Hall sensor confusion

Ebikeotis

10 mW
Joined
Apr 1, 2020
Messages
31
Hi guys. New to E-BIKES. I’ve built myself an ebike conversion which went brilliantly. Now it has a problem. Unfortunately as I’ve modified it I can ring the supplier (oops). I did squash 2 wires with the bearing which I have now fixed. The screen has a motor fault icon and when I try to use the throttle it jolts then nothing. I’ve used the multimeter to test the wires from the hall sensor wire and getting 0-5v on one but the other 2 are stuck at 5 and no change when I turn the wheel. Do the sensors have to be a certain distance from the magnets as one is close and 2 seem further apart. When the wheel was going round before it did have the sound of a misfiring car. The other thing I noticed is in between the gaps on the stator it’s being held together by wooden sticks. Is this right? Seems a bit cheap
The green tape isn’t the all the time ha ha

I have pulled out my hall sensors and looked at the codes. Two of the codes are the same saying 41H7KAG and the other has a different code on it 41H7CG. I have looked on the net and nothing is coming up with what I’m looking for.
This is my wheel https://www.ebay.co.uk/itm/324039515056

And my battery https://www.ebay.co.uk/itm/223033162309
 
Thank you for the link goatman. I have read and tested as it all says but I’m having trouble finding the correct sensors. If I type in the code for the ones I have in (I have 2 different kinds in mine. Don’t know if that’s right) they don’t seem to come up anywhere. What should I be looking for to replace them?

Many thanks
 
I have some Honeywell halls from grin, let me see if I can get a number for you.
s41 733a1 is the number I have, heres grins page where you can buy them

https://www.ebikes.ca/shop/electric-bicycle-parts/motor-hardware/hallsensor.html
 
Nice one. I’ve just emailed them to see if they can find a match or something similar. How would you know which ones to use if you can’t find the right code that is on the ones you want to replace?
 
pretty much all the hall sensors in ebike motors are (clones of) the honeywell ss411a. sometimes the ss41a. functionally the same for our purposes.
 
Positive its not just the wires to the halls?
 
Yeah. I’ve checked all the combinations that I’ve seen on the net. I’ve checked all the continuity on the main wires. I even reworded them and cut out where all the kinks were just incase they had breaks in
 
Honeywell are best.
Replace them 3, even when only one is fried.
Use thermal specific epoxy, JB Weld, 3M...
 
MadRhino said:
Honeywell are best.
Replace them 3, even when only one is fried.
Use thermal specific epoxy, JB Weld, 3M...

I’ve ordered 6 as like I said before I have different sensors in them. It probably doesn’t make a difference I just like things to be the same. Thanks for the advice
 
So when you test the halls,, you made connections to them inside the motor, and they still did not work?

Just saying, as you take it apart to put in the new halls, check em again before you rip them out of the motor. You will be cutting the wires to do this anyway, but if they are good, you can still just rewire. If your cut wires applied 48v to a hall, then for sure it fried them.

This is where a motor/ controller tester becomes a real time saver.
 
dogman dan said:
So when you test the halls,, you made connections to them inside the motor, and they still did not work?

Just saying, as you take it apart to put in the new halls, check em again before you rip them out of the motor. You will be cutting the wires to do this anyway, but if they are good, you can still just rewire. If your cut wires applied 48v to a hall, then for sure it fried them.

This is where a motor/ controller tester becomes a real time saver.
A tester would be so much quicker. Might even look for one now and see how much they are. I need to rewire my halls anyway. The cable is now to short
 
I put it all back together again last night checked and rechecked before I put it all in. Once it was all plugged in I did the rest again with turning the wheel. Yellow wire was on off but the blue and green just stayed on
 
are you testing with the motor controller connected and powered on?

if not, the halls won't get the pullup voltage required to have valid hall signals (unless you do this yourself with a resistor).

also, if you had cable damage, especially at the axle, and both phase and hall wires were damaged, the battery-level voltage on the phase wires can short to the low-level hall signal wires, which aren't just connected to the halls--they also go, unprotected against such things, directly to the mcu inside the controller. that kind of voltage on those pins often destroys the controller.

but the halls should still work correctly even if the controller failed, as long as the controller's lvps (12v/5v) wasnt' damaged, and the pullups on the signal lines inside the controller are intact.

if you want to eliminate the controller, you can use a 5kohm to 10kohm resistor (any wattage) from the hall signal line under test to the hall 5v line, and use antyhign form 4.5 to 6v (3 or 4 aa or aaa batteries in series) to power the halls from. then leave the ocntroller unplugged from the motor completely.



you could also still have shorts inside the cable if there was damage to wire insulation. those you can check for using the multimeter on continuity (beeps equal shorts) or ohms (under 1kohm could be a short), with the motor disconnected from the controller.



oh, and this is the ebike tester being talked about
https://endless-sphere.com/forums/search.php?keywords=ebike+tester&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search
 
amberwolf said:
are you testing with the motor controller connected and powered on?

if not, the halls won't get the pullup voltage required to have valid hall signals (unless you do this yourself with a resistor).

also, if you had cable damage, especially at the axle, and both phase and hall wires were damaged, the battery-level voltage on the phase wires can short to the low-level hall signal wires, which aren't just connected to the halls--they also go, unprotected against such things, directly to the mcu inside the controller. that kind of voltage on those pins often destroys the controller.

but the halls should still work correctly even if the controller failed, as long as the controller's lvps (12v/5v) wasnt' damaged, and the pullups on the signal lines inside the controller are intact.

if you want to eliminate the controller, you can use a 5kohm to 10kohm resistor (any wattage) from the hall signal line under test to the hall 5v line, and use antyhign form 4.5 to 6v (3 or 4 aa or aaa batteries in series) to power the halls from. then leave the ocntroller unplugged from the motor completely.



you could also still have shorts inside the cable if there was damage to wire insulation. those you can check for using the multimeter on continuity (beeps equal shorts) or ohms (under 1kohm could be a short), with the motor disconnected from the controller.



oh, and this is the ebike tester being talked about
https://endless-sphere.com/forums/search.php?keywords=ebike+tester&terms=all&author=&sc=1&sf=titleonly&sr=topics&sk=t&sd=d&st=0&ch=300&t=0&submit=Search

if a person has an automotive 12v turn signal bulb, not led, can they use that instead of a resistor? or if you have a non led flashlight bulb?
 
Many thanks guys. I re wired the broken hall cables where the cables got damaged. I also tested for continuity. I did run a test with the phase wires too which was part of the guild that someone posted.

Definitely going to get a tester. Didn’t realise they were so cheap
 
Continuity testing the wires to the halls would confirm they still work to connect to the actual halls.

I assumed you were testing like this, 4v battery cell connected to red and black, then see if voltage toggles on and off on voltmeter connected to black, and either blue, green, or yellow wire. You spin the wheel, or just pass a strong magnet past the hall sensor if the motor is apart.
 
goatman said:
if a person has an automotive 12v turn signal bulb, not led, can they use that instead of a resistor? or if you have a non led flashlight bulb?
if the resistance of the bulb under the conditions present is fairly high, in the 5-10kohm range, then sure.

if it's so low as to permit high currents beyond the rating of the hall sensor to sink (usually a few ma, maybe up to a few dozen ma), then you can damage the hall sensor. that's part of how these things fail when shorted due to wiring faults.
 
dogman dan said:
I assumed you were testing like this, 4v battery cell connected to red and black, then see if voltage toggles on and off on voltmeter connected to black, and either blue, green, or yellow wire. You spin the wheel, or just pass a strong magnet past the hall sensor if the motor is apart.
the voltage doesnt' toggle unless you *also* have a pullup resistor from the hall signal line under test to your voltage source.

halls like in the motors do not provide any output voltage.

they only ground the voltage already present on the signal line (provided by the pullup resistor inside the controller, normally).
 
i did a test. I took a hall sensor and soldered red wire to left leg, black wire to center leg and green wire to right leg. took a 18650 that read 3.96v.
connected red wire to positive, black wire to negative

I connected dmm to battery red lead positive and black lead to green hall wire, no resistor, then I waved a magnet(north pole) and voltage went from 0v to 3.96v then waved south pole past hall and it switched to 0v. hall sensor is working.

I kept the dmm connected and hooked 12v bulb from positive battery to green wire of hall. hall was switching but light wouldn't work. then I serialed 2-18650 to 7.5v and got same result, hall would switch but bulb wouldn't turn on. if I connect bulb straight to battery it lights up.

I mention this because im getting a different result than whats been said. im pretty sure I wired the hall correctly.
I got dogmans result except its red+ to green not black- to green
 
I used the picture of the mxus v3 hall sensors in this link as my reference for red, black,green
https://kinayems.com/Archive/Products/DriveTrainsMotors/MXUS_3K-Turbo
 
Hall, might work backwards, but pretty sure its red + black -, and the colored wires are sending power back to controller.

But looks to me like that green hall sensor works. So replacing it won't change your problems. Next look for if the controller actually powers up the halls, on the red and black wires.

Problem can still be in the phase wires as well, including any repairs you did to them. Continuity says they connect, but they might connect weakly, and not be able to handle any power.
 
from what I can tell, if you are testing with the controller connected and turned on where the green signal wire from the controller is sending the positive and the hall is switching the ground circuit off/on.

once you disconnect the controller and use a battery instead as the power source, you need to connect the dmm to positive red wire and green signal wire to test if the hall sensor is working

I just went to grins troubleshooting halls page to see if that was the case and if you look at the probes of the DMM they use, the pos and neg probes are shaped differently

https://www.ebikes.ca/documents/HallSensorTestingFinal.pdf

all im saying dogman is, I found the way you suggested to test just using a battery works but you connect the dmm to green wire and red wire not black wire and green wire
 
goatman said:
i did a test. I took a hall sensor and soldered red wire to left leg, black wire to center leg and green wire to right leg. took a 18650 that read 3.96v.
connected red wire to positive, black wire to negative

I connected dmm to battery red lead positive and black lead to green hall wire, no resistor, then I waved a magnet(north pole) and voltage went from 0v to 3.96v then waved south pole past hall and it switched to 0v. hall sensor is working.

then you may have hall sensors without the open-collector style output that is common for such hall sensors. most of the ones i've worked with are open-collector (like the honeywell ss411a), so they require a pullup on their output. these types give a better signal to the controller over a long wire run than the ones that output their own voltage, becuase there is no voltage drop for the high signal at the controller, as the high signal is *in* the controller. so the controller only has to look for a low signal, and it knows that anything below a certain point (say, 2.5v) is a low, and anything else is a high.

with ones that output their own voltage, rather than open-collector (oc) types, the signal strength at the controller is dependent on the quality of the wiring, and of the induced currents (and thus voltages) in the hall wires from the phase wires that run parallel to them in the motor cable, and all the other environmental noise.

so they don't typically use taht type in motors--they give terrible results, unreliable operation, without signal conditioning or special wiring, both of which are either expensive or complicated or both, and neither of which is going to happen in a cheap chinese ebike controller or motor. ;) it's much much simpler and cheaper to use open-collector-output type halls, and a pullup in the controller.


with open-collector halls, under some conditions, you mgiht get what appears to be a valid voltage when it isn't active (grounded), but if you don't have the pullup resistor (5k to 10kohm) from the signal wire to the 5v line, you can't really tell if the hall is working correctly or not.

this is becuase sometimes without that resistor, it will appear that the hall is working correclty at all, and so is going low-high-low-high, but if you put the pullup on there you find the hall isn't actually able to ground the signal correctly, so it never goes low, or that it is not releasing the ground properly and it never really goes high.

so if you want *definite* valid results with open-collector output halls, like the ss41 and ss411 (and clones of those) common in ebike motors, you have to use the pullup on the signal line.

without it, you can't tell if it's really working as it should, even if it "looks" like it is.

it's like a bms that has shut off it's output fets. there will still be a voltage there--but it is not "really" there, because if you put any load on it, it goes away.


I kept the dmm connected and hooked 12v bulb from positive battery to green wire of hall. hall was switching but light wouldn't work. then I serialed 2-18650 to 7.5v and got same result, hall would switch but bulb wouldn't turn on. if I connect bulb straight to battery it lights up.
the hall sensors cannot sink or source enough current to run an automotive bulb.

typically they can only do a few ma, to a few dozen ma.

the ss41 and ss411a only do about 20ma max.

an automotive bulb usually requires a few *amps* (about a thousand times as much as the hall sensor can handle). it might damage the hall sensor, but it can't light up the bulb.

you mgiht be able to use a low-power led, that uses no more than 20ma *max*, but almost any incandescent i know of (including little tiny old grain-of-wheat bulbs) will use up to several times that, and even if they work, they're overloading the hall's output and damaging it.





so...what it comes down to is that if you *really* want to knwo if the hall is working, either connect the signal lines to the powered-on motor controller so it's pullups will do the job, *or* use an external 5kohm to 10kohm resistor from hall signal wire under test to your 5v power source.


if you don't do that, you can't *really* know for sure that the halls are working correctly.
 
Back
Top