What's wrong with my BMC motor?

flyinbrick

100 mW
Joined
Apr 4, 2015
Messages
46
Location
Fremont CA USA
Hi,

My BMC motor failed suddenly the other day on my commute. The motor is a BMC 600W. Best I can tell, it is a v2 speed version.
I got it used, then put about 2k miles on it at about 1500W - gradual starts with CA3 programming.
Symptom: when I throttle, I hear the motor spinning, but no propulsion. I can pedal freely with no problem. When I flipped the bike over and applied throttle, the wheel moved forward every so slightly, but something was slipping very badly. My guess was the planetary gears were barely making contact with the case teeth.

I disassembled the motor and this is what I found:
1. Clutch has green composite gears - some teeth chipped. The teeth did not appear to be ground down.
2. With the motor and clutch out of the case, everything appeared to spin freely with throttle application.
3. After pulling clutch out, the bearings on the planetary gears were in bad shape. Two were sloppy, one was binding (or maybe damaged while pulling out?)
4. The clutch appears to be functional - spins freely in one direction and I cannot get it to slip with my hands on the other direction.
5. put the clutch directly in the case - the planetary gears and case teeth meshed pretty tightly - teeth did not appear to be ground down.

So the planetary gear bearings are definitely bad, but the failure does not make sense to me. If the clutch and all the gear teeth appear OK, then what is causing the slipping? If the motor works and the spur gear is OK, by deduction, it must be something in the clutch assembly. But I cannot understand the failure and I don't want to install a new clutch to find that this is due to a different problem.

Can anyone help?

IMG_4379.jpg
 
You have the older green gears which makes them no newer than 2012 or 2013 so they definitely can have some miles on them - they don't owe you anything at this point, but should work fine from what I can see in the photo :) . The newer blue gears are tougher but those green ones are quite respectable.

The planetary bearings are always sloppy, they really don't carry much load and just vaguely keep the gears spaced apart in orbit around the sun gear. The frozen bearing is not good though and you may be able to free it up or pop in a new one if you are handy. These tiny bearings are virtually impossible to remove from the clutch assembly without damage since there's no good means to access the inner race and prying them off by the gear+outer race is an invitation to damage. So - leave the gears mounted up unless you are going to replace them for sure.

Anyhow - on to the problem at hand - I'm wondering if you were fiddling with the phase wires prior to this incident because what you describe is exactly what happens when the motor runs backwards. All you hear is a quiet ZZZZing sound as the motor spins in reverse preventing the clutch from engaging. Interchanging two phase wires could cause this as could shorting FWD/REV controller wires if you have them.

FWIW - BMC parts are fantastically expensive and pretty rare these days but the entire clutch/gear assembly from a MAC or eZee will fit as a direct replacement. Over the years there have been minor variations in the thickness of the gears and clutches in the BMC/Mac/eZee lines so you typically don't want to mix parts - even within brand. However, the overall outer dimensions are the same, so replace the whole assembly if you go that route. The gray Mac gears are quite tough - I use Mac parts in one of my BMCs - they're inexpensive, too ($55 from EM3EV). I have some notes on the gears in my Yuba build thread and there are several other related posts on ES as well.
 
good theory on going in reverse. But,
1. This failure happened in the middle of a ride
2. When I flip the bike over, with throttle application, the wheel moves ever so slightly in the forward direction with a clicking sound coming from the motor
However I suppose there is the unlikely possibility of the Rev shorting with gnd. I will check that.
Too bad I already cut all the wires to remove the clutch. If accidental reverse is likely, then I'll connect everything back and test in forward/reverse.
 
flyinbrick said:
good theory on going in reverse. But,
1. This failure happened in the middle of a ride
2. When I flip the bike over, with throttle application, the wheel moves ever so slightly in the forward direction with a clicking sound coming from the motor
Hmmm - the clicking sound is unlikely for reverse and the 'middle of the ride' thing tends fo make the reverse theory NG as well.

If you can hear the motor running then there's an outside chance the ring gear is slipping in the case. This used to happen occasionally back in the day (around when you motor was made). If so, you can either epoxy the gear in place or drill/file some intersecting holes between gear and case and pin with a couple short 1/16" rolled tension pins from the hardware store.
 
If the hall board in your motor is like the one in my old Ezee (v1?), and if its wires are soldered like mine are, just the tips soldered to the surface of the thru-hole pads rather than actually run thru the holes and soldered, then one or more of yours mightve done what all three of mine did: vibrated away from the pads, and not made sufficient connection to get the hall signal to the controller.

Its easy to test; just use a multimeter on teh halls while hand-spinning the rotor (if its still disassembled; otherwise spinning the wheel or motor case backwards), with the halls plugged into the cotnroller and power on (but no throttle). If the voltage toggles between near zero volts and near 5v,, on each hall sensor wire at the connector, then the sensors and wiring are ok.


Another mechanical problem you could have is a cracked clutch, where manually turning things may allow it to appear to work, but putting the torque load of the motor against the wheel (even off-ground, but especially on-ground) the clutch then slips. This happened to my first Fusin geared hub, and I used a radiator hose clamp around the clutch as a temporary fix to let me ride home.
 
flyinbrick said:
2. When I flip the bike over, with throttle application, the wheel moves ever so slightly in the forward direction with a clicking sound coming from the motor

I took this to mean that the clicking was continuous because the motor was running and not just a single click each time you applied power - not sure that was the proper way to interpret the description.

flyinbrick said:
The ring gear seemed stable against the hub housing. I could not move it with my hand. Now I'm really stumped.
Me, too. Seems you have run the numbers for mechanical issues. If you are certain the motor is running when you apply power and the wheel just won't turn, then you are pretty much left with the clutch as the culprit regardless of how it appears to work with only hand pressure.

On the other hand, if the motor is not really running (i.e. I misinterpreted you remark above), then this might just be a hall issue as the motor won't start if parked on a dead hall or it might be a controller issue (blown FETs). This can be a little weirder than with DD since you can't directly push-start the motor to get it going - you have to apply power and manually spin the wheel backwards to get the halls to wake up and spin in the proper direction. Unfortunately without connectors some of the easy tests are not available. That said, hall and controller tests are still doable. If your motor was not running in your experiments above, I would pursue testing the halls and controller.

  • I'm a big fan of the cheapie ebike testers outfitted with hall/phase/throttle connectors that match your build (get one from EM3EV if you order a gear assembly there). Adding connectors to the basic tester is good rainy afternoon project that lets you test your controller and halls in just a few minutes by plugging in the tester.
 
Not sure this applies here but for completeness, another way a motor can "run" and not generate much power is if the permanent magnets become unglued from the backing. Probably not the case here, but when it happens they slide, I would think that could not happen until there is some load. This sounds more like a clutch problem, so it is puzzling the clutch seems okay, but perhaps there is some difference between operating in the motor and after disassembly, or perhaps the power transfer to the clutch is somehow broken.

I was advised on the BMC to keep the power under 1500W, and really the max motor current should be limited to avoid damaging the motor. Most controllers don't have good mechanisms to limit motor current, so under some circumstances they will generate too much torque even when the power is well under the limit.
 
On mine, the clutch failed where the two parts are press fit together, like in your picture, between the inner and outer circles with the knurling peeking out. It was hard to notice at first.
 
Voltron said:
On mine, the clutch failed where the two parts are press fit together...
Good catch!
I hadn't even noticed that this clutch has the really old pressed two-piece design. Later models of this general style (riveted side plate) went to single piece construction for exactly this reason. Worth checking out.
 
The motor spins with throttle application. So it seems even though the clutch seems fine from off-motor inspection, clutch failure is the logical conclusion.
Hopefully, a new clutch will be the fix. Thank you all for your help!
 
luckily, i took a video with motor/clutch assembly removed from housing rotating with light throttle. It does seem to spin in reverse direction if I’m not mistaken. Hmm, will have to verify. Thanks.
 
flyinbrick said:
luckily, i took a video with motor/clutch assembly removed from housing rotating with light throttle. It does seem to spin in reverse direction if I’m not mistaken. Hmm, will have to verify. Thanks.
Seriously? A video?
Seems reverse operation was proposed as the probable cause back in post #2...
 
It's more an issue of many posts running down other possibilities when all that was required was to look at a video to see if the motor was running backwards. Not to be snippy, but a waste of poster time.
 
Well to be fair, Teklektic, we sorta ruled out spinning backwards in posts 3 and 4. However, Alan B brought up that it could happen during a ride, which is why I went back to my videos. Also, apparently, cannot upload videos here anymore, so I just learned how to post on Youtube:

UPDATE: I bought a MXUS motor from Grin. Controller is a Grinfineon, so should be plug-and-play. Hooked it up and ... motor spins forward!

Motor appears to try to spin wheel forwards:
https://youtu.be/QmKxR2Bypi8

With motor out of the case, it appears to be spinning backwards if I work it out correctly. The freewheel is facing away from the camera. The green gears are spinning clockwise relative to the person holding the motor. So with the bike flipped over and facing the freewheel, wheel spins ccw - backwards (?)
https://youtu.be/i-u7ojmz5uw
 
Yep - well, as I said, I was trying not to be snippy, but my point was that having a video in hand wasn't mentioned and it's a very powerful bit of evidence to try to resolve your problem instead of looking at/explaining other more roundabout symptoms. Gear motor problems can be troublesome to diagnose because there's so much happening out of sight - a PITA to do in person and more so to explain the various clues to look (or listen) for.

Anyhow - the video is in hand, so, onward...

Here is a Youtube animation of planetary gears that illustrates how the clutch works when viewed from the freewheel side (opposite side from your 2nd video). The first sequence where planetary carrier is not allowed to turn is the 'clutch engaged' mode of the motor where the carrier is locked to the axle and the ring gear turns the wheel forward. The other sequences are the 'clutch disengaged' modes of normal freewheeling and motor running backward.

[youtube]ECljAo1q1RQ[/youtube]


For the bike to be powered forward, we want the clutch engaged and the sun gear attached to inner magnet bowl to rotate counter to wheel rotation. I think this is the way your motor is turning in the second video which would make the behavior consistent with the first video which seems to show the wheel trying to turn forward. This seems to show the motor is turning in the forward direction in this test.


The clicking sound is very odd and at first made me me wonder if this is just a broken shaft key on the clutch, but the planetary carrier isn't revolving in the 2nd video so it's not that. Also, the carrier (outer part of the clutch disk) isn't moving at all in the 2nd video so the click doesn't seem to be from the clutch either. It almost makes me wonder if the clicking is a second issue that is separate from the wheel motion problem. That said, coincidental appearance of the sound appearing at the same time as the loss of drive power is difficult to believe.

In the category of things you've probably already examined:
I might also dismount the gear cluster and try running the motor without the gear cluster mounted to try to isolate the clicking, but I appreciate that the wiring has been severed so this is inconvenient. Rotating the magnet bowl by hand w/o the gear cluster also might reveal it is somehow related to the clicking.

You might look in the edge of the magnet bowl for a loose magnet. You can also slip a thin piece of cardboard or 3x5 index card in between the magnets and stator and then rotate the bowl with and without the 3x5 card to find/sweep up any bit of trash in there.

With the motor turning the proper way, the shaft key apparently not the culprit, the ring gear securely in place and not even present for the open-motor click demo, and a weird clicking when there is no motion in the clutch, I'm puzzled. Time for a shotgun gear cluster replacement and enjoying the Mxus while the parts are in the mail...
 
Hmm - there's really nothing in the motor that should be clicking. As I mentioned above, the OP's clicking looks like a separate issue from the drive problem and as you observed, probably benign. Unfortunately the motor is turning too fast in the video to relate the sound to the positions of the moving parts to help localize the source. Just puzzling...
 
One problem I had with a BMC 600 hs is the cotter key rolled. I got a fat key way and filed it down to hammer in with new mac clutch from em3ev. It works. Just to many watts.
 
The shaft key was in place, and it looks good. I programmed in a relatively slow ramp on CA3 so I never put too much torque on it. No debris and no loose magnets. With the clutch out, I was not able to duplicate the sound spinning by hand. The bearings seemed good. Amberwolf suggested possibly bad hall connections, but motor was spinning forward (?). The mostly likely failure still seems to be the clutch. But since there's no smoking gun, I don't know if I want to sink $80 and more time on something that may not be the fix. I already have the MXUS working now, so this will go into a box for now. Thank you all for your help - and sorry to disappoint you that there is no resolution. Between family and work, I just don't have the time to tinker like I used to. Sadly, it is now easier to just to buy new parts.
 
flyinbrick said:
With the clutch out, I was not able to duplicate the sound spinning by hand. The bearings seemed good.
...
The mostly likely failure still seems to be the clutch. But since there's no smoking gun, I don't know if I want to sink $80 and more time on something that may not be the fix.
...
Sadly, it is now easier to just to buy new parts.
Yep - it's too bad there wasn't a clear indicator in hand, but as you say - you have a new motor, etc and there is no pressing reason to press on in the Springtime when the bike is already road-ready. As unsatisfying as it may be just now, this might better be postponed to a winter project. In the worst case, you could later sell off the motor as is to folks who have parts and time in hand.

Enjoy your revamped build!
 
Back
Top