On The Bench Today, Job Numbers 180057 and 180055

 The Aurora watch, job number 180057, has not been running reliably and is hard to wind and set. I think the main reason (there may be more than one) for its not getting through most 24 hour tests is that it is just not fully wound. It's really rough.

The trouble is that all the parts are quite crude and just barely mesh, push and pull the way they are intended to.

While going over the keyless works again, I thought I'd also mention the regulator. Notice the index of the regulator in this first image. There is a big screw with a crow slot. This screw is not holding anything, The index is held in place by the two small crews at either end. The big screw in fact turns freely. What is it?

Well it's part of a micro-regulator set that is not there. It's missing a sliding part that would fit between the index and the screw. This part would have teeth on one side so that turning the screw would cause it to move from side the side, taking the arm of the regulator with it.

The thing is though, that I think they sometimes shipped them like this, with the part missing. I have seen others missing this part, having no marks or other signs that it was ever there.  If you do an image search and look as close as possible at other movements, a lot of them look like this. This is personal conjecture of course, but when you work on these early American products they really feel like the products of factories that operated in a pretty frantic, ad-hoc manner.

Aurora used, I don't know, 5 or 6 regulator designs, and held valuable patents on them.

On to winding/setting...

The gears on this really grind. They just don't fit well, and it shows in the wear on them which only makes it worse. I don't want to loosen the cannon pinion to make it easier to set unless it's the only option. The first thing to try is some oil along the cannon pinion's grooves. 

Also, one thing I hear sometimes about lever set movements is that they do not pop into setting mode easily. On a modern watch you snap the crown into setting and it is in setting, instantly.

These are not modern watches.

The first image shows the lever tucked in. The movement is in winding mode. Turning the crown  turns the ratchet wheel just outside the top of the image and winds the mainspring.

This second image is what is looks like when the lever has been pulled out. Pulling the end of the lever away allows a cam to in turn allow the small idler gear to pivot (the cover over all this, the part that pivots, is called a "vibrating arm"). The gear then engages the minute wheel (removed in this image) to set the time - or well, it almost does. Actually, the pivot is not quite far enough.

In this state, turning the crown clockwise will do nothing. Setting is not fully engaged.

A small turn of the crown counter-clockwise creates a little torque and pivots the arm all the way. Note in this image how the cam has popped fully over the extended edge of the vibrating arm, forcing the setting gear further over and engaging the minute wheel (now in place).

All this is governed by two springs, which are visible in these photos. One pushes the vibrating arm into winding position when the cam is out of the way (lever in). The other pushes the cam into setting position when the end on the lever is out of the way (lever out).

These springs work against each other so their relative strength is important. They are also 100 years old. And on this watch one closely resembles a bent piece of wire which the other looks like part of an old mainspring. If I saw these in an Elgin, Waltham or Hamilton, I'd say some watchmaker of old had a broken spring to deal with and had done their best to rig something up.

So, long story short, yes these mechanisms can be a little sloppy. We do the best we can.

One last note about this watch, for today, It has a very heavy duty dust cover ring that slips over and around the outside to cover the space between the upper and lower plate. It's quite snug. This is a suspect in the watch stopping. I think it is putting enough pressure here and there to cause shifts in the plates, the pivots and thus cause gears and pinions to shift and bind a little. I am re-seating the cover to make sure I have it fully down. I'll also be careful not to over-tighten the case screws as that can sometimes cause the same issue. I'll leave them "just enough". A movement that runs fine not cased, but mysteriously stops when fully assembled is not uncommon.

On to this one...

Job number 190055 is one I thought was going to have a really nice result, once we found the needed escape wheel. Indeed, on the machine, it times not just well but dang near perfect with an extremely low beat error. It responded nice and smoothly to adjustment, just like it should. It's running great, or seems to be.

In practice it gains in the ballparks of 20 minutes per 24 hours. This is a mystery. I'll be checking a few things out later...