I was looking for magnet chips, not wiring issues. That or the hall sensor I used isn't suitable in some way. The SS41s will be here in a few days I hope anyway.

[Thuktun]

When I tested it it was with the motor apart and holding a magnet in my hand, I really should ha e tested it with the stator in place and the rotor magnets.  That's how you learn, by screwing up.

[Thuktun]

Opened the controller, getting a signal from the three hall sensors so that part is working.

[Bikemad]

Are you sure that the phase wires were refitted in the correct order?

Look carefully on the controller PCB and see if you can see the colours printed for each of the phase wires.

Alan
 
This might help:

[Thuktun]

Yes, took pics when I was taking it apart to be sure.

Should I be seeing 2 on at a time? I have the controller open now and am moving it a bit then checking and occasionally the yellow will be on with the green or blue, I haven't seen the blue and green on together. Yellow is the one I replaced.

I'm going home for the day now, tomorrow I'm going to open the other side so I can see the magnets and move it one at a time writing down which is on or off, see if there is a sequence or just a jumble of signals. Kinda hard to do that when I'm guessing where they are lol.

[Bikemad]

Perhaps the Hall sensor you have used works with opposite magnetic polarity? If so, it might be closed when it should be open (and vice versa). 

Alan