Success....well, mostly, but with a slight twist !!
It was a long (HOT....102+ degrees) fight, but I finally got the bike running with the timing--both at idle and at ~3000rpm (via the timing light)--dead-on....well, except for an odd
ghost in the left cylinder (more about that below).
But it was only after I reinstalled the notched nut the way it was originally (flange side in) that I was able to get it timed and running. As it turns out, that screw-thingy doesn't come off; the advance backing plate, once levered free, slides over it and off. Once I got the backing plate off I could see the pin behind it and the slot in the rear of the backing plate that that pin fits into. And with the pin in place, the backing plate only fits on one way and, on this bike at least, the engraved line on the backing plate does NOT point due north or due south when the rotor's at TDC.
Also, I cleaned/relubed the advance inner rod and made absolutely sure it rotated like butter and that nothing was binding as I turned it while watching the points move up-and-down as they should. Definitely, no problems with that inner rod.
As I reassembled the advance unit, I made sure that all the moving parts moved freely; and that the weights had sufficient space between them and the stoppers on the backing plate and that they rotated smoothly on their pins and retracted all the way when I pulled them to their maximum deflection point and then let them go.
After everything was re-assembled, I reset the points to .35mm and then began timing the right cylinder with the timing light. It was nearly perfect right-out-of-the-box and the screwdriver indent in the points backing plate (for setting the timing) is now more centralized than it was when the notched nut in the advance unit was installed with the flange facing outwards. So that seems more correct.
Similarly, when I revved the engine to ~3000rpm, the rotor mark now moves properly forward towards the Advance line on the stator. I can't honestly say if it stopped right at the line, but it was close. Definitely, behaving differently than yesterday when the timing mark just sat on the 'F' no matter how much I revved the engine.
The only thing that didn't look right is that when I time the left cylinder, the timing mark on the rotor appears twice, at the same time! That is, I'm seeing two images of that timing mark with the timing light. One appears where it's supposed to be (between the 'F' lines on the stator) and the other image appears just to the right of the right-hand 'T' line. I'm also hearing a slight misfire in the left cylinder, which I think was the original problem I've been having and why I originally said the idle was "slightly rough".
As I kept the timing light running (engine at idle), I noticed that every 4-5 seconds or so, the ghost image would disappear, momentarily, and I'd see a single timing mark dead in the middle of the 'F' lines. So, while there's still something wrong somewhere, I figured the timing isn't gonna get much better than that until I can figure out what's causing that ghosting in the left cylinder with the timing light.
Here are photos of what I saw. Before I began, I set the crank to TDC and checked the orientation of the engraved lines in the advance backing plate and center spacer. As you can see, they're definitely not pointing due north or south:
I ran the crank around again 'til TDC to check the other side. As you can see, the line's rotated 180 degrees, but still not pointing north or south.
Once I got the advance backing plate off, the inner pin obviously hasn't fallen out:
And you can see the cut-out in the rear of the backing plate where that pin resides, so I don't really see how you can possibly install the backing plate incorrectly with that cut-out (for the pin) there:
While I had it out, here's the advance inner rod. You can see the cam at the points end and the pin at the advance unit end. The rod is new (as of late last year) and looked fine, so I just cleaned and relubed it before reinstalling:
In case they're useful, here're a couple of shots of the inside of the cam shaft with the rod removed. From the points side:
From the advance side:
Once I reinstalled the backing plate, I moved the crank to TDC:
....and re-checked the orientation of the engraved line on the backing plate to see if it was now pointing either north or south. As you can see, it's just the way it was when I took it apart:
Here's a photo of the assembled advance with the weights in their normal (closed) position:
Here, I'm gently pushing the weights apart to illustrate the arc of movement. It looks pretty normal, so I think the apparent tight fit from yesterday's photos were just due to the missing cir-clips:
And here's a short video showing the movement of the advance weights while I'm manually turning the advance rod (from the points side) after I got everything back together:
http://triumphpc.com/work/xs2/final.avi
I haven't driven the bike since I got it back together (late, dark and I was too beat), so I'll reserve final judgement for a bit. But, from what I saw just revving it in the parking lot after I finished this evening, I think I'm kinda back where I was a couple of days ago where the engine's running great except for a slight roughness in the idle (likely related to that ghosting I was getting with the timing light when I timed the left cylinder).
Hopefully, if my description of the ghosting sparks someone's thoughts as to what might be causing it, I can finally get this long thread closed and use this heat wave for something productive (Yamaha'ing).