Cafe Racer Forum banner

521 - 525 of 525 Posts

·
Registered
Joined
·
1,865 Posts
Discussion Starter #521
My father-in-law was a research chemist. One of his favorite aphorisms was, "In theory, there's no difference between theory and practice. In practice, there is."

It turns out that my design for the the threaded-together cross-shaft is fundementally flawed. Pushing your foot down on the shifter can exert a remarkable amount of torque. Instead of a nice, positive stop, the limit of pedal travel instead devolves into an indistinct, mushy feel as threads begin to distort. After a number of upshifts and downshifts, either a) one of the threaded studs backs off and the whole assembly goes loose, or b) the lever rotates so far out of its original alignment the other way that the mechanism binds before upshifts or downshifts occur.

I tried adding blue Loctite on the threads; it might as well have been apple butter for all the strength it added. Then I tried a dollop of JB Weld in the bottom of the internal shaft threads, thinking that would provide more positive resistance for the studs and magically stop the thread deformation. That was utterly ineffective, and one of the studs broke free within a few gear changes. So, as of right now, the whole shaft is basically glued together with red Loctite 271 — effectively turning the self-aligning flange bearings and shaft into a single, non-disassemblable unit. I'm letting the Loctite set-up and cure for a couple of days before I really mash on the shift lever, but I am confident the whole thing will eventually need a re-design. Maybe this would not be an issue if I were dealing with a new Japanese bike that has a slick-shifting, low-effort, extremely precise shift action. Bultacos, however, are cantankerous, stone-age things with high-effort, balky shifting that requires a LOT of pedal travel. It might work marginally smoother when the engine is running and the transmission shafts are turning, so I am going to continue with the next steps and revisit this once I have a running machine, but I expect to revisit gear selection at some point.

[IMG]

Here you can see a whole raft of index marks on the backing plate. Between upshifts and downshifts, I need the whole range of travel between when the rear of the pedal hits the footpeg and when the front rose joint binds against the bracket going forward.

[IMG]

The bent-up linkage between the lower arm and the stub of the original shifter is just a quickie proof of concept. The real component will be an alloy clevis threaded to a spherical joint at the bottom, which should eliminate some free-play.

[IMG]

I had to re-bend the shift pedal just so. I discovered that there's barely enough clearance for it between the arc of the kickstarter and the shift linkage, but it all clears as it's positioned now.
 

·
Registered
Joined
·
1,865 Posts
Discussion Starter #524 (Edited)
This morning before work, I made up the clevis for the right-side shift linkage out of 6061-T6511 alloy 3/4" square bar stock. The upper end is attached to the Bultaco shift lever using a socket head shoulder bolt (5x0.8mm thread, 6mm shoulder dia.). On the other side is the original rose joint that came installed on the Ducati stub arm I'm using for the lower linkage. It still needs filing and polishing, but there is no noticeable play, and the geometry of the linkage arms is much improved. Even without the shaft bearings installed, I can already tell this will work much more smoothly and precisely than my earlier mockup.









I am still working to resolve my shaft-spline interface issues. I think the ends are good candidate for brazing, rather than welding. I ordered some MIG brazing wire, so I will see if my little Clarke MIG welder is up to the task once that arrives. I have not decided whether to simply braze the ends on with the flange bearings already on the shaft (at which point they will become no-removable), or to go with some sort of (as-yet undetermined ) split bushing arragement. I have some Delrin stock coming this week, but I am not sure I will go through the hassle, especially if the brazing wire gets here first.
 

·
Registered
Joined
·
1,865 Posts
Discussion Starter #525
After letting this shift shaft cause me bunch of unnecessary headaches — playing around with all sorts of solutions such as split bushings, pinned joints, and the like — I finally decided to use the KISS method and simply have the splined ends welded onto the shift shaft with the existing flange bearings already on it. A quick trip by my favorite local welder and I'm in business.


The bearings are now non-removable, which is admittedly a little bodgy, but in reality it won't make any difference. After all, having bearings as a captive part of a larger assembly isn't unheard of in the engineering world and, given this bike's experimental nature, the chance this linkage ever getting sufficient use to warrant replacing these bearings due to wear is a pretty far-fetched notion. At whatever future point these bearings are no longer serviceable, I am sure the bike itself will have been long-ago disassembled, or a rusty and corroded hunk lying in a field somewhere.


BTW, the discoloration on the shaft is NOT excessive HAZ from the welder. I'd previously blued the ends using a polypropylene torch to un-do the red thread locker.
 
521 - 525 of 525 Posts
Top