Starter clutch failure

[Dogma]

A little late to this thread but i will ask anyways.
Dogma, in your original post you said...."i used to get the occasional kickback"
How or what does this entail? Is there a metal on metal grinding, a clunk, a delayed start to the starter motor spinning?

Sometimes when the engine fails to start, there is some reverse rotation. This causes the starter clutch to engage in a rather dynamic way that hammers the components. It would be a thunk after you let off the starter button. It can happen during cranking too, but I'm not clear on how that sequence goes. There are a couple electrical interventions to reduce or eliminate this. Basically, power to the coils doesn't come on until the engine has turned over for some short time. The method I have installed is not available anymore, so I won't bother recommending it.

It occurred to me recently that I may have cause this failure by starting the bike in gear with the clutch out in situations that were not broken-clutch-cable emergencies. I thought I was only adding some wear to the starter motor brushes. I shall have to behave myself in the future, if I don't see a way to prevent these loads from going back through the pins. Right now, the design of the clutch seems faulty to me, but I may have to revise that opinion once I see it for myself. I suppose I could abstain from forming opinions until I have the facts, but where's the fun in that?

[Fjbj40]

Starter kickback is ugly! I eliminated it by putting in the ignition retard lever.

It was frequent due to the high compression Pistons in my 1085.

It can sound like a big hammer hitting something inside the engine and the object being hit moves a little bit, your teeth actually feel it!

[Dogma]

OK, that was by far the longest I've applied a rattle gun to anything. I've never had a socket get warm before. That was some good thread lock they put in there.

The threads in the hub measure Ø14.8 by my calipers. That's consistent with M16x1.25. I don't have thread gauges, but an M8x1.25 matched up perfectly in it. It's 26 deep to the end of the crank, and not much of a shoulder in there because of the M12x1.25 threads in the crank. I'd better find a way to protect the M12. Hmm. That hole seems to go all the way through the 1st throw, so I can't push on the bottom of it.

Hub OD threads measure M35x1.5, in case anyone is keeping score.

Now where am I going to get an M16 fine thread screw? McMaster, of course. I need to order some stuff from them anyway. They'd probably have it here tomorrow, if I had that other shopping list ready.

But I don't like that there's such a narrow shoulder for an M16 to push on at the end of the crank. Does the crank need any cleaning up after this process?

[850 Combat]

Man I love McMaster Carr and Grainger. I got all the rod ends to rebuild the steering and front suspension in my old Polaris Snow Machine at Grainger. Better quality and with zerk fittings for less than half the cost of OEM. Lots of the OEM were NLA in the parts book. It isn't like Polaris made their own.. Of course they were still available. Just had to figure out what they used. I think I use Grainger more often than McMaster Carr.

[Dogma]

Man I love McMaster Carr and Grainger. I got all the rod ends to rebuild the steering and front suspension in my old Polaris Snow Machine at Grainger. Better quality and with zerk fittings for less than half the cost of OEM. Lots of the OEM were NLA in the parts book. It isn't like Polaris made their own.. Of course they were still available. Just had to figure out what they used. I think I use Grainger more often than McMaster Carr.

My needs are a little different. One cool thing about Grainger is that you can often drive (or send a gofer) there to pick up what you're after, if you can't wait for next day delivery. I wonder if you could hire an Uber car to go get stuff from Grainger. Maybe they have a delivery truck now. It's been 20 years since I had dealings with them.

[steve murdoch]

..."a thunk"....you say. uh-oh, that sounds familiar.
I will be looking at the 1150 engine, pronto.

Sometimes when the engine fails to start, there is some reverse rotation. This causes the starter clutch to engage in a rather dynamic way that hammers the components. It would be a thunk after you let off the starter button. It can happen during cranking too, but I'm not clear on how that sequence goes. There are a couple electrical interventions to reduce or eliminate this. Basically, power to the coils doesn't come on until the engine has turned over for some short time. The method I have installed is not available anymore, so I won't bother recommending it.

It occurred to me recently that I may have cause this failure by starting the bike in gear with the clutch out in situations that were not broken-clutch-cable emergencies. I thought I was only adding some wear to the starter motor brushes. I shall have to behave myself in the future, if I don't see a way to prevent these loads from going back through the pins. Right now, the design of the clutch seems faulty to me, but I may have to revise that opinion once I see it for myself. I suppose I could abstain from forming opinions until I have the facts, but where's the fun in that?

[Dogma]

Also, three sheared screws. Only one of the pins move freely enough to shove the roller all the way over.

https://goo.gl/photos/hf892MFTiPrhA6q2A

Quite easy to remove once I got satisfactory tools together. I didn't want to risk damaging the end of the crank with an M16 bolt, so I ordered a puller from RMStator's eBay page for $18. I figured I could order an M16 fine screw from McMaster, or order the tool for only three times as much. Perfect fit for the threads on both the OD and the ID of the hub. What good is that? I tried using the M16 screw from the puller in the way Chuck suggested, but it was just ... crunchy. Fortunately, the visible damage was on the screw, and the threads in the crank were OK enough. I tried putting the original screw back in the crank, but the puller isn't long enough to assemble over that. Good grief. So this afternoon I went scrounging at the local Ace hardware. I found an M12x1.25 x 50 stud. I slotted it with the Dremel and ran it into the crank. I mounted up the puller and kablam! It was off and rolling across the floor. If I had to do it again, I'd just get the two screws.

Edit: The two screws have the same pitch, so they might have just turned in together if I had tried that instead of the puller. It was the difference in pitch between the M35x1.5 and the M16x1.25 that made it work.

I have a possible donor from an 850, but I ran out of time to dig in and compare the clutch parts.

[chuck hahn]

What grade bolt did you get??? Mines a grade 8 and never have felt anything other than rotors coming off. Used many times and the threads are just fine still.

[Dogma]

The screw came with the puller. I don't think it was an issue of grade so much as geometry. It has a bit of a dog point on it that fit into the M12 hole, so the M16 threads pushed on the end of the crank. It was starting to plow the threads off a little at a time. If it had a blunt nose like a regular screw, it would have engaged flat to flat, a ring of contact maybe 1mm wide. If the hardware store hadn't had that stud, I probably would have modified the M16 screw to have a flat end on it. Well, as flat as I could make it.

Even though you have experience with that method, I wasn't fully comfortable with it, especially with the hardware I had on hand. There are a lot of ways a screw I might find could be different from the screw you've been having success with. So I was happy to play it safe and arrange to have a couple cm of M12 in the crank working against the M35 on the hub OD.

The puller had reasonable build quality, but the design of the thing was inadequate. Whoever designed it didn't look closely enough at what it needed to do. It needed to have an M12x1.25 screw instead of the M16x1.25, or the outer part of the puller needed to be another cm deep so you could use the original M12 screw under it to push on the crank. Or it could have just been an M16x1.25 like Chuck recommended, with a nice flat end to engage the crank.

Either way, I should have thought it through to realize that it was going to pop off onto the floor when it let go. If it hadn't landed on my oil catch pan first, I imagine it could have cracked a magnet hitting the concrete floor.

[Dogma]

...
I have a possible donor from an 850, but I ran out of time to dig in and compare the clutch parts.

For future reference, this interchange doesn't work. The gear and clutch on the 1000G are larger than the 850.