The ugly dog worked well. Turned and threaded. I left a couple or three on the fits.
Now we get down to the sticky bit. A first year apprentice could get this far.
I have to massage out the sins created by my class three spindle bearings. 0.0003 run-out means 0.0006 variance in the fit. It probably will translate to about. 4 tenths out of round due to the tool radius.
So, I decided to do some layout on the register face before I pulled it.
I set up the nice sharp threading tool I used for the bearing retention thread, pointed it at the face. I applied layout dye. With some back lighting and magnification, I lined the point up with the 3 & 5/8 OD.
I put a good 1 inch travel indicator on the cross slide, zeroed the dial, and dialed down 0. 421. It was good to see that the machine's cross slide dial and the indicator agreed within a couple of tenths. I scratched the face at that point.
I checked the line's diameter two ways. I set up a pair of good 6" dividers with a Starett rule under magnification, and it matched the diameter. I got out a pair of new digital calipers and checked again. Looked good.
Laying out a bolt circle with 3 holes can be a pain. I went ahead and applied trig, and used the Pythagorean equation to get my chord length. (There is a table in Machinery's Handbook).
I marked a radial line, pricked it. I set my dividers under magnification, scratched two chord marks. I keep those dividers pretty sharp. By feel, I picked up one of the intersections, and scratched to the other unpricked mark. Checked under magnification, and bingo all 3 lines crossed together. I pricked both crossed lines.
Wish that happened every time...
Thanks for bearing with me..
Lastly before I yanked it out of the headstock, I used a small square to scribe tangent lines on the OD. Some very sticky locating to do there later.
I did this layout here/now because I will end up doing the milling work in the lathe. I don't have a way to grab and hang the dividing head waaaay back there to drill and bore the face of the new spindle. I am limited by the working envelope of what I have.
I'm still sorting out how I will drill and bore the cam holes. I think I have the way to locate them accurately.
OK, so through the generosity of some nice people, I have a copy of the ISO standard, and now know exactly where to poke holes for the cams.
My inferred placement was 3 degrees off, and 0.020 too far from the register face. By turning the retention pins out one turn on the gage line, it worked. However, it did not feel right. I will massage the register face back a bit. Overall, I am pretty happy with the test fixture.
Here is the setup I used to precisely mark the spindle.