Manual wind watch runs for a few hours then stops

[Francis]

I'm new to this forum and watch repair.  I've started to explore the subject with an inexpensive hand-wound Molnija pocket watch that I cleaned and serviced.  Amplitude, etc were all good at the end of the process and it ran well for a few weeks.  But lately, it just stops after a few hours with plenty of power reserve. The balance starts when I shake it and then works again for a few hours and stops.  I have two questions:

1) What might the problem be that makes it stop (is it a mainspring that somehow gets stuck)?

2) Is it usual to have some (older?) movements needing that initial shake after winding to get started?

[Klassiker]

If you have a timegrapher, you could compare the readings now with those you had immediately after the service, and report any differences. Most likely the problem is nearer the escapement end, not in the mainspring barrel, and most likely you did not do a perfect job of cleaning and lubricating, so that a spot of high friction or a blockage has developed. Also, if the balance is way out of beat it can stop due to the slightest disturbance and needs a shake to get moving again. If you know that the beat error is below 1ms, then you can rule this out for sure, and in answer to your 2nd. question, no; well adjusted older movements will self-start just as new movements will.

Failing an indication of the cause from looking at the timegrapher traces, your only option is to disassemble, inspect carefully, clean, re-assemble and lubricate. See it as an opportunity to inspect your own work and learn from any mistakes you find.

[Francis]

Thanks, Klassiker.  That's helpful. 

Timegrapher shows the readings haven't deteriorated - but the balance is way out of beat (it always was and is on a couple of other Molnija's I've since dealt with).  It may be beat-error as you suggest, but the watch stops without disturbance (ie sitting on the shelf) - so I wonder if it's the friction issue that you raise.  Could it be the pallet fork/escape-wheel lubrication - I can't remember if I had the knack of it back when I did this rebuild.  

Still, probably time to strip down and re-do as you say.  And/or face the next challenge of correcting the beat error in the absence of a mobile stud holder as found on new movements.

Again, thanks.

[Nucejoe]

If way out of beat, it will stop when impulse delivered by the fork gets weak enough. 

Best chance to adjust  the beat  is when you had everything but the cock and oscilator off the mainplate, turn the hairspring until impulse jewel comes to rest in midpoint of the banking pins.

Did you remove and clean the setting?  Check the pivot end under good magnification?  Check the end stones and hole jewels? 

Regs 

[Francis]

Everything looked fine under magnification - except the beat error.  I've avoided playing with it, but I see I need to bite the bullet because it's way off. I like your suggestion of using the main plate with just the balance cock etc for adjustment.  

Thank you

[Michael1962]

Welcome to the forum Francis.

Any photos to share with us?

[Nucejoe]

On 9/30/2021 at 5:31 AM, Francis said:

 I like your suggestion of using the main plate with just the balance cock etc for adjustment.  

Mark Lovic teaches this in one of his vids, so the wstch will run, final adjustments, however, best done on timegrapher. 

In case stopping is not associated with the period of or correct multiples of wheels period. I would take a close look at pallet jewels and escape teeth for wear. 

Did you oil the escape teeth?

[Francis]

On 11/30/2021 at 7:34 PM, Michael1962 said:

Welcome to the forum Francis.

Any photos to share with us?

 

Hi Michael1962:

Thanks for the welcome – and from a near-neighbour too (I’m in Fairhaven, VIC).  I’m new at this watch game (started with covid lockdowns!) – and it’s been an interesting and occasionally frustrating journey.  I tried to fix the beat-error on the Molnija as suggested and totally stuffed it up – it’s not the easiest thing to do.  But I don’t think that was the problem.  The phenomenon of working for while and then suddenly stopping happened on another watch.  When I pulled it out and simply removed the hands, dial, hour wheel etc without any further disassembly – it worked as new.  So it was something that I'd got wrong in the final assembly.

Anyway, you asked for pictures.  Here are two showing what I have been trying to achieve.  I wanted to rescue old beat-up pocket watches and convert those with unsavable or no cases to wristwatches.  In the first is a panel of three - a modern[ish] swiss conversion on left and pocket-watch-derived Molnija and Hamilton wristwatches middle/right).  And a picture of the same Hamilton wristwatch from the front and a restored pocketwatch version.

 

Enjoy your summer, Michael.

Cheers

Francis

 

 

[Michael1962]

Not bad at all Francis.

Where are you finding the watches that you are using? Australia or overseas?

[Francis]

Just now, Michael1962 said:

Not bad at all Francis.

Where are you finding the watches that you are using? Australia or overseas?

It's a combination - but mainly overseas and mainly on eBay.  The molnija's from eastern block countries - they are in the price range of the cheap Chinese movements that some people use for practice.  I've bought, fixed and sold a few of these during our (seemingly) never-ending lockdowns.  They look pretty nice all cleaned up - whether wrist or pocket watch.  That molnija wristwatch case is from Ukraine - but there's a Chinese one for the Swiss Unitas 6497/8 that works well (it's actually the left one of that group of three).

The Hamiltons are quite a bit more expensive, but there's a lot of the small ones where scrappers have kept the case for gold and sell the movements.  Again eBay (US), although harder to find for a good price with postage and duties - but I've got lucky a couple of times.  That complete pocketwatch was a non-worker with a broken mainspring and screwed up keyless works - easily fixed and reasonably priced.

Still love to know what I've been doing wrong with that final assembly.

[Klassiker]

8 hours ago, Francis said:

Still love to know what I've been doing wrong with that final assembly.

I think you are right to narrow it down to the final assembly. Matching up "foreign" movements, cases, dials and hands is tricky, because the clearances are so small. You need to carry on with your process of elimination, removing one possible area of collision at a time (hands to crystal / hands to hands / hands to dial / hour wheel to dial) until the problem disappears.

Beautiful-looking watches by the way.

[JohnR725]

On 9/29/2021 at 3:24 PM, Francis said:

Timegrapher shows the readings haven't deteriorated - but the balance is way out of beat (it always was and is on a couple of other Molnija's I've since dealt with).  It may be beat-error as you suggest, but the watch stops

I don't suppose you can give us a picture?

 

[KarlvonKoln]

I don't know if anyone has asked this, but have you noticed if it stops at the same time, each time?  Or if it stops after a particular duration, such as a certain number of hours, or hours plus plus a certain number of minutes?  That kind of stoppage can direct you to either a place of friction in the gear train, or between the hands when they reach a particular configuration.
There are many reasons a watch might stop, but a process of elimination, as Klassiker says, will usually locate the problem. 

[Francis]

21 hours ago, Klassiker said:

I think you are right to narrow it down to the final assembly. Matching up "foreign" movements, cases, dials and hands is tricky, because the clearances are so small. You need to carry on with your process of elimination, removing one possible area of collision at a time (hands to crystal / hands to hands / hands to dial / hour wheel to dial) until the problem disappears.

Beautiful-looking watches by the way.

 

Thanks for the nice complement on the watches, Klassiker.

By trial and error I was able to isolate the problem to the hour wheel and assume it was too much friction between it and cannon pinion.  I swapped for another hour wheel and it ran fine, swapped it back and it wasn't happy. 

My understanding is that too tight a cannon pinion (ie too much friction with the centre post) doesn't affect the way a watch runs, only the way it winds.  But I'm assuming friction between the hour wheel and cannon pinion is a no-go because they're supposed to run at different speeds.  Do I have that logic correct?  

I have no idea why it was OK for a while and then stopped.  But I'm beginning to feel that old watches have rules of their own.

[JohnR725]

On 12/3/2021 at 3:45 PM, Francis said:

The molnija's from eastern block countries - they are in the price range of the cheap Chinese movements that some people use for practice.

they may be in the same price range but are they in the same quality range?

41 minutes ago, Francis said:

By trial and error I was able to isolate the problem to the hour wheel and assume it was too much friction between it and cannon pinion.  

the hour wheel should fit with basically zero friction. This is why commented above if there is a quality issue like tightfitting components that can be a problem or if you start swapping components especially on older watches that can be an issue because they may be manufactured the slightly different tolerances.

then did you lubricate between the hour wheel in the cannon pinion?

 

 

 

 

[Nucejoe]

Malnija were produced at two plants in different locations,  some parts wont interchange between same calibers  produced at the other location. ,  I wonder if the issue you were facing is one such case.

I have twenty NOS Malnija pocket watches, some fresh out of box refuse to run, have a bunch of used ones that seem to run smoother, seems to me designers though parts best polish themselves as the wach runs,  Nevetheless  providing regularely serviced they work for a life time.

 

 

[Klassiker]

2 hours ago, Francis said:

My understanding is that too tight a cannon pinion (ie too much friction with the centre post) doesn't affect the way a watch runs, only the way it winds.

No effect on the winding, but setting the watch will be difficult and transmit too much torque down the gear train.

[Francis]

2 minutes ago, Klassiker said:

No effect on the winding, but setting the watch will be difficult and transmit too much torque down the gear train.

Sorry, I miswrote.  Yes, setting.

[anilv]

1. Most Russian watches have quite weak mainspring. Having said that they will usually keep time for at least a day. Did you remove the mainspring from the barrel? If it is 'set' then you need to source a replacement. Its not easy to source new parts for russian watches but I can usually find one that fits in my stock of swiss parts.

2. What I usually do to check the train is to give the barrel a wind and see if it wind downs smoothly. But one thing to check is the barrel holes in the mainplate and barrel bridge. If they are worn the barrel will stick under load.

3. An oft overlooked area is the mainspring arbor and how smooth it moves in the hole. It is recommended to install the arbor into the barrel with the mainspring and check for free movement, If the lid or barrel is slightly bent it will not allow the mainspring to release fully. I must admit that this is quite rare and I usually skip this when reassembling a watch but if I have power transmission issues this is one area I check.

Anilv

[Nucejoe]

On 12/5/2021 at 9:51 AM, Francis said:

I  have no idea why it was OK for a while and then stopped.  But I'm beginning to feel that old watches have rules of their own.

   Did you treat pallets with epilame?   if you didn't, oil has migrated form the face of pallet stones therefore weakening the force of impulse ,  eventually stopping the watch.

 

 

[Francis]

43 minutes ago, anilv said:

1. Most Russian watches have quite weak mainspring. Having said that they will usually keep time for at least a day. Did you remove the mainspring from the barrel? If it is 'set' then you need to source a replacement. Its not easy to source new parts for russian watches but I can usually find one that fits in my stock of swiss parts.

2. What I usually do to check the train is to give the barrel a wind and see if it wind downs smoothly. But one thing to check is the barrel holes in the mainplate and barrel bridge. If they are worn the barrel will stick under load.

3. An oft overlooked area is the mainspring arbor and how smooth it moves in the hole. It is recommended to install the arbor into the barrel with the mainspring and check for free movement, If the lid or barrel is slightly bent it will not allow the mainspring to release fully. I must admit that this is quite rare and I usually skip this when reassembling a watch but if I have power transmission issues this is one area I check.

Anilv

Thanks, anilv.  On reflection – I reckon it was the mainspring not releasing for some reason.  I dismantled the barrel assembly, cleaned everything and put the spring back by hand.  I’ve done a few now and they can vary – probably the most variable part of these watches.  I can’t remember the state of this one, but I wouldn’t be surprised if it wasn’t perfect (I’ve yet to find one that is) or if the mainspring was on its last legs.  Unfortunately, I took the advice about fixing the beat error and did that first – but with a non-adjustable stud carrier it was beyond me, so I slipped and wrecked the hairspring.  A pity.  But it’s all part of the learning experience.

[Francis]

8 minutes ago, Nucejoe said:

   Did you treat pallets with epilame?   if you didn't, oil has migrated form the face of pallet stones therefore weakening the force of impulse ,  eventually stopping the watch.

 

 

No I didn't treat with epilame - so maybe.  But some more advice if you don't mind.  Before the last step of fitting the balance assembly, I make sure that I'm getting power through to the pallet fork.  I look to see if the pallet fork "flips" with the lightest of touches - and that it does so in both directions.  Does that provide some indication as to the force of impulse?

[Nucejoe]

15 minutes ago, Francis said:

 Does that provide some indication as to the force of impulse?

Yes it does , but how do we know the force would remain strong enough and for how long? 

Oil gets brushed off the pallet face with every stroke of escape teeth, epilame fixodrop keep the oil put.  

What I don't quite understand is that I see a similar behaviour when  I don't lube pallets at all, runs good for a few days or weeks then stops. I suppose escape teeth starts scartching pallets, the accumulative effect of which stops the watch, not sure though.

Good luck pal

 

 

[JohnR725]

1 hour ago, Francis said:

I took the advice about fixing the beat error and did that first – but with a non-adjustable stud carrier it was beyond me, so I slipped and wrecked the hairspring.  A pity.  But it’s all part of the learning experience.

To help you in the future I'm curious as to what you're procedure was it didn't work?

 

[jdm]

1 hour ago, Nucejoe said:

   Did you treat pallets with epilame?   if you didn't, oil has migrated form the face of pallet stones therefore weakening the force of impulse ,  eventually stopping the watch.

That is not possible. It is just fine to not epilame pallets, as it has been done for more than a century. Watches run by virtue of precision mechanic, lubrications is just a factor to further smooth and preserve that precision.