First try at temperature compensation

[steve1811uk]

Hi All, I just wanted to post something about my first try at temperature compensation adjustment.

 

Basically I have been working on my first pocket watch from 1935 that turned out to have a cal 19.71 movement. I had got this running very well with amplitude DU/DD of 310 degrees at full wind and 285 degrees after 24 hours. I did have to change the balance staff as that was terribly worn.

I did notice that there was a variation in rate due to temperature.

Normal +3 SPD

Hot -2 SPD

Cold +9 SPD

It appeared that the split balance was under compensating. I read the books that I had on the subject (De Carle and Fried) that suggested that If I could move weight to the free end of the balance arms then that would give more correction at temperature. The books even gave a guide on which screws to move to achieve different levels of compensation. I proceeded with caution and moved the screws indicated. I was then thoroughly amazed at what this 90 year old watch was achieving:

Normal +3 SPD

Hot +1 SPD

Cold +1 SPD

I have noticed that the poise is just a little out now (as it was before) but I should be able to adjust that with considered adjustment of the 4 meantime screws as that is where the heavy spots appear to be. That's a job for another day as I am still in shock. Steve.

 

 

Edited April 6, 2023 by steve1811uk
Added image

[JohnR725]

3 hours ago, steve1811uk said:

I have noticed that the poise is just a little out now (as it was before) but I should be able to adjust that with considered adjustment of the 4 meantime screws as that is where the heavy spots appear to be.

ideally the mean time screws should all be even unlike two of yours that are very grossly out of adjustment.

and yes it's amazing what a railroad grade pocket watch will do if it was properly adjusted at the factory and hasn't been played with by inexperienced people which typically is what happens to watches over time. That includes things like Rolex watches that come in that are out of poise because people adjusted the weights inappropriately.

 

[steve1811uk]

31 minutes ago, JohnR725 said:

ideally the mean time screws should all be even unlike two of yours that are very grossly out of adjustment.

and yes it's amazing what a railroad grade pocket watch will do if it was properly adjusted at the factory and hasn't been played with by inexperienced people which typically is what happens to watches over time. That includes things like Rolex watches that come in that are out of poise because people adjusted the weights inappropriately.

 

Thanks, they are a lot more even now after a bit of dynamic poising. There was also a big timing washer on the screw next to the meantime screw at the end of the balance arm. The one that is screwed right in. I thinned that washer right down. Now the balance is poised and the meantime screws are pretty even all round.

7 minutes ago, steve1811uk said:

Thanks, they are a lot more even now after a bit of dynamic poising. There was also a big timing washer on the screw next to the meantime screw at the end of the balance arm. The one that is screwed right in. I thinned that washer right down. Now the balance is poised and the meantime screws are pretty even all round.

I've also got the regulator centred. With lift angle set to 41 degrees after 1 hour I am getting:

DU/DD 0 spd 305 deg

PU 0 spd 245 deg

PL, PR, PD around -7 SPD 245 deg

I could try and improve the slower positions but I think that would make PU slower. As it's a pocket watch worn mainly PU or left DU on a desk would it even be worth attempting to get it even better, or even possible?

 

Edited April 6, 2023 by steve1811uk
Added a bit

[JohnR725]

11 hours ago, steve1811uk said:

As it's a pocket watch worn mainly PU or left DU on a desk would it even be worth attempting to get it even better, or even possible?

realistically it looks fine but if you want to be obsessed about it wait 24 hours and run your timing again and see if there's any problems there. Then ideally you should run the timing compared to a master clock over several days. The timing machine just points you in a direction of it looks really good. But if are aiming for a chronometer standards or really precision standards then you time over several days. Or if you're like the watch companies run at several days at whatever temperatures they were testing the compensation at. But still it looks really good I would just leave it alone.

[steve1811uk]

41 minutes ago, JohnR725 said:

realistically it looks fine but if you want to be obsessed about it wait 24 hours and run your timing again and see if there's any problems there. Then ideally you should run the timing compared to a master clock over several days. The timing machine just points you in a direction of it looks really good. But if are aiming for a chronometer standards or really precision standards then you time over several days. Or if you're like the watch companies run at several days at whatever temperatures they were testing the compensation at. But still it looks really good I would just leave it alone.

Wise words I think. Thank you for your input. The rates have become more equal overnight due to the reduction in amplitude. I'm going to get the dial and hands back on and wear the watch for a while to see how it goes. Incidentally, I do see that the amplitude of the watch is affected by handling more so than any wristwatch I have worked on. For example if you pick the watch up and move it around then put it back dial up on the timegrapher then the amplitude can drop to say 240 degrees and then take a good 15-20 seconds or so to build back up to the 300 degrees. It's my first pocket watch project so I do not know if that is normal behaviour or not. I suspect that it is, due to the heavy balance and slowish best rate 18,000 bph. Any views on that?