What could have caused my motor to break?

[CGameProgrammer]

Today I installed a new rear 5303 motor and took it for a short ride. After 10 seconds of very gentle throttle, I suddenly felt resistance from the motor, and two seconds later it stopped responding completely. It now no longer works at all. But it didn't budge in the rear dropouts, the bolts are still perfectly tight, and the wires from it didn't move at all. I had a piece of duct tape holding them to the frame, with a bit of slack, but the duct-tape never budged and basically it's obvious they weren't pulled on, so I can't imagine what's wrong besides the motor just being generally defective.

I connected the controller to another motor of mine and it worked, so the problem is definitely the motor. Anyone have any ideas on things I can look at or test?

 

[Ypedal]

Do you have another throttle to swap and test with ? and did you test the voltage of the battery pack upon cut-out ?

 

[RatoN]

...sounds familiar.

My motor died on me also, but i was asking for trouble.

Did it make a ratling noise before dying? Was is right after taking off from a stop? What voltage do you run?

 

[CGameProgrammer]

I connected the controller to my other motor without disconnecting *anything*. I literally had the working bike upside-down next to the broken bike, and I unplugged the phase and hall cables to the broken motor and plugged them into the adjacent working motor. I did not remove the controller, batteries, or throttle from the broken bike -- they stayed on, fully connected.

I did further testing. I plugged the hall cable to the broken motor and the phase wires to the working motor. The working motor did not spin.

Then I did the opposite -- plugged the hall cable to the working motor and the phase wires to the broken motor. The broken motor spun correctly. So the problem is with either its own Hall cable or the Hall sensors themselves. My guess is the sensors were installed very badly and the vibration from riding shook them loose. I'll try to order a replacement motor.

What exactly are the Hall sensors used for? Is there any way around them?

Raton: No rattling, it happened when going around 20 mph (at a guess) while accelerating, and I was using 48V. I was using exactly the same system that worked reliably on my other motor; the only difference was the motor itself, and the bike was 100% steel this time.

 

[RatoN]

well friend that make two of us who suspect the hall sensors in the hub. After 3 days of testing that is the conclusion i came to.

Lucky you, you have extra components to test the hub with. Can't wait to read your findings and solutions.

There is a useful link provided by 29a for testing the hall sensors.

http://www.crystalyte-europe.com/ -see left hand side/scroll down/about motor controller/check hall effect.

Did all the test and my voltage reading does not change as I spin the wheel by hand, as it should.

And no power comes out of the leads of the controller, probly due to faulty sensors in the hub.

Good luck

 

[CGameProgrammer]

I can't tell which numbers correspond to which colors -- I have a 5-pin mini-DIN (round) Hall connector. My colors are red, green, blue, black, yellow, not necessarily in that order. Anyone know? Only the female connector is labeled, but it's impossible to see which colors plug in to those.

 

[RatoN]

Yes, you will have to partly undo the hall plug to see the colours.

Take a small screwdiver, partly undo the two screws.

Then you should be abble to unscrew the plug and see the wires.

To test though is from the controller.

Black & yellow
Black & green
Black & blue

-Hall pluged in
-Power on
-Hub leads unpluged

 

[Lessss]

Man you have bad luck with motors.

Take this ooportunity to upgrade to better electronics!

Hall sensors are CHEAP. So instead of replacing crappy chinese electronics with more of the same spend 10 bucks and get some good hall sensors. That's what I did but I made e-ride pay for em.



It might also not be the sensor but the cable connection to the sensor.

 

[Lessss]

What exactly are the Hall sensors used for? Is there any way around them?

http://en.wikipedia.org/wiki/Hall_effect_sensor

No, not with a brushless motor.

 

[Lowell]

What exactly are the Hall sensors used for? Is there any way around them?

http://en.wikipedia.org/wiki/Hall_effect_sensor

No, not with a brushless motor.

For what it's worth none of my brushless RC motors have hall sensors. Ebikes speed controllers should be more like RC controllers. For example, 50V x 110amps in a 150g package:

http://www.castlecreations.com/products/phoenix_hv-110.html

I love being able to program my CC controllers with a USB cable.

 

[Lessss]

Ok let me rephrase. Not with cheap hub motors. :lol:

 

[Lowell]

Ok let me rephrase. Not with cheap hub motors. :lol:

Isn't the hall sensor requirement an issue with the speed controller?

 

[fechter]

There's no reason you really need hall sensors, they're just the cheapest way to build the controller. As pointed out, many RC controllers do not require hall sensors.

The hall sensors tell the controller when to switch to the next phase. It's possible for the controller to sense the back EMF from the magnets to detect the switching point. This generally requires a microprocessor-based controller with software. These are called sensorless controllers.

I wish somebody would make a nice cheap 5kw sensorless controller. It would sure make swapping motors a lot easier. Another one of those "I'll have to do it myself" things.

With most 'cheap' controllers, there's not really any way to bypass the need for the hall sensors.

Some controllers use an additional hall sensor to detect pedal input for a pedal-first mode. Don't confuse these with the ones inside the motor.

BTW: you might follow Raton's thread about trouble shooting and see what you find. Perhaps there's a manufacturing or design defect trend here.

 

[CGameProgrammer]

I wish somebody would make a nice cheap 5kw sensorless controller.

Does someone make a nice expensive 5kw sensorless controller? I really loathe Hall sensors now and would gladly fork over cash if such a controller could be used without major modifications.

 

[Lessss]

http://www.marotta.com/electronic_brushless_sensorless_controllers.htm

 

[Lowell]

http://www.marotta.com/electronic_brushless_sensorless_controllers.htm

That package is a little large for an ebike. Also seems on the heavy side.

This one looks interesting, but only 50 amps:

http://www.utmc.com/ProductPages/PH_motor.cfm

 

[Lowell]

Speed controller cooling ideas:

http://www.mcmanis.com/chuck/robotics/projects/esc2/speed2.html

And how speed controllers should be built

http://www.trossenrobotics.com/store/p/3199-AmpFlow-2-channel-Controller.aspx

 

[CGameProgrammer]

Nothing overheated; the controller works perfectly. But this is twice now that the hall sensors in the motor itself went to hell, and this time it wasn't my fault.

 

[safe]

Nothing related to your motor, but I just overheated my 750 Watt Unite. One of the little contact patches that the brushes touch simply melted. Not much to do but replace it. I was testing some new gearing and went up a steep hill in a tall gear surrounded by traffic and didn't want to stop. By the time I crested the hill the motor was smoking...

 

[fechter]

That water cooling looks like a bit of overkill. We don't want the FETs to be dissipating that much heat in the first place.

Depending on how much voltage you want to run, some of the RC sensorless controllers are reasonably priced. It's easy enough to interface a standard throttle to a servo tester.

http://www.castlecreations.com/products/phoenix_hv-110.html

This is very close to what a brushless controller ought to look like.

 

[RatoN]

Fechter,

What is the component inside the controller responsible for the hall sensors signals ''in'' and ''out'' (or is it just ''in'' maybe)?

Could that be the problem in this case and mine?
How can I test this component(s) on the board?

 

[fechter]

The halls hook up to the uPC1246 inside the controller. I believe this also supplies the power for the halls.

Since the hall supply voltage was shorted when the halls were hooked up, but the voltage went back up to 14v when you disconnected them, I don't see how the commutator chip could do that (although I suppose one of the inputs could be shorted somehow).

To test that, you would want to connect the red and black wires to the halls, but disconnect the other 3 wires (requires removing the pins from the connector). If you connect the red and black wires only and the voltage does not drop to near zero, then it might indicate a bad commutator chip.

 

[CGameProgrammer]

The ElectricRider guys called me back today; turns out the wire to one of the Hall effect sensors had come loose from the soldered connection to it, and running the motor caused a short that killed that sensor. They replaced the sensor and wire.

 

[fechter]

So much for quality control :?

 

[CGameProgrammer]

Yeah it's really annoying; wasted time and $50 shipping. But I haven't heard of this problem happening before, and my other motor seems very solid, so hopefully that's the last time this'll happen.