Free oscillations test off the rails.

[mbwatch]

A beat up Vostok 2214 with a rusty but still nice dial didn't move as well as I wanted during an initial free oscillations test manually turning 180 and letting it spin. So I gave it a little puff of air, and learned this is not a good idea on a wristwatch with an overcoil.

Time to call it a night and unless I already have another balance complete in a box somewhere, or I get lucky untangling it, this one is probably headed for the parts bin.

[caseback]

I don't see any sharp bends, so you might be lucky 

[nevenbekriev]

How this actually happened? May be the cock was not fixed with the screw and was lifted by the air jet???

It is important to understand, that when the pallet fork is not in place, there is nothing to limit the amplitude of the balance, so it can make several full turns, even to the point where the hairspring will get deformed. In rotation direction where the spring 'unwinds', the spring may increase it's size enough to go under the balance wheel, and this is eased by the fact that it will no longer want to stay in plain. So, never blow that strong!

[Neverenoughwatches]

4 hours ago, mbwatch said:

A beat up Vostok 2214 with a rusty but still nice dial didn't move as well as I wanted during an initial free oscillations test manually turning 180 and letting it spin. So I gave it a little puff of air, and learned this is not a good idea on a wristwatch with an overcoil.

Time to call it a night and unless I already have another balance complete in a box somewhere, or I get lucky untangling it, this one is probably headed for the parts bin.

Bloody unlucky mb, it should unravel as it didn't suffer direct contact with anything but it does look tough with it wrapping over the wheel like that. Get your hairspring head on matey.

[rehajm]

5 hours ago, caseback said:

I don't see any sharp bends, so you might be lucky 

Yah- I see lots of tangles but no creases…just maybe

[mbwatch]

5 hours ago, nevenbekriev said:

How this actually happened? May be the cock was not fixed with the screw and was lifted by the air jet???

The cock was firmly screwed down, but something caused the hairspring to snag the balance screws. And the return oscillation pulled the other side of the hairspring in underneath the balance or snagged an opposite screw. Even though the hairspring looked flat, I suspect it was tilted very slightly and when it opened up on huge amplitude it just hooked itself over the balance screw. Or the spring was flat but the continued force of air from my blower pushed a coil down to hook onto the balance wheel.

 

4 hours ago, Neverenoughwatches said:

Bloody unlucky mb, it should unravel as it didn't suffer direct contact with anything but it does look tough with it wrapping over the wheel like that. Get your hairspring head on matey.

 

52 minutes ago, rehajm said:

Yah- I see lots of tangles but no creases…just maybe

If there is any serious damage, it was probably done while removing the balance cock from the mainplate. I don't think I can work on it again until Saturday but I'll let you all know what becomes of it.

I really enjoy learning these lessons in practice and then showing you all my mistakes. And this is why I still never open up a watch worth more than $30.

[Neverenoughwatches]

27 minutes ago, mbwatch said:

The cock was firmly screwed down, but something caused the hairspring to snag the balance screws. And the return oscillation pulled the other side of the hairspring in underneath the balance or snagged an opposite screw. Even though the hairspring looked flat, I suspect it was tilted very slightly and when it opened up on huge amplitude it just hooked itself over the balance screw. Or the spring was flat but the continued force of air from my blower pushed a coil down to hook onto the balance wheel.

 

 

If there is any serious damage, it was probably done while removing the balance cock from the mainplate. I don't think I can work on it again until Saturday but I'll let you all know what becomes of it.

I really enjoy learning these lessons in practice and then showing you all my mistakes. And this is why I still never open up a watch worth more than $30.

Good lessons to learn mb, avoid persuading high amplitudes on isolated balances especially if they can snag up on something such as wheel components, wonder if it would have happened on a plain wheel ? If you'd be so kind as to purchase another balance....plain wheel this time and perform the same exercise to provide us with a comparison....that'd be great ...cheers sunshine.

[mbwatch]

I'm thinking about what happens to a coiled garden hose or an old telephone cord, where if a coil is opened too far it will tend to want to twist downward until it flips underneath. Wonder if a hairspring would try to behave the same way, given enough room to breathe (even though it is concentric rather than vertically coiled).

33 minutes ago, Neverenoughwatches said:

you'd be so kind as to purchase another balance....plain wheel this time and perform the same exercise to provide us with a comparison.

for science.

[Neverenoughwatches]

1 hour ago, mbwatch said:

I'm thinking about what happens to a coiled garden hose or an old telephone cord, where if a coil is opened too far it will tend to want to twist downward until it flips underneath. Wonder if a hairspring would try to behave the same way, given enough room to breathe (even though it is concentric rather than vertically coiled).

for science.

Always in the name of Science and becoming more educated about watch repair mb.

[Nucejoe]

 I don't see any sharp bend in the coil.

Stud restricts maneuvering  hairspring free , I unpin the spring from stud, then open regulator boot, lift the balance with tangled spring.

In case you have no high magnification tool such as  strong loups or microscope, take pix of balance and spring on your smart phone,  enlarge the pix on your screen to clearly see all details and decide how to go about untangling  the spring without creating bends, take your time to find out how to go about untangling , once untangled, you can see if the spiral is a goner or not. 

Mild intermitant  tangential  to the rim of the balance air blow is risk free aproach to start oscillation, and continue to blow to increase the amplitude. 

Good luck.

 

 

 

 

 

[mbwatch]

10 hours ago, nevenbekriev said:

here is nothing to limit the amplitude of the balance, so it can make several full turns

Because I mostly work on pocket watches, I have made the habit of blowing  hard on the balance to test that it will oscillate down for at least 60 seconds. But the old pocket watch blued steel hairsprings are so much more robust than this fine modern wristwatch hairspring and I had never seen one unfurl enough to deform.

But this experience will make me stop using the blower forcefully. It was only in the last 1 or 2 months I read one of your posts somewhere else about turning the balance 180 degrees and counting to about 100 oscillations. Much gentler than a full air blast. I actually started with your method here, got only 30 oscillations, then moved to the blower.

[mbwatch]

Well my friends, it is a Friday-Before-Black-Friday miracle. I have got this watch running again. This is my first significant hairspring repair and I feel like I just leveled up.

When I removed the balance complete from the cock, I found it had 2 coils swung underneath and looped around the roller table. I might have made things harder for myself by lifting them over in the wrong sequence.

But when the hairspring was removed, it didn't look that bad. The overcoil had several issues but no kinks. Concentricity looked very decent and flatness was a little out, but I discovered while fixing the flats that what looked like concentricity defects were actually tangles. There were two tangles which I could not see well under a loupe (have no microscope) without a 3D view.

The first two photos are what I started with:

I did not know the strategy for removing a tangle I first did one in the outer 2 coils by manipulating the stud through them - which worked but was very challenging and probably dangerous.

The second tangle was much farther in and I didn't see any feasible way to pass the stud through there and have it pop out untangled. I figured I could maybe get some oilers in between and just push tangle around and around until it made its way out to the stud where I could lift it through. This worked.

With the tangles out, I was left with not too bad a state, 2 or 3 minor concentricity issues and the overcoil pointing down a little. Also the collet was twisted at a slight angle.

Because I don't know any theory on forming overcoils, I decided I would just try to wing it. I mounted it back to the balance cock and pushed here and there until it was pretty well centered. Any additional adjustments to the overcoil I have to make while the whole balance is in the watch, because I can't work out the 3D relationship. Someday I will learn to do this the right way but for now, it's just YOLO and poke at the running balance until it looks right.

I took the opportunity to check its poise, which was fine on this tool. I probably won't go to the trouble of dynamically posing this watch.

And now it's running in! I have it at 1.4ms BE and honestly, I think I may leave it here. After my first service attempt from a year ago, it was at >3.0ms BE but running flat in all positions. How I got myself into this mess in the first place was by going back in to dial down the BE and reclean the balance jewels, then my ocillations test and my tangled up hairspring.

Right now, the verticals are very good (>250°!). I'm okay with the BE since there's no mobile stud carrier (and this is a vostok not a patek which has already caused me a lot of trouble). The horizontal traces are noisy a bit irregularly wavy. I will go back in to clean those jewels once again, because at one point before messing up the hairspring I swapped the lower from a spare movement while end shake seemed too high. I can swap that back and I expect to see some improvement.

 

 

Thanks for all the encouragement. I didn't expect I would actually fix this one rather than toss it in my "USSR box"

[nickelsilver]

Impressive work, contrats!

[mikepilk]

I found this videos so useful, I bookmarked a link. It's the man himself :  Mr Henry B Fried ! 

The first part is untangling - he is quite "robust" with the spring!  Then he goes on to removing kinks and bends.
Sometimes a hairspring looks a total mess, but if only tangled, just one simple manipulation and it's fixed. 

 

[mbwatch]

3 hours ago, mikepilk said:

The first part is untangling - he is quite "robust" with the spring!  Then he goes on to removing kinks and bends.

Yes, I think the main lesson I just learned from HB Fried's demonstratoin is that next time I work on a 16 or 18 size pocket watch, I can twist and yank and pull on the hairspring with abandon and nothing bad will happen

Edited November 23, 2024 by mbwatch

[mbwatch]

On 11/22/2024 at 4:46 PM, mbwatch said:

And now it's running in! I have it at 1.4ms BE and honestly, I think I may leave it here. After my first service attempt from a year ago, it was at >3.0ms BE but running flat in all positions. How I got myself into this mess in the first place was by going back in to dial down the BE and reclean the balance jewels, then my ocillations test and my tangled up hairspring.

Right now, the verticals are very good (>250°!). I'm okay with the BE since there's no mobile stud carrier (and this is a vostok not a patek which has already caused me a lot of trouble). The horizontal traces are noisy a bit irregularly wavy.

Examining the horizontal positions with the best magnification I have (which is not good at all) I could see there were times when the overcoil was rubbing at high amplitude. I could see that was because it was angling slightly downward near the stud, which pushed the Breguet bend up where it could touch the balance cock. I was able to fix that and it took all the noise out of the traces.

Of course I could not let it stay with 1.4ms beat error... It took way too many tries to get it down below 1ms but now it is as low as 0.3ms (or as high as 1.0ms, which I think is an imperfection still in the terminal curve). Final product: a Vostok 2214 from the scrap bin that now runs like a demon.

I had to redo the shellac on the pallets, and ultimately my only mishap today was dumping my container of shellac flakes into the carpet.

Pictured DD, while the verticals are a little faster but narrower at +30/255°/0.3ms. I had planned on selling this but now I may have to hold onto it as a trophy.

Edited November 24, 2024 by mbwatch