Sabvoton hall issue

[marcos_88]

Hi guys,
I just purchased a Sabvoton 72100 and i cannot make it work with my setup. I get hall fault error even when i rotate the wheel by hand. If i try the trottle it moves just a little bit and than the same error. It doesn`t matter if i unplug the throttle, display, pas sensor and so on.
I tryed all the 36 combinations for the hall and phase wires without any luck. When i tested again the motor with my old kt controller it works just fine. With a motor tester all hall sensors seem to work.
I measured the voltage from the controller to the hall sensors, i get a full +5v on red wire, +5v on yellow and +2,95v on blue and green wires. If i conect the motor, i get the same voltages that drop to 0 when i spin the wheel by hand.
I also tryed setting up the hall angle, the wheel spins slowly only to give me a fail .
What can i try to make it work ?

Thanks.

 

[DogDipstick]

Brand new motor?

 

[marcos_88]

No, the motor is 2-3 years old. i upgraded only the controller and battery.
It worked and it works with the old 50 amps kt controller.

 

[amberwolf]

I measured the voltage from the controller to the hall sensors, i get a full +5v on red wire, +5v on yellow and +2,95v on blue and green wires.

if those voltages are without the motor connected, just at the controller, then the controller is not working correctly. it should have around 5v on all of the hall wires except ground, as there are internal pullup resistors from each signal line to the hall 5v power, inside the controller itself. (teh sensors in the motor only ground this voltage whenever a magnet passes them, so without the motor connected to the controller, the controller wires will all be at 5v).


if the pullup resistors are simply missing inside the controller, you can add your own (5k-10kohm) from each hall signal to hall 5v.

if the resistors are there, but the controller doesnt' have 5v on both sides of the resistor without the motor connected, then there's a short between the signal and something else, dragging the 5v down, which unfortunately could be the mcu itself (damaged, not replaceable).

 

[DogDipstick]

I ask because I have a similar issue, but... mine still works, too , like your hub still works with the old controller....

Been having "Hall Errors" and getting a reset every now and again. I have the same results when I back probe (similar reading).... (motor connected)

A(5v-0v-5v) / B(3.2v-0v-3.2v) / C(5v-0v-5v) / (repeat ABC ) ...

I get very good and in sequence 0v signal from the freewheeling the hub by hand... Bike works, just errors every now ad again.

....from the Hall sensors on my QS hub. Powered, backprobing, with the powered system wheel turning by hand ( freewheeling). No real idea as to the solution. The machine still works ( went 16 mi yesterday) and ony gave one error... reset fixxed. I will be opening the hub soon, looking for (short, open, grounded, burnt or otherwise damaged) reason.

My hub is relativly new, so the sensors are young.. but I have been a little abusive, and hot running,.... so... IF I do NOT have a BAD sensor ( intermittent), I probally DO have a SOFT short... ( or maybe a hard short forming over time and motion from bike and pedaling... )

A "soft" short is a leak of electrical current through something that's partially conductive, rather than a "hard" short such as what occurs when one electrical wire contacts another. I will be looking for both soon. I do not know if the 3.2v reading is right ( should be 5v? Visa Versa?)... but I have reliable operation. I do not know if this is the controller side, or wire run side.

I will have replacement sensors here too, and will replace them, after confirmation of no compromised wiring. I will let you know if this changes the signal output ( mystery 3.2v? Error? Working good... )... I am willing to bet it may be in the wire yet, must confirm.

I also wonder if all Hall sensors should be at the very same voltage.... I hope you find the issue...

6 Mo. old hub for me.

 

[amberwolf]

read my post above for testing / how they work.

if it's not clear from that, the hall sensors don't put out any voltage. they just ground the voltage provided by the controller.

 

[marcos_88]

The controller i got is brand new and it didn`t work from the start, it could be defective. But the seller tells me the 3-3.2 volt on the comunication wires from the hall sensors are normal. I get those reading with the sensors connected and disconnected, it doesn`t matter.
I get on the yellow hall +5v, but when i turn the wheel by hand it goes down only to 0,7-0,8 volts, so maybe this is the problem. I`ve tested with some brand new hall sensors (HONEYWELL SS441A) and the voltage goes down to +0,02 volts, so maybe i got a defective hall sensor.
I will open up the motor this days and i will change all the hall sensors, hope that solves the problem ....

 

[marcos_88]

I changed all the hall sensors with HONEYWELL SS441A and everything works nice now. Even the hall angle test passed in my first try
The voltages are the same, +5v on the yellow hall wire and about +3v on the green and blue wire, but considering everything works fine it may be normal.
The strange thing was that the old hall sensors worked fine with the old kt controller, on a hub motor tester everything was ok, but they refused to work with the sabvoton ...

 

[amberwolf]

its' good that it's working now, but it means there is something strange about it's design, and it's likely that it is not actually operating correctly (even if the hall stuff appears to work now).

The controller i got is brand new and it didn`t work from the start, it could be defective. But the seller tells me the 3-3.2 volt on the comunication wires from the hall sensors are normal. I get those reading with the sensors connected and disconnected, it doesn`t matter.

it's not normal. if it doesn't go all the way up to the pullup voltage, something is wrong inside teh controller. (either bad design or defect in parts or manufacturing)

all three hall signal lines from the controller should have the same pullup voltage when not connected to teh motor halls. if they don't, the controller has a problem.

I get on the yellow hall +5v, but when i turn the wheel by hand it goes down only to 0,7-0,8 volts, so maybe this is the problem.

that's completely normal. they have a transistor in the hall that pulls the signal to ground when teh sensor detects a magnet. this transistor in many sensors has a voltage drop of around 0.7v. some of them use a type of fet instead, and that has a lower voltage drop.

so that's the difference you're seeing.