Watch Adjustment

[Alaskamick]

Hi, I'm asking this question more out of interest and learning rather than to solve a problem.

I have a good collection of Omega wrist watches but I started out with pocket watches.

A lot of pocket watches are marked ADJUSTED, the more expensive ones are often marked with ADJUSTED 4 , 5 positions and some TEMPERATURE.

The only adjustment I know of is at the balance wheel.

If I get that spot on how else can it be adjusted without disturbing the initial setting?

Can someone more knowledgeable than me explain it please.

[Chopin]

I believe temperature has to do with the watch being able to maintain time accuracy in more extreme temperatures. (has something to do with the materials used, I guess, but also to design)

4-5 positions means, again, that the watch is able to maintain time accuracy regardless of the angle. Normally time keeping is influenced by the angle of the watch (because of gravity, inertia, weight of parts, etc.). But it depends on what type of movement (quality, design, vintage or modern, etc.).

Someone might explain this better, this is just a quick answer.

[CaptCalvin]

Entire books can be written on the topic. But quick and dirty of it is that the screws closest to the cuts in the rim would be used to adjust for temperature.

[JohnR725]

A related question in the US would be why are Swiss watches sometimes marked unadjusted or limited number adjustments versus American pocket watches for instance? That was because of the early days to protect the American watch industry there were tariffs and things like adjustments the number jewels marked on the plates push the cost up. So the watches were all adjusted or regulated they just weren't marked that.

The terminology is kind of misleading? The simple answer is they verified that the watch would keep time in various positions and temperatures. Then the other thing that happens with watches that are marked multiposition temperature etc. the quality the watch improves.

I'm attaching a PDF it explains what they expect just not how they did it.

So regulated of course means you can move the regulator watch keeps time. Ideally on a railroad watch the regulator is in the middle timing screws were adjusted to get the watch to keep time. Adjusted to position means they took time to poise the balance wheel. The more adjustments the more time they spent to get that right. Adjusted the temperature is usually a bimetallic balance wheel that they moved the screws around to find the proper position where the movement of the bimetallic balance wheel would cancel out the effect of temperatures.

 

 

 

Hamilton 992b 950 timing.PDF

[Walsey]

I found this blog to be very helpful in understanding watch adjusting.

https://adjustingvintagewatches.com/what-is-watch-adjusting/

 

[CaptCalvin]

A perfectly poised balance wheel is only the most basic step to start positional adjustment. If all you need is a poised balance wheel then stating the number of positions adjusted is irrelevant. Most of the work is done at the hairspring. Introducing poise errors to the balance wheel to counteract errors caused by the hairspring is not an ideal solution because these manipulations would have variable effects at different amplitudes. Such method should only be done to take out small errors after the balance is perfectly poised to begin with and hairspring adjusted to as close as you can.

[Michael1962]

Ok, I realise it is as older post, but I have the same questions regarding Adjusted & Unadjusted. I understand the import issues, but reading the adjustingvintagewatches.com link above, I have come up with another question.

Is the movement being left in various positions for a period of time to get the +/- secs per day numbers or are these numbers taken from a timegrapher machine now that they are so accessible?

If they are taken from leaving the watch sitting in a position for a period of time, what is the +/- sec per day number compared against? How would we know if the timepiece used for comparison is accurate? Unless some people possibly have access to an atomic clock?

@CaptCalvin As you have said above, if I understand you correctly, you are saying that we should not adjust the balance any further if it perfectly poised prior to fitting. How many hairspring manipulations can there be though if there are different positional timing errors?

[nickelsilver]

In the past watchmakers had to  spend a lot of time adjusting precision watches, sometimes weeks. With a timing machine things are much faster. In chronometer certification, COSC, Besancon etc., they are cased and dialed with white test dials, and photographed with a camera that takes the image at a very precise time, triggered by a precision (nowadays atomic) clock. The images are taken for the various positions and various states of wind, temperature, etc, over a period of days.

 

As to poising and further adjustments, in a modern watch the balance is poised very precisely by an automated machine, the hairspring is attached to an engineered collet that is also "poised", and there is typically very little adjustment needed once assembled. With a perfectly poised balance, and a well adjusted hairspring system (true at collet, regulator pins well adjusted, flat, etc), there may be a little bit of residual error in the vertical positions that can be adjusted by dynamic poise. But as we are taught in school dynamic poise is done on a watch that is in perfect running order.

 

All that said sometimes I will correct quite large poise errors with dynamic poising, for example I did a LeCoultre 145 pocket watch a couple of weeks ago; this watch is 1.38mm thick, and I didn't like the idea of pulling the hairspring and sticking the balance on a poising tool. It had over 100s delta in the verticals, by stroke of luck one added timing washer brought it to 15s, which is fantastic for such a piece. Other than the vertical delta it was in terrific shape so it was an easy decision to dynamic poise.

Edited April 1, 2021 by nickelsilver

[JohnR725]

12 minutes ago, Michael1962 said:

Is the movement being left in various positions for a period of time to get the +/- secs per day numbers or are these numbers taken from a timegrapher machine now that they are so accessible?

If they are taken from leaving the watch sitting in a position for a period of time, what is the +/- sec per day number compared against? How would we know if the timepiece used for comparison is accurate? Unless some people possibly have access to an atomic clock?

Because it's faster usually use a timing machine. If you look at the various watch companies they usually have recommendations for timing specifications. In other words you rotate the microphone to a position allow it to stabilize like Omega typically 30 seconds then they have a measuring time. If you have a really fancy witschi machine It will  do all that for you use programs at times push the button and you get the numbers.

You could do it by looking at the hands comparing at the something else but because problematic because in other words like if you had one hour per position and you did six position timekeeping it would take six hours. This means the effect of amplitude between a fully wound up watch and six hours later we change the timekeeping. So would probably suggest that you would wind the watch up let it run 15 minutes compare it to time source and timer for one hour in each position and repeat that for it would take you forever timing machine would be better.

Or I suppose if you're really obsessed wind it up and put it in a position and let it run for 24 hours. For each of the positions.

20 minutes ago, Michael1962 said:

If they are taken from leaving the watch sitting in a position for a period of time, what is the +/- sec per day number compared against? How would we know if the timepiece used for comparison is accurate? Unless some people possibly have access to an atomic clock?

I was one set a lecture where the person explain their frustration with precision watches. It was not the problem with the watch it was a problem with the owner of the watch or this case several owners. They would compare their watch time standards like now ancient history VCR clock. A variety of other clocks not known to be very precise at all. So you'd have frustrated customers using improper time standards and wondering why their watch didn't keep time.

As far as a time reference goes in a variety of countries around the world there are or at least were shortwave radio broadcasts. There is also lower frequency broadcasts were your clock will synchronize to those. But it looks like a lot of them are now replaced by GPS timekeeping. You can actually in a special GPS receiver for timekeeping purposes and even a regular GPS is not be so much better than your watch especially if you timer watch over a period of time.

Then understand how the timing machines if you have a really fancy one come up with all the numbers I've attached a PDF otherwise you can do the math by hand.

12 minutes ago, nickelsilver said:

In the past watchmakers had to  spend a lot of time adjusting precision watches, sometimes weeks.

I once gave a lecture on timing machines. One of the things I was looking for was how did they do it before timing machines so I found an interesting book. The person explained everything they did and it took a couple months. That's because each adjustment would take several days to verify whatever the adjustment was to take effect. True there were some breaks where they took weekends off but it took some time. One of the things especially in the early days they would deliberately introduce errors in to fix other errors in the watch.

Otherwise if you aren't so obsessed you could set the watch to a time standard like most towns would have a railroad train station they should have a decent clock. Set the watch to your clock time come back in a week see what it's doing or come back every day. Adjusting without a timing machine is going to take you some time. Even with the timing machine uses still verify that the watch actually does whatever the timing machine says it's doing.

witschi X-D-DVH-Di-Im-N_EN.pdf