Hall sensor issues

[cohberg]

My hub motor / controller exhibited problems after climbing a long and steep hill. When I got back home I checked out the motor and found large portions of the phase wires to be melty. Additionally all the hall sensors were outputting erratic/no voltage on their output pins. I figured that I had cooked the hall sensors. I changed out the hall sensors and replaced the phase/hall wires with magnet wire.

I should say at this point that i'm positive that the hall sensors currently work. Short of scoping them, i've tested each one thoroughly and found that they all output ~1-4V with varying voltages depending on position like they should. However when the sensor outputs are plugged into the controller (vs just the 5V supply) two out of the three hall sense wires constantly read ~.4V with no variation when spinning the wheel.

I took apart the controller and I don't really understand what is happening with the circuitry. Could someone explain why there is a cap between the gnd and the signal from the hall? Did I fry the cpu when one of my phases (maybe) shorted with one of the hall sense wires? Are the capacitors shot and constantly trying to pull the pin to gnd? Or are there other explanations for why the hall sensor outputs become .4V when plugged into the controller?

 

[Hugues]

hi,

some expert will come in for sure, but what i remember from my problems with Halls (on infineon board) is that the Hall is sinking the sensor input to zero when turning the wheel. So if it stays close to zero when you connect them, maybe you have some kind of short on the controller side. You should be able to check that with a multimeter i guess. Maybe that cap has blown up ?

just guessing here.

 

[ddk]

the resistor/capacitor functions as a signal conditioner for taming any spikes emanating from the hall sensor.

It's unfortunately far more likely your cpu's magic smoke has been released

 

[dogman dan]

Did you replace the wiring to the halls too? Maybe it's shorting in the axle.

Duhh, yes you did.

 

[dnmun]

did you ever diagnose what was wrong originally before you replaced the halls? it seems obvious you don't know how to verify their operation so it would help to tell the whole story so we have an idea of what has happened.

we can start there.

 

[dogman dan]

At this point, if you had a motor, controller, throttle tester you could save yourself some time. Lyens sells them, and I've seen em on ebay too.

 

[cohberg]

hi,

some expert will come in for sure, but what i remember from my problems with Halls (on infineon board) is that the Hall is sinking the sensor input to zero when turning the wheel. So if it stays close to zero when you connect them, maybe you have some kind of short on the controller side. You should be able to check that with a multimeter i guess. Maybe that cap has blown up ?

just guessing here.

ok, yeah thats what i was guessing. i don't have a scope, so i guess i'll try to charge the cap with a 5V supply with an amp meter inline and see if its constantly pushing amps or stops?

the resistor/capacitor functions as a signal conditioner for taming any spikes emanating from the hall sensor.

It's unfortunately far more likely your cpu's magic smoke has been released

on just 2 hall sensors/cpu pins? It still spins the wheel: albeit only with that single working hall signal and very roughly. Also its still booting up and outputting blink codes. i mean yeah the 200+ momentary phase amps that potentially could have entered might have done some damage, but it just took out 2 hall sensor input pins instead of the entire cpu?

did you ever diagnose what was wrong originally before you replaced the halls? it seems obvious you don't know how to verify their operation so it would help to tell the whole story so we have an idea of what has happened.

we can start there.

edit: it seems obvious that you're an ass. I spun the wheel and watched the voltage levels on all the sensors they all toggled correctly.

At this point, if you had a motor, controller, throttle tester you could save yourself some time. Lyens sells them, and I've seen em on ebay too.

what would that tell me that a couple multimeters wouldn't?

 

[Hugues]

hi,

some expert will come in for sure, but what i remember from my problems with Halls (on infineon board) is that the Hall is sinking the sensor input to zero when turning the wheel. So if it stays close to zero when you connect them, maybe you have some kind of short on the controller side. You should be able to check that with a multimeter i guess. Maybe that cap has blown up ?

just guessing here.

ok, yeah thats what i was guessing. i don't have a scope, so i guess i'll try to charge the cap with a 5V supply with an amp meter inline and see if its constantly pushing amps or stops?
...

i had in mind you just disconnect your halls, then test for continuity between ground and entrance of hall signal in the controller. It should not be grounded. Unless i'm mistaken.

 

[dnmun]

first thing the guy shoulda done is test them with the controller to see if the hall sensors were working. no need for the lyens tester or any special tools. just check to see if they toggle. i don't see anywhere where he did this.

 

[hardym]

Hall sensors need 5V and ground to work. Each sensor needs a pull-up resistor. The hall output typically uses a RC for filtering. See the attached.

So to test a hall, you need to apply power AND have a pull up resistor.

 

[cohberg]

first thing the guy shoulda done is test them with the controller to see if the hall sensors were working. no need for the lyens tester or any special tools. just check to see if they toggle. i don't see anywhere where he did this.

Additionally all the hall sensors were outputting erratic/no voltage on their output pins.

 

[cohberg]

Hall sensors need 5V and ground to work. Each sensor needs a pull-up resistor. The hall output typically uses a RC for filtering. See the attached.

So to test a hall, you need to apply power AND have a pull up resistor.

I do know that. what i want to figure out is why my controller is pulling 2 of my sensors to gnd

edit: for clarity, i'm just checking the hall output voltages straight out of hall. Just 5V, Gnd, and my multimeter on the output pin, no additional resistors. If there is something inherently wrong with that test procedure then i don't know what to say. The voltage is "pulsing", for lack of better term, correctly when the hall's aren't plugged into the controller. But when they are, only the blue hall output still modulates: the green and the yellow stay fixed at .4V.

 

[dnmun]

that is what they do for a living. i really cannot understand why you don't just test to see if they toggle. that's why i don't think they were bad to begin with.

 

[cohberg]

that is what they do for a living.

cool. +1 for vague they and 0 contribution to the discussion.

i really cannot understand why you don't just test to see if they toggle.

When the hall sensors are not plugged into the controller they toggle.
I have tested it many times, you can stop saying the same thing over and over.
If you're not going to read the thread don't bother commenting

that's why i don't think they were bad to begin with.

None of the halls were outputting any voltage before the replacement how the hell is that not "bad to begin with"

 

[cohberg]

i had in mind you just disconnect your halls, then test for continuity between ground and entrance of hall signal in the controller. It should not be grounded. Unless i'm mistaken.

Hrm, you're definitely onto something. The blue hall input is the only one working and its not being pulled to ground at any point. The yellow and green both exhibit the image's behavior.

 

[dnmun]

hall sensors do not put out voltage. they sink current coming from the controller. they are an active device that is powered by the 5V red wire and when turned on by the magnet passing by, they sink the current that is supplied through the resistors inside the controller which creates a voltage drop at the microprocessor sense pin in order for the controller microprocessor to know where the magnetic phases are.

they always have either 2 on and one off or one on and two off. that is normal. but we don't even know if you have a DD motor or a geared motor, so we don't knw if you actually were able to move the magnets inside the hub, just all the info has been smothered in claptrap and i never saw where you had actually tested the hall sensor wires to see if they were toggling.

how can anyone help you if you have already decided what is wrong? if you know what's wrong, then fix it. but you already replaced the halls from what i understand and it still does not work.

 

[Spicerack]

Aside from the current to-ing and fro-ing about the halls, someone asked a while back what point there is in using a lyen/eBay motor and controller tester.

Well I have decent electronic testing skills but for 25 bucks you get a tester delivered to your door which shows via bright red LEDs if your wheel and controller are working. Really easy and quick. Even if you only use it once it's worth it. Yes you can do tests with a multimeter but with this, connect the phases and hall wires, turn it on, spin the wheel and straight away you can see with bright red flashes the halls toggling and the phases working. Likewise you plug the tester into the controller as if it were a hub motor and again it tells you straight away what functions are working as you open the throttle.

Anyone who plays with an electric bike should have one.