Hub Motor Axle Nut Tighteness and Corresponding Performance Issues

[carltoza]

I have noticed now after several ebike builds that if I tighten the axle nuts too aggressively then the hub motor seems to have trouble rotating as freely, draws more current, and sometimes even gets very hot while not in use.

This seems to indicate some kind of short but I'm not sure where or how. I am typically running bigger controllers pushing the motors at 2x - 4x their rated capacities. I have poured over countless forums and articles before posting this but I haven't seen any pages or literature address this problem directly. It has happened with various controllers as well with both the geared 250w bafang hub and the common (non geared) 1000w conversion hub motor, front and rear.

any and all advice is greatly appreciated!

 

[motomech]

On a set-up like yours, it's common to have to put a flat washer between the bearing(or axle shoulder) on the end of the axle and the inside of the drop-out. If it's not there, as the axle nut is cinched down, the rotating part or the cassette or free wheel is rubbing on the chain stay.
When this is right, it's possible to snug the nut down, but don't over-tighten.
Lot's of med. strength Loc-Tite.

 

[E-HP]

I have noticed now after several ebike builds that if I tighten the axle nuts too aggressively then the hub motor seems to have trouble rotating as freely, draws more current, and sometimes even gets very hot while not in use.

Can you provide some more detail about this? Does that mean when it's parked, the motor starts getting hot?

 

[calab]

motomech nailed it, could be rubbing causing drag causing extra juice to be used, had a similar issue with the wide mxus 45h 3kw and had to dremel a large washers OD and ID. I can see how that could be missed, I barely caught it the first time (right away) myself, of course we SHOULD catch it when we tighten things up.

 

[carltoza]

I have noticed now after several ebike builds that if I tighten the axle nuts too aggressively then the hub motor seems to have trouble rotating as freely, draws more current, and sometimes even gets very hot while not in use.

Can you provide some more detail about this? Does that mean when it's parked, the motor starts getting hot?

Correct, it also seems to happen while the bike is parked, after I have spun the motor up a couple times for testing with practically no load. I'm not sure if the two issues are the same, all I can say is that the motor generates a ton of heat and doesn't really seem to start cooling down until I unplug the battery. I wonder if these are two seperate issues? What motomech mentioned about the washers and seating of the axle to the dropouts is a good lead and I am going to work on that soon, but I don't know how a friction issue would also cause the motor to draw current in a stopped position.

 

[E-HP]

Correct, it also seems to happen while the bike is parked, after I have spun the motor up a couple times for testing with practically no load. I'm not sure if the two issues are the same, all I can say is that the motor generates a ton of heat and doesn't really seem to start cooling down until I unplug the battery. I wonder if these are two seperate issues? What motomech mentioned about the washers and seating of the axle to the dropouts is a good lead and I am going to work on that soon, but I don't know how a friction issue would also cause the motor to draw current in a stopped position.

Sounds like an incorrect hall sensor wiring combination. If you got the stator hot enough, it will continue to release heat to the case for a while, which may by why it seems to heat up while not in use. If there's not enough washer spacing between the frame and freewheel, causing binding, you wouldn't be able to pedal, and would jam up if you tried to pedal backwards. Is that the case?

 

[carltoza]

Correct, it also seems to happen while the bike is parked, after I have spun the motor up a couple times for testing with practically no load. I'm not sure if the two issues are the same, all I can say is that the motor generates a ton of heat and doesn't really seem to start cooling down until I unplug the battery. I wonder if these are two seperate issues? What motomech mentioned about the washers and seating of the axle to the dropouts is a good lead and I am going to work on that soon, but I don't know how a friction issue would also cause the motor to draw current in a stopped position.

Sounds like an incorrect hall sensor wiring combination. If you got the stator hot enough, it will continue to release heat to the case for a while, which may by why it seems to heat up while not in use. If there's not enough washer spacing between the frame and freewheel, causing binding, you wouldn't be able to pedal, and would jam up if you tried to pedal backwards. Is that the case?

That's possible. I believe the hall sensors and phase wires are in the correct combination, I always check to make sure that the colors of the wires are the same on both ends.

I still seem to be able to pedal, or in this case use the throttle. What seems to happen though when I apply the throttle is that it accelerates much slower and if punch the throttle too hard then it makes a loud squeaking/squealing sound and doesn't go. This happens particularly in the middle of the rpm band.

 

[amberwolf]

Correct, it also seems to happen while the bike is parked, after I have spun the motor up a couple times for testing with practically no load. I'm not sure if the two issues are the same, all I can say is that the motor generates a ton of heat and doesn't really seem to start cooling down until I unplug the battery.

If you simply turn the system on, do nothing with it, and the motor heats up, there is a significant system problem in the controller, possibly hardware, possibly just software.

If the heat only occurs after using the motor, and if you let it sit for several hours after doing so and the temperature drops (as measured with a non-contact IR gun type, for instance) over time until it cools to room temperature (even with battery connected), then it is probably a hall/phase combination problem or a controller setting problem



If you are using an FOC controller then if it is a type using a continous signal within the phase wires (even when motor is stopped) to sensorlessly determine rotor position, if this signal is too large or the motor properties (phase resistance, phase inductance, kV, etc) are not set correctly within the controller, it can cause excessive heating of the motor simply by being powered on and doing nothing at all with it. (this happened to me when beta testing the SFOC5 due to some firmware issue, resolved with an update, but it heated the phase wires so much the insulation melted on them, doing nothing at all but sitting there powered on).

 

[amberwolf]

That's possible. I believe the hall sensors and phase wires are in the correct combination, I always check to make sure that the colors of the wires are the same on both ends.

This is not the way to determine hall/phase combination. Wire color literally means nothing, regarding that, as even with a "kit" where all parts come together in a box they may not match, and with parts from different places it's a fair bet they won't..

You need to use the "self learn" procedure specific to your controller, if it has one, to get this correct.

If it does not, then you need to manually try the six different combinations of phase wires (leaving halls alone) until the motor spins with no undue noise, at the lowest possible battery current even at full throttle no load (wheel off ground).

If the motor is not spinning the correct direction at this point, then if there is a reverse wire on the controller you can simply connect that to ground to fix this. If it does not have one, then you try the six different combinations of the three non-red/black hall wires (leaving phases alone) until the motor spins the correct direction, with no undue noise, at the lowest possible battery current even at full throttle no load (wheel off ground).

 

[amberwolf]

I have noticed now after several ebike builds that if I tighten the axle nuts too aggressively then the hub motor seems to have trouble rotating as freely, draws more current, and sometimes even gets very hot while not in use.

If tightening the nuts causes this, but nothing else does, then the most likley problem is there is something missing or wrong in the mounting hardware, allowing the frame to contact the wheel in some way and apply friction to the wheel and prevent motor from turning freely. This will be obvious when turnign the wheel by hand.

If that extra friction is constant during rotation, it is probably at the axle area (such as freewheel pressing against dropout and then pushing against motor cover).

If the extra friction is periodic, once per rotation, it's probably at the tire or brake area.

If it's periodic, regular, many times per rotation, it's probably the phase wires being compressed together at the axle exit until they short together or to the frame. This would also cause excessive current to flow all the time the motor is powered by the controller. The damage to the cable may not be externally visible, it can happen entirely within the outer jacket, but there will usually be marks on that jacket to show where it's being compressed or bent.

 

[carltoza]

That's possible. I believe the hall sensors and phase wires are in the correct combination, I always check to make sure that the colors of the wires are the same on both ends.

This is not the way to determine hall/phase combination. Wire color literally means nothing, regarding that, as even with a "kit" where all parts come together in a box they may not match, and with parts from different places it's a fair bet they won't..

You need to use the "self learn" procedure specific to your controller, if it has one, to get this correct.

If it does not, then you need to manually try the six different combinations of phase wires (leaving halls alone) until the motor spins with no undue noise, at the lowest possible battery current even at full throttle no load (wheel off ground).

If the motor is not spinning the correct direction at this point, then if there is a reverse wire on the controller you can simply connect that to ground to fix this. If it does not have one, then you try the six different combinations of the three non-red/black hall wires (leaving phases alone) until the motor spins the correct direction, with no undue noise, at the lowest possible battery current even at full throttle no load (wheel off ground).

I have checked and tuned the motor using the self learning feature and everything seemed fine. Under no load (spinning freely) the motor seems to be spinning fine with little to no noise. It is only when I start riding the bike that the problem becomes apparent.

 

[carltoza]

That's possible. I believe the hall sensors and phase wires are in the correct combination, I always check to make sure that the colors of the wires are the same on both ends.

This is not the way to determine hall/phase combination. Wire color literally means nothing, regarding that, as even with a "kit" where all parts come together in a box they may not match, and with parts from different places it's a fair bet they won't..

You need to use the "self learn" procedure specific to your controller, if it has one, to get this correct.

If it does not, then you need to manually try the six different combinations of phase wires (leaving halls alone) until the motor spins with no undue noise, at the lowest possible battery current even at full throttle no load (wheel off ground).

If the motor is not spinning the correct direction at this point, then if there is a reverse wire on the controller you can simply connect that to ground to fix this. If it does not have one, then you try the six different combinations of the three non-red/black hall wires (leaving phases alone) until the motor spins the correct direction, with no undue noise, at the lowest possible battery current even at full throttle no load (wheel off ground).

I have checked and tuned the motor using the self learning feature and everything seemed fine. Under no load (spinning freely) the motor seems to be spinning fine with little to no noise. It is only when I start riding the bike that the problem becomes apparent.

I also don't believe that the issue is with the controller since I plugged it up to a second hub motor that I replaced the front wheel with and it ran perfectly, quieter than the hub on the rear.

 

[carltoza]

I have noticed now after several ebike builds that if I tighten the axle nuts too aggressively then the hub motor seems to have trouble rotating as freely, draws more current, and sometimes even gets very hot while not in use.

If tightening the nuts causes this, but nothing else does, then the most likley problem is there is something missing or wrong in the mounting hardware, allowing the frame to contact the wheel in some way and apply friction to the wheel and prevent motor from turning freely. This will be obvious when turnign the wheel by hand.

If that extra friction is constant during rotation, it is probably at the axle area (such as freewheel pressing against dropout and then pushing against motor cover).

If the extra friction is periodic, once per rotation, it's probably at the tire or brake area.

If it's periodic, regular, many times per rotation, it's probably the phase wires being compressed together at the axle exit until they short together or to the frame. This would also cause excessive current to flow all the time the motor is powered by the controller. The damage to the cable may not be externally visible, it can happen entirely within the outer jacket, but there will usually be marks on that jacket to show where it's being compressed or bent.

Okay, this is more along the lines of what I was suspecting. Considering that I am running about 3x the power that the motors are rated for, my thoughts are that something inside the motor is shorting, and it may not be shorting all the time but something definitely doesn't seem right. I do plan to gut the wires inside the motor and replace them with heavier gauge silicon wrapped wires regardless of what I find during the diagnostics, but I hope that once I open it and look at the wires there will be some evidence for a short. I have checked the phase wires on the outside of the hub and while they have gotten hot, I wouldn't say that they were in the melt zone. However, I wonder if for the internal wires, the heat generated by the phase current coupled with the higher temperatures inside of the motor casing is enough for them to melt and short...

 

[amberwolf]

THen the most likely issue, assuming the heating occurs without ever even using hte motor, just powering it on and leaving it, with the axle hardware tightened, is that it is crushing the phase wires and shorting them to each other or the frame.

I also don't believe that the issue is with the controller since I plugged it up to a second hub motor that I replaced the front wheel with and it ran perfectly, quieter than the hub on the rear.

I have checked and tuned the motor using the self learning feature and everything seemed fine. Under no load (spinning freely) the motor seems to be spinning fine with little to no noise. It is only when I start riding the bike that the problem becomes apparent.

 

[amberwolf]

It's probably not isnide the motor. It's probably inside the cable at the axle, if it only happens when the axle hardware is tightened, and would require extra axle hardware to prevent whatever cable damage has occured to fix this.

Note that changing the phase wires for larger ones is not generally necessary unless you are using the higher power levels more than just intermittently, such as for acceleration.

in my case I use a "500w" DD hubmotor at about four times that power for up to a few seconds during every startup from a stop, but at only about it's rated power otherwise, and heat is not an issue especially in it's wiring. Keeping the wiring short will help as much or more than using thicker wiring, unless you are greatly increasing the crosssection of the conductors (not typically easy and sometimes not possible with the narrow space available thru the axles of most motors).

If the phase wires had melted and shorted, the short woudl be permanent, and probably sufficient to blow up the controller, which would prevent it from working at all, and also cause the motor to be extremely hard to turn by hand.

A phase wire short to the frame thru cable damage at the axle exit could be a high-resistance short, allowing the system to work but causing the symptoms you see.

Okay, this is more along the lines of what I was suspecting. Considering that I am running about 3x the power that the motors are rated for, my thoughts are that something inside the motor is shorting, and it may not be shorting all the time but something definitely doesn't seem right. I do plan to gut the wires inside the motor and replace them with heavier gauge silicon wrapped wires regardless of what I find during the diagnostics, but I hope that once I open it and look at the wires there will be some evidence for a short. I have checked the phase wires on the outside of the hub and while they have gotten hot, I wouldn't say that they were in the melt zone. However, I wonder if for the internal wires, the heat generated by the phase current coupled with the higher temperatures inside of the motor casing is enough for them to melt and short...