Magic pie 3 motor won't run(starts and after a second stops)

[MagicJacob]

Hello,
Will do my best to give you all the details.

All started with a faulty throttle that was working half way and was giving 50% of power(back then everything was working)

I purchased another throttle, and because the throttle wasn't from golden motor I had to determine the colors of the wires and to do some soldering. By mistake I probably crossed some wires and damaged the controller. It locked the wheel every time the battery was connected and I wasn't able to connect it to my PC with a USB cable(the software couldn't find it), Plus the the +5V wire in the throttler was showing  a zero reading

So, I have purchased another, new controller, connected it the same way the old controller was connected, used the usb cable to set it(I can read and write information without any problem)

Without connecting the throttle or the control harness I decided to check if the battery-controller-motor connection was working.

Using the picture that is attached to this post, I connected 2 wires and a multimeter  and got a 5V reading, next I crossed the 2 wires, which should have made the motor spin at full speed but nothing happened.

I tried to set my controller differently, mostly changing the  max current settings(not more then my battery could handle, 48V, 20AH, 60A BMS) and the acceleration.

I was able to make the wheel spin with a "burst" of energy for about a second, when my acceleration was set on 100%, the wheel would "kick" and immediately stop.
Putting the acceleration to even 85% would not spin the wheel at all.

Also I have decided to check if the regen brake were working, I set them to 50% and connected the control harness, again with the multimeter got a 5V reading. I connected the brakes and tried using them, but nothing happened and the wheel was moving by hand with no resistance.

I am also attaching a picture of my controller settings so you can see what they are.
*I don't have a paddle asset but I don't see any way to turn it off(only 3 options, Low, Med, High) could this be my problem?

I would really appreciate any help, I am ruining out of diagnostic ideas.

[Bikemad]

I'm pretty sure that your problem has nothing to do with the pedal assist not being fitted, I would suggest that it's much more likely that one or more of the three hall sensors in the motor may have been damaged (just like the original controller) by battery voltage being accidentally supplied to the +5V connection on the controller.

I don't think the MPIII controllers have a sensorless mode like the earlier MPI &II controllers had. In which case, if the sensors are damaged, I wouldn't expect the motor to run.

Unfortunately, you will need to remove the controller to access the hall sensor connections to allow testing to be carried out using a voltmeter as follows:
Be very careful not to allow the controller board to touch against the metal controller housing, spanner or axle etc. while the battery is connected.

1) Connect the Black meter lead to the hall sensor Black wire or (if possible) wedge the meter's Black probe into the battery's negative (-) power outlet connection to leave more room to access the rear of the hall sensor connector for testing the signal voltages with the Red probe.

2) Connect the Red meter lead to the hall sensor Red wire and check for approximately 5V (with the battery turned on).
   
3) Move the Red meter lead to the hall sensor Yellow wire and very slowly move the axle using a 10mm spanner on the flats (with the battery turned on).
    You should notice the meter reading alternate between 0 and 4-5V as the axle is moved very slowly, indicating that the Yellow hall sensor is working correctly.
    This video shows the expected voltmeter readings quite well.

4) Repeat step 3 for both the Blue and Green sensor wires.

Here is a photo showing the relevant connections on the hall sensor plug, but as this plug will be connected to the controller during the tests, you will need to access the individual pins from the rear of the connector plug where the wires exit:



Alan
 

[MagicJacob]

Thank you very much for the  new diagnostic direction!

Will be doing my testing in a few days(got a vacation coming on).
To speed things up, were, online do I get the new hall sensor, and what is the specific type I need?

[Bikemad]

The Honeywell SS41 latching bipolar hall sensor is commonly used as a replacement in many hubmotors and can be purchased from most electrical component suppliers, but if you can't find them at your local supplier, check out this post on Endless-Sphere to purchase hall sensors from Edward Lyen.

Alan
 

[MagicJacob]

Thank you very much for the link, already ordered me 3 sensors.

Going to do the testing this weekend
Wanted to ask, Do I have to connect the controller to the 3 phases or I need to connect just the halls sensors plug?

+is there any test I should run regarding the 3 phases?

[Bikemad]

The phase wires are unlikely to be faulty, but with the controller removed you can easily carry out a very simple check by disconnecting the phase wires from the controller and then shorting together two phase wires at a time and then turning the axle with a 10mm spanner. If the phase wires are OK it should be very stiff to turn the axle with each pair shorted together, and it should turn much more easily with all three phase wires kept apart.
Do this with all three combinations (Green/Yellow, Yellow/Blue and Blue/Green) and check that a stronger force is needed to turn the axle for all three combinations.

To check the hall sensors, it will not matter whether the phase wires are connected to the controller or not, just ensure that the hall sensor plug is connected and that the battery is turned on during the test.

Alan
 

[MagicJacob]

Thanks again for all the help.

I didn't got the chance to check the hall sensors yet(need an extra pair of hands and a multimeter I can trust..will get one this weekend).

What I was able to check is the phases.

I done a few tests.

1.while the controller connected to the hall sensor plug and the 3 phases and the battery disconnected I crossed the 2 wires(black and red) that I use to connect the battery. I felt a resistance while trying to rotate the axle. 

2.while the controller connected to the hall sensor plug and disconnected from the phases(and the battery) I crossed the phases wires with each other(3 pairs in total) and rotated the axle, this time I felt much more resistance and the moment wasn't smooth(like at the first test), this time I felt little "jumps" while I was rotating the axle.
*did the same test and this time disconnected the controller completely from the wheel and got the same results.

3. while the controller completely disconnected from the wheel I did a connectivity test for all the phases( 3 pairs in total) with an old multimeter. Got connectivity for all of them(multimeter was beeping loud and clear).
*did the same test(battery disconnected form the controller), with the controller connected to the hall sensors plug just to be sure and got the same results.


Are the "jumps" I got in test number 2 are normal and is there any more test regarding the phases I should run?
*what about measuring resistance(Ohm's law) in the phases?

[Bikemad]

Are the "jumps" I got in test number 2 are normal and is there any more test regarding the phases I should run?
*what about measuring resistance(Ohm's law) in the phases?

Yes, the jerky movement is perfectly normal as it is producing strong dynamic braking on only one of the three phases, it's a bit like turning over a three cylinder engine which has lots of compression on one cylinder and very little compression on the other two.

If the mechanical resistance feels the same for all three pairs of shorted phase windings there's not much point in checking the electrical resistance with a meter as you have already discovered that all three phase wires have continuity.

Alan
 

[MagicJacob]

Thanks for the confirmation Alan.

Today I was finally able to test(using the instructions in this thread and the video) the hall sensors using a Fluke 115 multimeter 

Test was made while the phases disconnected and the hall sensors plug connected to the controller and the controller connected to my battery.

Black wire and red wire gave me 5V

Black wire and blue wire gave readings of 5.022V-0.018V(and nothing in between)
Black wire and green wire gave readings of 5.017V-0.017V(and nothing in between)
Black wire and yellow wire gave reading of 4.529V-0.018V(and nothing in between)

The readings kept alternating between the maximum minimum readings while I was turning the axle, which should be normal because the hall sensors are not linear as the hall sensor in the throttle.

[Bikemad]

It looks a if the motor is OK as the hall sensors and the phase wires all seem to be working correctly.

Using the picture that is attached to this post, I connected 2 wires and a multimeter  and got a 5V reading, next I crossed the 2 wires, which should have made the motor spin at full speed but nothing happened.

Was the battery still connected while you joined the wires together or was is only turned on after joining the wires?
It is very important that the wires are not joined together before the battery is turned on, or the motor will not run.


Using the above diagram for reference it might be worth trying a few simple tests to confirm that the Throttle, Cruise and Brake signal wires are not shorted to ground somewhere inside the motor harness or on the controller connections, as grounding any of these signal wires could prevent the motor from running.

Connect the negative meter lead to the centre "Ground" contact and measure the voltage on the "Brake Signal" contact. It should read somewhere between 5V and battery voltage, if it reads 0V then something is wrong.

Leave the negative meter lead connected to the centre "Ground" contact and measure the voltage on the "Cruise Signal" contact. Again, this should read somewhere between 5V and battery voltage, if it reads 0V then something is wrong.

Leave the negative meter lead connected to the centre "Ground" contact and measure the voltage on the "Throttle Signal" contact. It should read less than 1V, if it reads more than 1.25v then the throttle test may not work.
Now connect the "+5V" to the "Throttle Signal" and check that they are both reading around 5V. If it reads 0V then I would suspect a short to Ground somewhere on the Throttle Signal wire.

If all three signal wires produce the expected results, I would say that the new controller is faulty (unless you have a battery or BMS problem that is not allowing enough power to run the motor when load is applied).

Here's an alternative throttle test using a 10K potentiometer:


I don't know what else to suggest as I am running out of ideas.

Alan
 

Edit: Incorrect diagrams replaced.

[Mrkrsh]

The Honeywell SS41 latching bipolar hall sensor is commonly used as a replacement in many hubmotors and can be purchased from most electrical component suppliers, but if you can't find them at your local supplier, check out this post on Endless-Sphere to purchase hall sensors from Edward Lyen.

Alan

Hi Alan,  are the Honeywell ss41 halls an more effective halls then The original one found in mp3. I am sure GM uses reliable halls.also. But a lot of riders like ss41. Is their a better hall then even the ss41?

[MagicJacob]

thanks again Alan, I feel that in the end will get my answer

Does it normal that the blue and green max reading is 5V and yellow is only 4.5?
*I do understand that in the end the sensor is giving 2 types of readings to the controller, closed for 5V/4.5V and open for 0V.

Was the battery still connected while you joined the wires together or was is only turned on after joining the wires?
It is very important that the wires are not joined together before the battery is turned on, or the motor will not run.

The wires were not joined before the battery was turned on.


Also I see by the diagram you added that when I was doing my testings(at the very beginning) based on the diagram I added at the first post at this thread , I measured the voltage between the +5V connector  and "not used" connector and after that I connected them. Is this the right way to make the wheel spin at full speed?
*asking because of the "not used" label.


Will do the rest of the voltage test hopefully today and see what I get.

[Bikemad]

Does it normal that the blue and green max reading is 5V and yellow is only 4.5?
*I do understand that in the end the sensor is giving 2 types of readings to the controller, closed for 5V/4.5V and open for 0V.

I'm guessing it's just a slight difference in the manufacturing tolerance of the hall sensor itself, unless there is something on the controller pcb that is putting a higher load on the one sensor. Swapping the yellow and blue leads in the connector should show whether it is the hall sensor or the controller board causing the voltage difference. But make sure you put them back in the correct positions after checking.

Hi Alan,  are the Honeywell ss41 halls an more effective halls then The original one found in mp3. I am sure GM uses reliable halls.also. But a lot of riders like ss41. Is their a better hall then even the ss41?

I don't know the exact differences between the GM hall sensors and the Honeywell ones, but the Honeywell sensors have been proven to do the job very well.
Although the GM sensors are bipolar, I'm not sure whether they actually latch like the Honeywell sensors do, but I don't suppose you would notice any real difference between the two.

Also I see by the diagram you added that when I was doing my testings(at the very beginning) based on the diagram I added at the first post at this thread , I measured the voltage between the +5V connector  and "not used" connector and after that I connected them. Is this the right way to make the wheel spin at full speed?
*asking because of the "not used" label.

Sorry for the confusion, I somehow copied the labels from the plug view instead of the socket view. I have amended the incorrect diagrams in my previous post so they are now hopefully correct. I really should stop posting on the forum late at night when I'm not fully awake!

Alan
 

[MagicJacob]

Thanks for the update Alan

Here are my readings:

voltage between "Brake Signal" and "Ground" is 4.814V

voltage between "Cruise Signal" and "Ground" is 4.815V

voltage between "Throttle Signal" and "Ground" is 0.001V you said it should be less then 1V, but should it be 0V?

voltage between "+5V" and "Ground" is 5.055V

voltage between "Battery+" and "Ground" is 53.52V(same as the voltage coming straight from the battery)

voltage between "Throttle Signal" and "+5V" is 5.050V

[Bikemad]

voltage between "Throttle Signal" and "Ground" is 0.001V you said it should be less then 1V, but should it be 0V?

0.001V seems reasonable (with the throttle unplugged) but I would expect it to read around 0.8V with a throttle connected.

Alan