3 phase wire combination

Hello,

Ive been lurking on this forum ever since I converted my my bike to a hub drive. My motor was buzzing/vibrating as soon as I finished the conversion, but i never thought anything of it. Well, 30 miles later I have a fried controller.

Ive read many previous diagnostic threads about this issue. There are obviously many posts about checking "all 36 combinations" of phase and hall wires. I understand that 3 of the phase combinations will drive the motor forward and 3 will drive it in reverse.
Im not posting here for diagnostic help on this issue.


For this question, lets forget about the hall sensors. Can some one explain why one phase combination would work better another?

Why would one phase combo fry a controller, and another works flawlessly?



Attached, is a pic of my old Trek that I converted. Its 2wd with the front motor on a switch so I can turn it off to conserve energy. It was such a blast when it worked.

Sent from my SM-G965U1 using Tapatalk

My understanding of it is pretty primitive, i was an ag major not E engineer. But if your motor runs forward but vibrates, its real likely that you have it running on two of the phases, with the third fighting it. This overheats the controller. wrong combo may run forward, but not too smooth, or with the power you expect.

On the other hand, your wiring may have been correct, but the controller was defective from day one. Again, overheating the functioning two wires and the fets running them in the controller.

Lastly, a poor connection of phase wire, or more likely one hall wire, will also result in running on "two of three cylinders". That one is the single most likely reason an e bike won't run, or run right.

The phase wire firing order coming from the controller has to match the order the coils (and sensors) are installed in on the motor.

1,2,3, from the controller, must match 1,2,3 on the motor.

If one side or the other is not in order, you get weirdness that needs to be addressed right away.

...but the controller was defective from day one.

Click to expand...

Maybe, it's happened to me w/ genaric controllers.
When the "Cute" motors came out years ago, they never seemed to match the controller colors.
Anyhow, I've tracked down the right combo at least a dozen times and never damaged the controller.
I think something else is going on.

Thanks for the responses.

Follow up question:

What in the controller does this fry?

Sent from my SM-G965U1 using Tapatalk

Can some one explain why one phase combination would work better another?

Click to expand...

If you match the phases incorrectly then you get the wrong firing order as people have already stated, it can be compared to a combustion cylinder firing too early or late, fighting the rotation direction and the other cylinders.

What is also lost is the controllers ability to control since motor stutters/misfires and jerks. This means that there will be high current and voltage peaks created that can blow a number of components like fets, caps and internal dcdc converters etc.

What blows then? I can’t really say, depends on the controller design. Seems fets are among the common ones but i guess this is partly because it’s so easy to find when they’ve exploded.

Easy to find only sometimes with a visual because there are visual clues like black bits and physical damage, but otherwise you'd have to break out a digital multimeter and measure the mosfets high/ow side. Replacing them can be quite the challenge, just by the soldering job alone and I dont know how closely matched the fets have to be. I've seen posts of people using the oscilloscope to match the same (?cliff?bump?) trace amongst the exact same mosfets. That might be the case with good quality controllers, because the generic controllers would all use cheap generic mosfets, but who knows if they closely match those at the controller factory or its just random and doesnt need to be closely matched.

https://ebikes.ca/learn/troubleshooting.html

Please confirm you are using 2 controllers; one for each motor.

Assuming you're driving each motor with it's own controller, then having them enclosed in a bag with no airflow (they're not visible on the bike so I'd assume they are in that triangle bag) can damage them from overheating, even if the failure is not immediate it can age the parts faster so that they fail (much) sooner than normal. Mounting them out in the airflow is a good idea, and can be required for some systems.


As for what could fail in a controller and why, when the wrong phase/hall combo is used, it is typically the FETs that will fail first because they are being overloaded and overheated, and voltage spikes from incorrectly timed changing currents in the motor windings can kill them too. The gate drivers are often also destroyed by the battery voltage passing back thru the internally-shorted FETs, and sometimes the gate drivers can fail in a way that feeds that back to the MCU, killing the brain chip of the controller, too. In some cases the FETs don't directly die, but the phase wires or the motor windings actually melt their insulation and allow shorts between phases, and then that kills the FETs. (this usually leaves the motor much harder to turn than usual, even with the controller disconnected).

Why it happens is that when the position sensors (halls) are not setup in the right order to match the phases, the timing of the pulses sent to the motor are wrong, and it takes a lot more current to do the same job, and feedback from the motor's phases to the controller is not timed correctly so both controller (FETs) and motor (windings) and phase wires all heat up (much) more than they would if timing were correct. That's what the "false positive" references are to. Other wrong combinations may not even allow the motor to spin.

If you like you can look up "advanced timing" and "retarded timing" on motors, as this is what is happening with a false positive combo. It's just much more extreme in this case than when it is typically used in a system on purpose, and so it has much greater detrimental effects.