Need a replacement controller for Bafang front hub

Irae

1 mW
Joined
Jun 8, 2020
Messages
12
Location
NOVA, USA
Hello Sphereists,

I'm about at the end of my rope. My son and I built him an electric kickbike a couple of years ago using a Bafang FR G020.500.D front hub in the rear. We run a P850c display, and run a 52v battery. Everything worked great until I decided to try to internally route the 1-to-4 cable and pinched it between the deck and the frame. Must have shorted something out, as the display showed a 08H error (hall sensor, but I think it's also a general 'motor not found' error).

My first troubleshooting step was to replace the motor extension cable I'd damaged, but I got the same result. Panicking that I'd damaged the motor, I bought a replacement and swapped the internals into the 20" wheel. Same result. So it must be the controller, right?

Here are the specs of the controller that came with our kit:
Voltage: DC48V
Undervoltage: 39+/-1V
Brake Type: P-
Type: YCSV126-017-48T
Motor Phase Angle: 120deg
Current Limiting: 22+/-1A

I've now been through several replacement controllers, but it seems that they either don't support 52v, only support a KT display, or just refuse to talk to the motor. This was the case with my most recent purchase after I spent quite a lot of time discussing my needs with the vendor. He detuned a 750w controller from 25 to 22A peak, and replaced the 3-pin connector with the 9-pin that matches my motor. I was sure that would do it, but, alas, when I connect it all up the display shows no errors, I blip the throttle, the wheel moves a few mm, and then stops with 08H again.

It seems like my options are to ditch the P850c and get a KT display and controller, or find a new one that's compatible with my motor and display and battery. I'd very much prefer the latter, as I don't really care for the KT displays. If anyone has a lead on a controller that will work I'd be enormously grateful!
 
Irae said:
My son and I built him an electric kickbike a couple of years ago using a Bafang FR G020.500.D front hub in the rear. We run a P850c display, and run a 52v battery. Everything worked great until I decided to try to internally route the 1-to-4 cable and pinched it between the deck and the frame. Must have shorted something out, as the display showed a 08H error (hall sensor, but I think it's also a general 'motor not found' error).

My first troubleshooting step was to replace the motor extension cable I'd damaged, but I got the same result. Panicking that I'd damaged the motor, I bought a replacement and swapped the internals into the 20" wheel. Same result. So it must be the controller, right?

If the entire motor core including it's cable has been changed but still have same results, it's probably not a damaged motor or cable or hall sensor as those all are different with the new core.

Otherwise I would normally say in this situation that the cable pinch could've damaged a hall sensor.


I've now been through several replacement controllers, but it seems that they either don't support 52v, only support a KT display, or just refuse to talk to the motor.
It may help you find a solution to know how some fo this works:

FWIW, for that motor, there is no communication as such between motor and controller.

The only things that happens are:

--the hall sensors in the motor detect where the rotor (magnets) are in relation to the stator (windings), and their outputs change in a detectable pattern that the (any sensored) controller sees on it's hall inputs.

--the controller then sends phase signals on the larger phase wires in the motor cable to power the windings in a pattern that makes the motor spin in the right direction, which changes the hall signals which makes the controller send the next part of the pattern and continues the spin, as needed and as commanded by throttle/etc.

If the controller can't read the hall signals, or they are damaged or miswired between controller and motor so it gets the wrong pattern, it won't send the right signals and the motor won't spin correctly.




This was the case with my most recent purchase after I spent quite a lot of time discussing my needs with the vendor. He detuned a 750w controller from 25 to 22A peak, and replaced the 3-pin connector with the 9-pin that matches my motor. I was sure that would do it, but, alas, when I connect it all up the display shows no errors, I blip the throttle, the wheel moves a few mm, and then stops with 08H again.

Do you have a link to the specific sale page for that controller? It may have info on it that will help us figure out what it requires.

If the controller that was used only had 3 pins to go to the motor, then it should mean it is a sensorless controller, and so there is no feedback (halls) from the motor. This means that it will work with any motor at all, it just sends phase signals to the motor until the motor spins in a direction. If that is the wrong direction, you swap any two of these three wires to spin the right direction. (or use the self-learn function if it has one, or the reverse function if it has one).

But if the controller expects hall signals, and does not have them wired up to the motor, it can't operate the motor. It would require connecting the five hall sensor wires (power, ground, three signals) from motor to controller to make it work. If they are not connected, you could get an error from the controller that it can't read the signals.

Assuming it is just a sensorless controller, then if the controller won't spin the motor, it means there is a problem in the controller itself, or there is damage to the motor or it's cable preventing it from operating (usually a short between two or more phases, but you can feel that as extreme resistance when spinning your geared hubmotor in reverse--forward it will always feel easy as it has a freewheeling clutch inside to make that happen).

The problem could even be an open-circuit between controller and motor, anywhere between the controller and the motor. Could be broken wire anywhere along there, or a bad contact connection in the plug, or a contact not properly crimped or soldered to the wire it's for. Testing for this is easy if you have access to the inside of the controller, as a continuity test should read a short on every phase to every other phase with the motor plugged in. If no access to inside of controller, can test motor and motor cable the same way, and should read the same thing. Testing controller would be the same as testing FETs via phase wires; if you get the same readings from every phase to every other phase then all the wires are good.
 
Thanks so much for the help.

This is the controller I bought: https://www.aliexpress.us/item/3256804429801507.html Looking more closely it appears the Bafang fat hub motor connector still has phase pins, but three central pins vs the single pin for the 500w hub. I wonder if the pinout for the halls varies between the motors.
 
Irae said:
This is the controller I bought: https://www.aliexpress.us/item/3256804429801507.html
Looks like a typical sensored controller, and if so, it would need to have the five hall wires connected as well as the three phase wires.

As long as power and ground are in the right place, then if the controller has a self-learn function, it will figure out which order to use the phases and halls in. If power and ground for the halls isn't right, then the halls won't operate and the system won't run.


This image is the motor connector from the controller sale page,
https://ae01.alicdn.com/kf/Sbecfb8fbb7144f4e8df9d1f092e8679f2.png
Sbecfb8fbb7144f4e8df9d1f092e8679f2[1].jpg
this page
https://ebikes.ca/learn/connectors.html
shows one that looks like it that is the Julet Z916
This octagonal style motor plug from Julet has higher gauge wire and larger pins than the Z910 connector, but less current capability than the L1019. It is quite popular for generic mediium power ebikes and kits on the market, including many motors in the Rad Power bike lineup. The connector has 3 phase wires in the middle and 6 signal lines around the perimeter, allowing for an additional motor speed sensor but not a separate speed sensor and temperature sensor like the L1019 plug.
so it would support hall sensor wires, so the controller itself probably does too.


But at the moment, you've got to trust that the seller wired the new connector correctly--this is not guaranteed.

If you're up for it, you can test whether it is wired right or not, by opening the controller and doing a continuity test between each pin on the connector and the wires from that cable to the points on the board each goes to, while either drawing it out on paper and labelling it, or labelling pictures of the connector and inside of controller, etc.

Note that you may also need to verify the wiring of the original plug on the motor, to compare it to, but this can be checked using the orignal controller rather than opening up the motor again (or using the old motor core that should be wired the same).



Looking more closely it appears the Bafang fat hub motor connector still has phase pins, but three central pins vs the single pin for the 500w hub.
I don't really understand what that means. The hubmotor has to have three separate phase wire pins (as does the controller). If there's only one pin, no current can flow anywhere.

Do you mean that the pin layout / spacing within the connnectors are different? If so, they wouldn't line up and the connectors, if they even plug into each other, wouldn't make electrical connection, so the controller couldn't drive the motor.

If you mean something else, you'll need to post good clear pictures of each thing you're describing so we can see exactly what it is.


I wonder if the pinout for the halls varies between the motors.
It certainly isn't unknown for that to happen. ;) But in your case, if there are only three wires (phases) on the controller, and no hall wires (which is what a 3-pin connector would mean), it's not relevant because it isn't connected to the halls at all at the controller side anyway. I don't think that's the case, as noted at the top of this post, however.
 
amberwolf said:
But at the moment, you've got to trust that the seller wired the new connector correctly--this is not guaranteed.

That's the problem with the single motor connector. There's no ability to trouble shoot the wiring, and I'm not too confident that vendors do things right, either on the controller or motor end.

Installing my current controller, that doesn't have self learning, I matched all the wiring by color, hit the throttle and nothing; stuck motor. Swapped the blue and green phases and everything runs perfectly. Couldn't do that with the single connector cable, without breaking out the wire cutters, soldering iron, and shrink tubing.
 
Or opening the controller and desoldering / swapping / resoldering the wires at the PCB (which I've done before to prevent having to cut into cabling). But it's still a PITA.
 
Ugh. This has been so frustrating. Sent a video to the controller vendor at their request showing the connections and startup/error. Got half a mind to just get a sensorless controller. How rough are they?
 
I have a 500W bafang front hub, currently using a KT sensorless controller. It is measured at 21A from my bms with a 52v battery and works very well, smooth startup and good top speed and acceleration. I use a throttle only so not sure about PAS.

https://bmsbattery.com/home/849-18288-s09p-bl-36v48v-450w-controller-for-with-bluetooth-ios-app.html#/497-voltage-36v
 
That may just be my salvation! This project is a kickbike, so no drivetrain, and no PAS. Thank you!

Oh, what kind of controller do you run?
 
It is the 48V version of the S09P-BL at the link. I don’t use a display but you can get one to match. Been totally reliable so far, runs cool. Need to disconnect the cruise control wires or it locks the throttle on after 5 seconds, which is exciting
 
Thanks. I apparently had AM brain, as you'd even linked me your controller. What I meant to ask what which display and throttle you're running. Currently have the Bafang display, but it looks like that controller takes a few specific ones. Do you use the standard Bafang thumb throttle?
 
Irae said:
Thanks. I apparently had AM brain, as you'd even linked me your controller. What I meant to ask what which display and throttle you're running.

drink more coffee. he just said "I don’t use a display but you can get one to match"
 
Ok, I'm sufficiently mortified. Won't post again until I've had 2 coffees and shaken off the AM meetings. Thanks, again, all!
 
Irae said:
Ok, I'm sufficiently mortified. Won't post again until I've had 2 coffees and shaken off the AM meetings. Thanks, again, all!

If you have several replacement controllers, then you may want to sacrifice one, cut the motor cable so you can spice different combinations of phase and hall wires. If you have a "stuck" motor, that doesn't move at all or barely does when you twist the throttle, then often just swapping two of the phase wires (the fat ones) will get the motor spinning, but you may need to swap the hall sensor wires in order to get it to run in the right direction or run smoothly.

If you can slice open the housing, without cutting the wires, then you could just cut two of the phase wires, and swap the blue and green (for example), and see if you get a response from the motor, and go from there. If it runs smoothly in the right direction, you're done. If not, you may need to cut and swap some of the thinner hall sensor wires.
 
Back
Top