What's Wrong with my Watch?

[Marv]

I just got this Bunn back from service. It was my great grandfather's watch, so I had it cleaned and lubed.

I hooked it up to my home-made timing grapher and got this (both watches are dial up):



Now, the amplitude is good, I think, but the beat error is poor. However, the signal in the "oscilloscope" looks like crap. No amount of gain setting can fix this.

Here is the same watch with the tracks:



So, here is my personal Bunn that I just "serviced" the balance wheel. I'm a complete newbie, but the signal on the timing oscilloscope looks more like what I would expoect:



So, why the stark difference in signals? The cases are different between the two watches, mine is a 1911 18s and my great grandfather's is a 1921 16s.

 

[roughbarked]

So have you tried getting into better beat?

 

[Marv]

So have you tried getting into better beat?

No. This is straight from the service shop.

I would guess that the hairspring collet needs to be turned slightly to do that?

But all that hash in the oscilloscope trace wouldn't be from beat error and it doesn't seemto do it when the pendant is up, only dial up.

 

[roughbarked]

Yes you do turn the collet to get it into better beat.
I'd probably be asking whoever serviced it why they couldn't set it better and check it better?

 

[Skutt50]

The Beat Error is a bit too high but the watch should run without any problems (except a slightly lower swing.) Now, your swing is 328 degrees and with a zerored Beat Error this may increase. Keep an eye on if the swing gets close to knocking. If so you might consider having a slightly weaker mainspring installed.

it doesn't seemto do it when the pendant is up, only dial up.

If this happend ONLY when you have it Dial Up it may be linked to the hairspring beeing too close to the balance cock and it actually touches the cock.

Try listening with your ear. Do you hear a different sound when holding the watch Dial Up and Dial Down?

Also check the end shake of the balance.

Anyway I would contact the repair person to get his/her oppinion. If you work on it yourself he/she may refuse any warranty claims.

 

[gmorse]

Hi Marv,

If you only get the noisy trace in a horizontal position, it may be that the balance spring is rubbing slightly. Does your software allow you to hear the actual escapement noises? It doesn't appear to differentiate between the three principle noises of unlock, impulse and drop.

Putting it in beat is certainly the first step, but there are more outliers in one pallet than the other, so there may be something amiss there. In any case, these days I'd expect that anyone taking money to service watches would have some form of timing equipment.

Regards,

Graham

 

[Marv]

Hi Marv,

If you only get the noisy trace in a horizontal position, it may be that the balance spring is rubbing slightly. Does your software allow you to hear the actual escapement noises? It doesn't appear to differentiate between the three principle noises of unlock, impulse and drop.

Putting it in beat is certainly the first step, but there are more outliers in one pallet than the other, so there may be something amiss there. In any case, these days I'd expect that anyone taking money to service watches would have some form of timing equipment.

Regards,

Graham

Thanks, Grahm. Thanks to the others that took the time to respond.

I can easily listen to the audio. My setup is DIY. I built the sensor pad, preamp, and run it into a USB audio interface. Electronics is part of my real job.

Let me know what you think.

 

[Skutt50]

To me that metalic sound, like someone is beating on a steel rod, sound a bit different between the two positions.

I have heard similar sound from the hairspring on a couple of movements that ran very well.

I still think you have a problem with the hairspring beeing too close to (rubbing against) the balance cock. The balance (and specially the outer coil of the hairspring) will move a tiny bit closer when you place it Dial Up, which that would explain the difference.

 

[gmorse]

Hi Marv,

I think Skutt is right, but here are some images from Audacity and eTimer.




The eTimer has a high-pass filter enabled, which is why the traces are much cleaner, but they both definitely show some ringing in the horizontal position, which is usually down to the balance spring brushing against the underside of the cock.

If you have Audacity or a similar editor, try slowing it down to about 0.2x and listen to the extra noises.

Regards,

Graham

 

[Marv]

Hi Marv,

I think Skutt is right, but here are some images from Audacity and eTimer.

View attachment 647290 View attachment 647291
View attachment 647292 View attachment 647293

The eTimer has a high-pass filter enabled, which is why the traces are much cleaner, but they both definitely show some ringing in the horizontal position, which is usually down to the balance spring brushing against the underside of the cock.

If you have Audacity or a similar editor, try slowing it down to about 0.2x and listen to the extra noises.

Regards,

Graham

Thank you for taking the time to run my audio files!

What you and Skutt are saying makes sense to me.

 

[gmorse]

Hi Marv,

This file has been slowed down to 0.1x speed but the pitch remains unchanged.

Regards,

Graham

 

[Marv]

Hi Marv,

This file has been slowed down to 0.1x speed but the pitch remains unchanged.

Regards,

Graham

Thank you doing that. It sounds like like something wiping over the coils of spring to me. What a neat diagnostic trick.

Do you own a full copy of eTimer? How do you calibrate the software for any latency in the computer's analog to digital time base? I would expect that the crystals used for the A/D's time base are probably ±100 ppm, maybe 50 ppm.

 

[SpringDriven]

The lift angle is probably not correct, which is providing an incorrect amplitude. To determine lift angle you wind the watch until you see that the balance arms are rotating 180 degrees in both directions. Then adjust the left angle on the timing machine until the amplitude reads 180. Pocket watches in my experience have a lift angle in the 40s.

The hairspring collet would need to be rotated to bring the watch in beat, assuming the hairspring coils are concentric. This is a balance out job, and should be done in the inspection phase before running through the cleaning machine. It is not difficult while the watch is fully apart to adjust the bear error, or get it very close so when you have assembled and are in the timing phase you can make a tweak if you need to.

Lubrication or lack thereof on the escapement can be rough looking and noise. And my experience servicing pocketwatches is that escapement lubrication can be a problem if the pallet fork is not clean enough it is very very easy for the grease to creep up the sides of the stones...

As far other reasons that would have to be observed. The comment on the hairspring rubbing would create a lot of noise, and kill amplitude...

There are a lot of variables and without being the watchmaker that serviced the watch it would be difficult to say.

 

[gmorse]

Hi Marv,

Do you own a full copy of eTimer? How do you calibrate the software for any latency in the computer's analog to digital time base? I would expect that the crystals used for the A/D's time base are probably ±100 ppm, maybe 50 ppm.

Yes, I do now, after a long period using the demo system. I can best answer your question by reference to the discussion on stability on the Delph website, and recommend that you try the demo version.

Regards,

Graham

 

[Marv]

The lift angle is probably not correct, which is providing an incorrect amplitude. To determine lift angle you wind the watch until you see that the balance arms are rotating 180 degrees in both directions....

How do you "see" an amplitude of 180°? This sounds very subjective.

I guess the next question is, what constitutes 180°? If I think of a circle with a dot at the 0° position (say 12 O'Clock on the circle), how far clockwise will the dot need to travel in the clockwise direction? Is it to the 6 O'Clock position, then swing back (counterclockwise), through the 12 O'Clock point and to the 6 O'Clock point? That's ± 180°.

Or is it first to the 3 O'Clock position and then counterclockwise to the 9 O'Clock? ±90°.

Sorry for the newbie question, but I really want to learn.

 

[Skutt50]

Make a dot and wind slowely. When the dot stops in the same position in both directions the dot will look like one dot. Thats where you want to be.

 

[SpringDriven]

How do you "see" an amplitude of 180°? This sounds very subjective.

I guess the next question is, what constitutes 180°? If I think of a circle with a dot at the 0° position (say 12 O'Clock on the circle), how far clockwise will the dot need to travel in the clockwise direction? Is it to the 6 O'Clock position, then swing back (counterclockwise), through the 12 O'Clock point and to the 6 O'Clock point? That's ± 180°.

Or is it first to the 3 O'Clock position and then counterclockwise to the 9 O'Clock? ±90°.

Sorry for the newbie question, but I really want to learn.

I use the position of the balance arms to note amplitude. From your perspective that would be a single point travelling 12 to six, and then six back to 12, repeat. So long as the balance arms are not moving less than or more than those two points, that would be 180 degrees of amplitude. Considering that the balance should be moving a similar amount of amplitude in each direction. You could only observe 12 to six in one direction and that would be the same.

A dot is much easier to see and follow as suggested, but you have to be able to remove that dot.

 

[gmorse]

Hi SpringDriven,

A dot is much easier to see and follow as suggested, but you have to be able to remove that dot.

A tiny spot of artist's white acrylic paint is quite visible and is easy to remove.

Regards,

Graham

 

[Marv]

New puzzle for me.

The beat error jumps every 1 hour and 5 or so minutes (see green trace). Thought to myself, well, the two hnds must be coinciding when that happens! Brilliant, right? No! The two hands are 90° apart when that happens.

I'm puzzled why this happens greater than 1 hour apart.

 

[roughbarked]

Clearly you are seeing more than your watch repairer saw?

 

[SpringDriven]

Is the shellac that is holding the impulse stone loose, so that it is moving? Or a loose roller table? Loose hairspring collet?

But the above problems would have an impact of the amplitude... Beat error should not change? Maybe your timing machine?

 

[Marv]

Is the shellac that is holding the impulse stone loose, so that it is moving? Or a loose roller table? Loose hairspring collet?

But the above problems would have an impact of the amplitude... Beat error should not change? Maybe your timing machine?

Could be. Let me throw another watch on my home-made contraption.


Well, the results are inconclusive. It did not behave exactly the same, but did get funky after an hour. So, I suspect the something with the measurement system is amiss.

 

[Marv]

What should my next step be?

I am not confident with the guy that worked on it. To me, he appears to limit his work to cleaning and oiling and has little interest in digging deep into problems to the level I would like to see.

The watch runs 35 seconds/day fast and the regulator was adjusted fully slow, so, something is still amiss with the watch.

I know the beat error is bad, but would that be the cause for the fast running issue?

Also, there is the hairspring noise in the dial up position that needs to be resolved. How does that impact performance?

 

[Skutt50]

What should my next step be?

The first thing you have to decide on is if you are going to try fixing it yourself or if you are sending it to a watchmaker you have faith in!

If you decide to give it a try yourself the first thing to do would be to inspect the outer coil of the hairspring with a loupe. Is it very close to the balance cock? If so you could try to lower the hairspring stud a tiny bit. That may just fix the clinging noice.

Then you could address the Beat Error. You would have to inspect where the balance wheel is when the mainspring is fully would down. Turn the balance so the pallet fork is right in between the banking pins and mark the balance with the position of the hole for the hairspring stud.

Now you remove the balance from the balance cock and turn the hairspring collet so the stud is positioned where you marked the balance wheel earlier. Put everything back and check your timing.....

Pay attention to the swing and timekeeping. If the swing increases closer to 350-360 you need a weaker mainspring!

The timing issue can be dealt with using the timing screws on the balance cock but leave that for later.

 

[Marv]

The first thing you have to decide on is if you are going to try fixing it yourself or if you are sending it to a watchmaker you have faith in!

If you decide to give it a try yourself the first thing to do would be to inspect the outer coil of the hairspring with a loupe. Is it very close to the balance cock? If so you could try to lower the hairspring stud a tiny bit. That may just fix the clinging noice.

Then you could address the Beat Error. You would have to inspect where the balance wheel is when the mainspring is fully would down. Turn the balance so the pallet fork is right in between the banking pins and mark the balance with the position of the hole for the hairspring stud.

Now you remove the balance from the balance cock and turn the hairspring collet so the stud is positioned where you marked the balance wheel earlier. Put everything back and check your timing.....

Pay attention to the swing and timekeeping. If the swing increases closer to 350-360 you need a weaker mainspring!

The timing issue can be dealt with using the timing screws on the balance cock but leave that for later.

Thank you for your help. I really appreciate it.

Well, I think I need to first fully understand the problem to make an informed decision on whether this is something I want to even try or within my grasp.

Let me ask you a few questions.

First, think back to when you first started. If you could go back to that time, would this be something you would encourage yourself to do?

Second, what tools would be needed for the job?

My tool chest consists of the following:
• Bergeon screwdrivers
• Tweezers
• Horotec hand levers 1mm & 2.5mm
• Bergeon # 5395-75-N Gel Pad
• Loupes 5X, 10X; Plus a Stereo Zoom microscope (I do a lot of surface mount PC board work)
• K&D 600 Inverto Jr.

 

[Skutt50]

First, think back to when you first started. If you could go back to that time, would this be something you would encourage yourself to do?

Not on the first 2-3 watches. There was no need to.

To clear the hairspring by repositioning the stud, yes. Nothing special here. You need to loosen the stud to remove the balance so you can clean the jewels and when you put it back you need to position it.

To adjust beat error took me a bit longer because I had no timing machine to make me aware of bad beat error. I visited a friend some months after I got started and he had just bought a timing machine. We played around with it and we did adjust the beat error on some watches.

To check the hairsprig vs the balance cock is just a visual inspection with a loupe.

To reposition the stud you only need a loupe and a screwdriver.

To adjust the beat error you need the same tools. (and a marking pen) One of the smaler screwdrivers can be inserted in the slot of the collet so you can turn it. I have re-shaped an old oiler to create a flat tip which I use, but a small screwdriver will do the same job. And unless you have good vision close up, you might want to use some magnifier/loupe......

You did state in the beginning that you had serviced a balance wheel and you already have a timing device, so i asumed you had some basic experiece.

The most common beginners mistakes in my oppinion are to break a pivot/jewel or to slip and damage the hairspring..... If you feel uncomfortable doing the job and if you value the watch, send it to a watchmaker. If you want to get into watchmaking give it a try!

EDIT: You have a Stereo Microscope! Wow, I whish I did.... That will make it much easier for you to see close up what is happening....

 

[Marv]

Not on the first 2-3 watches. There was no need to.

To clear the hairspring by repositioning the stud, yes. Nothing special here. You need to loosen the stud to remove the balance so you can clean the jewels and when you put it back you need to position it.

To adjust beat error took me a bit longer because I had no timing machine to make me aware of bad beat error. I visited a friend some months after I got started and he had just bought a timing machine. We played around with it and we did adjust the beat error on some watches.

To check the hairsprig vs the balance cock is just a visual inspection with a loupe.

To reposition the stud you only need a loupe and a screwdriver.

To adjust the beat error you need the same tools. (and a marking pen) One of the smaler screwdrivers can be inserted in the slot of the collet so you can turn it. I have re-shaped an old oiler to create a flat tip which I use, but a small screwdriver will do the same job. And unless you have good vision close up, you might want to use some magnifier/loupe......

You did state in the beginning that you had serviced a balance wheel and you already have a timing device, so i asumed you had some basic experiece.

The most common beginners mistakes in my oppinion are to break a pivot/jewel or to slip and damage the hairspring..... If you feel uncomfortable doing the job and if you value the watch, send it to a watchmaker. If you want to get into watchmaking give it a try!

EDIT: You have a Stereo Microscope! Wow, I whish I did.... That will make it much easier for you to see close up what is happening....

I wish I could see in stereo! I have a form of strabismus that makes it hard for me to use two eyes in binoculars or microscopes, but you learn to adjust.

Yeah, I did pull the balance wheel from a different Bunn and cleaned and reinstalled it. Might not have been my brightest moment, but as the Klingons say, "It's a good day to die!"

Back to reality. So, using a screwdriver to adjust the collet, I am thinking that you give it a slight twist to open the slot of the collet so that it can be removed from or moved about the staff?

I take it that you must remove the balance assembly and place it on some kind of anvil with a hole it it so the balance wheel lays flat?

I actually have a manual milling machine and lathe in my garage (used for developing prototypes for my business), so making some tooling is not out of the question for me.

 

[Skutt50]

Back to reality. So, using a screwdriver to adjust the collet, I am thinking that you give it a slight twist to open the slot of the collet so that it can be removed from or moved about the staff?

Don't try to remove it. Just put the tip of the screwdriver in the slot of the collet and turn. It should move around the balance arbor quite easily.

To make a holder you could use a flat 2x3 piece of hard wood with hole drilled into it. (I made one from oak and drilled holes in steps of 0.5mm ranging from 1mm to 10mm. Useful for many jobs!)
You could also hold it with your fingers if you are careful. Just don't put pressure on the ends of the cuts in the wheel.

A lathe is extremly useful for a watchmaker. I would feel lost without it but I don't think you need it for these repairs.......

 

[roughbarked]

It isn't wise to open the collet with a screwdriver. If you want to adjust the beat, the screwdriver flat goes in the side of the collet to only hold it while you turn the balance to reposition the hairspring.
You'd be better off to lift the collet off with the screwdriver by placing it under the collet and twisting slightly. Works better with two srewdrivers.

All of this will be in the book about watch repair that I hope you have.

 

[Marv]

Don't try to remove it. Just put the tip of the screwdriver in the slot of the collet and turn. It should move around the balance arbor quite easily.

To make a holder you could use a flat 2x3 piece of hard wood with hole drilled into it. (I made one from oak and drilled holes in steps of 0.5mm ranging from 1mm to 10mm. Useful for many jobs!)
You could also hold it with your fingers if you are careful. Just don't put pressure on the ends of the cuts in the wheel.

A lathe is extremly useful for a watchmaker. I would feel lost without it but I don't think you need it for these repairs.......

Okay. That makes sense now.

The lathe I have is too large for watch work. It has an 11" swing and 5C collets that range from 1/16" (1.6mm) to 1-1/8" (28.6 mm). It's like a lorry compared to a scooter.

 

[Marv]

It isn't wise to open the collet with a screwdriver. If you want to adjust the beat, the screwdriver flat goes in the side of the collet to only hold it while you turn the balance to reposition the hairspring.
You'd be better off to lift the collet off with the screwdriver by placing it under the collet and twisting slightly. Works better with two srewdrivers.

All of this will be in the book about watch repair that I hope you have.

Thanks for the clarification. I'll get back to my books and read.

 

[Marv]

Well, I am going to give it a whirl and try to adjust the beat error on the watch.

However, I'm going to practice first. I found this cadaver watch on eBay. It's one year newer than my great grandfather's watch, but the balance looks the same. So, hopefully, the procedure will be the same. If I muck it up, at least it's not that important.

The surprising thing is, the cadaver watch runs better than my great grandfather's watch! Also has the 160 hour motor barrel.



Here is the timing graph:

 

[Skutt50]

That looks like a nice movement.... 21 jewels .... adjusted in 6 positions and temperature......

Have you considered keeping this and finding a much simpler movement in worse condition to practise on? Or perhaps there are other problems like damaged dial, broken shafts etc...?

 

[179]

I don't know why you call this a cadaver watch. It is a type 2 60 hour Bunn Special ( the most desirable variant of the 60 hours) double marked. This is not a watch to practice on, it should be restored. From the photo i see no major issues. Many collectors would be thrilled with this watch

 

[Marv]

I believe you are right. The watch movement is in very good shape and my wife thought I should restore it as well. Beautiful watch.

I need dial screws, dial, watch hands, and a case (with crown) to make it complete.

What does "Double Marked" mean?

 

[179]

It is marked motor barrel 60 on the barrel bridge, and sixty hour on the train bridge. Thus double marked type 2. The lettering needs to have the black refilled, simple job.

 

[Times]

It is marked motor barrel 60 on the barrel bridge, and sixty hour on the train bridge. Thus double marked type 2. The lettering needs to have the black refilled, simple job.

May I ask how do you do this simple job of “black refill”?

 

[Skutt50]

May I ask how do you do this simple job of “black refill”?

Try this thread:

https://mb.nawcc.org/threads/repairing-enamel-on-watch-cases-and-bridges.133088/

 

[179]

I use black Testors enamel, applied with the tip of a toothpick. Let dry for a few minutes and remove excess with a clean toothpick. It would be nice to have a junk mvt. to practice on. Good luck.

 

[Marv]

Thanks for the information about the "double marked".

Very nice idea about the paint. Thanks, everyone.