Q100 Front Hub Motor Single-Phasing?

[Kevin Crawford]

Hey everyone. I have a Q100H front hub motor on one of my bikes that is almost exclusively used for commuting, about 100 miles per week. I have used this build for many thousands of miles - maybe as many as 40,000! It's been rock solid and I'm very happy with it. I've almost always ran it, with a 48V battery, at 5A or less - I've found that any more current just doesn't seem to do much other than help with climbing a little, and is not worth additional wear and tear on the battery (or any other component). I'd probably be just fine with running it at even 4A (~200W) or even as low as 3A (~150) to eek out the last bit of use out of my battery packs.

Up until about two years ago I stored the bike indoors, but there was a period of about 15 months where I had to keep it outside.

I just started having an intermittent problem that seems to me like single-phasing. The power will suddenly drop off dramatically, and I've noticed that my power output (I'm using a Cycle Analyst from Grin Tech) indicates exactly a third of the normal power. It seems like vibration may cause the issue. But every time I do any testing, I can't find anything wrong. I'm getting 39VAC out of my controller on each phase-to-phase connection, and I'm getting near-zero resistance across each pair of phase wires on the motor, which seems to be normal. Then again, it's difficult to tell if the problem is occurring with the bike in a stand, because I don't get normal power readings with the wheel spinning in the air.

I suspect that one phase lead is losing connection inside the motor. The motor is older than my controller. I've never done anything to the inside of the motor. I've been trying to find out how to take the thing apart, but all I've found so far is how to take off the side opposite of the wires, where the gears are. I'm wondering whether replacing or reterminating a wire is even possible inside the motor.

Anyone else ever have an issue like this or have any tips? I suppose I could also take my controller apart and have a look to see if there's an intermittent connection somewhere. The only other idea I have is to check for a short to the hub shell - does anyone know where I could put my meter lead on the shell to detect such a short?

Thank you.

 

[amberwolf]

I just started having an intermittent problem that seems to me like single-phasing.

What is "single phasing"? (not a term I've seen before in relation to ebikes, only house/etc AC wiring)

The power will suddenly drop off dramatically, and I've noticed that my power output (I'm using a Cycle Analyst from Grin Tech) indicates exactly a third of the normal power. It seems like vibration may cause the issue.

Does it only happen at specific times or points in a ride, or under other specific conditions? Or just generically when there is more vibration than other times?

But every time I do any testing, I can't find anything wrong. I'm getting 39VAC out of my controller on each phase-to-phase connection, and I'm getting near-zero resistance across each pair of phase wires on the motor, which seems to be normal. Then again, it's difficult to tell if the problem is occurring with the bike in a stand, because I don't get normal power readings with the wheel spinning in the air.

I suspect that one phase lead is losing connection inside the motor.

A connection fault is unlikely to be inside the motor; it's almost always at the connector between motor and controller, or the wiring to that connector, usually wherever the wiring bends the most or enters/exits a connector or a housing.

If it's a sensored motor, you could have a connection fault to a hall sensor instead of a phase. Same things apply, except that if water got into the motor there could have been corrosion on a hall sensor leg that caused it to break, and that could now be connected most of the time, but sometimes not.

These both usually cause the motor to run roughly, making them feel like they're vibrating (on a geared hub it might feel like riding over pebbly surfaces instead of smooth road, and may even be audible).

The motor is older than my controller. I've never done anything to the inside of the motor. I've been trying to find out how to take the thing apart, but all I've found so far is how to take off the side opposite of the wires, where the gears are. I'm wondering whether replacing or reterminating a wire is even possible inside the motor.

In a geared hub, to access the motor itself, you may have to remove a C-clip from the outside of the bearing on the non-gear side (usually the left side of the motor), so the axle can then pull thru the bearing.

Could simply require pushing down on the motor housing (or wheel if it's laced into one) with the axle resting on the floor, to push the housing down past the motor. Normaly that's easy in a geared hub (but can be quite hard on a DD hubmotor).

Keep in mind that the axle wiring usually goes past that bearing, and can be damaged in the process of getting the motor out of the shell, or back into it.

Anyone else ever have an issue like this or have any tips? I suppose I could also take my controller apart and have a look to see if there's an intermittent connection somewhere. The only other idea I have is to check for a short to the hub shell - does anyone know where I could put my meter lead on the shell to detect such a short?

If there's a short anywhere to the motor casing, the entire casing is metal so it would be shorted everywhere, and would also be shorted to the axle and every other metal part connected to those including the bike frame and everything that's bolted to that.

When there *is* a short from a phase to the motor mechanicals, it's almost always in the axle where the wiring passes thru it, and almost always right at the part where the wiring enters it from outside, and almost always actually visible damage exposing the conductors inside the insulation of the cable right there.

When it is not that, it is almost always a motor winding shorted to the stator laminations, and this is almost never visible and is also almost never a short except under power, when the voltage applied to the windings is high enough to arc across the gap. Finding the actual short would require unwinding the motor, which is impractical and kinda pointless, but they can be fixed in large part by soaking the windings in CoronaDope or similar high-voltage-insulating lacquer to ensure it gets between the conductors and the laminations. Even a tiny thickness of such lacquer usually insulates to hundreds of volts, so this almost always fixes this specific problem.

(if it's a dead short all the time then it can't usually get between the two to do the insulating, so it doesn't usually work on those).

 

[Kevin Crawford]

What is "single phasing"? (not a term I've seen before in relation to ebikes, only house/etc AC wiring)




Does it only happen at specific times or points in a ride, or under other specific conditions? Or just generically when there is more vibration than other times?





A connection fault is unlikely to be inside the motor; it's almost always at the connector between motor and controller, or the wiring to that connector, usually wherever the wiring bends the most or enters/exits a connector or a housing.

If it's a sensored motor, you could have a connection fault to a hall sensor instead of a phase. Same things apply, except that if water got into the motor there could have been corrosion on a hall sensor leg that caused it to break, and that could now be connected most of the time, but sometimes not.

These both usually cause the motor to run roughly, making them feel like they're vibrating (on a geared hub it might feel like riding over pebbly surfaces instead of smooth road, and may even be audible).



In a geared hub, to access the motor itself, you may have to remove a C-clip from the outside of the bearing on the non-gear side (usually the left side of the motor), so the axle can then pull thru the bearing.

Could simply require pushing down on the motor housing (or wheel if it's laced into one) with the axle resting on the floor, to push the housing down past the motor. Normaly that's easy in a geared hub (but can be quite hard on a DD hubmotor).

Keep in mind that the axle wiring usually goes past that bearing, and can be damaged in the process of getting the motor out of the shell, or back into it.



If there's a short anywhere to the motor casing, the entire casing is metal so it would be shorted everywhere, and would also be shorted to the axle and every other metal part connected to those including the bike frame and everything that's bolted to that.

When there *is* a short from a phase to the motor mechanicals, it's almost always in the axle where the wiring passes thru it, and almost always right at the part where the wiring enters it from outside, and almost always actually visible damage exposing the conductors inside the insulation of the cable right there.

When it is not that, it is almost always a motor winding shorted to the stator laminations, and this is almost never visible and is also almost never a short except under power, when the voltage applied to the windings is high enough to arc across the gap. Finding the actual short would require unwinding the motor, which is impractical and kinda pointless, but they can be fixed in large part by soaking the windings in CoronaDope or similar high-voltage-insulating lacquer to ensure it gets between the conductors and the laminations. Even a tiny thickness of such lacquer usually insulates to hundreds of volts, so this almost always fixes this specific problem.

(if it's a dead short all the time then it can't usually get between the two to do the insulating, so it doesn't usually work on those).

Thank you Amberwolf, that’s a much appreciated and thorough response. I did suspect the motor to controller connections first, and that seems to be fine. I haven’t thought of checking the wires right where they exit the axle. I did think of the hall sensors, but I had such a problem before and the symptoms were different.

I did check the resistance across each pair of phase wires while bouncing the wheel on a desk, and never saw a failure. I should try the same while testing wires to the axle.

I’m starting to suspect an intermittent connection in the controller.

Single phasing is what results when one of the three phase wires in a 3 phase motor loses connection. Only one of the three windings is then ever energized.

 

[gromike]

The motor won't start single phased (ie. two phases), but it will run, poorly. Your three phase issue could be from the making of them to the consumption of that power.

 

[amberwolf]

Your three phase issue could be from the making of them to the consumption of that power.

What does that mean? (doesn't make sense to me)

 

[gromike]

Excuse my clumsy language. From the production of the three phases in the controller to windinings of the motor. As you know, anything along that line can make a phase weak.

 

[amberwolf]

I’m starting to suspect an intermittent connection in the controller.

While possible, it's less likely than in the cabling or connector. (because connectors are always a weak point and *the* most common point of failure in a wiring system, and wires themselves the next point especially where bent or where secured to things).

If there is an intermittent phase connection inside the controller, it would almost certainly be where the phase wires solder to the board itself.

Single phasing is what results when one of the three phase wires in a 3 phase motor loses connection. Only one of the three windings is then ever energized.

Ah. In that case, actually two of them would be, and only the same two of them, since they are almost always connected in Wye. Current would flow in first one direction, then the other, and then would not flow, each for 1/3 of the time, so current is still flowing 2/3 of the time. (not counting PWM off time).

Just one of them *never* would be energized, or have current flowing thru it, if that phase is disconnected.

So if you really wanted to use a term for it it would have to be "dual" phasing, rather than single.


If it was wired in Delta, then it would still be two, but each could be energized separately rather than always together since they aren't used in series as in Wye.

 

[Kevin Crawford]

While possible, it's less likely than in the cabling or connector. (because connectors are always a weak point and *the* most common point of failure in a wiring system, and wires themselves the next point especially where bent or where secured to things).

If there is an intermittent phase connection inside the controller, it would almost certainly be where the phase wires solder to the board itself.




Ah. In that case, actually two of them would be, and only the same two of them, since they are almost always connected in Wye. Current would flow in first one direction, then the other, and then would not flow, each for 1/3 of the time, so current is still flowing 2/3 of the time. (not counting PWM off time).

Just one of them *never* would be energized, or have current flowing thru it, if that phase is disconnected.

So if you really wanted to use a term for it it would have to be "dual" phasing, rather than single.


If it was wired in Delta, then it would still be two, but each could be energized separately rather than always together since they aren't used in series as in Wye.

I disagree with your assessment of what would happen when losing one phase with a Delta-wired motor. Refer to the drawing below. If for example the L1 wire was disconnected, we couldn’t have current flow between L1 and L2 nor between L1 and L3. Current would only flow between L2 and L3 - one phase.