MPIII suddenly goes DOA

[dallasdick]

I recently had to replace my internal controller, after a CB failure. It worked great after the repair, as I rode my KHS XTeam to school (about 9mi.). Coming home from school it was running fine and suddenly just quit. I cycled the power switch and it acted like it wanted to go, but I could feel the force of the wheel stop as I put the tire back to the pavement. Now when I cycle the power off-on the internal controller "clicks" but the wheel won't turn at all when the throttle is applied.  This MPIII kit is under 2 months old and in less than 3 weeks it has failed again.  I am very disappointed with the reliability of this set up and hope that someone can shed some light on what the failure and it's cause might be.  Thanks for any assistance you can offer.  I can TS the problem with a little guidance.

MPIII/LiFePO4 48V/20aH(16cell) mounted on a KHS XTeam full suspension MTB frame, custom built from original factory spec.

DallasDick

[David]

Hello DallasDick,  Please tell me your order reference number and your email address to wyh@goldenmotor.com, It is most likely that the controller is damaged.  Now we further optimize the controller, and enhance the stability. I will send a new controller to replace it.

BTW, would you like to return the broken controller to us?  so that we can find out root issue, and constantly improve our product quality.

please email to me ,wyh@goldenmotor.com

sorry for any inconvenience caused


thanks

[dallasdick]

Thank you for the response David.  I will be taking the MPIII apart tomorrow to troubleshoot. I will email the pdf. of my invoice for my purchase ref number from BMS Battery.

[dallasdick]

Well after taking the pie completely apart and finding very small pieces of broken magnet chips had scored the stator surface and the field surface, I spent two hours ensuring all the fragments were removed, the surfaces had no sharp edges, and the scores were sanded flat.  I then put the pie back together with the new controller David provided and viola! It's running again.  The question now becomes: what caused the fragmentation of two magnets side by side? Did something come loose from the controller and drop into the stator area?  None of the screws were missing, no foreign bodies were discovered, just magnet chips that wedged between the field and stator, stopping it cold.
 Heres and interesting finding as well...when I connected to the software and set the regen for 50%, as I have done in the past; I took it for a spin and my output current was severely limited.  Then I notice that when I release the throttle and the reapply throttle the regen kicks in all by itself, slowing me down until I get deeper into the throttle where it will pick up speed but limits the output. Never had that happen before.  I decided to forgo using regen altogether for now, which sucks because it really enhances short stop braking.  Needless to say I have no confidence in this system at all, and that's a big problem for me with my commutes to work and school.  My old Wilderness Energy bike is bulletproof at this point.  Never lost a controller, only problem I ever had was realizing I would need to waterproof the controller. I have 100% confidence in the WE, none in the MPIII. Also, has anyone figured out what the optimal perameter set up is on the software?  Anyone found an easy way to squeeze some more rpm out of this system?  My WE runs an honest 31mph with the 48v /20aH LiFePO4 battery and that bike is 8lbs heavier than the KHS with the MPIII, which barely hits 26mph with the wind.

[Bikemad]

The question now becomes: what caused the fragmentation of two magnets side by side?

As the gap between the stator and the magnets is very small, it could be possible for the axle to flex slightly under very high impact forces that typically occur when riding into potholes in the road or mounting kerb edges etc.
This type of sudden high energy impact could easily cause damage to the fragile magnets if the axle deflects enough to allow the stator to make contact with them.

Regarding the speed issue, if you want more speed from your 48V pack you will need to use a different controller that produces a higher rpm.
Some of the the Infineon controllers can be programmed (and then switched) to 125% for a faster rpm. I seem to recall that some of the Kelly controllers also have this high speed function too.

Another alternative would be to use a higher voltage controller and battery pack for an even higher top speed.

Heres and interesting finding as well...when I connected to the software and set the regen for 50%, as I have done in the past; I took it for a spin and my output current was severely limited.  Then I notice that when I release the throttle and the reapply throttle the regen kicks in all by itself, slowing me down until I get deeper into the throttle where it will pick up speed but limits the output.

Did you try using the controller (and regen) before you adjusted the regen setting with the software, or could it be that the current is simply lower compared to your previous controller?

Any chance you can load the current configuration from the controller and post a screen print?
(Press "Alt" and "PrintScrn" together to capture just the open program window instead of the entire desktop)
Paste the captured image into MS Paint (or your preferred drawing package) and then save it as screenshot.jpg
(or something similar) and then attach it to your next post.

For some strange reason the regen operation on your new controller does not seem to be working correctly, perhaps it's an adverse side effect from the controller being further optimized to enhance the stability?

Alan
 

[Leslie]

When pulling you wheel apart it is easy to drop a washer or screw into the motor and miss it.  Once a foreign object gets utterly chewed by the motor, it would be hard to notice any difference between magnet fragments and the foreign body. Im not saying this is the case with you, but it is good to advise people that do repairs of this anyway, so double check tripple check when doing repairs as usually after something gets chewed up in the field may be hard to tell.

The regen you maybe experiencing may not be regen at all.  Check your controller phase wires are not damp, or there is any connection between them. Look for insulation wear on all the wires, or bits of magnet that may of got onto the controller. Check all your wires for moisture.

To be honest the GM controllers are crap IMO.  Too many people who ride Ebikes with much experience agrees, I hate coming to these forums and reading failures.  For very light use they maybe ok.

But I've been a member here on and off with different accounts for years and and put heaps of effort in helping people here. Yet I still see no effort to fix these controller issues.  I've seen people come and go, make requests in design and GM has delivered, however when I asked for 100v 12 fet controllers as the default for 48v systems and suggested making a replaceable internal controller part of their battery and not the motor, GM has always failed to deliver.  IMO Longer phase wires makes a lot more sense than longer power conductors in order to make regen more forgiving. Electrically, layout wise, one set of conductors, be it, phase or main power cables, is going to have some length to them, they got it upside down IMO.  And DC is more efficient transferred over longer cables via PWM and hall timing events than pure DC from a battery.

It all might be cost effective for GM but for $50 or so more for a kit, wouldn't the missing headaches be worth the extra expense?

GM can attract plenty of business with its awesome design looks, it a good selling point, but people are prepared to pay more for reliability.

[Leslie]

Phew now I got that off my chest.

What is the torque like on the GM wheel?  I have an old MP v1.2 and I had to get a whole different controller to experience the torque it can deliver.  Otherwise go for a different wound motor for speed, or a higher voltage battery. By the sound of your usages it I would guess that a 60v 10-15ah pouch pack would make you happier.

Because of the type cargo bikes I require for my uses, and the weights I drag around, I could burn a "WE" motor out very easily and it would be smokin.  The MP is a very cool running powerful motor that also loves higher volts. 

The MP has more potential than a lot of smaller diameter motors.  Potential is the definitive word here.  Whether youre getting that with your controller and battery is a different story.

[Morgen 3Eman]

Hi Les,

So we finally agree on something!   Can you imagine the market Golden Motor could command if they built reliable stuff! Gorgeous and fragile are hard to sell. 

I just ordered a Lyen 12 FET controller.  Any tips for installing it? 

TTFN,
Dennis

[Leslie]

Yes, Blue and green are swapped.  I think.

You can get more current into the motor with shorter phase wires and use some connectors or solder  the controller side close to the hub, this makes it easier as you can use a thinner gauge into the motor and then connect heavy gauge from the controller to your new wires.  Unless your pulling over 50A cont, a smaller gauge out of the motor, just over a foot long, connected to a heavier gauge, then onto the controller should be adequate and much easier to do..

I used an old PC mouse cable for hall sensor wires.  Its damned reliable. Red for +, Sheild for -, white for yellow, green for green, and black for blue.  I have a new standard USB cable to Ebike hall sensor color conversion.

Make sure you strip/burn or how eva you figure, all the shellack off motor winding ends properly before you solder the new phase wires in place.

And in my experience don't run the program cable while the controller is on, I found the program cable might FAILZ to work anymore.  The program cable and software GM offers is much better than my infineon get up.   They got that one right.

Oh and if you cant get the software to work, look for an XP PC.  I had nightmare with WIN 7 and the EB312 software/

[Bikemad]

Yes, Blue and green are swapped.  I think.

Blue->yellow
yellow->blue
green->green

Try a simple swap.  This is how the Infineon controllers hook up to a Magic Pie.

Hi Spacey,

As far as I remember, the Infineon controller wires match up the same way as the magic controllers.


Yellow-Yellow

Green-Green

Blue-Blue

Same for hall wires, Oh except I think the power wires on the halls were different. I think it was red-red
and maybe brown? I forget. I have it all covered up now in water proof tape!

Lyen Controller to Golden Motor MagicPie motor wire color lining (5/19/2011):

Controller: Motor:
HALL:
Yellow Green
Green Yellow
Blue Blue

PHASE:
Yellow Green
Green Yellow
Blue Blue

Now this is really getting confusing.

Dennis, I would try the Green and Yellow swap first as this is the one that sticks in my mind.

Alan
 

[Morgen 3Eman]

Now I'm getting scared!

Here is what I discern from your generous array of comments:  When swapping to an Infineon controller, the MP3 will rotate backwards if the sensors and phase wires are hooked up with matching colors, so I need to swap two Hall Effect sensors and two phase wires to make it all work the normal way.

Is that correct?

TTFN,
Dennis

[truly_bent]

The quote from endless-sphere is correct. My Lyen/Infineon controller required the yellow and green wires for both the phasing and hall effect sensors to be swapped.


Edit: I just noticed that you bought a Lyen controller Dennis. I think you'll like it. The only thing I haven't quite sussed out yet is the regen. The software setting uses a feedback voltage level (mine is set for 55V). The result of that setting is a pretty abrupt deceleration (which is okay cause my brakes suck bigtime). I guess I just have to play with it, but it's not quite as intuitive as the GM controller's setting of percentage. What options did you order with it?

[Morgen 3Eman]

Hi Truly,

When I was doing my research on which controller to buy, I found a site that mentioned ESB as a setting on the programming panel that selected regen braking force.   I have no idea what the effect might actually be....What does the selection of 55V mean in practice?

Throttle, programming cable, test box, connectors, multifunction switch were ordered to make it all his stuff.

TTFN,
Dennis

[truly_bent]

This regen business gets stranger the more I look at it. The dropdown menu has three voltage listings; 55, 60, and 75. You cannot enter any other values. I'll definitely have to find more on this. The good news is that Edward is pretty good at answering any questions.

What does your optional "test box" do? I didn't notice those when I was buying.

[Morgen 3Eman]

here is the test box link

http://endless-sphere.com/forums/viewtopic.php?f=31&t=20413

TTFN,
Dennis