how to test a hall sensor on the shelf?

[izeman]

some may have followed my thread with my broken mac/bmc motor which had not properly working hall sensors even though lyen's style ebike tester showed them working correctly.

i installed new hall sensors and everything was working again. until the next long ride. near the end of a long steep hill speed became lower and lower and those well known symptoms came back: a weird noise from the motor like a mechanical defect inside the motor. at a speed below 10km/h i stopped because the motor was producing no more power. i tried launching the bike again and it had big problems (me pushing it uphills - not riding it!!). the wheel just didn't start turning from a full stop.

after reaching the top of the hill i tried riding the bike again. the motor 'caughed' a bit but the higher the speed the better it was. it was like cutting and reengaging the power several time a second. after i reached 25km/h the ride was almost totally smooth again and the bike pulled like before until it's normal top speed of 45km/h.

so i don't know why the original honeywell ss41 halls got damaged again (i guess ONE of them) but it something killed it - maybe the controller did it. a new one is on it's way

to make a long story short: i now have those 3 halls laying in front of me and all look good. how can i be 100% sure which one is working and which one is not. they ALL react as the should: with GND and VCC attached they switch between those two when a magnet is moved along and switch back when the poles of the magnet are swapped and done again.

do i need an osciloscope? or are some hundreds or tenth of a volt difference enough to make one hall sensor a good one and one a bad one?

maybe someone can help!
thanks!

 

[salty9]

http://www.rcgroups.com/forums/showthread.php?t=307133

Describes using a Hall sensor to make a Gauss meter. Should be able to substitute the sensor in question to test it. Interesting read also.

 

[amberwolf]

to make a long story short: i now have those 3 halls laying in front of me and all look good. how can i be 100% sure which one is working and which one is not. they ALL react as the should: with GND and VCC attached they switch between those two when a magnet is moved along and switch back when the poles of the magnet are swapped and done again.

If all of them are properly sinking current at the signal output, then they are all working. Halls don't really output a voltage so much as they turn the output's ground on and off. So you need a pullup resistor on the output pin, from it to hall power. If that is in place, and the hall output still goes to nearly zero volts when it is activated by a magnet, and then goes back to hall power level when the magnet is removed, then the hall is working.

If so, the problem is more likely to be a connection failure somewhere, either a wire that is broken inside the insulation, or a wire that has broken insulation that is shorting to something else, perhaps the axle or stator, or another wire inside the axle/motor.

 

[izeman]

it really seems i'm more the mechanical guy than the electronics one. it's getting stranger.
i wanted to see if it's the halls, and removed one hall sensor cable. the motor did start turning, accelerated to full speed, than suddenly stopped. the controller is dead now. no movement nothing.
seems something burned inside. i opened it, but couldn't find anything. it's an infineon eb209.
can a controller really be damaged only because on hall sensors signal cable comes loose?
i tested the motor with a smaller controller i have as spare and it turns fine. any ideas?

 

[ZOMGVTEK]

Can the motor spin free with the controller connected? If its very hard to spin the motor with the controller connected, you blew some FET's in the controller. If it spins free, you likely have another issue.

Could be the 5V line dropped out on the hall sensors? Is the wire good there, they have 4.xV on the red/black wire going to the halls? Do you have all 3 halls connected, but you suspect one is bad?

You should be able to test halls quite simply. Power them with the rated voltage, which tends to be 5V, and measure the output as you pass a magnet across its face. Flip the magnet over and its output shout toggle between roughly 0.xV and 4.xV...
Grab the spec sheet below and take a peek at the pinout on the drawing in the lower right.
http://www.honeywell-sensor.com.cn/prodinfo/magnetic_position/installation/p88781_2.pdf

If the halls are installed in the motor, you can very simply probe between each of the 3 colored hall wires, and ground, while spinning the motor. They should toggle between high and low voltage, otherwise replace.

 

[izeman]

thanks for the hint with the mosfets. the first one i tested was dead. all other 8 behave normally.
can i swap the 4310 it for a 4410 which i have lying around?

the halls seem ok.

what i saw as well: the two 'big' caps both are close to blow up. they are not leaking, but not long and they will crack open.

 

[ZOMGVTEK]

It's not a very good idea to run different MOSFET's in parallel. You can replace all of them, or get an identical replacement for a single one.

Are the caps running near the rated voltage?