First time powering on!

[jameswalker]

Afternoon all,

I have just finished piecing together my build, but the wheel does not turn (properly).

I have a fully charged battery at 66.4V. It is able to offer decent current draw, I have tested this. 20A through a high power light bulb, no problem.
I am using a Lyen controller, 12 FET 3077 MOSFET Mark II controller LYEN edition.
The motor is a BMC V4-T.

The motor is proven to be OK; I span the wheel reverse and took 3 phase voltage readings between different phases. All gave up to 4V AC reading, indicating that the motor is wired correctly and behaving itself.
The throttle in the off position reads 0.8V and rises to 3.6V fully open. The positive throttle feed is appx 4.5V.

When I open the throttle, the motor makes a high pitched whine for 1 second then cuts out (the cutout time is software controlled). Additionally, the wheel does twitch but it does not rotate.

I have read various posts where people try various permutations of phase wiring (yellow to green etc) and the associated hall sensors wiring, but nothing has worked so far.

What should I be investigating next?
Will the controller turn the wheel if there is a problem with the hall sensors / hall sensor wiring? I havent tested these at all; I assume they are OK as they are new, but no proof.
Is there something in the programming of these controllers that I am missing?

Thanks,
James

 

[wesnewell]

Sounds like phase and hall wires out of sync. Normally you have to swap the hall wires if you swap the phase wires around.IOW's, yellow to green on both phase and hall wires, etc.

 

[jameswalker]

I will try deterministically tomorrow all of the combinations

Thanks,

 

[cal3thousand]

First, is that a sensored controller?

Second, you need to test the hall sensors on the motor with an ebike tester spinning the motor backwards. Or lookup how to rig a tester with a breadboard.

 

[ken1645]

I'm just guessing as this information I found is for a different controller, but I'd start by trying all these combinations of the phase / hall wiring.

"
Phase/Hall connections for different motors. *=verified.
Crystalyte motor wire color lining:

Controller: Motor:
HALL:
Yellow Yellow
Green Blue
Blue Green
PHASE:
Yellow Blue
Green Green
Blue Yellow
_______________________
Ebikekit 9C motor wire color lining (around 2A unlaced freewheel at 44.4v):

Controller: Motor:
Hall:
Yellow Yellow
Green Blue
Blue Green

Phase:
Yellow Yellow
Green Blue
Blue Green

or

Controller: Motor:
Hall:
Yellow Yellow
Green Blue
Blue Green

Phase:
Yellow Blue
Green Green
Blue Yellow
_____________________
* Golden Motor 901/902/MagicPie.
Controller: Motor:
HALL:
Yellow Green
Green Yellow
Blue Blue

PHASE:
Yellow Green
Green Yellow
Blue Blue
"
from: http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=29469

-Kendall

 

[John in CR]

I will try deterministically tomorrow all of the combinations

Thanks,

That is never necessary, because there are 6 valid combos, 3 forward and 3 reverse. Every phase combo has one valid hall combo, and every hall combo has one valid phase combo. The most common mistake is changing halls and phases at the same time. Keep the halls or phases static, whichever is more difficult to change, and try all 6 combos of the other using small pulses of the throttle. You will get a valid result in those 6 possibilities, though it could be reverse. If reverse, then swap 2, only 2, any 2, of the wires you were keeping static, and go find the right combo of the wires you were swapping before.

With a geared hubbie, a valid reverse is the motor spinning smoothly inside, but the freewheel means the wheel doesn't spin.

To verify a proper wiring, the motor starts and spins smoothly and quietly with a low no-load current. Be alert for false positive results where the wheel does spin, but the wiring still isn't right. They are generally somewhat noisy, especially on startup, and have very little torque if you try to ride it. All false positives I have seen are opposite direction from the valid combo for that set of static wires.

Take your time and be systematic. Anyone who suggests 36 combos and using a spreadsheet doesn't understand the relatively simple logic/math problem that finding correct wiring represents.

John

 

[DAND214]

BMC V1 & MAC is yellow / green on phase and halls.
On an Infinion controller. So should be the same.

Woked for me.

Dan

 

[jameswalker]

cal, yes the controller is sensored.
i do need to test the hall sensors. I guess that is something I need to look into.

John, I actually did cover those 6 phase swaps last night keeping hall wires consistent. I guess this goes back to the first point of testing the halls?

Dan, That is the first thing I tried (having read it on a website) but sadly it didnt work.

Finally, would faulty halls cause the controller to go into overload shutdown? As I only hear the whine of the motor for a second or so. Or does another fault point to that?

Thanks,
James

 

[cell_man]

You shouldn't really be using that controller at so high a voltage. It likely has 63V Caps, it has 75V fets. The equivalent controller I supply as approx 50V nominal (maybe a little higher, but not much), 60V Max input voltage. Any higher than that is a bad idea IMO.

Swapping yellow and blue on hall and phase is likely required (there are other combination that will work too), that is a common swap for most motors to work with the Lyen Infineon. The BMC uses a pretty common hall and phase combination so far as I know, same as Mac, Bafang and most motors I have tried.

 

[jameswalker]

Cellman, you are correct the fets are rated to 75V. I am running at 66.4V max. The caps are not likely to be only 63V, as I explicitly asked Lyen if the controller was suited to run at 67V. He said it wasn't a problem and I will be fine so long as I do not exceed FET max voltage. I can only assume the caps were changed. Furthermore, I am not running high current by any stretch, so the power draw will remain far below the FET maximum, or even recommended spec.

Thanks for raising the point though, I will have an inspection tonight to see if they are indeed 63V caps. Certainly another point to check.

I will try the phase swap

Thanks,
James.

 

[dogman dan]

And if what John in CR said to do doesn't work for you, find the bad connection at a plug that is the real problem. His method works great. You don't need 36 tries to find it.

I like to pick a phase order, make damn sure those contacts are fully into the plug housings, then with the three hall sensor wires removed from the housings, start making combos of the halls.

With the Y B G halls wires out of the housings, you can be sure they are making good contact on each try.

A bad hall sensor is possible in a new motor, but unlikely.

 

[jameswalker]

Connector integrity; the phases use a non standard connector from Germany. They are extremely high quality, better even than the EC5 connecrots that I use for power. I dont question there is a problem with those at all. The hall sensor conenctor though... it is the standard one offered with the controller. That might be a cause of problem. If following Johns advice, then testing the hall sensors and their supply yields nothing (and having a peek into the controller for component ratings) I will remove the hall sensor connectors and make sure manually that there is a valid connection.

What happens if I connect the 3 phases but no sensing wires at all?
Can I run this infineon controller in sensorless mode at all?

Thanks,

 

[jameswalker]

Update: controller has 100V caps, so no problem there.

I tried the phase combos as described above, yet no rotation still.

Onto the hall sensors. the red line has 4.5V, each of the sensor outputs measures 4.9v (pulled up internally to the controller). Rotating the wheel in either direction, the voltage does not change for any of the 3 hall sensors. It is as if either no magnet is in front of the sensor, or the pullup is too strong within the controller. Or of course, the hall sensors are dead...

I will assemble a quick test rig with a bench supply and a varying pullup resistor to see if anything changes.

What on earth could cause all 3 hall sensors to be dead?? It is a new motor....

 

[999zip999]

I had a bmc 600 h.s. and it would NOT work with a Lyen 12 fet. Ilia at ebikessf had problems with the Lyen controller working with some bmc's. I did have it work on a different bmc motor. So don't no why they wouldn't work with certain bmc motors. Had to use a crytalyte analog controller. ???

 

[jameswalker]

I don't suspect there is anything wrong with the controller. I think the hall sensors are dead. Bench testing the hall sensors and they still gave no readings.

I am going to strip the motor down a bit, to see if everything looks OK. I should probably individually test the hall sensors too, off the PCB. A bit of a pain... I don't have a cassette tool handy to remove it! :x

 

[dogman dan]

Cover should come off with the freewheel attached.

Odd that a new motor would have 3 bad halls. I wonder who put overvolt into the hall sensors wires.

 

[jameswalker]

Ah yes thats true, it should come off as one. I will give it a go later. I might not get around to testing it until tomorrow... i've had enough for one night I think!

So... reading conflicting forum posts stating which hall sensors to replace these with. Anyone care to comment?

Thanks,

 

[jameswalker]

I have found the problem. The +5 (red) wire to the hall sensors had snapped at the point where the cable meets the axle. Not good for a new motor!

Pretty much worst case scenario. Repairing the single cable is going to be EXTREMELY difficult. I could either try and repair it, taking a couple of hours or strip the lot out and re-wire it all. Thoughts??

 

[ken1645]

I have found the problem. The +5 (red) wire to the hall sensors had snapped at the point where the cable meets the axle. Not good for a new motor!

Pretty much worst case scenario. Repairing the single cable is going to be EXTREMELY difficult. I could either try and repair it, taking a couple of hours or strip the lot out and re-wire it all. Thoughts??

Contact who you bought the motor from and ask for an exchange?

 

[jameswalker]

Sadly I am in the UK and the motor was from USA. Postage is too high to replace it. Its also laced to a wheel

 

[John in CR]

The stock phase wires are too thin anyway and waste power and battery on every ride, so I say do a new and better harness for a better than new repair. If you simply want to get on the road, then push the harness into the motor far enough to expose the break and fix it.

 

[jameswalker]

So with the sensor supply fixed, the bike moves! After the repair, I did a full bench test at 5A and the wheel span nice and smoothly. A couple of hours later, the wheel is installed on the bike and it moves! Still running 5A (16s, so thats still 300W). Time to kick it up to 15A (~900W) to see what happens!

Thanks for the assistance all. Oh and John, I will indeed rewire the entire hub. But for now, I need a morale boost by test riding I will probably source some stainless braid to protect the wires from anything in the future.