Placing hall sensors on vectrix motor

[Thibaudcnlc]

Hi, i've bought 2 vectrix vx1 scooters.
They are quite old scooters but they have big motors so i would mike to use them with standards controllers such as fardriver for example.
And here's the problem. They don't have hall sensors. No sensors at all in my case, the original optical sensors are no longer on the motor.
So my question is, how to install hall sensors on my motors.
I've already tried but my fardriver tell me ''hall sensor error '' so i think the placement isn't right.
Could someone give me an idea please?
(I'm sorry if my English isn't right it's not my native language)

 

[eee291]

From what I can tell you spaced the sensors 40° apart Wich should be correct for 6 pole pairs. What I find strange though is that one large gap between 2 magnets, could that cause it to fail? Also, are the sensors even close enough to the magnets? It looks like you mounted them on the outer edge of the stator, usually they sit deeper or flush with the steel laminations.

 

[Voltron]

Just a guess, but usually the Halls are mounted right into the laminations, which would amplify the field fluctuations?
Most motors people see are out runners, where the core is stationary with the magnet ring rotating around it, so there might be a lot of guessing with yours. But Halls closer to the laminations might help. Maybe too close to the ends of the coils where the magnet field isn't really sweeping past the sensors at an angle?

And yes, that gap is weird. Maybe they used it as a timing sensor in conjunction with the optical sensors?

 

[Thibaudcnlc]

Yes the gap is weird, but the motor is working with a sensorless controller, witch means that it can work no?
The sensor position is right i think, the sensor is sensing correctly, the problem is when. A sensor that can be mounted on the shaft of the motor could work? Because the shaft is accessible from both side so i have the possibility to mount it on one side

 

[amberwolf]

Some possible issues:

Are you using a bipolar latching hall sensor? (like the SS41 or 411) If not, the controller may not be getting the kind of signal it expects. If they are linear (analog, variable output) sensors they won't operate the way most UVW-type controllers require.

Sensorless may work if it's either dumb sensorless (that just feeds a waveform to the motor, and hopes it makes it spin), or something like FOC where it actually monitors phase current changes, so that regardless of whatever odd timing htere might be, as long as the MCU can respond in time to the current change it will still drive the motor, at least without a load. (under load can sometimes be different).

But the sensored mode may not work if the magnets are not spaced the same as the phases, equally around the rotor. Everytime it reaches the different gap, it may cause a timing fault, or a mistiming for the phase signals and drive the motor wrong or fault out. Depends on how the controller does it's sensing and processing.

Another issue can be if the magnets are all the same polarity; most hall sensors used for this stuff require magnets that alternate in polarity in order to turn them on and off in sequence.

However, as pointed out by others, if the halls are not close enough to the magnets they may not be correctly triggered by them.

The small face of the hall is also the one that should be used to face the magnets, as close as possible to them.

 

[Thibaudcnlc]

The hall sensors type and placement seems good, i get a good detection of 1/2 of the magnets according to my hall tester. It's the orders of switching that isn't right.

 

[amberwolf]

The hall sensors type and placement seems good, i get a good detection of 1/2 of the magnets according to my hall tester. It's the orders of switching that isn't right.

If the order is wrong, then you'd need to swap pairs of hall signal wires until the motor "runs right"--it might not be the right direction, though. If so, then swap pairs of phase wires until it runs right in the right direction. Then verify the no load battery current is low (a couple of amps, usually) at full throttle with the wheel offground.

But, as noted, the timing can still be wrong enough because of the magnet placement so the sensors just won't supply the correct timing. It may require a separate magnet ring just for the halls that *is* evenly spaced.

Or it could be that the magnet placement of this motor will not work with the FD's software.


Also: I think Fardrivers can be had in either UVW hall type, or "encoder" hall type (some form of resolver, I think, which comes as a module with the controller, to mount on the shaft of a non-hubmotor). Not totally certain it's that brand, but you should verify your FD is a UVW-type.

 

[larsb]

I can’t see that motor working well, the magnetic fields will be asymmetrical/unbalanced and the hall switch timing will be off at the big gap. How can they build it like that.. Are all the motors this way?

 

[Voltron]

It looks like there's a circle in the rotor right at that magnet gap that's maybe a balancing weight, so maybe the magnetic wobble isn't enough to be noticed, but sends a simple rotation feedback.