Went over a bump: noisy hub now

[rust]

A little background first:

I've been struggling with my bike for a while now. First the original controller decided to die a silent death. No magic smoke was produced. Just working and then not. Not even pfft. That was 2012. Aw crap, so I ordered a YK85S 36V/48V 250W/350W off of eBay. MISTAKE. It was really confusing to get right, but once it got into its "self configuration" mode, it was "OK" except that it has a horrible "low voltage protection" circuit, that for the 36V configuration kicks in at 31.5 Volts. If I recall correctly, it cuts out at 42V on a 48V config. It really cut the autonomy down to a couple of kilometers from much more than that. Add on top of that, the batteries went a month or so afterwards. Then, I had money problems and couldn't buy new batteries until just recently. I got those installed into the carrier and hey presto, I was riding again, but the demon "low voltage protection" took over.

I still had a little money left over and ordered a CON61 from GBK. They sent me a CON611 which is a little bit different with lots more connectors (PAS, cruise, 3 speed throttle, double brake, brake sensor), but most importantly has a 28V low voltage cutoff. Amy the rep sent me a wiring diagram which answered a lot of questions I had. I took out the YK85S and put in the CON611. I hooked up the power, turned the key, twisted the throttle and groan groan groan, the phase was wrong. I consulted ES and honed up on phase alteration (which I learned about back in 2013 when I first got my YK85S) *thanks to SpinningMagnets for informative posts here on ES* … in any case, I swapped around the motor phases until I got a smooth reaction: but it was in reverse. That would be G-G, B-Y, Y-B. So I figured I'd have to swap around the Hall leads. Fortunately (or rather, coincidentally) a moron had decided he'd have to CUT the motor lead when he changed my wheel after I damaged the rim on some steel in the road. I had replaced his soldering job with some barrier strips, so I unwrapped them to expose the Hall leads. I swapped them a few times when the controller just up and failed. Wierdness, considering I would unplug the power pack and turn the key off for each test. I wrote back to Amy at GBK to let her know, but they're on a national holiday (Mid Autumn day), so I won't hear from her until probably Monday.

So I re-installed my YS85S and of course, it worked "perfectly" as it had before with the crappy low voltage cutout - working just "fine" on a test run. Nothing out of the ordinary: performed just the same as it has always.

This morning I was going to finish up the wire wrapping and crapola when I realized that I needed to get cat food and some human food. I hopped on the bike, got the cat food, went over to the butcher for a pork chop, hit a small bump I've gone over hundreds of times (the kind that is made from concrete where a driveway meets the road, about an inch or less- 15-25mm) and the hub started making a horrible groaning sound. I stopped after 10 seconds of this, got my pork chop, turned the bike around and limped back to the house. It was stiff going.

I inverted the bike (which is something I very rarely do) and examined the yet unwrapped connectors: everything looked good, no loose wires down near the motor. Everything looked fine, except the hub makes a horrible groan as it runs, getting worse as it gains speed. It has always been silent, sometimes just enough to know that it's still there.

Here's some readings:
Batteries show 38.7VDC (1212 sla)

Ω to black (GND)
Hall: 4.7K all
Phase: Y 2.87K B 2.89K G 2.87K

start twist throttle (low) Voltages to GND Motor phases
G 36-40
B 36-40
Y 42-60!! Yes, I checked that several times

Full twist throttle (high) Voltages Motor phases (and I had no idea of this figure until I measured it)
G 16
B 16
Y 16

I've seen some videos (like this |one: https://www.youtube.com/watch?v=xBrJ_aXWRCk ) where operating a hub motor in this condition results in magic smoke. I'm not made of money so, I'd prefer to DIM (err DIY?).

So, what's YOUR opinion. Should I tell the YK85S to reconfigure itself by running the config mode? Minor hassle in that it's inside an aluminum case now, and would require some unscrewing of nuts, bolts and screws. I think that it may be worth the small hassle.

Or should I go with major hassle and disassemble the hub motor (remove the rear wheel, set it up on a bench and attempt the feat of opening the motor) in order to inspect to see if there is some foreign body grinding inside?

EDIT:
Here's some additional troubleshooting: through hand cranking, I can definitely detect some kind of vibration happening on the hub, where normally it would crank the wheel smoothly and spin away, it's definitely doing something that I can feel through the transmission from the hub to the cassette, through the chain and crank into my hand and hear quite low and increasing while it spins down. I think this means I'm going to open the hub motor.

EDIT 2.0:
Between tossing and turning, this hub motor business has me going in a tizzy: cooler heads prevailed this morning. I uncovered the nuts on the rear axle and gave 'em a drop of oil to make sure there'd be no trouble turning when it came to removing the wheel. As I stood in uffish thought, I slowly rotated the wheel and noticed a detent every 7 degrees or so. So, I counted the detents, got 48 in a full turn (that'd make 7.5° between each). So I guess that's the physical resistance as the magnets pass the coils while rotating - or 16 for each phase. If you, Dear Reader, have a better idea, by all means, tell me about it.

 

[d8veh]

After what you did with all the motor wires, I don't think that its fair to involve GBK, who sold you that controller in good faith.

 

[rust]

Oh really? What did I do with the motor wires?

I'd love for you to explain to me how testing for the proper configuration after finding the DEFAULT YY-GG-BB combination (which SHOULD have worked from the first moment) produced an incompatible situation using the CON611.

GBK's product was probably defective: I should not have had to fiddle with ANYTHING. The YK85S was drop and go, do a config routine and done. I think GBK is super, and they'll probably replace the controller. As Amy explained it to me, this is a NEW product that isn't even in their website yet. In any case, I'll wait on her reply.

GBK's controller had nothing whatsover to do with the hub motor's current condition. Re-read the post if necessary: it's pretty clear that the hub motor started its vibration problem with the YK85S installed. It's likely coincidence that their controller failed the previous day, and has nothing to do with the hub motor.

 

[dogman dan]

If the cogging goes away when you unplug the controller from the motor, you do have a blown controller. If it stays, then the problem is in the motor.

 

[rust]

I agree with Dogman's logic.

The vibration still goes on when the controller is not in use so I have to believe that it's a non-electrical problem with something inside the hub giving it grief.

I'm going to watch Neil Paisnel's video a couple of times more (https://www.youtube.com/watch?v=g-pcvtl73fo) to get a better feel for what I'm about to do.

Any recommendations on my course of action once I have the hub opened?

I plan on going through an excruciating process simply to wipe down the outside of the armature assembly (if that's what you call it) and the inside of the outer rotating ring with a rag dampened with WD40; if it works it was worth the effort. If not...

 

[rust]

Holey Underwear!

So I've got the wheel 99% off the frame, just removing the barrier strips for the connections when I notice that the YELLOW motor phase wire was loose! The problem was when I was testing that the wires were pretty tight in their position, and a little inaccessible (prior to being wrapped completely) and that when I was checking the voltage on the yellow phase I was doing it blind, I couldn't actually see the condition of the connection. It turns out I imagine, it wasn't actually reaching the motor, which is why it was 42-60 volts -- NO LOAD!

Thank dog. DISASSEMBLY ABORT!

Updates later ...

 

[dogman dan]

Partially disconnected phase wire can be a bear to troubleshoot. runs smooth as silk no load, hop on to ride and Kak kak kak kak.

 

[rust]

Yep. It was a loose wire which explains why the hub was vibrating the way it was. Push push coast push push coast push push coast TIMES 16 for each revolution. In any case, I got to know my install a little better. The phase wires loosened up when I removed the tape from the barrier strip I guess. That explains that the bump I went over gave me grief, allowing one of the phase wires to stop contact.

I think what I'll do however is to solder the hall sensors wires again, as they are pretty fine, what is that? 18 gauge stranded? Or is it finer? Maybe a 20 gauge? I'd say a barrier strip is too clunky to ensure a good connection. I'm going to heat up the soldering iron, that cheap piece of crap, do the soldering, then tape them up against the elements. My plan was to wrap an old inner tube to keep out the rain. A barrier strip is ok for the phase wires, but I think I'll solder them just the same.

Lessons learned:

Take a REALLY good look, put on your magnifying lenses if you have to, and examine connections well before resolving to disassemble your hub motor.

If you see that one of the phases measures a whole lot more voltage than either of the other two, you've got a disconnected phase wire.

 

[rust]

I complained to GBK about the CON611 controller and they refunded my money plus shipping. That was very good of them.

I intend to send it back to them and I asked them to provide me with a return address.

My curiosity really got the better of me, so I took a philips screwdriver and opened up the controller to inspect it. Sure enough, I followed down the Hall sensors to the solder points on the board and the Yellow and Green were apparently swapped. The Phase wires at the MOSFETS are labelled with U, V & W (Yellow, Blue & Green respectively). The Hall sensor connector is grouped together with +5V, GND and U, V & W labels. Unfortunately, U is GREEN and W is YELLOW. OOPS! I guess that's why the controller made the wheel turn in reverse when the Phase wires were swapped around and why it groaned horribly when it was connected without any swapping.

 

[amberwolf]

It's been my experience with controllers I've opened up and some posted here on ES by others, that the order of phase and hall wires can be random, even with two "identical" controllers from the same company.

The same is true of the motors, and so that's why we have the articles on the forum and the wiki about how to find the right combination.

 

[Alan B]

While trying to find the right phase/hall combinations, and any time these combinations are wrong, avoid giving more than a tiny bit of throttle, as too much throttle can in some configurations cause very large motor currents that will damage the FETs.

When searching for the right combinations never place a load on the wheel, spin it in open air, and ramp the throttle up very slowly and carefully. If the wheel turns smoothly and the current/power remains low you can gently ramp up to full speed and verify that it remains smooth and the power/current remain low.

If the rotational direction is wrong then change the directional control input, or swap one pair of motor phase wires. Then the job becomes finding which pair of hall wires need to be swapped to compensate for the motor phase wire swap and find the same condition in the forward direction.

 

[d8veh]

Some of these controllers, like KU63 and KU93 from BMSBattery (I thought Con61 was the same as KU63) are fully automatic. IT doesn't matter which sequence you connect the hall and phase wires, they always run perfectly.

Recently, I was trying tomake a Q128 motor run in reverse with a KU93. I could put the wires in any random sequence, but still it ran perfectly forwards.

 

[Plaster03301]

I had the same problem and it was from the magnets being a loose like even just a couple of them were loose and this was a new hub motor too . Also noticed a small negative wire touching positive which was causing a shortage to motor I fixed at same time . I used a high temp apoxy on all the magnets took them all off numbered them and apoxied the shit out of them . Careful not to break magnets they are very brittle . Not sure but did you check those out? It solved my bike