Watch runs well dial side up, but not dial side down

[swordfish]

I'm stumped, fellows. The first watch I ever built is this FM97. It worked well for several years, but recently has been stopping and restarting unpredictably. I took it apart, cleaned the plates and gears, inspected the pinions and teeth, and inspected the jewel holes.

I fond that in the dial down position, it will run for 5-10 minutes, then stop. A gentle nudge on the fourth wheel gets it ticking again. Also, I noticed—and I think that this is an important diagnostic clue— that if I loosen some screws on the wheel bridge, it will start ticking again, only to stop some minutes later.

In contrast, when I put the watch dial up, the watch ticks merrily until the mainspring energy is depleted.

The beat error is only about 0.4 milliseconds.

I'm beginning to suspect that one of the pivots is too tight in the lower (dial side) plate, despite my not finding problems there when I looked.

I know that the crown wheel is absent in this photo. That is not the problem. I'm in the process of replacing it. I can wind the watch with a screwdriver placed on the ratchet wheel screw.

Any recommendations?

—Eric

Ohio, USA

Sent from my iPad using Tapatalk

[WillFly]

This has happened to me on many occasions - and to others here - and, apart from the odd cracked jewel (which you don't have), it's nearly always been a pivot problem for me. They're not the only causes, of course, but your guess at a pivot causing the imbalance would fit.

Cheers,

Will

[clockboy]

Agree with Wilfly. It is normally is a pivot issue, the first thing to check is end shake on the wheels if OK then with a nice strong loupe/microscope check the ends of the pivots for wear/straightness etc. Also check the jewels again with good magnification for wear of the jewels normally the suspects are the larger wheels. If confident that the wheel train is OK then check the balance for end shake & check the cap jewels to make sure they are spotlessly clean & lubricated. 
Finding this fault & faults is best solved with a slow but methodical procedure 

[vinn3]

good info,  i agree.  my guess its in the balance.     let us know what you find.

[Alastair]

Pivot/jewel/plate bush worn, wheel fouling plate or another wheel, particularly mainspring-barrel, or barrel/bridge.

Guard pin fouling safety-roller.

Hairspring fouling balance.

Balance jewel-setting cracked, mis-aligned, staff bent, endstone worn/cracked, ill fitting.

insufficient endshake on pallets or escape-wheel. Centre-seconds wheel/pinion bent/fouling.

Damaged/worn teeth on wheel or pinion.

[MeticulousOne]

All of the above and sometimes I just clean them again just to make sure they are clean...

[anilv]

1. It ran well before.

2, It started to run erratically.

3. Cleaning didn't help.

Assuming that the cleaning was done well, this probably means something else has happened to cause the erratic running. This usually happens when the watch is dropped, resulting in..

a. Mushroomed pivots.. if the watch is dropped dial down, the pivot will mushroom at its tip. Perversely it will run ok dial down but when placed dial up, the mushroomed bit will 'hang' on the hole jewel and the watch will stop.

b.Hairspring got caught up on the regulator or the regulator has moved as a result of the drop.

c. One problem I encountered was where the screw for the pallet cock was not correct, the head of the screw was too thick so that when the watch was dial down the hairspring touched the screw head and resulted in the watch stopping. Worth checking on this. In your case, the original problem could have been a dirty movement but an oversight during the service may have created a new problem.

 

Go back to basics, assemble the train without the pallet fork and balance. put some wind on the mainspring..see that it winds down smoothly. On a handwind movement, you should see a bit of 'backspin' as the last ounce of energy from the mainspring is dissipated. If you nudge the escape wheel and it continues to spin another few rounds then your escapement needs attention.

Once you're satisfied that the escapement is fine, remove the train and fit the pallet-fork and balance (the escapement). At rest the pallet fork should be at the center of its travel with the balance installed. Your watch seems to have a movable stud holder so this is easy to do. On watches with a fixed stud holder the hairspring will need to be adjusted at the collet. Here you would also examine the pivots closely and triple check that the hole jewels are clean and clear. Endshake should be present but not measurable, The balance should move slightly when you move the movement (balance installed).

Next re-assemble the train wheel and pallet cock, wind the barrel a turn. using a trimmed down pegwood, manipulate the pallet fork to and fro while observing the interaction of the pallet jewels viz escape wheel. Its hard to explain but one stone should catch immediately as the other releases... no dead space.

If after all this and it still doesn't work, put the watch away and start with another.. this will relieve your frustration and allow you to come back with a clearer frame of mind after a decent interval.

Anil

[swordfish]

Gentlemen, we have a diagnosis!

On re-inspection, I noticed that the reflection of the escapement wheel pivot (in its jewel in the top plate) was variable as I manually caused that wheel to turn. I disassembled the watch—again!—and find that the upper escapement wheel pivot (on the pinion side of the wheel) is worn away to almost nothing. (See photo). So, indeed the problem was a pivot. But instead of the lower pivot being too tight, the upper one was too loose. I don't have the skill or lathe to create a new arbor for the escapement wheel, so I'm going to try to buy a replacement for it.

Thanks for your help, advice, and encouragement.

—Eric

Sent from my iPhone using Tapatalk

[StuartBaker104]

Well spotted Eric,

This looks to me like the pivot has broken off, not worn. It's quite possible that this happened during repeated disassembly / assembly, so dont be disheartened if you fit a replacement and still have the fault. However, a damaged pivot is more likely to catch and break when assembling or disassembling so may be the original cause.

Take a really good look for the part which has broken off! You don't want to leave anything in the patient that shouldn't be there!

Cousins has the escape wheel, but I'm sure you can find one in the US

S

[anilv]

well done!

[Cad101]

A nice spot of detective work there Eric well done.

[swordfish]

You make a good point, StuartBaker104. I see now that the pivot is broken, not merely worn. But why THAT pivot, eh? My working hypothesis now is the one you suggest, that I broke an already faulty pivot. I'll carefully inspect its jewel to make sure that the broken fragment isn't there. I have ordered a replacement part from a US supplier.

This reminds me of a case (patient, not watch case) I saw many years ago. A man had some aching in his humerus (upper arm bone) from undiagnosed myeloma (bone cancer). When he opened his refrigerator door, he heard a snapping sound coming from his arm. The mere force of opening the door had broken the already weakened humerus! He then came under my care. Alas, although watches are often difficult to cure, humans are more so.

Sent from my iPad using Tapatalk

[jdm]

You are exactly right swordfish. Staff pivots can act sometime, you can abuse them and they take it, but sometime align things the best you can, and just present the bridge screws and they break.

That's one reason why I think insisting in servicing mech. watches every 5 years is kind of crazy. Before or after something breaks because of manipulation. 

[swordfish]

Follow-up report: I inserted the replacement escapement wheel into place. I had a difficult time getting all three wheels aligned to go through their bridge. I found I had to take the pallet out to get a good view of the underside of the bridge. Now the gear train works.

I'm glad I removed the pallet. I noticed that the entrance jewel of the pallet is bent downward out of the plane of the pallet. (See picture). Of course I don't know if this was a pre-existing problem or an iatrogenic injury I inflicted on my poor trusting watch. My next step will be to gently heat the pallet to coax the jewel back into proper place. What a maddening hobby!

Sent from my iPhone using Tapatalk

[clockboy]

Ok the depth of the pallets is very important & setting them incorrectly it will not run. You will need a pallet setting tool much the same as below (the same as I use). I have also attached a doc I found on the net a few years ago but it explains what you need to achieve.

 

 

pallet folk jewl setting.pdf

[anilv]

maddening but now that you've positively identified a few issues don't you feel a lot better?

Anil!

[swordfish]

clockboy, you gave me good advice. I had been tempted to try resetting the jewel in a roller table warmer, but I now recognize that that would have given too much heat to the contralateral pallet stone. So I got a lovely old pallet warmer fairly inexpensively on eBay, and tried getting the stone into proper position.

I was able to restore the stone to the proper plane fairly easily, but when I returned the pallet to the watch, it STILL wouldn't work. Careful inspection showed that the entrance stone went too deeply into the escapement wheel. On removing the pallet yet again—[sigh]—I saw that that stone was not set deeply enough into its setting in the pallet. Tomorrow I'll try again to get it right. Wish me luck!

Sent from my iPhone using Tapatalk

[clockboy]

OK this might not be correct but what I do is set both pallets to the same height as each other & it always seems to work. If one pallet was already set then set the new pallet to the same height.

[jdm]

On 8/12/2016 at 6:01 AM, swordfish said:

Follow-up report: I inserted the replacement escapement wheel into place. I had a difficult time getting all three wheels aligned to go through their bridge. I found I had to take the pallet out to get a good view of the underside of the bridge.

Were you trying to install the train bridge with the pallet installed? It would impinge on the escape wheel and -as you found- make it very difficult and dangerous.

Regarding the pallet alignment, I think with that you are at the very hart of the Swiss lever system. Get it right it works wonders, get it wrong by hundred of a mm,.. well, you know. Seiko have inspection holes that allow to observe how stones engage the wheel but even that requires a loot of experience. The "proper" Bergeon tool is only GBP650 before VAT
https://www.cousinsuk.com/product/pallet-alignment-tool-bergeon?code=E9511

Good luck!   

[swordfish]

Update: it took me several hours—I'm a newbie, remember—but I was able to coax the entrance pallet stone deeply into its setting. Now the escapement works well. But I find that the watch will tick for only a few minutes (and with low amplitude as measured by eyeball), and then it stops. If I then slightly loosen one of the three screws on the gear train bridge, the watch will start ticking again for perhaps another minute or so, then stop.

Any recommendations?

Sent from my iPad using Tapatalk

[toptime810]

Sounds like you might have to push a jewel

[clockboy]

When you put a bit of a turn on the mainspring you should be able to move the pallet folk and the escape wheel should escape. If this part is now operating correctly then there could be another fault with then watch.

[vinn3]

i like that "bit of a turn"  to start the watch.    says you have done a good job.

 

[ricardopalamino]

On August 21, 2016 at 3:28 AM, jdm said:

Were you trying to install the train bridge with the pallet installed? It would impinge on the escape wheel and -as you found- make it very difficult and dangerous.

Regarding the pallet alignment, I think with that you are at the very hart of the Swiss lever system. Get it right it works wonders, get it wrong by hundred of a mm,.. well, you know. Seiko have inspection holes that allow to observe how stones engage the wheel but even that requires a loot of experience. The "proper" Bergeon tool is only GBP650 before VAT
https://www.cousinsuk.com/product/pallet-alignment-tool-bergeon?code=E9511

Good luck!   

I like how easily you said "tool is only GBP650 before VAT"....To us amateur newbies...

Edited August 27, 2016 by ricardopalamino

[jdm]

Just now, ricardopalamino said:

I like how easily you said "tool is only GBP650 before VAT"....To us amateur newbies...

I was being sarcastic...I'm more amateur than anyone else

I think there are very, very few watchmakers that both have real need for and are qualified to use it.