Timegrapher

[Treetopspider]

My cousin, knowing that I am a watchmaking hobbyist asked me to take a look at his watch after dropping it recently. It would work for a couple of hours then stop until shaken. Anyway I opened it up and took the balance out to inspect it but there wasnt anything I could distincly identify as the problem. I put the watch back together and now it seem to run without stopping. However, when I put it on my timegrapher I get very diiferent readings at different angles. From +5 second face up to -60 pendant down. at the same time its erratic in the sense that it can suddenly speed up or down for no apparent reason. what would be the likely diagnosis for something like that happening?

Thank you in advance

Johnny

[jdm]

Looks like you should properly clean, inspect and lubricate this mov't. Don't be surprised if you find some debris  or other dirt in some crucial place.

[m1ks]

A service is certainly a good suggestion but after a drop re inspect the seating of the jewels and shock springs, (you'll need to remove the dial) and re check the pivots for damage and straightness, check the hairspring isn't catching in any orientation.

That's a big positional error.

[saswatch88]

Assuming this watch is not antique and has a shock proof system on the balance i would say that would need to be checked first. It could mean a cracked balance pivot jewel even more so than a broken or bent staff. I would say this because of the changes in behavior in different positions, a watch should run slower at when standing vertical in all 4 positions but -60 is way out of the realm of acceptability. That fact that it speeds up and slows down by just sitting and not being touched could suggest a broken jewel as well, but it could also mean something else in the train, but i would say unlikely. In my experience shock proof or not any kind of drop usually results in damage somewhere in the balance. Either way its needs to be fully disassembled. You must put the balance jewels under a microscope as well at the balance staff. Check shakes but side shakes in particular. Make sure balance wheel is true and not weaving up and down. Remove caseback and observe the balance in 6 positions with movement out of the case.

[jdm]

Sorry I had missed that the watch had been dropped. If the watch has already been serviced recently then it can make sense to concentrate right away on the balance and escapement. In similar please post the relevant timegrapher pictures as pattern and amplitude actually do tell more than just rate values.

[Shane]

After thinking a bit about dropping what was a perfectly functional watch.  I would think that most problems with the pivots would average themselves out over the full rotation of its particular wheel and be less dependent on orientation (run fast, then slow) as the fit between its wheel and the wheel/wheels it meshes with changes.  Severely bent pivots turn a conical path and with enough end shake may not have clearance to freely "orbit" around the inner circumference of it's jewel without interference.  This should be most significant face up or face down.  On the other hand, a problem such as a cracked jewels should be totally dependent on orientation, giving more or less clearance to the defect from position to position (run ok in one position and slow/possibly not at all in another.  Depending on the cyclic rate of the symptom, it might suggest one wheel over another.

It sounds like your possibly experiencing multiple problems.

Please feel free to let me know if this sounds plausible.

[Shane]

After reading my own post, I didn't mean that it would "run fast, then slow".  I ment was (run at its normal rate, then slow).

[saswatch88]

7 hours ago, Shane said:

After reading my own post, I didn't mean that it would "run fast, then slow".  I ment was (run at its normal rate, then slow).

 

7 hours ago, Shane said:

After thinking a bit about dropping what was a perfectly functional watch.  I would think that most problems with the pivots would average themselves out over the full rotation of its particular wheel and be less dependent on orientation (run fast, then slow) as the fit between its wheel and the wheel/wheels it meshes with changes.  Severely bent pivots turn a conical path and with enough end shake may not have clearance to freely "orbit" around the inner circumference of it's jewel without interference.  This should be most significant face up or face down.  On the other hand, a problem such as a cracked jewels should be totally dependent on orientation, giving more or less clearance to the defect from position to position (run ok in one position and slow/possibly not at all in another.  Depending on the cyclic rate of the symptom, it might suggest one wheel over another.

It sounds like your possibly experiencing multiple problems.

Please feel free to let me know if this sounds plausible.

What you said is plausible, and now that you clarified that it runs normal then slow makes suspect a jewel even more so. A watch can def run with a cracked jewel, but as you said depending on where the crack is and how bad the crack is can slow the watch even at stand still because the pivot does have a bit of play and depending on when and where the pivot Is rubbing along side that crack, you will see a loss of time, higher beat error, and change of amplitude. Jewels tend to crack on one side which could confirm why there is such a loss in PD position. Again remove balance and observe jewels.

[nickelsilver]

What caliber is it? Did you uncase it to take the balance out?

 

My guess is a shock jewel came unseated, and taking the balance out and putting it back reseated it or at least got it better. It is possible if the shock was enough to unseat a shock jewel it also slightly bent a balance pivot. This would possibly explain the variations in timing in different positions, and even a very tiny amount of bend can create a large positional error- you may need a microscope to see it. If it is suddenly speeding up and slowing down (by how much?) it could be the hairspring is out of true and touching somewhere depending on amplitude. Keep in mind that instantaneous readings on a timing machine can move around a lot, while the average is quite stable. I don't know what machine you have but I set mine for a 10 second window generally.