Unclutching (fixing / welding up) a Yamaha XV starter clutch

 So imagine this, your beloved steed (in my case a Yamaha TR1, but the same will also apply to XV700s, late 750ies and of course 1100s), instead of greeting you with the much beloved sound of grinding and crunching starter gears, greets you with this noise

So it's disconnect the battery and take the left sidecover off time.

The name starter-"clutch" might be a bit misleading as it really is a one-way-clutch and as you can see in the clip below, the clutch in this engine is more an "open for everything" clutch. (Which is a great mindset in general, but a bit problematic, when the clutch is supposed to drive the engine in only one direction.)

As you can see, if you look at it closely, I already welded up the clutch once, but this didn't quite yield the results I wanted. (The tacks on the back sheered off.) 

In order to fix this mess, we need a starter clutch (I am using a broken one, which luckily I didn't throw away last time, a shaft for alignment of the parts (your starter will do just fine, I had the shaft from a broken starter, at least an angle grinder and a welder with some chest hair as the parts are relatively thick. I will do it in a slightly more elaborate way on the lathe, but that's really totally unnecessary as will come apparent over the course of this post. (But it's been like four or five years since I modified my last starter clutch and the first time to do it this way.)

All the parts as outlined above and an 18 tooth (XV1100) starter clutch, for a regular TR1 with the original starter, you'll need the 19 tooth version. 

The whole starter clutch is held together by the folded over lip on the back, which I carefully removed on the lathe. A simple cut through the thin metal shield perpendicular to the axis of rotation with an angle grinder would achieve exactly the same result in a lot less time. 

Peeling the metal shield back reveals the actual retainer plate and because everything is rather worn on this clutch, the gear can be pulled out. 

 

Plate, springs and pins inside can be discarded and the gear and the clutch housing be reinstalled on the shaft, in order to trim it back substantially, so it can be welded, where the two parts meet up.

At this point I have to admit, I used my trusty lathe as a rotary head and cut the housing with an angle grinder as it is STUPIDLY hard.

 

And yes, I cut it too far towards the end initially, but we'll call this a lightening cut to reduce rotational mass and improve starter acceleration. 😅
 

After cleaning the two parts can be welded together, obviously while assembled on the shaft, so they are in proper alignment. I did it with TIG and quite a few amps, but probably it could have done with even more and probably some quenching in between, because I think I properly tempered the part. (Undid the heat treating/hardening of the steel.)

 

If you trim it down a little further, so it is the same diameter as the gear on the back (or less), you can make it small enough so in theory the start can be removed w.o. taking off the cover. (Not yet tested, but in my mind none of the other starter components in there can fall out...)

 

Additionally I had to ream the bronze bushing inside the gear of mine, because it always had been an overly tight fit on the starter shaft and while this is a good thing for alignment, having to install it by hammering it onto the shaft certainly isn't for either the starter OR the starter clutch. (Normally this isn't necessary, but the clutch shown was one of the cheapest ones I could - and apparently for a reason.)

 And that's what it looks like, reinstalled. 

 Reinstall the cover, fill up with oil, hook up the battery and all the other bits you removed, cross your fingers, hope for the best ... you get the idea.

 And then... nothing.

 It does help to hook up the wire to the actual solenoid. Probably to the great dismay of a bunch of Yamaha engineers, it works just fine without the extra clutch between starter and flywheel. 

 

... but boy oh boy does the old girl sound fine, when she "just" fires up right. (Added bigger pilot jets and turned the air-screw in half a turn and she just starts like it's an EFI bike... perhaps even a bit better.)

 

So it does work remarkably well with the welded up / one piece drive sprocket. Even the initial repair would probably have been sufficient, if I am honest, but a few carb issues made starting a lot more difficult and the constant backfires eventually broke the welds. Not sure what would break this time, admittedly I am not overly keen to find out... 

New 160mm chuck backplate for the Matra

 Martha II came with a generous selection of original three-jaw chucks, made by very, very reputable manufacturers. Unfortunately they all shared one commonality: They were at least 70 years old and hopelessly worn out. 

To my big surprise Din800 M33 is still a pretty common standard on woodworking lathes. As cutting M33x3.5mm threads by hand is at the best of times "no fun", I was more than delighted to find out that cheap steel fixture plates are commonly available, but usually only up to around 125mm, with the odd one being available in 150mm/6". 

 

So in fact only the center section with the actual thread was really useful to me (for now) and a cheap 63mm (2 1/2") holesaw made relatively short work of the mild steel plate.

 

Trying the same trick on some actual tool-steel finally killed the poor holesaw and left me now other chance than to first drill and then bore the center hole to size.

 

 

Bit of TIG-magic later and a hole for a C-spanner later:

 

The thing with proper tool-steel and the right inserts is, you end up with a really lovely finish.

As the whole "using-a-holesaw-to-drill-the-center-out"-thing really backfired, I decided to do something I had put off for ages. A long while ago, I found a Morse Taper 2 reamer made in the Soviet Republic of Ukraine, still in its original wax coating and wrapper. Based on the writing the outside, I assume it was manufactured in February 1986. So, the brown stuff is almost forty year old wax and not rust. Also it is sharp enough to immediately draw first blood...

 

 

The bore cleaned up absolutely beautifully, even though I might have to face the barrel off by a milimetre or two as all the tool sit in there quite a bit deeper than before.

 

So that's it and I have to admit, even a relatively inexpensive 160mm chuck is a massive improvement over a worn out top-of-the-shelf one from more than 70 years ago. 

Dre-XT-Stück - Cylinderheadboltthreadsshouldhavedonethemrightthefirsttime

It started like those stories always seem to start - when things are actually going well. Or to put a date on it, on May 24th, 2024 after enjoying a lovely little rideout on the old girl, the front left of the cylinder was covered in oil. 

So I did what any sensible person would do: clean it off, do nothing, hope for the best and get the inspection out of the way and then as the problem didn't magically solve itself, mumble something and park the old girl in the shed and ignore it for about half a year. Which is roughly when I realized that the old girl would come in a bit handy right about now.

Apparently stripping down the engine never goes quite without a bit of extra hassle, so the opportunity arose to fix all the remaining camshaft journal threads as well to improve my chances of making it through a full season without a rebuild of the engine. 

 

 

The piston has sucked in quite a bit of oil through the engine breather and was covered with a nice crust of burnt in oil.

... and this is from right after I rebuilt the engine and it felt a bit doughy.

 What looks like a fairly usable headbolt thread actually isn't and is the reason we're having all this fun now.

The solution will be the same as used on the my dad's XT500 4-valve engine - stainless steel M12-to-M8 adapters with an overall length of 27mm. Why not 30, you ask? Well because the stock threads are only 26mm long and with the steel inserts they should be very much strong enough. Also 25Nm on M12 is so far away from the maximum torque that is permissable in aluminium, I was actually going to use threadlock just to make sure they won't rattle loose. The other, quite valid point is: why not use a commercially available insert? In short: I tried Helicoil and they failed as they are too short and thus too weak and the top most is a commercial product, but again only 15mm long AND about half of it is slotted for installation, meaning it will most likely not be able to carry the full load.

 Installation is very straight forward, center on the old thread with a 6.5mm drill lock the table on the mill and clamp the cylinder down tight and go at it with a 10.5mm drill. 

That's not quite looking right, is it? It is not indeed. For some reason the guy in charge of tightening up the drill chuck didn't do his job very well and the drill came loose, wandering about and making a very interesting tapered 11.2 to 11.8mm hole. Not good. 

After some thinking, I decided to make an oversize insert as luckily I had started in the one corner that had plenty of meat around the hole and settled for M14x1.0.

Now the cheap thread die left a remarkably bad finish, which in hindsight was a good thing as it was rather unpleasant to thread in, which in turn most likely meant it would be rather unlikely to come out as well. (Or as someone else said: cross-threading is a kind of thread locker.)

Measured the thread depth and decided where to part it off, so it would end up flush or slightly below the headgasket surface.

Then it was just a matter of doing the other three WITHOUT cocking it up again. 

Basically do this until you're done and test it with a headbolt.

 Because the department of fun and games hinted at the fact that on more M6 thread was stripped and at least another one was feeling a bit dubious, I decided to install timeserts on all of them.

Obviously not without making another mistake: I didn't quite countersink the first one deep enough as I thought the chamfering tool had already bottomed out.
 

Having gotten this far and not being sure just how well that insert was seated, I decided AGAINST any attempt of getting it back out. Instead, I decided to work around the issue. 

First I put the cover back on, centered it with two additional bolts and tightened it down.

This left an imprint (hardly visible in the picture) on the boss.

 Which was then chamfered to provide the necessary clearance so the cover could be tightened all the way down. 

Gave the cylinder a quick hone and then wondered whether installing it without a base-gasket would improve the squish gap. For those interested: Nope, stuff gets too tight (again), between 0.45 and 0.55mm of the piston sitting low in the bore. With an admittedly very pliable base gasket I ended up with 0.70 to 0.80mm, which has to be added to the 1.00mm of the cometic head gasket. 

To quote a certain repair manual: Assembly is the reverse of disassembly, with the two most important things to bear in mind being that the cylinder base nuts need to be tightened to 40Nm (42 in the manual) and the headbolts to 25Nm. (29 in the manual which is just too high and the cause for many a stripped thread!)

 

 

First start was rather anti-climactic - not much smoke just a two or three kicks, pretty much standard procedure as if the old girl would have just been parked over the Winter.