Specs Way Off After Reassembly

[Lawren5]

First time newbie here. I’ve been playing with some old watches and reassembling them but one has given me trouble. Prior to teardown, the readings were:

Rate: 75.1 s/d

Beat error: 0.1 ms

Amplitude: 244 deg

BPH: 18000

After teardown, cleaning, lube and reassembly:

Rate: 612.0 s/d

Beat error: 17.8 ms

Amplitude: unobtainable

BPH: 18000

I disassembled and reassembled the watch again but with the same results. It appears to run smoothly and I couldn’t find any areas where it might be sticking. It doesn’t appear to have an adjustment for beat rate. The watch is a 1950’s era from the Berman Watch Co. of Switzerland.

Would appreciate any suggestions on what the problem might be.

[khunter]

Photos would be helpful, did you remove the hairspring? It could be the hairspring has shifted on the balance, or even possibly the roller table and or roller jewel is out of alignment or the jewel is no longer perpendicular to the table. More information is really required to give any advice. 

[digginstony]

First time newbie here. I’ve been playing with some old watches and reassembling them but one has given me trouble. Prior to teardown, the readings were:

Rate: 75.1 s/d

Beat error: 0.1 ms

Amplitude: 244 deg

BPH: 18000

After teardown, cleaning, lube and reassembly:

Rate: 612.0 s/d

Beat error: 17.8 ms

Amplitude: unobtainable

BPH: 18000

I disassembled and reassembled the watch again but with the same results. It appears to run smoothly and I couldn’t find any areas where it might be sticking. It doesn’t appear to have an adjustment for beat rate. The watch is a 1950’s era from the Berman Watch Co. of Switzerland.
Would appreciate any suggestions on what the problem might be.

Without an amplitude reading, you can't really trust what the timegrapher is saying. Have you still got two trace lines going across the screen and are they close together.
If so it may be that you've been too enthusiastic with the oil.

Sent from my SM-T585 using Tapatalk

[Lawren5]

I removed the balance cock assembly but did not disassemble it. The roller jewel looked fine but I'll try to get a picture of it. The attached photo is about as close as I can get before it starts blurring.

I don't have a timegrapher so I've been using the Watch Tuner app on my iPhone. Looks like I'll need to invest in one.

[khunter]

It's really hard to tell, but is the hairspring inside the two regulator pins? If not that can cause erratic readings, rubbing, etc....Inspect everything closely and make sure nothing is out of place or distorted.

[Lawren5]

I removed the assembly and confirmed that the hairspring is between the two regulators. There were also no signs of rubbing or anything unusual that I could see. The readings are very erratic so it still may be the hairspring or related components. I'll try to get some photos.

[khunter]

In removing the balance it is very easy to distort the hairspring. If the coil is deformed it will cause beat errors and erratic behavior. All other things being equal I'd start looking there.

[Lawren5]

Here's another photo.

[JohnR725]

Does your timing app have a graphical display if so what does it look? Personally am suspicious of all your numbers the beat is so extreme. For instance the modern timing machines the maximum beat error they will measure is 9.9 ms. You're almost twice that. Not picking up the amplitude suggest is a problem with the signal.

This is where having a timing machine that allows you to listen to the ticking sounds would be nice. If they hairspring is rubbing or touching bumping into something it usually makes a nice clanging sound. The extra sounds confuse the timing machine. 

That it be nice to have an entirely different picture that's closer and much much sharper. But even then your problem might not be there may only be in the watch. So maybe my imagination but the hairspring doesn't quite look centered in the picture? Almost looks like they hairspring could bump into the back side of the regulator pin and maybe the terminal curve isn't quite right but the picture just I'm guessing perhaps. So put it back in the watch because that's where the problems occurring. Look very very carefully and yes it's very hard to see hairsprings at times. Follow the hairspring from the stud to the regulator pins make sure the rest the coils are bumping into that when they open up they shouldn't be anywhere close. Then make sure the hairspring looks flat really isn't a lot a clearance if the things not flat it can touch the balance wheel arms.

[digginstony]

Personally I would dismiss those readings. I enquired about the trace lines to determine if indicate a beat error or if they have become scattered. If so has the movement been fully demagetised.

Sent from my SM-G920F using Tapatalk

[jdm]

53 minutes ago, digginstony said:

 If so has the movement been fully demagetised.

Such a major fault it can only be a balance / hairspring issue, not magnetization.

[jdm]

1 hour ago, JohnR725 said:

 Personally am suspicious of all your numbers the beat is so extreme. For instance the modern timing machines the maximum beat error they will measure is 9.9 ms. You're almost twice that.

The thing is, for the OP it was measuring fine before, but not after. Basic logic says, blame the mov.t not the instrument.

[digginstony]

The thing is, for the OP it was measuring fine before, but not after. Basic logic says, blame the mov.t not the instrument.

It's not inconceivable for a movement to become magnetized during service. The OP clearly states, visually the movement is running smoothly. With the stated beat error, if accurate, it's unlikely the watch would even run ?

Sent from my SM-G920F using Tapatalk

[jdm]

2 minutes ago, digginstony said:

It's not inconceivable for a movement to become magnetized during service. The OP clearly states, visually the movement is running smoothly. With the stated beat error, if accurate, it's unlikely the watch would even run ?

Trust me on this one. OP is a beginner and doesn't have the eye trained enough to tell if balance is swinging good. Facts:

• When a machine can't detect a steady amplitude, all other values can be discarded.
• Machines that can pick a signal before work, will pick it again after.
• Magnetization can cause timing issues, not lack of amplitude and crazy beat errors.

 

[khunter]

8 hours ago, Lawren5 said:

Here's another photo.

Your hairspring is distorted, it should be a smooth curve from the stud to the regulator pins. It's also not centered due to the distortion, in this photo it's shifted towards the lower left, assuming the pivot is in the jewel.

Have a look at this video. 

 

Edited April 9, 2018 by khunter

[JohnR725]

7 hours ago, jdm said:

Trust me on this one. OP is a beginner and doesn't have the eye trained enough to tell if balance is swinging good. Facts:

 

19 hours ago, Lawren5 said:

It appears to run smoothly

So it appears to be running smoothly is an interesting statement. Plus not having a trained eye? On a modern balance wheel how easy is it to tell whether it's operating correctly or not? This is exactly why I purchased a timing machine a real timing machine that's reliable they can tell me things that I may or may not be able to see.

So one of the problems with a lot of these timing applications are are they really reliable? So the indication is you have a problem visually you think everything is fine then let's find out? The only real test of whether a Watch is fine or not is to put the hands-on wind it up and see what it does 24 hours later.

 

[digginstony]

So it appears to be running smoothly is an interesting statement. Plus not having a trained eye? On a modern balance wheel how easy is it to tell whether it's operating correctly or not? This is exactly why I purchased a timing machine a real timing machine that's reliable they can tell me things that I may or may not be able to see.
So one of the problems with a lot of these timing applications are are they really reliable? So the indication is you have a problem visually you think everything is fine then let's find out? The only real test of whether a Watch is fine or not is to put the hands-on wind it up and see what it does 24 hours later.
 

Exactly and I personally would be monitoring the watches behaviour prior to taking further action. Also am I missing something ? Is the Ops watch running + or - the figures given. Even without a trained eye, I doubt the balance would oscillate with the figures given and the OP could form a valid opinion, probably.
I remember once servicing a pocket watch with quite a healthy balance rotation and after my over oiling, the performance was very poor. It did recover from my abuse. I've since learnt one drop of oil can service many movements.
As a hobbiest I also have the luxury of running a watch after servicing for several days and allowing the watch to settle, prior to any assessment on my timegrapher.
Also I constantly demagetise as I believe this can give a whole range of false readings other than running fast.
All said, it's probably a hairspring issue but I'm learning not to make assumptions.


Sent from my SM-G920F using Tapatalk

[jdm]

36 minutes ago, JohnR725 said:

So one of the problems with a lot of these timing applications are are they really reliable? So the indication is you have a problem visually you think everything is fine then let's find out? The only real test of whether a Watch is fine or not is to put the hands-on wind it up and see what it does 24 hours later.

I for one like to use a physical machine but unlike you, I'm not so dubious about timing applications.. Normally these are developed by intelligent people using simple well know formulas that you've also published in the past, and which have a genuine interest in having what they do to work well. By the way we also have here one of these developers with an interesting on-going thread.

If you can find documented examples of applications giving wrong or different values from a physical machines please post, I'll be happy to learn and discuss, so far we have none.

Again, it's really pointless to shift blame to the instrument when it was reading good before, but not after the timepiece was manipulated.

[digginstony]

I'm not so dubious about timing applications like you seem to be above. Normally these are developed by intelligent people using simple well know formulas that you've also published in the past, and which have a genuine interest in having what they do to work well. By the way we also have one of these developers here.
If you can find documented examples of applications giving wrong or different values from a physical machines I'll be happy to learn and discuss, so far we have none.
Again, it's really pointless to shift blame to the instrument when it was reading good before, but not after the timepiece was manipulated.

I don't think anyone is blaming a particular type of machine or application but that other factors can influence the readings given

Sent from my SM-G920F using Tapatalk

[jdm]

1 minute ago, digginstony said:

I don't think anyone is blaming a particular type of machine or application but that other factors can influence the readings given

Fine, but that is a bit of thread hijacking. Let's start again with the known basic facts:

Prior to teardown, the readings were:
Rate: 75.1 s/d
Beat error: 0.1 ms
Amplitude: 244 deg
BPH: 18000
After teardown, cleaning, lube and reassembly:
Rate: 612.0 s/d
Beat error: 17.8 ms
Amplitude: unobtainable

https://en.wikipedia.org/wiki/Occam's_razor

 

[digginstony]

The OP states he's a Newbie but this reference is to the forum and also has successfully serviced other movements. So he's not a complete novice.
I think a practical observation of the watches behaviour over 24 hours would be prudent.

Sent from my SM-G920F using Tapatalk

[jdm]

15 minutes ago, digginstony said:

The OP states he's a Newbie but this reference is to the forum and also has successfully serviced other movements. So he's not a complete novice.
I think a practical observation of the watches behaviour over 24 hours would be prudent.

Once again, I'm only reporting the summary of known facts. We all know that even the more experienced repairer can have a problem with a watch after servicing, nothing to be ashamed about. And it will be interesting to hear from the OP if the good old methods can prove that the app or machine he's using can go into the bin.

Just let me state that isn't very helpful to derail discussion going after ghosts instead of trying to pinpoint a real possible issue, for example a capstone upside down, a bent pivot, or end-shake issue.  

[digginstony]

Once again, I'm only reporting the summary of known facts. We all know that even the more experienced repairer can have a problem with a watch after servicing, nothing to be ashamed about. And it will be interesting to hear from the OP if the good old methods can prove that the app or machine he's using can go into the bin.
Just let me state that isn't very helpful to derail discussion going after ghosts instead of trying to pinpoint a real possible issue, for example a capstone upside down, a bent pivot, or end-shake issue.  

Wasn't aware I was suggesting chasing ghosts. Looking at the known readings I can't see reference to + or - 75 or 612 respectively. Considering it's an old movement, minus 75 initially then moving to minus 612 after servicing would make more sense to me. Most old movements, neglected, dirty, with no lubrication, usually run very slow but normally don't even run. Applying too much oil can make a slow movement run even worse. Does the app used give a + or - ? ? If it is fact running + then a slightly spiral hairspring or a touching hairspring may well be the cause.
But once again with readings as shown I personally wouldn't do anything without a practical test run of the watch. Ok it might be a bit obsessive but I demagetise before, during and after servicing. It's never my intention to de-rail a thread ?

Sent from my SM-G920F using Tapatalk

[Lawren5]

Thanks to all of you for your responses. I appreciate them and it shows me just how much I have to learn. I will order a timegrapher within the next day or two so hopefully by next week I’ll be able to provide some more meaningful data. In the meantime, I have several questions and comments concerning the information provided.

1)               The movement was demagnetized but there was no appreciable change in operation.

2)               The video on hairspring adjustment was very informative but I’m not quite ready to attempt correcting it. This is a beater watch so at some point I will, if needed.

3)               The balance wheel jewels were removed and lubed but they looked like they were just one piece. Could this be correct or would these jewels likely have caps?

4)               The movement was running fast prior to teardown and is much faster after reassembly. It also takes more winding of the crown to get the balance wheel turning.

5)               You keep referring to me as the OP. I’m afraid to ask what it stands for.

[khunter]

OP stands for Original Poster....