LMX Motor - Kelly Controller - Hall Problem?

derb

10 mW
Joined
Oct 30, 2015
Messages
30
Location
Germany - on the Trail
I would like to use a 2018 LMX Motor for my Etricks R01.
As a controller, I have a Kelly KBS 72181E.

The problem is:
The Hall sensors make short pulses before and after the main pulse. As a result, it runs rough and needs too much power (4-5A at 72V at full throttle without load) and fluctuates in the speed while I give constant throttle.

Tested / made so far:
- flawless function test with Infineon 12 FET controller
- All 36 combinations Hall / Phase tried.
- Auto tune
- Diodes in the yellow, green, blue Hall sensor line (tip from the ES Forum works fine: the engine runs better/smoother than before)

Any ideas what I can do?

[youtube]https://www.youtube.com/watch?v=a2RBGpUWKDA[/youtube]


View attachment 12.jpg
 
It's always nice to see how forgiving Infineon controllers are in terms of that. They just work and make no trouble.

Doesn't LMX use same Kelly controller? If yes i would ask them for advice.
I don't know if 4-5A at 72V or 350W is to much, but for a motor this size it should be quite normal. What was the RPM?

What i would do is replacing the sensors with Honeywell SS413A.
If you take a look at the spec sheet of mentioned sensors, you will find a circuit diagram that recommends to use use capacitors near the sensor and pull up (or pull down) resistors. Thats probably only needed for extremly high switching frequencies, but it cannot harm to add those parts.
 
That has to be caused by the magnetic field the sensor is seeing. Something is wrong about the senor placement.
Can you find a picture of how the halls are placed in the motor?
 
Wouldn't hurt to pull the cover off the motor and check the hall mounting. When I got my BHT motor I had to do this to get it settled into the right spot. The mounting of the hall sensors was not exactly rock solid on that motor. They just slid into grooves if I recall. I think they had a dab of silicone to help hold them in place. One of them was slightly crooked. Don't know why one controller would work better than the other if this was the case though. Curious to know what you figure out I just ordered one of them for my wife's bike. :D I know LMX has done due diligence to get these built with good bearings and up to their specs etc. Have you reached out to them? Maybe it is something they have seen before.
 
Have you tried powering the halls off the 12v port on the Kelly instead of the 5v port? I just solved an issue I was having with the GNG motor halls by changing the thin red hall sensor power wire from the motor side from pin 5 (5v source) to pin 11 (12v source) on the kelly. Motor runs smother, hall sensor errors went away. Runs perfect now. Its a very similar motor to the LMX.
 
the motor will run harsh with no load with those hal sensors but will run fine with load. if you want / need smooth spinning with no load change the hall sensors to higher gauss point like 466
 
Hi,

I have a simular problem. First, I bought a Kelly KLS7218S, which I couldn't get to work with my engine. In the end I got it to work by feeding 12v(Pin11) to the halls.
That changed everything, identification angle process now succeed right away, and my scooter drove perfectly.. For 1 day.. A loud noice occurred at full speed, and that was the end of that Kelly controller.

In good faith I took this loss and reordered 2 new Kelly(2 engines on my scooter) controllers, stronger ones then the 7218s, 2x the KLS6035s.
Seller was about 100% sure they would work. And me also, I thought I would do the same trick, putting 12v(pin11 instead of pin5) to the hallwires.

But this time that didn't work. I cannot get it through identification angle process, because the hallwire error, they are not recognised.

So now I am stuck with 3 Kelly controllers, lol..
Yes, now I can smile a bit about it, but I can assure you that wasn't the case when they just arrived and I spent 2 whole nights to get it to work.. After long long waiting for this 2 arrive.

I decided to no longer go on and frustrate myselve with not working Kelly controllers, and ordered controllers from different brand. They worked straight away, so I can drive my scooter again.

But I still have this frustated, uncomfortable feeling in my stomach about the Kelly's in my barn. That I not succeed in making it work. Did I try everything? Maybe I gave up too fast with last 2 Kelly's? What if.. Maybe it is only this.. to make them work? ;)

So here comes the question: Maybe those Kelly's will work if I replace the Hall Sensors in the wheels with the sensors Honeywell SS413A? Like suggested in post 1 of this topic?

So I ordered them silently, a few days ago. They were very cheap less then 1 euro, so I guessed I could miss that, after spending about 700euro's on not working Kelly's ;)

Only thing: When I replace this sensors, my new controllers which are running now, might not work anymore with the new Honeywells SS413A?

Is it possible(I think not, but you never know) to use external hall sensors to make the Kelly's work? That would save me A LOT of work.

Anybody knows?

I will open the wheels
and replace the sensors, if there's a chance it will succeed, because that 1 day of driving with the Kelly did make me a fanboy.
Although their service towards customers is big fat crap, in my country the company wouldn't live for a week with this customer service..

Greetings Tjap
 
Hi..

JP and Sabvoton are working. Here you can see the Kelly settings with the 7218s..
[youtube]https://youtu.be/8WF49v8FRpg[/youtube]
My scooter: [youtube]https://youtu.be/nafboW0hZY4[/youtube]

2 Engines, 2x2800w
60v/42Ah powerpack. Fully charged 67v
 
I have a similar oddity with my lmx motor. My cycle analyst records speed correctly when throttle is engaged but shows speed exactly 3x higher when throttle is released as if there are 3 pulses coming from each sensor. Not sure what causes it. Also with Kelly KBS 72181E
Be glad when my Nuc 12f turns up.
 
brumbrum said:
I have a similar oddity with my lmx motor. My cycle analyst records speed correctly when throttle is engaged but shows speed exactly 3x higher when throttle is released as if there are 3 pulses coming from each sensor. Not sure what causes it. Also with Kelly KBS 72181E
Be glad when my Nuc 12f turns up.

The abnormal hall signals are most likely caused by the rotor laminations between the magnets as you can see on picture below (the triangle shaped part between them).
At this area the fields are cancelling out or fluctuate between north and south (field coming from the back iron not the magnets) which could make the halls "crazy".

file.php


On this similar motor they did not let them go until the end of the rotor probably to solve the problems, at least thats my guess:

Competed%20motor%20rewire_zpssglhyq2d.jpg
 
So I guess it would be worth a try, to open up the wheel, and see if anything could be done. I can always open it up and have a look.. I will open her up, and post some pictures here..
A few months ago I ordered a second bike, the exact same model and engine as the first one, also with 2x 2800w engines.
I intent to keep that one completely stock, except for the headlights. So motorwise, I won't be changing anything on the new one. I like it a lot like this, too :)

I choose to take a very nice battery pack for the new bike, 60v48Ah Panasonic Powerpack, the old bike has a 60/42Ah Powerpack.
Controllers were originally placed on both sides, hidden behind steel covers. But all new controllers I played with, didn't fit in there, so attached them to the bike without cover.. And painted them blackkk ;)
 

Attachments

  • 20190728_013924-1038x979.jpg
    20190728_013924-1038x979.jpg
    136.6 KB · Views: 1,889
Is anybody familiar with this kind of motor? Does anybody know what kind of engine this would be? Please see the pics..
I choose the parts and bought the bike directly from factory in China. It was delivered without much documentation, none to be precisely.
They are very very carefull with charing their company motor secrets, it seems. Now I have 2 bikes and can compare easily, they are loosening up, it seems ;)
 

Attachments

  • 20190728_175645-1612x784.jpg
    20190728_175645-1612x784.jpg
    138.1 KB · Views: 1,884
  • 20190728_175734-1612x784.jpg
    20190728_175734-1612x784.jpg
    168.4 KB · Views: 1,884
I opened up the back wheel. Interesting.. I read 120. That would be the Hall Angle? This grey stuff on the copper, is that grease?

EDIT: I found another topic about it. https://www.endless-sphere.com/forums/viewtopic.php?f=30&t=9061
This might be more difficult then I expected..
 

Attachments

  • 20190728_191900-2016x980.jpg
    20190728_191900-2016x980.jpg
    313.9 KB · Views: 1,877
  • 20190728_191734-980x2016.jpg
    20190728_191734-980x2016.jpg
    153.6 KB · Views: 1,877
  • 20190728_191759-980x2016.jpg
    20190728_191759-980x2016.jpg
    174.1 KB · Views: 1,877
Ok.. I just found this.. Cable is very damaged, you see the inside of the yellow hallwire.

This makes me doubt.. Could this the reason be Kelly has problems identifying the hallwires?

But it works with the other controller though.. But the wheel did react a little strange.. When the wheel is in the air, at my house, at 55km/h I heared a loud bang.. Everytime when wheel is in the air, at high speed/fast spinning.
Maybe I should fix the wire, and try Kelly again ;)
 

Attachments

  • 20190728_203130-1008x490.jpg
    20190728_203130-1008x490.jpg
    66.3 KB · Views: 1,867
  • 20190728_203219-2016x980-1008x490.jpg
    20190728_203219-2016x980-1008x490.jpg
    71 KB · Views: 1,867
  • 20190728_203204-1008x490.jpg
    20190728_203204-1008x490.jpg
    40 KB · Views: 1,867
Major breakthrough!!

I tried one of the Kelly kls 6035s controllers to the front wheel, instead of the backwheel.. I tried both controllers on the back, not the frontwheel, because of the wiring.

So I tried it with the frontwheel, and the frontwheel went through identification angle process straight away, and after about 20 secs the 1 red light started blinking! I knew instantly, it completed that identification angle thing, so set the ignition off and on and... BAM.. 1 green light and the wheel is silently spinning like a charm..

So it must be that damaged cable!! Strange that other controllers are still working with it being damaged.. Only the Kelly not.
So now we know the cause, how do I best fix a difficult brake in the wires, which are strong and tense? The cable is already quit short..
It seems only the big yellow fase wire is damaged, but far from being cut in the middle. Maybe an open wire gives some kind of wrong magneting field?
 

Attachments

  • 20190729_013720-980x2016.jpg
    20190729_013720-980x2016.jpg
    77.9 KB · Views: 1,856
I was lucky.. Lucky the cable was hit on that side, the side with the big fat fasewires. And not the side with the thin hallwires.
Now only the yellow fasewire is just a little damaged.. Allmost not damaged, but a little 'flat'.. This blow on the other side of the cable would probably have cut off 3 or 4 thin hallwires..

So I guess I just need to isolate that yellow wire, just some isolation tape might already do the job?
 

Attachments

  • 20190729_015100-2016x980.jpg
    20190729_015100-2016x980.jpg
    152.8 KB · Views: 1,854
Back
Top