Pallet jewel depthing?

[spectre6000]

I'm working on an early-40s Certina KF304. It seems to have had a whole host of issues, and I think I've resolved all of them but the crystal (we'll see what shows up later this week) and the pallet fork. The watch has poor amplitude, is running extremely slow, and is extremely variable across positions.

The pallet fork has a lot of shellac on it, which makes me think it's been messed with before. Looking at the engagement, I think another byproduct of this fettling is the entry jewel is too deep. Exit seems fine. I examined the pallet, and there is room for the jewel to be sunk a little deeper.

First question: do the photos below bear out my diagnosis in the minds of those who have a clue (I do not).

Second question: Assuming my diagnosis is correct, I'll need to melt the shellac and sink the jewel deeper into the fork. I looked on eBay for pallet jewel setters, and I'm seeing a variety of tools. None of them are obvious in how or how well they work. Any advice on the best/easiest tool for the job? Can it be done with just tweezers and pegwood or something if it just needs to be set all the way? Does it just need to be set all the way?

 

[nickelsilver]

Can't really tell just from the picture, but the entry stone could be a little deep. The lever escapment is complex, and really needs to be approached in a logical step by step manner. If you move that entry stone in, you will reduce the drop lock on the exit side, which may or may not be ok. If you then move that stone out to increase its drop lock (and total lock), you will increase the drop lock on the entry. Or, it won't unlock anymore at all, as the fork hits the banking pin before the escape tooth can come off the impulse surface of the exit stone.

 

Adjustments to stone depth are often on the order of 0.005mm (5 microns).

It's good to have a way to measure how far you've moved a stone; Bergeon makes a tool for this of course. I don't like it, I use a measuring microscope, and heat my fork on a brass plate and move the stone with with either an oiler or clean pegwood; oiler to move it out, wedging it behind the stone in the slot and twisting, pegwood to push in, holding the fork with stout nickel (or brass) tweezers. The plate has holes in it to accommodate the arbor. Fork upside down. I use an adjustable heating plate under the brass plate (bergeon), but you can do it with an alcohol lamp like most books show. Mark has a video on it.

I recommend Jendritski's book Watch Adjustment for a good description of the checking and adjusting procedures.

 

But to get you started-

With the balance in, power on the movement

-stop the balance with a finger (with finger cot), and rotate it slowly until an escape tooth drops off one side of the fork to the other

-observe the depth of lock at exactly that point. It should be, as a rough rule, about 1/5 the width of the pallet impulse face (just as a length reference). If you move the fork with an oiler toward the balance, the escape tooth should not come off the locking surface. If it does, the stone needs to come out

-continue turning, the lock should increase, this is the run-to-the-banking, or slide, as the fork continues to the banking. Check the fork horn freedom along this time, this is the fork horn hitting the face of the impulse jewel (the safety action before the guard pin takes over). The total lock now should be roughly 1/3 the width of the pallet face.

-check that you have freedom between the guard pin here, using an oiler to move the fork from the banking. Continue around the safety roller moving the balance and the fork

-do this on both sides, entry and exit. If you have good looking drop lock on both, but excessive total lock, move the bankings in to reduce. If you have excessive drop lock, move one or both stones in as needed, keeping in mind moving one has the same effect on the other (for drop lock). If the locks all look good, but you have no or little guard pin clearance on one side, you may have a bent pin. This goes for the fork horns too, but that's rare to have a problem there. If the guard pin looks nice and straight, and the safety roller isn't damaged, it may be that the bankings have been shifted, both, in one direction and the stones moved to adapt, but now the pin is contacting the roller sometimes.

 

For a quick cheat- and this is really a cheat, meant for folks who know how to otherwise check things-

- move the balance until the escapement drops. Check the fork safety (fork horn clearance). This is, or at least should be, a little larger than the guard pin clearance, so the maximum "safety action" the escapement will see. You can move the stone in until you are just on the verge of unlocking when checking this clearance. Most of the time this will work, especially on a modern piece, and you will have safe escapement function with the stones in the fork as far as they can safely go.

[clockboy]

pallet folk jewl setting.pdf

Not easy job see PDF. Also see this from a previous post which includes a pic of the tool I use.

 

[nickelsilver]

Here's another link with some practical info.

 

https://mybulova.com/sites/default/files/file/Joseph Bulova School of Watch Making - Unit 10a.pdf

[spectre6000]

7 hours ago, clockboy said:

pallet folk jewl setting.pdf 2.09 MB · 2 downloads

Not easy job see PDF. Also see this from a previous post which includes a pic of the tool I use.

 

I saw that and looked to see if I could find one (and how to use it). Google doesn't have much to say on the subject. I found a small selection currently on eBay. Some are self evident in their use.

https://www.ebay.com/itm/Watch-Pallet-Jewel-Setter/264799720564?hash=item3da74b1474:g:LgcAAOSwLPlfEj3i

https://www.ebay.com/itm/Vintage-Unbranded-Watchmakers-Pallet-and-Roller-Jewel-Setter-Watch-Tool-used/143656504047?hash=item2172984aef:g:490AAOSwLE9fDQij

Others I can't quite tell if they're accurately listed.

https://www.ebay.com/itm/VINTAGE-COMBO-ROLLER-PALLET-JEWEL-SETTER-WATCHMAKERS-WATCH-JEWELERS-TOOL/254655828623?hash=item3b4aab928f:g:WMwAAOSw6~de4qqk

8 hours ago, nickelsilver said:

 It should be, as a rough rule, about 1/5 the width of the pallet impulse face (just as a length reference).

No way in hell... The exit jewel is about that, and didn't look suspect to me at all. The entry jewel depth is closer to half the impulse face than a fifth. 

Looking at the movement in its entirety, the mainspring was very wrong and causing problems; rectified. The train is excellent all around. There is significant beat error, suggesting someone messed with the balance. The balance was not spinning freely until I backed a few turns off the screw and popped the cock back a bit. I put a shim under it, and now that's happy. It still has low amplitude, and investigation turned up a significant disparity between the locking depths of the pallet jewels. It makes sense to me that the additional force required to overcome that significant extra lock could easily cause my low amplitude.

There is also a significant positional error, and the timegrapher trace is generally non-linear (the rate wanders a bit). This could have a lot to do with the low amplitude, and may disappear once that is rectified.

Could the pallet lock depth/disparity cause the issues I'm seeing (excluding the beat error)? If I fix this, what are the odds I end up faced with yet another problem that's more likely causing or caused by some other issue?

[JGrainger]

Basically what Nickelsilver said..

I actually have one of the tools with the graduated arms.. they were quickly removed after purchase and put somewhere better than on the actual tool (anywhere safe) because there's no real way to accurately calibrate and use them - I use a microscope to measure with. The hotplate is a great idea too, can also be used for convenient bluing of screws.

[nickelsilver]

Excessive lock will certainly cause a drop in amplitude. If it's on the entry side it is worse than the exit. It will show up more in vertical posts as well.

That you have (as I recall) 2 vertcals that are particularly bad I'm guessing there's very much an issue with a stone, both stones, bankings, possibly guard pin (looked a little off in a pic, but hard to tell), or any combination of the above. The checks I described and which you'll find in the link will help you narrow it down.

[rodabod]

It doesn’t look bad from the photos, but I would always check that the banking pins are sitting at the correct angle. 

[spectre6000]

Banking pins look fine to me, but I have to admit this is the only pair of banking pins I've examined closely. They appear perpendicular to the plate and parallel to each other, and that's my understanding of their intended geometry.

I read the Bulova literature, and I'm still trying to wrap my mind around the geometry behind retracting one jewel causing a decrease in the depth of the opposite jewel. I think the procedure outlined is what I need to do to rectify this escapement.

I have had nearly zero minutes to dedicate to that purpose though, as I'm flying solo with a 6 1/2 month old who has recently learned to crawl, and is hell bent on peeling up the edge of the foam play mat and eating it. It's been a challenging evening... For both of us (as she erupts into tears after being pulled away from the tasty vittles for the umpteenth time). I'm such a terrible daddy... :'-(

Edited July 22, 2020 by spectre6000