How to reverse brushless motor, even if it wont.

[dontsendbubbamail]

On my 80-100 Turnigy with 120 deg hall effects, leaving the hall effects in place and rotating the phase leads changed he direction of rotation. A-B, B-C, and C-A.

Bubba

 

[Hyena]

Hyena,
Maybe it's a 60°-120° hall phase angle conflict between the motor and controller.

It's unlikely, I ordered 2 direct from GM, one worked fine out of the box but the other has the above problems. Plus the controller is set to auto to handle both.

What everyone really needs is one of Lyen's ebike testers.

Yeah I'll have to get me one of those.

 

[Kingfish]

@GrayKard:
I was intrigued by your chart, and having a similar issue last night I decided to map out the Lyen's Mini-Monster. Initial inspection confirmed that the colors for Channels B and C were swapped as compared to my 9C 2806 Hub/ebikes.ca 25A Controller. I resolved this by replacing the terminators with the correct AAP connecter color. However the hub wouldn't turn when I hooked it up. Unknowingly I had forgotten to swap the Halls wires on the Controller as well; figured it out after 6 tests and lots of ES-reading. Subsequent tests - especially #7 confirmed this.

However I am quite intrigued by the two super-fast Reverse configurations.

Digressing...

• With my original setup at 37V nominal, my Freewheel speed was about 32/33 mph.
• With the Mini-Motor, I am running 74V nominal and expected pretty much a doubling in speed - which is confirmed by Test 7.
• If I have time before I leave for holiday (taking the newly minted ebike) I would like to configure the last two tests to determine the utility of using a super-fast forwards. In honesty I was afraid to spin the rim beyond 85mph; there was more throttle left to go! Therefore I only have recorded what the CA MaxS stated. It was a wicked experience feeling the gyroscopic effects of a hub spinning above 70mph

Enjoy, KF

 

[GrayKard]

However I am quite intrigued by the two super-fast Reverse configurations.

Digressing...

• With my original setup at 37V nominal, my Freewheel speed was about 32/33 mph.
• With the Mini-Motor, I am running 74V nominal and expected pretty much a doubling in speed – which is confirmed by Test 7.
• If I have time before I leave for holiday (taking the newly minted ebike) I would like to configure the last two tests to determine the utility of using a super-fast forwards. In honesty I was afraid to spin the rim beyond 85mph; there was more throttle left to go! Therefore I only have recorded what the CA MaxS stated. It was a wicked experience feeling the gyroscopic effects of a hub spinning above 70mph

Enjoy, KF

Those super fast configurations will only run with no load. The second you try it with weight it stalls. Basically what you are getting is very high speed with almost no torque.

Gary

 

[whatever]

an old table I have shows all combos tested, motor was x5, controller a non popular brand
got two smooth forwards, and three smooth reverse, shows which ones ran but roughly also.
seems odd two forward and three backward that work, I would have thought three forward also, maybe I made a mistake, will see if I have some more tables filled out and post

 

[whatever]

heres another one x5 with infineon
2 forwards and 4 reverse!!

 

[rkosiorek]

Mightyvolt,

good that you found the correct connections to get your motor to work properly, in the chart above i have taken the results from you chart of all of the working combinations that you found to have the motor work in reverse. i have just rearranged the results ao that on the controller end the A,B,C aka Green,Yellow, Blue sequence is kept constant in the first column. then i listed the working reverse combinations you found for tests 1,2,3 and 4. notice the neat pattern that results for tests 1,2 and 3. all of them are the essentially the same sequence except that in test #2 the start is shifted by one position. and in test #3 it is shifted one position from #2. all of them end up being the same a,b,c sequence.

the only problem is test #4 it does not folow the A,B,C sequence. even though it probably does run in reverse. i would bet that it is a conbination that ran hot and used excessive current.

rick

 

[John in CR]

the only problem is test #4 it does not folow the A,B,C sequence. even though it probably does run in reverse. i would bet that it is a conbination that ran hot and used excessive current.
rick

My money would be on a test error (maybe a false positive like you said), since there's a missing forward that is the same sequence as the other two. Thanks for the effort Rick, because your arrangement clearly displays why there are 3 good forwards and 3 good reverses with the vast majority of the motors. That's because the 3 are actually the same combination, just a different starting point, and the starting point doesn't matter with a circle.

John

 

[rkosiorek]

another thing convinces me that test #4 has an error. if the sequence A,B,C is the firing order. how can you have the same firing order for both forward and reverse? in test #4 Mightyvolt shows good results for both forward and reverse. one of those has to be incorrect. here are the results of tests 4, 5 and 6 presented as correct forward combinations



now these make sense and fit a pattern. Test #4 is actually a good Forward combination and a bad reverse combination. since it fits the pattern test #6 must be included in the results for foward combinations. i have confirmed this using a 9C motor and an infineon controller. the 9C is also a 120 degree motor so the results should be the same as for the Crystalite or any other motor that uses 120degree hall spacing.

till now the charts show the halls in the sequence A, B, C. for the Reverse this results in a downward staircase on the chart. for forward this results in the opposite or an upward staircase.

to show the same or downward staircase for forward the chart just has to reordered so that the A,C,B (same as C,B,A) firing order is used for the halls.

 

[whatever]

nice work, my guess is a 60deg will give exact same patterns, 3forward/3 reverse

 

[njloof]

This thread was awesome. The spreadsheet in particular helped me find the reversal combination much faster -- if you use it as a "truth table", and cross out the rows and columns that won't get you to the right solution based on the results you get, you can reason your way to the right combination in no more than 6 tries. I did it in five

I needed to do this in order to get the disc brake on the "correct" side of my Schwinn Sting-Ray, which for some bizarre reason was designed to be on the right (meaning the brake caliper is at some weird angle in order to use a left brake caliper on the opposite side!). So I cut the motor cable in half, crimped and soldered spade lugs on the wires, and used this as my cross-connect:

http://www.radioshack.com/product/index.jsp?productId=2103229

Made it easy to see what I was doing, and when I found a working combo I just stuck the whole thing in a small project box and sealed it up.

So thank you guys, I was lucky to stumble onto your forum from The Googles.

 

[Tyro65]

I'm trying to find out if this motor can be reversed. But to tell you the truth I don't even know if it is sensorless or has hall effect sensors. Help!
http://www.czbrushlessmotors.com/750w-48v-electric-trike-conversion-kit-p-60.html

Thanks.


Stan

 

[njloof]

I'm trying to find out if this motor can be reversed. But to tell you the truth I don't even know if it is sensorless or has hall effect sensors. Help!
http://www.czbrushlessmotors.com/750w-48v-electric-trike-conversion-kit-p-60.html

Just judging by the 6 wires coming from the motor, it has Hall effect sensors.

You should be able to find a reversing combination for that motor using the ideas in this thread.

 

[Tyro65]

Thank you for confirming that. I will try some of the suggestions here to see if I can reverse it.

 

[Tyro65]

I see that the motor I'm considering can be reversed. But what kind of switch would I need? A double-pole, double-throw switch?

 

[njloof]

I see that the motor I'm considering can be reversed. But what kind of switch would I need? A double-pole, double-throw switch?

Triple pole, no?

 

[Kingfish]

I just realized this material has been covered before. With 2WD, we are running in both directions at the same time.

Two threads both relate to the method that I used last:

• RUST: DIY Hub Motor Repair & Restoration; idea born here.
  Known to Work Brushless Motor/Controller Wiring Diagrams; documented formally here. This thread should become a sticky - or included in a sticky (if not already) as a good reference.

The premise, instead of using "color" (which can be arbitrarily defined) is to use the actual identity of the items in question and correctly apply the mapping - regardless which of three objects (Controller, Motor, HE), as long as two items are paired we can ascertain the correct mapping of the third. Color then becomes unimportant and a distraction.

Take care, KF

 

[bobc]

I have the feeling that a number of factors are confusing an essentially simple issue....
1) with 3 phase induction motors or a sensorless controller, just swap any 2 of the 3 motor wires & it'll go backwards.
Because of this I suspect folk reason that to go backwards you have to swap 2 power wires. This is not true.
With a sensored motor, you will ONLY get proper operation when the hall sensors are in the correct (same) phase sequence compared with the motor power leads.
To reverse the torque direction you actually need to simply alter the phase relation between the magnetic field (as reported by the halls) and the motor current by a fixed angle.
If your hall sensors are physically moveable, this is easy to do (by physically moving them).
If not, rotating the motor leads (eg A-B, B-C C-A) will give you a 240degree phase jump, and this could work pretty good (180degrees for the torque reversal and a couple of 30degree phase advances).
It should be possible (I've not tried it) to come up with a combinational logic circuit that will do a similar "rotation" to the 3 hall sensor signals.
Hope that helps somebody ;^)

 

[justin_le]

It should be possible (I've not tried it) to come up with a combinational logic circuit that will do a similar "rotation" to the 3 hall sensor signals.
Hope that helps somebody ;^)

The logic is very simple, you just need to invert all 3 hall signals with a NOT gate. What you want is a 180 degree rotation in order to drive current the opposite way through the windings for a given hall position, and inverting the 3 halls will do that. Advancing an additional 30 degrees will advance the timing far past the point of being beneficial.

 

[Matador]

https://endless-sphere.com/forums/viewtopic.php?f=14&t=98771&p=1457184#p1457197

Any advice on this one ? I'm down to three combination out of 36... Which one of the three should I choose ?