5KW HPM BLDC Air-cooled Golden Motor runs only with interruption

[Hansbaer18]

Hello togehter,

i have an air-cooled HPM 5KW BLDC Golden Motor (12 coils and 8 magnets) The motor is in a Trialbike (Escape) from Electric Motion.
The motor always "interrupt / stuck" at one point per revolution. It's always exactly the same place. When driving very slowly under load, you can provoke it to such an extent that the motor actually no longer turns at that point.
At higher speeds you can hear and notice a small interruption and a loss of power.
The whole thing happened suddenly and occurs since then constantly.

The following has already been checked and done:

1.) Engine tested in a vehicle of the same construction, no change in the fault pattern. This means that the fault can be restricted to the engine.

2.) Resistance of the coils checked, these were the same for all 3 phases, but very low. Individual coils were not tested, only the 4 combined coils on one phase against the other phases.

3.) Hall sensors voltage tested approx. 10V each, slight fluctuations when tested with an oscilloscope.
Hall sensors exchanged on suspicion.
The error is still identical, no change whatsoever.

Has anyone an idea what could be the failure?
What else could I check?


Thanks and regards
Thomas

[Bikemad]

Hi Thomas andto the forum.

John (Electronic Nut) experienced a very similar problem just over five years ago and it turned out to be the controller.

The new controller and old motor are working fine.  It has never failed to start since I replaced controller.  It has been running the air conditioning for my home for over a week with no problems.  The old controller would never make it over 2 days without failing.  Thank you very much for your help

Did you use the same controller when you tested the motor in a different vehicle?

There is a very good video by Justin (ebikes.ca) showing how to test Hall sensors using a voltmeter.

Alan
 

[Hansbaer18]

Hi Alan,

fist of all thanks for welcome me to this forum and your quick reply.


i have only changed the motor, so controller, throttel and all other electric was all from the other bike. Because this im very sure it must be something on the motor.

But if it woul be a Hallsensor or a damaged wire of a coil the failure would be present not only on time per turn, then it would be four times?


Thomas

[Bikemad]

But if it would be a Hall sensor or a damaged wire of a coil the failure would be present not only one time per turn, then it would be four times?

The same logic should also apply to a faulty controller, but John's replacement controller definitely seemed to cure his problem of one dead spot per revolution.

I think it would be highly unlikely, but the only scenario I can envisage that might cause a single dead spot per revolution would be if one of the eight magnets was weaker than the other seven and one Hall sensor was less sensitive than the other two. 

Incidentally, I'm presuming that you must have previously dismantled the motor to replace the Hall sensors, in which case, I'm surprised that you say it has 12 coils and 8 magnets as I was convinced that the HPM5000 motors had 24 coils (as each magnet appears to be spanning 3 coils) as shown below:



Measuring the Hall sensor voltages for all 3 Hall sensors in all 24 positions over a complete revolution might highlight a slight variation in one of the 24 positions (if there is one).

Alan
 

[Hansbaer18]

ok, I did some measurements. When is turn the engine I have 24 steps wich you can feel proper and you can repeat them also.

Hall sensor voltage is in every "switched on" position between 10,77V an 10,78V. So there is no difference.


So I did this test an found a failure, the red marked "1" shuld be a "0"
(i just copy a second one under the first to make it more visible)

   G   B   Y
1   1   1   0
2   1   0   1
3   1   0   1
4   0   1   1
5   0   1   0
6   1   1   0
7   1   0   0
8   1   0   1
9   0   0   1
10   0   1   1
11   0   1   0
12   1   1   0
13   1   0   0
14   1   0   1
15   0   0   1
16   0   1   1
17   0   1   0
18   1   1   0
19   1   0   0
20   1   0   1
21   0   0   1
22   0   1   1
23   0   1   1
24   1   1   0
1   1   1   0
2   1   0   1
3   1   0   1
4   0   1   1
5   0   1   0
6   1   1   0
7   1   0   0
8   1   0   1
9   0   0   1
10   0   1   1
11   0   1   0
12   1   1   0
13   1   0   0
14   1   0   1
15   0   0   1
16   0   1   1
17   0   1   0
18   1   1   0
19   1   0   0
20   1   0   1
21   0   0   1
22   0   1   1
23   0   1   1
24   1   1   0


i will open the motor tomorrow again.

Regards Thomas

[Bikemad]

As the fault occurred suddenly, I suspect that part of a single magnet may have broken off as it is obviously affecting all three Hall sensors equally.

It will be interesting to to see what you find when you dismantle the motor.

If there is no obvious sign of a missing part from a magnet, you may need to use another magnet to check the magnetic polarity across the face of each of the magnets until you locate the leading 1/3^rd section of the particular magnet that is causing the problem.

Alan
 

[Hansbaer18]

Here we go! I think I got the failure. I is like you thougt. At one Magnet there is a a small spot where the polarity is changed, it is exactly where the sensors are locatet . So this would explane why all hall sensors give at one point a too long signal....

Now is the good question where to get a new magnet... I think I only can get it from goldenmotors.
Do you know anybody there which I could contact directly?

My problem is, that is already need my bike urgent in 1 1/2 Week . 


And thanks a lot Alan for your help, you were exact on the right way.


Thomas

[Bikemad]

Hi Thomas,

I'm glad you've been able to locate the cause, but I don't know whether you will be able to obtain a magnet separately or even a complete rotor assembly on its own.

Getting a replacement magnet in just 1 1/2 weeks is not going to be easy.
If you can get hold of Philip Yao (sales@goldenmotor.com) he may be able to supply a replacement magnet (if they are available separately) but I don't think you will get it from China in time for your deadline.

You might be able to buy a complete motor from a local GM dealer, but I doubt that any of the dealers will keep separate magnets (or even complete rotors) in stock, unless they have a returned motor with a failed stator that still has a good rotor. 

Perhaps you may be able to find a suitable round magnet of a suitable thickness that you could inset and glued into the failed magnet by drilling/grinding a hole in the correct position to trigger the hall sensors correctly. 

I've never actually tried drilling/grinding magnets so I don't know whether this would even be feasible, but it might be worth trying as you have nothing to lose.

Unfortunately, I don't know what else to suggest to help you further. 

Alan