TSM-A5 opinions & impressions

[gothicsurf]

After seeing posted here a conversion done with this motor with an integrated torque sensor, I thought of asking what other people are thinking about this choice for a build. Would the integrated sensor detract interest for some due to this part being less accessible for service / replacement? Would the premium price sway few people from a standard DD motor paired with a torque sensing bottom bracket?

 

[gromike]

Well, kinda late to the party, but I had my tsm-a5 motor fail recently. The motor has a torque arm that is keyed on to the hub on this (what I assume is a pressed fit) boss:

It appears that the boss has shifted and pinched the phase wires:


The motor had about 2k miles on it, and used Grin loaded regen levels. It's sad for me, this motor cost some bucks, and the controller is specific to this motor. At least the EMV3ev battery is good.

I'll get around to emailing Grin, but in reality, me thinks this motor is toast.

 

[gromike]

I was able to pull the bearing boss that the torque arm keys to up to expose the wire damage:

A direct short of the phase conductors.

I was able to lift the boss up higher, and I think I have a path to repair:
I'll rotate the boss back into alignment, repair the wires' insulation in place, push the boss back down and spot weld it to the shaft.
It's a nice looking motor and it would be shame to scrap it, but it's integrated torque arm has a critical point of failure. I'll try to fix that with the spot welds.

As to the original poster's question: I can not recommend this motor. A moot question, as Grin has discontinued selling the motor. If someone has this motor with regen in use, it will certainly fail, but maybe a couple of spot welds can fix that.

 

[amberwolf]

I was able to pull the bearing boss that the torque arm keys to up to expose the wire damage:
View attachment 381771

Can you get any pics of the insides of that boss from the motor side? I'm curious how they designed it to stay in place?

I'll rotate the boss back into alignment, repair the wires' insulation in place, push the boss back down

Between the wires and the boss, I recommend adding a piece of heat resistant cloth as is commonly used in motors (not sure if they use fiberglass or what, but you've probably seen those cloth tubes over the winding-phasewire junctions; that stuff. )

Otherwise the same problem is likely to happen again, even if the boss does not shift, simply from pressure and warmth over time.

Kapton tape can help with the temperature, but it is physically fragile and easy to cut thru or tear, so it won't do as good a job, even in multiple layers.

and spot weld it to the shaft.

Which shaft? The steel(?) axle? Or the aluminum (or whatever) cast metal stator support?

If the two things are not the same metal, welding either won't work or may cause a catastrophic failure later of the welded area. Depends on what the metals are and the weld procedure, depth of interpenetration, etc.

A key or pin or set screw into a dimple would be a better idea.

 

[gromike]

It's a press fit on the knurled part of the shaft, no keys, etc..

I'm going to spot weld the boss to the axle at the free end. That's all some sort of iron.

 

[Papa]

Can you reposition or remove the phase wires to allow rounding-off the sharp exit corner of the boss that sliced into the wires?

 

[gromike]

The failure was due to the boss rotating under torque loading and pinching the wires, not from wire routing in it's normal position. My fix is to find a way to fix the boss in position. I repaired the insulation and pounded the boss back on to the shaft, so no die grinding para mi'.
More pictures when I do the fix.

 

[Papa]

Yea... I get that the boss rotated, just initially puzzled as to why the wires were lanced on the boss side. I guess i need to upgrade my spectacles.

I agree with tack welding . I would however, try to figure out a way to pre-heat both the shaft and the boss prior to welding, in an attempt to minimize thermal cracking in the HAZ. But of course that means unlacing the wires, which would undoubtedly be a major hassle. If you hear a 'ping' sound immediately after welding, you'll know instantly your tacks 'let go'.

 

[amberwolf]

Yea... I get that the boss rotated, just initially puzzled as to why the wires were lanced on the boss side.

FWIW, from the specific damage visible, I don't think they were cut by it rotating so much as heat deformation of the insulation along wiht pressure of the wires pulling up against that edge.

Not possible to tell for certain from the pics of the wires; there could also be cut damage.

Either way, I'd still recommend putting that cloth previously mentiond between any wiring and the boss sleeve.

 

[gromike]

I ended up pinning the bearing/torque boss to the motor's spyder:


It was a bitch to drill those two holes. They only had three carbide bits of the wanted size at my local hardware store, I needed four.

Wired the motor up, and ran the motor Autotune function in the Phaserunner program. No faults were detected. Spun the wheel on throttle, so the motor runs. My final judgement awaits turning the bike over and riding it.


As the the pins are in at an oblique angle, the boss will never pull up. So this might be a one and done repair, and it's time to save up for a new motor. The two phase wires were trapped when the boss turned, my insulation repair is more robust than the original. A couple or more years of use would be a win.

The motor sounded great when I bought it. A direct drive wheel motor with an installed torque sensor and integrated torque arm, but this motor has some serious design defects. I wonder if any of these motors are still in use?

 

[Papa]

It was a bitch to drill those two holes.

And why I avoided suggesting pining - you dun good

 

[gromike]

Now that the motor is repaired I might get a couple more years out of it, though I think I might need another freehub body (already replaced the chain and sprockets). It's a bit chewed up, but I'm unable to source it. The torque sensor fits inside. I guess I'll need to see what Grin says. If anyone has any idea, here's a picture of it.

 

[Papa]

My first thought was to shim (or eliminate a cog) to shift the pack outwards slightly - just enough to align the cogs inner teeth with fresh, undamaged splines on the freehub body?... maybe?

 

[A-DamW]

Now that the motor is repaired I might get a couple more years out of it, though I think I might need another freehub body (already replaced the chain and sprockets). It's a bit chewed up, but I'm unable to source it. The torque sensor fits inside. I guess I'll need to see what Grin says. If anyone has any idea, here's a picture of it.

I've seen someone remove enough material off the damaged area of the freehub to insert steel shim stock as a buffer.
Should nearly eliminate cassette bite.

 

[Papa]

I've seen someone remove enough material off the damaged area of the freehub to insert steel shim stock as a buffer.
Should nearly eliminate cassette bite.

Your idea definitely has merit, but unless you whittled-off the exact amount off each spline, you'd end up with some splines engaged and likely not others. Perhaps?

 

[gromike]

Yeah, it would be a lot of time spent with a file to buttress the splines on the freehub.
I've spent something like ten hours working on/in this motor. It's a beautiful motor, but there comes a time to cut your losses.

The motor is now all together and on the bike, so I'll ride it a bit and see how it goes. And I'll also email Grin, for sh*ts and giggles, to see about a freehub body for this motor.

 

[gromike]

I put it all together and went for a ride. The torque sensor PAS didn't work, no speed reading or motor temp, and for some reason regen didn't work, but it ran well on throttle. So much so that I was thinking I love this bike. If those small wires were bad I was screwed being how the pins went in tangentially, and I wouldn't be able to pull the bearing boss for access in a rewire.
Then the motor failed.

Did a failure analysis, and took the motor apart. This is what I found:

The bad is the pins broke thru the hub and the bearing boss once again shifted. The good is that I can pull the boss off now and redo the wires and then weld the shaft to the boss. But that's bad because I'll waste more time on this motor. But that's good because I get to practice rewiring the motor's cable.
Nothin' left to lose.

 

[gromike]

Just to wrap up my attempts to repair this motor. I built up the spider boss with epoxy, drilled a third pin hole, drilled all holes deeper into the spider, and epoxy glued the bearing boss on:


Put it back together with a replaced motor cable. It runs on throttle, ran it at 1450w for awhile. Torque sensing PAS, and regen don't work. All the wires had been nicked before when the boss turned. Wires are now good, but maybe there is a controller issue. The Grin controller is specific to this motor, and like the motor is discontinued.
I'll do a bit more troubleshooting, but this motor is dead as far as in an ebike, though it still functions well on throttle.
I should just give up and harvest the rim and spokes, and order an all-axle motor.
I think that it's running sensorless through the Baserunner controller.