@Bob Tascione wrote:
Great job zeroing in on the problem Tom and really good, clear description on how you moved through the check list!
I am a little confused on one part though. I’ll get to it in a second.
Yes it is possible that the hole in the jewel is too large or that the pivot is too small. Also possible that the jewel setting has been replaced with one a bit taller which might move the cap jewel out too far. How’s the end shake look?
Other than turning the thing upside down and trying to observe how much play is in the vertical, I can’t really tell about the end shake. I looked at the side shake and I observed it to be less than the 5 degree limit. The end shake doesn’t appear to be excessive, but that’s based only on a visual observation. It’s difficult to see much when you have it assembled. I would like to be able to calculate the dimensional difference between the balance staff length and the assembled components, but we’re talking hundredths of a millimeter, so I’m guessing that would be a futile exercise with the equipment that I have. I did try shaking it while listening, but that didn’t reveal anything either.
@Bob Tascione wrote:
Things do sometimes get replaced with the wrong parts over the years. Other possibilities too like pivot broken off or ground down too short. Sometimes hard to tell.
When you say “The safety roller appears to be a replacement and looked to be seated squarely onto the balance cross arm/balance seat and seemed to be installed correctly. The safety jewel seems to be well secured and nicely perpendicular to the safety roller table”. I’m not quite sure what you mean.
The last person to repair the watch scribed the date “1983” on the balance wheel cross arm, so other than that and the fact that it looks to be newer than the rest of the components, I just assumed that it was not original to the watch. The balance staff and safety roller look newer as well. Again, I’m only going on what I can see, so I could be wrong about that. The comment about the roller jewel being perpendicular to the roller table was to note that the jewel wasn’t tilted out of the veticle plain and thus would be unlikely to strike anything when assembled into the watch.
@Bob Tascione wrote:
When you say safety roller are you referring to a double roller or the roller jewel? It’s a single roller.
If a double roller is it possible that it isn’t an 1883 movement? What are you seeing that makes you feel that the roller might be a replacement? Is it the bottom part of a double roller (safety roller) that’s dragging or is it the roller jewel? If it’s a single roller and the jewel is hitting then the jewel may be too long or the table is sitting too low.
I did try to look to see it the jewel was hitting the plate, but I couldn’t really tell. However, where the thing seems to be hanging, there isn’t anything that the jewel could collide with because it’s hanging when the jewel is on the back side of the rotation.
@Bob Tascione wrote:
I think as you suggested I would check that the roller is seated all the way up against the hub. …snip
Yes – I think I’ll just put it on the staking set and give it a few light taps just to make sure.
@Bob Tascione wrote:
The 1883 uses two different staffs. One is the 1365 (small waist) and the other is the 1364 (large waist). The taper for the roller is slightly different. If your watch was replaced with the wrong staff then the roller may possibly be sitting a tad lower than it should be and could sit slightly off axis. George Townsends staff interchangeability list shows the major dia. of the small waist staff being .60mm tapering down to .53mm and the large waist staff as .59mm down to .57mm measured over same length…hence more taper on the small waist staff. When I measure the two staffs here I’m not getting that much variation between the two. It’s something to consider though.
Bob
I did see that on the OFrei site and in some of the catalogs that I have. I guess replacing the staff will be the last resort. If it comes down to that, I suppose I’ll need to buy one of each and try them both until I solve the problem.
Thanks for responding with the great suggestions. I know you all get tired of my inane questions. 
Later,
Tom