Re: How to Determine the Wiring for a Brushless Motor

My undying thanks to Fechter for his awesome post; finding it turned my Schwinn Sting-Ray Spoiler e-bike conversion from a mechanical nightmare (who the heck designs a bike with the front brake rotor on the RIGHT side?) to an easy electrical hack (flop the hub and run it backwards!)

As a bonus, his truth table made it MUCH easier to zero in on the right combination; for example, if you have a known "runs good forward" combination, you know ten motor/Hall combinations you can safely skip. When you find a runs-rough combination you can kill that motor phase combination *regardless of the Hall setting* because runs-good motor phase settings don't have any hall settings that run except the correct one.

Now all I have to do is eke out a few more millimeters of clearance between the hub bolts and the brake caliper and I'm in business...

Not sure what you mean by kill phase settings and runs good phase settings. All phase combos are "good". They are either reverse or forward, and each phase combo will have one good hall combo to match it. Net result of 36 possible combos of hall/phase, 6 are good. 3 are forward and 3 are reverse. If you have a reverse, then just swap 2 phase wires and find the new good hall combo and you'll have your forward. Where most run into trouble is swapping too many wires at once, leading to unnecessary confusion.

Hrm, his original post is now missing, which doesn't help here.

If you look at his truth table (attached), you see that certain motor combinations will *never* produce smooth running in either direction. "Runs good" motor wirings produce either good running or "NR" -- no rotation -- across all of the possible hall sensor wirings. "Runs rough" motor wirings only produce rough running or "NR" across the hall sensor wirings.

I'm not making an electrical argument here; I have no idea why that is. But if the chart is valid, then you should abandon the motor combo if you encounter rough running.

njloof said:

Hrm, his original post is now missing, which doesn't help here.

Click to expand...

Do you mean this one?
http://endless-sphere.com/forums/viewtopic.php?f=16&t=3484
I'm guessing the reason it appears that his post is missing is that you perhaps tried to reply to his post, and couldn't, because his is in a no-discussion forum (technical reference area). So then you made your post is in a different forum than his original, and not connected to the original thread?

Ah, likely true. Thanks.

njloof,

It's much easier than all that. If you have a good forward wiring combination, simply swap 2, any 2, only 2 phase wires, and try the other 5 combinations of the hall wires until you get the right one. 5 attempts max is all it takes, so forget the spreadsheet and do it the easy way.

Does that imply that the spreadsheet is incorrect? Your version would suggest there should be 3 valid forward-running combinations and 3 backward-running, since they are just rotations of one another.

njloof said:

Does that imply that the spreadsheet is incorrect? Your version would suggest there should be 3 valid forward-running combinations and 3 backward-running, since they are just rotations of one another.

Click to expand...

Correct, there are 6 valid combo's, 3 forward and 3 reverse. The only exceptions to that are a few BLDC motors that require a special controller, but those with some odd timing wouldn't run in reverse. The correct firing sequence for forward is for example A-B-C, which since it repeats is the same as B-C-A and C-A-B. ie all end up A-B-C-A-B-C-A-B-C-... just with a different starting point. You just have to match up the same hall order and sequence. Swap any 2 phases and you get reverse C-B-A, which since repeating is the same order as B-A-C-B-A... or A-C-B-A-C...

It's by far the most commonly made difficult but actually easy thing in DIY ebiking. The most common mistake is swapping around more that 2 wires at a time....well that and not starting with the realization that every phase combo is correct and you just need to find the right hall combo of only 6 possibilities. The converse is also true, but that just leads to more confusion, so I never mention it.

Well, then the job was easier than I thought.

Regardless, the discussion set me on the right path, since I would otherwise have wondered why swapping the same two colored wires for motor and sensor didn't work.

Welcome to the forum, and best of luck with your build.
Brian L.

It might be good to rewrite and optimize the "standard recommended" procedure. Saying there are 36 possibilities to explore is doing a disservice to newbies.

The way I approached it is to analyze which wires were easy to change (the phase wiring is easy to switch on my setup) and which were hard (the hall sensors are small fragile difficult to swap pins). So I want to minimize the switching of the tiny hall wire pins.

Assuming the +5 and ground of the hall wires are correct that leaves the three hall signals to play with. Only two combinations of those three wires are necessary to explore. One corresponds to forward rotation and one to reverse. We don't know which direction corresponds to which wire configuration so we may have to try both. The two combinations differ by the switching of one pair of wires. For each directional hall setup we have to try all combinations of the phase wires of which there are six. If none of those runs forward smoothly at low current then we have to swap one pair of hall sensor wires and again try the six phase combos. Twelve to test. Done.

Perhaps there is some special motor/controller combo this doesn't work for, but testing all 36 combinations should be a last resort, and a simpler procedure similar to the above tried first.

That sounds much more like the process I went through.