Intermittent motor problems after high load and heat.

[barboma]

I likely have a controller that is burned out or about to be. It has been working for about 1500 miles but today I did some mountain biking up some pretty steep inclines and had the controller under max load a lot (dumb, I know). Anyway, about an hour or so in, the motor, pedal assist and throttle all started acting strange in that they would work, but intermittently. The throttle and pedal assist would get the motor up to about 100w and stop to. The display would also turn off sometimes. After checking all the connections on the trail, my gut says its the controller. MOSFET maybe? I dont know enough about that. Is it advisable for a novice to fix the controller?

Anyway, as I do not have a replacement controller on-hand, I can either ask the seller, Electric Bike Outfitters for a warranty (this is the second controller on this bike kit.) Or look for a better controller as their controller is not very good (fine for a road user I am sure). I have been drooling over the ebikes.ca Baserunner but not sure I want to shell out that kind of dough for the Baserunner + Cycle analyst.

Specifics:
*36v 11AH battery in a dolphin case.
*KT-LCD3 and of course some KT compatible controller in battery case (which I like).
*Julet connectors (would prefer them to be plug and play, but am willing to solder/splice).
*Rear geared motor.

If they don't honor warranty, I think its like 80 bucks for a replacement which is likely going to the cheapest plug and play option but then how far in the future am I going to be at this again?

If anyone has some insight or suggestions, I am happy to include any pertinent information I may have left out.

Thanks

 

[amberwolf]

If the display turns off, then the whole system probably lost power.

The most common reason for that is the battery's BMS shutdown to protect he cells, either from overdischarge or overcurrent (or possibly overheating, but most don't have thermal sensors).

So the first step is to test your battery; there are a lot of threads where people have power-cutout problems where they get help troubleshooting the battery, so I would start by looking thru those threads for steps to do so.

An 11Ah 18650 battery doesnt' usually have a high current (A) delivery ability, so if it's used hard, especially continuously for long rides, it may heat up significantly and / or have quite a lot of voltage sag, especially as it gets more toward empty than full.

The heat problem is exacerbated by having hte controller inside the battery, becuase the controller heats up under load too, and all that heat is bottled up iinside with the cells. How much heat depends on the load, the current, the voltage, and the efficiencies of the parts, but it can be a lot of heat if conditions are poor. Most likley this is not your problem...but it could be, so if you don't find any other issues you might want to put temperature sensors (like a BBQ thermometer remote) inside the case.


When FETs fail in a controller, they usually fail shorted, which means the motor won't run at all anymore. If it still works after "cooldown" period, or after recharging the battery, it's probably not the controller, or motor.

 

[barboma]

When FETs fail in a controller, they usually fail shorted, which means the motor won't run at all anymore. If it still works after "cooldown" period, or after recharging the battery, it's probably not the controller, or motor.

Letting everything sit for a few hours worked. Everything seems to be working again. The BMS is a Speedict Neptune, which has a temp sensor and logging capability. I will download those logs and upload them soon. In addition to temp in the case, logs have cell voltages with some other goodies. Should be interesting. I don't have temp readings at the controller as it sits in a section behind the battery.

Thanks!

 

[barboma]

If the display turns off, then the whole system probably lost power.

I do not know the correct terminology but the controller is wired directly behind the battery so that the BMS can turn off the controller but have the BMS remain on. This is how it is wired:

https://static.wixstatic.com/media/bf36dd_1d74327bbe1944508f290104c3c84f51~mv2.png/v1/crop/x_0,y_0,w_1449,h_557/fill/w_923,h_355,al_c,q_80,usm_0.66_1.00_0.01/neptune%20diagram%20(20170820).webp

The most common reason for that is the battery's BMS shutdown to protect he cells, either from overdischarge or overcurrent (or possibly overheating, but most don't have thermal sensors).

The BMS is set to a max discharge Amp of 17A, looking at the data in the excel file, I get up the 15A range but not that much higher. I do not think the BMS tripped due to high Amps. Although I notice in the quick states below, it did peak at 17.03A.

So the first step is to test your battery; there are a lot of threads where people have power-cutout problems where they get help troubleshooting the battery, so I would start by looking thru those threads for steps to do so.

An 11Ah 18650 battery doesnt' usually have a high current (A) delivery ability, so if it's used hard, especially continuously for long rides, it may heat up significantly and / or have quite a lot of voltage sag, especially as it gets more toward empty than full.

I do see a bit of voltage in but the average across the cells when I think I started to have problems was only 3.5V

The heat problem is exacerbated by having hte controller inside the battery, becuase the controller heats up under load too, and all that heat is bottled up iinside with the cells. How much heat depends on the load, the current, the voltage, and the efficiencies of the parts, but it can be a lot of heat if conditions are poor. Most likley this is not your problem...but it could be, so if you don't find any other issues you might want to put temperature sensors (like a BBQ thermometer remote) inside the case.

This is what I think the problem is, but of course, I could be wrong. The motor was warm to the touch with it being about 50* outside, so I am sure the controller was baking as well, being enclosed with no air flow like you said. Unfortunately, I do not have temps at the controller. But suffice it to say, I am sure it was very warm. The BMS only has one temp sensor input so I may have to go with what you suggested, the BBQ thermometer.

Could it be possible that the controller being hot, could explain the problems I describe or that there was some kind of thermal rollback going on? I do not know much about the controller but the closest thing I could find on Aliexpress is below. In fact, I wouldn't be surprised if this is what they use.

https://www.aliexpress.com/item/36V250W-6Mosfets-max-current-15A-controller-for-electric-bike-down-tube-battery-dolphin-case-controller/32813853990.html?spm=2114.search0104.3.2.335d28c8FufEU2&ws_ab_test=searchweb0_0,searchweb201602_3_10065_10130_10068_10890_10547_319_10546_317_10548_10545_10696_453_10084_454_10083_10618_10307_537_536_10059_10884_10887_321_322_10103,searchweb201603_52,ppcSwitch_0&algo_expid=4b9bdb67-7246-4bc2-8ae9-81e8c73a4f5a-0&algo_pvid=4b9bdb67-7246-4bc2-8ae9-81e8c73a4f5a

When FETs fail in a controller, they usually fail shorted, which means the motor won't run at all anymore. If it still works after "cooldown" period, or after recharging the battery, it's probably not the controller, or motor.

Make sense, thank you for explaining that about the FETs and how they fail. I have tested with no load and everything is working as it should, including the pedal assist sensor and throttle. I did put under a slight load of about 10 A and all seemed well. That kind of still pushes my thought process to a heat problem since, after a few hours of being idle, everything came back to life. If you concur, do you think that anything got damaged? If you don't, I am eager to learn what next.

Quick stats from trip:
Voltage max: 41.86
Voltage min: 34.03
Voltage used: 7.83
Pack temp: 87*F
BMS FET temp: 89*F
Amps max: 17.03
Amps avg: 1.63
AH used: 5.88
Watts max: 650.2
Watts Average: 61.2

Time when I first noticed a problem: 1:01:55.

The full file is attached.

Thoughts?

 

[barboma]

As an update to this, I have had to get a new controller. The dead giveaway was when I opened the motor to look at the hall sensors and noticed the waxed twine that was used to wrap the coils had gotten burnt and so was some of the grease. There is no doubt that the controller got way too hot.

I now know that pushing the components that hard was a dumb thing to do. At any rate, the motor looks to still be good, although I am sure the life span was probably reduced. The coils and magnets do not appear to be damaged. It's good it was only about 40-50*F outside or I am sure the motor would be toast along with the controller.

I haven't got the controller installed for testing yet but I remember having controller problems in the past and many symptoms are similar.

Finger crossed and yet another expensive lesson learned.

 

[amberwolf]

As an update to this, I have had to get a new controller. The dead giveaway was when I opened the motor to look at the hall sensors and noticed the waxed twine that was used to wrap the coils had gotten burnt and so was some of the grease. There is no doubt that the controller got way too hot.

Hmmm.... The two things you are talking about are completely different, and completely separate.

The below is just to make sure you understand this, so you can clarify the above statement to let us know the real reason you chose to replace the controller (and to let you know that the reason given is not really a way to know the controller is bad).

The controller (at least in your case) is not in the motor (it's in the battery), and so the motor overheating hasnt' got anything direclty to do with the controller, unless the motor go so hot it damaged winding enamel or wire insulation, and shorted phases together (blowing FETs up in the controller.

The controller being inside the battery *could* damage the battery, or it's BMS, if enough heat is generated. If the battery is making enough heat to damage the controller, then the battery is *also* damaged.

So while your motor might be damaged from overheating, that doesn't mean you need a new controller, and looking inside the motor and seeing overheating damage does not tell you anything about the controller condition--there is no "dead giveaway" about the controller by looking inside the motor.

The controller could also be damaged from overheating *itself* if it's current limiting doesn't work or has been overridden, or it is hidden in a box or a bag (or battery case) instead of being out in open air to keep it cool.


FWIW, there is no need to open a motor to test hall sensors. See the http://ebikes.ca page, Learn tab, Troubleshooting, for how to test many common problems.

 

[barboma]

As an update to this, I have had to get a new controller. The dead giveaway was when I opened the motor to look at the hall sensors and noticed the waxed twine that was used to wrap the coils had gotten burnt and so was some of the grease. There is no doubt that the controller got way too hot.

Hmmm.... The two things you are talking about are completely different, and completely separate.

The below is just to make sure you understand this, so you can clarify the above statement to let us know the real reason you chose to replace the controller (and to let you know that the reason given is not really a way to know the controller is bad).

The controller (at least in your case) is not in the motor (it's in the battery), and so the motor overheating hasnt' got anything direclty to do with the controller, unless the motor go so hot it damaged winding enamel or wire insulation, and shorted phases together (blowing FETs up in the controller.

The controller being inside the battery *could* damage the battery, or it's BMS, if enough heat is generated. If the battery is making enough heat to damage the controller, then the battery is *also* damaged.

So while your motor might be damaged from overheating, that doesn't mean you need a new controller, and looking inside the motor and seeing overheating damage does not tell you anything about the controller condition--there is no "dead giveaway" about the controller by looking inside the motor.

The controller could also be damaged from overheating *itself* if it's current limiting doesn't work or has been overridden, or it is hidden in a box or a bag (or battery case) instead of being out in open air to keep it cool.


FWIW, there is no need to open a motor to test hall sensors. See the http://ebikes.ca page, Learn tab, Troubleshooting, for how to test many common problems.

The controller sits behind the battery in its own enclosed compartment, behind the battery. The BMS, however, is in inside the battery pack. The controller has no means of cooling itself down other than the bottom of the controller compartment being metal.

I do have means of seeing what is going on in the battery and the data provided does not show any significant abnormalities in my opinion. In the battery, there was no extremely high FET (BMS) or pack temp, no extreme voltage sag on the pack or cell group voltage.

The one time I noticed the display off, was I suspect, the controller because the BMS never apparently shut itself down either due to fault or protection. The BMS would require manual reset if that was the case. And, being that I could see no significant damage in the motor, the controller seems the next logical item for the problem. Coupled with experience of seeing what a bad controller on this kit does (flaky motor operation, not registering ebrakes, irregular pedal assist operation, etc, when said components are known to be in a good state), I made a guess. I wasn't trying to say, at least, conclusively, "aha, the motor got hot so the controller blew up, lets replace it" because that doesn't make a lot of sense. Really I am just using the non-technical tools I have, however unscientific to make the best guess I can. Its all hearsay at this point as the controller is not installed, but I will report back once its done. Not out of the woods, yet.

In the future, I will limit the current at the controller level just to be safe. I can also do this at the BMS level but it shuts down the BMS and thus the controller, instead of limiting. This is really only to keep me from doing something stupid as I almost always ride on flat surfaces. Not extreme hills like I was.

Sorry if I left out key details and thanks for the help.

 

[amberwolf]

The controller sits behind the battery in its own enclosed compartment, behind the battery. The BMS, however, is in inside the battery pack. The controller has no means of cooling itself down other than the bottom of the controller compartment being metal.

I do have means of seeing what is going on in the battery and the data provided does not show any significant abnormalities in my opinion. In the battery, there was no extremely high FET (BMS) or pack temp, no extreme voltage sag on the pack or cell group voltage.

Ok, that's good, so it means the battery didn't overheat, and most likely the controller itself didn't overheat either (depending on exactly what kind of heat transfer could occur between them thru the compartment walls, and where the temperature sensors are).

And, being that I could see no significant damage in the motor, the controller seems the next logical item for the problem. Coupled with experience of seeing what a bad controller on this kit does (flaky motor operation, not registering ebrakes, irregular pedal assist operation, etc, when said components are known to be in a good state), I made a guess. I wasn't trying to say, at least, conclusively, "aha, the motor got hot so the controller blew up, lets replace it" because that doesn't make a lot of sense.

Ah; well, that's what it sounded like--I've seen "unusual logic" used by a lot of people over the decades, when troubleshooting problems with equipment they may not really understand. (I'm guilty enough myself, when dealing with things I have little or no experience with).

So apologies if I sounded...off.

In the future, I will limit the current at the controller level just to be safe. I can also do this at the BMS level but it shuts down the BMS and thus the controller, instead of limiting.

Yeah, the best place to limit power is at the controller, so it doesn't overload either motor (by allowing more than the motor can handle, cuz the motor can draw much more than is safe for it under the wrong conditions) or battery (by pulling more than it can handle).

 

[barboma]

The controller sits behind the battery in its own enclosed compartment, behind the battery. The BMS, however, is in inside the battery pack. The controller has no means of cooling itself down other than the bottom of the controller compartment being metal.

I do have means of seeing what is going on in the battery and the data provided does not show any significant abnormalities in my opinion. In the battery, there was no extremely high FET (BMS) or pack temp, no extreme voltage sag on the pack or cell group voltage.

Ok, that's good, so it means the battery didn't overheat, and most likely the controller itself didn't overheat either (depending on exactly what kind of heat transfer could occur between them thru the compartment walls, and where the temperature sensors are).

And, being that I could see no significant damage in the motor, the controller seems the next logical item for the problem. Coupled with experience of seeing what a bad controller on this kit does (flaky motor operation, not registering ebrakes, irregular pedal assist operation, etc, when said components are known to be in a good state), I made a guess. I wasn't trying to say, at least, conclusively, "aha, the motor got hot so the controller blew up, lets replace it" because that doesn't make a lot of sense.

Ah; well, that's what it sounded like--I've seen "unusual logic" used by a lot of people over the decades, when troubleshooting problems with equipment they may not really understand. (I'm guilty enough myself, when dealing with things I have little or no experience with).

So apologies if I sounded...off.

In the future, I will limit the current at the controller level just to be safe. I can also do this at the BMS level but it shuts down the BMS and thus the controller, instead of limiting.

Yeah, the best place to limit power is at the controller, so it doesn't overload either motor (by allowing more than the motor can handle, cuz the motor can draw much more than is safe for it under the wrong conditions) or battery (by pulling more than it can handle).

No problem at all. You have always been helpful in the past to me and others, so thanks for that.

I will get the controller installed today, so fingers crossed. I did a long, unassisted, ride yesterday, and I am feeling it, haha.

 

[barboma]

FWIW, there is no need to open a motor to test hall sensors. See the http://ebikes.ca page, Learn tab, Troubleshooting, for how to test many common problems.

So, got the controller installed. Same problem as before, motor starts to spin very briefly and then stops. The motor connector is a HiGO type, so its not easy to probe the wires without damaging the plug. So I took the motor apart and probed the hall sensors individually and guess what, one of them reads 0v (the green wire, sensor "B"). I happen to have another damaged motor, so. I am going to look at using one of those sensors.

I guess there are worse things than having a spare controller laying around.

 

[barboma]

Double checking for feedback here because I have a bad habit of convincing myself of something without knowing what I am doing

I have attached a picture. When I test the negative pad and the A, B, C pad, I do not get a voltage on A or C, but do on B. Further, when I test the sensors where they are soldered, I get 5v on the first and last joint on B but only the first joint on A and C registers 5v. That should mean that the B sensor is shorted and needs to be replaced because it should not read 5v on the output pin without the magnetic field, is this correct?

I have ordered this, so hopefully this is what I need: https://www.amazon.com/gp/product/B07CVRYLPH/ref=ppx_od_dt_b_asin_title_s00?ie=UTF8&psc=1

 

[barboma]

I am really confused right now. I just tested the sensors again after thinking more about how they are supposed to work. When I moved the magnets, all three will toggle from 0 to 5v, so they are apparently good after all. When I put everything back together, the motor spun up as it should and everything apparently seems to be working. I put the motor under load via my hand and the rim and no problems.

Could it be possible that the new controller learned what the hall sensors were giving it when I spun the motor and it learned the motors position and that is why its working? Or could it be that there is still some underlying issue with the BMS or Cells, even though the BMS reports no irregularities? How would I check?

Another theory is that when the twine burned, the hall sensors dislodged as it was the twine that was holding the circuit board and them down in the motor. I did go back and hot glue them back into position (see pic in previous post). It seems sturdy, any real world problems with vibration or heat?

Any help would be appreciated. While I am happy its working, I just want to root out any problems.

Thanks.

 

[amberwolf]

Another theory is that when the twine burned, the hall sensors dislodged as it was the twine that was holding the circuit board and them down in the motor.

This would be the most likely issue--it doens't take much to put them far enough out of the magnetic field from the rotor magnets to fail to react.

Hotglue is ok, but if you push teh motor hard enough, especially with a geared hub that can't get the heat out nearly as fast as a DD hub, it may get hot enough to soften the glue and let the sensor boar dmove.

If it begins acting squirelly, that's the first thing I'd check. If it does turn out to be the problem, you can use dental floss as a replacement for the original twine (it's not as good but it can be better than hot glue under the right circumestances ).

 

[barboma]

If it does turn out to be the problem, you can use dental floss as a replacement for the original twine (it's not as good but it can be better than hot glue under the right circumestances ).

Thanks, I will have a chance to test it out fully today on an actual ride.

I will also use some dental floss to secure it further. I am not too worried about extreme heat because my riding habits are usually low load (70to 130 Watts) with a relatively low ambient temp (less than 100*F in full summer). I also need to repack the gears with grease as I noticed some noise at higher RPM's that was not there before. I probably removed too much grease when handling the motor. In another forum/website I read that Mobil 28 is a good one to use.

When repacking, obviously one removes all of the old grease but then do you reapply to the gears and everything on the geared side of the motor? I can't imagine you can damage the motor with too much grease. Better safe than sorry, that's why I am asking.

Thanks a bunch.

 

[barboma]

I am happy to report that after a 25 mile ride, where I stressed the motor & controller, everything is working as it should.

Thanks amberwolf for sharing the knowledge and pointing me in the right direction!