BBSHD permanently stops assisting

[BBSHME]

My BBSHD has a very odd issue and this is coming from someone who addressed several issues in the past on several different motors.
This motor has thousands of high intensity miles on it and it is about 6 years old.

I was recently riding in wet and muddy conditions when the harness cable became unplugged and the part that is connected to the motor got a lot of mud on it. Fortunately, after plugging it back in, I was able to ride the last 8 miles without issue.

The next day I took the bike out again and after 4 miles there was a strong vibration from the motor and then the pedal assist and throttle stopped working. I fiddled with a few things but the assistance only worked again after I walked the bike backwards a few steps then immediately pressed the throttle or started pedalling. At that point both the pedal assist and throttle worked as they should until I stopped either pedalling or pressing the throttle for more then a second.

That is the current state of the motor with the exception that it sometimes works fine for a few minutes after starting it up, before it goes back into throttle and pedal assist shut down. There is an odd jerky feel when the motor is under load but I am not so sure that wasn't there before.

I have disassembled the front plate of the motor and the gears looks properly greased. I have cleaned the harness wire with isopropyl alcohol the best I could but it goes right back into this state of occasional shutdown soon after.

My guess is that it could be a hall sensor issue or a harness wire issue, but I would like to try and eliminate all the easy possibilities before investing time and money into those more complex avenues.

 

[calab]

I'd redo the connectors for the throttle and pas with good quality connectors, or be lazy and hard wire them to bypass existing connectors.

Try putting the back wheel up, or put the bike on a stand and play with the setup unloaded and see what happens. Pin down the going backwards part, could be a loose wire somewhere.

 

[BBSHME]

I took the bike out for a ride today and the pedal assist only worked when i rotated the wheel backwards a little for the first part, then the bike worked fine on the way out. I hope the issue is working itself out but I can't imagine what is causing it.

 

[amberwolf]

Is there a sensor on the wheel you have to rotate backwards? Or is it just pulling the chain backwards, thus pulling the chainwheel to the BBSHD backwards?

Does it do the same thing if you just pedal backwards the same amount as the cranks are rotated backwards by rotating the wheel backwards?


Do you know if any of the mud/water got inside the controller casing itself? Or was it just inside the one connector that came loose?

Which specific connector was it? (knowing which signals are potentially disrupted may help diagnose the issue, or suggest possible repairs or workarounds)

If any of the contaminants remain on the connector pins inside (especially inside the open barrels that the pins fit into), it may increase resistance on them enough to cause signal problems without actually causing it to not work at all.

If any humidity remained inside the connector/pins when it was reconnected, it could have caused corrosion of the surfaces and increased their resistance like the above. If the connector internals are not totally sealed at the pin/connector interface (just sealed against external intrusion into the pin space) humidity could have gotten into the contact/wire interface (crimp, etc) and increased resistance there instead--this isn't visible as the other might be, but can still be tested for.

I'd first just try unplugging and replugging the cable at least once, and up to several times, to scrape the pins clear via friction with each other. This works more often than one might expect, but doesn't always fix things.

If that doesn't work, then to test for increased resistance or actual broken connections, you can set a multimeter on manual setting (rather than autorange) to 200ohms (or lower if it has the option) and use the red lead on the pin at one end, and wherever the wire connects on the other end of that cable inside the controller or display or other casing, as long as those are accessible to testing.

Given the later results that the bike doesn't always require this "reset" of rotating the wheel backwards, it's more likely to be something that can dissipate like humidity interfering with contact resistance, but here's other things to check just in case.

It can also be damage to the connector barrels, so they are spread open a tiny bit too much, and the pin doesn't make contact either at all, or just not enough to pass a good signal.

Or a pin could have bent over during reconnection, and no longer be going into the barrel at all.

 

[BBSHME]

Is there a sensor on the wheel you have to rotate backwards? Or is it just pulling the chain backwards, thus pulling the chainwheel to the BBSHD backwards?

Does it do the same thing if you just pedal backwards the same amount as the cranks are rotated backwards by rotating the wheel backwards?


Do you know if any of the mud/water got inside the controller casing itself? Or was it just inside the one connector that came loose?

Which specific connector was it? (knowing which signals are potentially disrupted may help diagnose the issue, or suggest possible repairs or workarounds)

If any of the contaminants remain on the connector pins inside (especially inside the open barrels that the pins fit into), it may increase resistance on them enough to cause signal problems without actually causing it to not work at all.

If any humidity remained inside the connector/pins when it was reconnected, it could have caused corrosion of the surfaces and increased their resistance like the above. If the connector internals are not totally sealed at the pin/connector interface (just sealed against external intrusion into the pin space) humidity could have gotten into the contact/wire interface (crimp, etc) and increased resistance there instead--this isn't visible as the other might be, but can still be tested for.

I'd first just try unplugging and replugging the cable at least once, and up to several times, to scrape the pins clear via friction with each other. This works more often than one might expect, but doesn't always fix things.

If that doesn't work, then to test for increased resistance or actual broken connections, you can set a multimeter on manual setting (rather than autorange) to 200ohms (or lower if it has the option) and use the red lead on the pin at one end, and wherever the wire connects on the other end of that cable inside the controller or display or other casing, as long as those are accessible to testing.

Given the later results that the bike doesn't always require this "reset" of rotating the wheel backwards, it's more likely to be something that can dissipate like humidity interfering with contact resistance, but here's other things to check just in case.

It can also be damage to the connector barrels, so they are spread open a tiny bit too much, and the pin doesn't make contact either at all, or just not enough to pass a good signal.

Or a pin could have bent over during reconnection, and no longer be going into the barrel at all.

Oh, thanks for your interest, I will do my best to answer your questions.
The connector that came out was the "harness" that is connected to the go between from the controller to the display, throttle, brakes. I was lifting the bike over logs and a branch must have caught it.
I do not use the wheel sensor for any of my bikes as they are too finicky, but after the issue arrose, one of my attempts to fix the problem was to connect my speed sensor and see if that helped fix the priblem and it was no different.
I have to rotate the chainwheel back 4 or 5 chainlinks in length before the motor will work properly again. If I took the chain off the motor, rotating the wheel back would do nothing. It is like those wind up toys where you have to pull them back before they will go forward. Rotating the crankarms backwards has no effect.
I have taken the crankarms off and looked inside the motor gearing to see if there was grease on the sensor, but everything looked fine in there.
I have removed the controller and disconnected every anderson connection and reconnected them, they all seemed fine and there was no contamination there.
My first attempts at fixing the problem were to disconnect the harness and controller wires and clean controller wire with rubbing alcohol and wd40 specialist, and that sometimes allows me to operate the motor normally without the rolling backwards for a couple minutes, I can even stop pedaling and the motor still works fine, but eventually it goes back to stall which needs to rotated backwards.
All the pins are there, but the connection between the wires was really stiff. Cleaning the wires made it slide on and off better but it only temporarily fixed the problem. I am going to get some dielectric grease today to see if that makes any difference, if I had had some in there to begin with, I might not have this problem.
I can't picture what you mean by testing the wires, but I bet that would produce some useful information.

 

[amberwolf]

I have to rotate the chainwheel back 4 or 5 chainlinks in length before the motor will work properly again. If I took the chain off the motor, rotating the wheel back would do nothing. It is like those wind up toys where you have to pull them back before they will go forward. Rotating the crankarms backwards has no effect.

Does rotating the crankarms mechanically do the same thing the chain pull by the wheel does, and turn the same chainwheel backwards? If not, then they won't be able to cause the same effect.

But manually turning the chainring backwards by hand, whether or not the chain is attached or the wheel is turned, would cause the same effect. (if it doesn't then none of what happens makes any sense)

If the crankarms do rotate the same chainring backwards (coincidentally also moving the chain if it's attached), then they also should cause the same effect. If they don't, then the difference in what they actually do should point to what is happening inside the system to cause the problem. It would mean that the crankarms are *also* causing some other thing that the wheel movement does not, in addition to the chainring movement. Whatever they are doing that the wheel doesn't, is directly related to why it restarts and probably why it stops.


What I suspect from the very small amount of movement required is that it is a hall sensor signal from the motor, which are used to tell the system what position the motor is rotated to so that it can know what currents to send on whcih phase wires to cause the motor to start spinning and keep it going correctly.

If it works until the motor stops spinning, then doesn't until the motor has been rotated slightly (which happens when you rotate the system output backwards), then it's likely that there is some position at which even with the bad signal it still lines up right enough for the controller to go "okay, I can work with this" and send startup currents that work.

Given the initial failure conditions, it is almsot certainly a wiring/connector problem if it's the hall signals from the motor. The other options are an actual hall sensor failure (or dislodged so not close enough to the magnets to trigger properly all the time), or a rotor magnet problem (perhaps dislodged laterally so not close enough to the halls to trigger properly, but still able to work with motor current to spin).

I have removed the controller and disconnected every anderson connection and reconnected them, they all seemed fine and there was no contamination there.

What about the non-andersons, with the multiple smaller wires? (the andersons are only used for the current-carrying connections, not the control and sensor signals, AFAIK. I think it is some form of PCB-mount JST (or similar) connectors for those).

It's not likely that they are a problem, as that is much more likely to be in the part that became disconnected and directly contaminated, but it is worth checking, because if it is a hall signal issue, it has to be in those connectors, or the halls themselves, or the motor magnets.

 

[BBSHME]

I took it apart again thanks to your getting my wheels spinning.
The crankarms do rotate backwards when the bike wheel pulls the chain backwards, but rotating the crankarms backwards has no effect on the motor coming back to life so to speak.
The idea that the motor only turns on when the sensor reads the magnet in a certain position doesn't quite work for an explanation because the motor shuts off and turns off super predictably. I can always know when it's going to turn off and exactly when it will turn back on, it is independent of where the rotation of the system is.
The hall sensor disc does float front to back and it might be the cause, but I feel like someone might have had this happen at some point and be chiming in as that seems like it would have popped up before, or there might be an error code indicating that is an issue.
I will try unplugging those controller wires next time I get in there, the only reason I didn't before is because they are held in with caulking that needs to be cut out, so it is unlikely they worked themselves free.

 

[amberwolf]

Ok. At this point I'm speculating based on previously posted info, and problems seen with various drive systems over the years. More thoughts below:

I took it apart again thanks to your getting my wheels spinning.
The crankarms do rotate backwards when the bike wheel pulls the chain backwards, but rotating the crankarms backwards has no effect on the motor coming back to life so to speak.

But does rotating the crank arms backwards also turn the chainwheel the same way pulling the chain backwards does?

If it does, then it must cause the same mechanical operation that pulling the chain backwards does, which turns the motor's rotor...so it must then *also* do something else (triggering a PAS sensor perhaps) that pulling the chain does not do, and the difference between these things should help lead to what is causing the problem.

The idea that the motor only turns on when the sensor reads the magnet in a certain position doesn't quite work for an explanation because the motor shuts off and turns off super predictably. I can always know when it's going to turn off and exactly when it will turn back on, it is independent of where the rotation of the system is.

Are the specific conditions that turn it off, and the specific conditions that turn it back on different or more detailed than what has been posted so far? If not, I am unsure how to proceed regarding that.

If the conditions are known exactly enough, it should be possible to use that info to narrow down the specific trigger that causes the behavior.

FWIW, it doesn't have to be that it is only in a certain crank position, because the gearing between motor and cranks is probably not a ratio that always leads to the same rotor position vs crank position.

It doesn't even have to be in only a certain rotor position, depending on the problem, but it is usually a set of specific positions relating to one phase or one hall sensor.


It's fairly common for something hall or phase related to cause problems starting a motor (in any drive system, but most often hubmotors) from a stop, and manually rotating the motor even a tiny bit may reposition the rotor to where the controller can now start it moving. Hence this troubleshooting path.

It might have nothing to do with your problem, but it does seem to fit the common symptoms.

The hall sensor disc does float front to back and it might be the cause, but I feel like someone might have had this happen at some point and be chiming in as that seems like it would have popped up before, or there might be an error code indicating that is an issue.

Where is this hall sensor disc? Are you referring to one that is mounted in the motor coils to detect the passing of hte rotor magnets? (because that is the set of hall sensors I'm referring to--if there are other sensor discs such as to detect pedal or gear rotation, they aren't what you're looking for).

I will try unplugging those controller wires next time I get in there, the only reason I didn't before is because they are held in with caulking that needs to be cut out, so it is unlikely they worked themselves free.

While it isn't likely in this case, the connector does not have to be loose or disconnected for there to be contamination intrusion that has caused a signal problem. There can even be a connection problem between the conductors in the wire and the contact in the connector (such as a broken wire at the back of the contact crimp, such that the insulation on the wire is still crimped in the contact but the conductor inside it is no longer fully (or at all) connected to the contact.

 

[BBSHME]

So I finally got around to trying a controller from a different bike and the problem was gone. I put the old controller back on and the problem is there so it must be the controller. I am considering replacing the connector on the black cable from the controller, replacing the Anderson connectors to a newer connection style, but I am most focused on hoping someone out there has a used controller with Anderson connectors that they want to sell me.

 

[BBSHME]

Soooooo I have solved one issue only to find another issue. The new spade connectors WORK with Anderson connectors. I installed a new controller (52v 1000w) with spade connectors and it is giving an error code 10 temperature too high. Is there a fix or is the controller not compatible with the motor? I have seen a few posts about a firmware update, but I have also seen a few posts saying they are incompatible and the only fix is a different controller. If that is the case, I am wondering if such a controller exists that will work

 

[Bengy22]

Soooooo I have solved one issue only to find another issue. The new spade connectors WORK with Anderson connectors. I installed a new controller (52v 1000w) with spade connectors and it is giving an error code 10 temperature too high. Is there a fix or is the controller not compatible with the motor? I have seen a few posts about a firmware update, but I have also seen a few posts saying they are incompatible and the only fix is a different controller. If that is the case, I am wondering if such a controller exists that will work

The firmware on your new controller likely expects a newer temp sensor and is confused and misreading it as way too high. I assume this because you mentioned you have had your bbshd for a while and I know bafang switched from a 1k ptc to 10k ntc a bit back (don't quote me on those I remember for sure though they were different). To fix this problem hardware wise you would need to swap out the temp sensor on the hall board. I believe you can find firmware though to fix this issue, if this is the case you would need firmware designed for this older version which might be tricky to confirm without testing. I know that Daniel on here has open source firmware that should be able to auto detect which sensor is being used and not have this issue, would be looking into

 

[BBSHME]

Th

The firmware on your new controller likely expects a newer temp sensor and is confused and misreading it as way too high. I assume this because you mentioned you have had your bbshd for a while and I know bafang switched from a 1k ptc to 10k ntc a bit back (don't quote me on those I remember for sure though they were different). To fix this problem hardware wise you would need to swap out the temp sensor on the hall board. I believe you can find firmware though to fix this issue, if this is the case you would need firmware designed for this older version which might be tricky to confirm without testing. I know that Daniel on here has open source firmware that should be able to auto detect which sensor is being used and not have this issue, would be looking into

thanks for your comment, that is what I was planning on trying, but I wanted some reassurance before trying. I downloaded Daniel's software and for some odd reason it isn't compatible with the computer even though it is a modern Windows operating system. I will post on Daniel's theead to try to get a workaround, or is there another approach?

 

[Bengy22]

Th

thanks for your comment, that is what I was planning on trying, but I wanted some reassurance before trying. I downloaded Daniel's software and for some odd reason it isn't compatible with the computer even though it is a modern Windows operating system. I will post on Daniel's thead to try to get a workaround, or is there another approach?

That's weird I haven't heard anyone experiencing that issue before. Are you using windows 11 (not sure if 11 is supported) and do you have a antivirus program? Might be worth making sure to have that turned off and maybe try on a different (windows 10?) pc? Hope it gets figured out for you