noisy vibration BUT already tried 36 hall/phase comb.

Sparfuchs

1 kW
Joined
Nov 22, 2020
Messages
305
Hello EV Friends,

i bought a new 1500W hub motor and connected it to a Kelly KBS48101X. I tried all the 36 possible hall and phase combinations with 120° hall angle and as usual three of them seemed to work fine on the assembly stand but i couldn't tell any difference at those three combs. Neither from the sound and behavior nor from the current shown on the cycle analyst.
When i tried to ride the bike the first time i noticed that the motor runs quite rough, noisy and even vibrates especially at start and slow speed. At higher speed its quieter and smoother. So i tried all the three combs that worked fine with the wheel lifted up but didn't get any better solutions. I also tried the "auto identify" function and switched between torque, speed and balance mode and tried the "quiet" settings but with the same result. I haven't checked the hall sensors of the motor yet but as it's brand new and i also got no hall error i assumed they are fine. Maybe it's worth mentioning that i connected a stand alone Cycle analyst V3 with the external shunt kit and connected the yellow hall sensor wire from the motor in parallel to the CA for the speed reading but have the throttle connected directly to the controller because i want to solve the problem first. As i have a bike with the same motor and a controller from the kit and a bike with a mxus 3k turbo and a sabvoton controller i know how smooth hubmoters usually run.

So i really don't know what else i could try or do to make it run smooth and need your help.

Thanks for your help.

Best regards
Sparfuchs
 
The KBS is a trapezoid controller, right? It will make more noise and vibration at low speed than a sinewave controller just so that you’re not expecting the same as the sabvoton output.. you can try the different throttle control settings to see if it makes a difference (torque throttle, speed, combined). I never used a KBS so i cannot judge what is normal for this controller. What’s the noload current at full speed?

Better check all connectors and give the motor and hall wires a jiggle and pull test when running the motor on a stand to check it’s not somethinge else in the wiring. Check all crimps, connector pins and soldering points.

Then: it could be a hall problem like poor connection, the extra wiring (like CA) causing some distorsion or short circuit, hall placement poor in the motor, one hall sending erratic signals intermittently etc.

Can you try the controller with another motor and the motor with another controller? It would rule out who is the culprit. If that doesn’t make a difference then i’d minimise your system, take away all connections that are not needed and try again.

If that doesn’t work and you have a oscilloscope then you can measure the hall signals to see if they’re looking good. Hall signals can be checked with a dmm also but in my experience you would see a hall error from the controller if it’s functioning poor enough to be noticed with a dmm.

I don’t know about the kbs but kelly kls can be connected with either 5V or 12V to the halls. 12V connection will improve noise on the hall signals. A bit of caution: most halls but not all can take 12V input.

As a last resort, open motor to check halls, wiring inside and isolation.
 
larsb said:
The KBS is a trapezoid controller, right? It will make more noise and vibration at low speed than a sinewave controller just so that you’re not expecting the same as the sabvoton output.. you can try the different throttle control settings to see if it makes a difference (torque throttle, speed, combined). I never used a KBS so i cannot judge what is normal for this controller. What’s the noload current at full speed?

Better check all connectors and give the motor and hall wires a jiggle and pull test when running the motor on a stand to check it’s not somethinge else in the wiring. Check all crimps, connector pins and soldering points.

Then: it could be a hall problem like poor connection, the extra wiring (like CA) causing some distorsion or short circuit, hall placement poor in the motor, one hall sending erratic signals intermittently etc.

Can you try the controller with another motor and the motor with another controller? It would rule out who is the culprit. If that doesn’t make a difference then i’d minimise your system, take away all connections that are not needed and try again.

If that doesn’t work and you have a oscilloscope then you can measure the hall signals to see if they’re looking good. Hall signals can be checked with a dmm also but in my experience you would see a hall error from the controller if it’s functioning poor enough to be noticed with a dmm.

I don’t know about the kbs but kelly kls can be connected with either 5V or 12V to the halls. 12V connection will improve noise on the hall signals. A bit of caution: most halls but not all can take 12V input.

As a last resort, open motor to check halls, wiring inside and isolation.
Thanks a lot for your reply larsb,

unfortunately i don't know what type of controller a kelly is, but i've also used many cheap controllers from conversion kits and they were all way smoother than the kelly in my case. I just can't imagine that a big company like kelly would sell a controller that makes the motor so unacceptable loud. I still think something has to be wrong here.

Noload current at fullspeed was at all of the 3 of 36 combinations that worked 0,57A. And they seemed to work all the same.

Like recommended i've minimized my system and double checked all the connections.

I've also already tried the different throttle modes. Maybe i'll also play around with the throttle voltage ramps to see if it affects the noise.

I checked the halls and made a discovery that seems strange to me. At first i connected the halls from the motor to +5V, GND and Hall A,B,C of the controller. When i measured from gnd to +5V i got 4,96V on the multimeter.. seems fine. But when i measured from gnd to the halls i got 0 or 11,2V depending on the wheel position. I expected a value changing from 0 to about 5V. Why is it 11,2V ?
When i disconnected the Halls but left +5V and gnd connected and checked from gnd to each of the disconnected halls of the motor i got 0V on each. When i did the same check from +5V to the disconnected halls i got a value changing from 0 to about 5V. When i checked from gnd to the hall connections of the controller i got 11,2V on each, but when i did the same check from +5V to the halls of the controller i got -6,25V each.

Then i changed the +5V to the +12v output and measured from gnd to +12V i got 9,7V.. seem to be ok ?
As long as the halls were connected all the values were the same than with the +5V output. But when i disconnected the halls they had a value changing from 0 to 9,7V (makes sense) when i checked from +12V to the halls and -1,46V from +12v to halls of the controller.

So what i got from this is that the halls of the motor seem to work and the change from +5V to +12V doesn't change anything on the value the controller receives. But is there anything suspicious to you ?

What else could i try ?
PS: already bought the same controller again and waiting for delivery.
 
With that low noload current it's hard to imagine something being wrong. Do you reach the expected rpm?

The KBS is a trapezoid controller, you'd need to get a KLS for sinewave.

The voltage measurements depend on the pullup resistors inside the controller so i think they are normal. My KLS controllers have been the same.

I think you should do the controller/motor combination switch with the sabvoton and the mxus turbo and see where that gets you. If the motor runs perfect with another controller then it's to do with the kelly or with motor/controller combination not suitable. If motor runs poorly also with the sabvoton it's likely the motor.
 
larsb said:
With that low noload current it's hard to imagine something being wrong. Do you reach the expected rpm?

The KBS is a trapezoid controller, you'd need to get a KLS for sinewave.

The voltage measurements depend on the pullup resistors inside the controller so i think they are normal. My KLS controllers have been the same.

I think you should do the controller/motor combination switch with the sabvoton and the mxus turbo and see where that gets you. If the motor runs perfect with another controller then it's to do with the kelly or with motor/controller combination not suitable. If motor runs poorly also with the sabvoton it's likely the motor.
But i've heard there are always 3 of the 36 combinations that work good but only 1 of them works best and to find out witch one it is you should note the current. But in my case the current was the same at the 3 good ones.

Unfortunately i can't disassemble my sabvoton/mxus build to check. But i could try the controller from the kit... but i allready know that it will work great because i know the kit and its quite unlikely that the brand motor has a problem. And i would really like to make it work with the kelly because i need the variable regen brake and all the other controllers are way to big.
And i still can't believe that the vibration is normal because then nobody would ever use a kelly kbs on a hub motor and everyone would know that that doesn't work.
 
Sparfuchs said:
i already know that it will work great because i know the kit and its quite unlikely that the brand motor has a problem.
And i still can't believe that the vibration is normal because then nobody would ever use a kelly kbs on a hub motor and everyone would know that that doesn't work.

In most cases controllers and motors match just fine but not in all - therefore you really can’t be sure if your brand new motor is OK. Chinese quality of production and designs.

I’ve had four brand new motors myself with issues. Two with hall sensor issues due to wrong position and two with isolation problems due to damaged windings. Then there are also some motor/controller combinations that just don’t work even if the controller is ok and the motor is ok paired with others. On the occasions that root cause was found (from my memory) it was due to the BEMF of the motor had a weird shape so the motor wasn’t designed in an optimal way. Some controllers could still drive that motor but the controller in question failed to drive it properly.
 
Back
Top