Accuracy expectations for different types of movements

[spectre6000]

I've recently come to realize that I don't actually know how accurate different types of watches typically are or can be. I have a collection of small fully jeweled French movements, and I need to have an idea for what to expect once serviced and running to know if I succeeded or not. I can't find service sheets, so I'm looking for a general idea.

My expectation is that a high end mechanical movement should be (manufacturer's spec) in the single digits seconds +/- per day, and can be within a second. What about a more run of the mill fully jeweled movement (thinking your typical A Schild or Eterna)? What about jewel free or single jewel movements? Fully jeweled modern Russian movements are -20/+60 S/d, which is higher than I would expect for a similar Swiss movement. What are the general rules of thumb? 

[VWatchie]

I can’t answer your question but there’s an interesting difference between accuracy and precision very well explained in the following video:

 

[VWatchie]

I’ve always been fascinated by accuracy and based on the video lecture linked to in my previous post, and without really being aware of it, precision as well.

So, my thought is that any healthy movement can be reasonably well regulated to be accurate to its owner, even in the case of a movement with a specification of -20 to +60 sec/day like my much-loved Vostok movements. As a matter of fact, I have a Vostok caliber 2409 which on average gains about 3 seconds per month, provided I wind it and wear it consistently not subjecting it to any extreme motions of my wrist. However, in the various horizontal and vertical positions it varies between -25 and +17 seconds per day which is sort of better than the specs but still far from chronometer grade. During a short interval of time, say a day. It can be off by several seconds but over the span of a month it averages out to about 0.1 seconds per day. It’s really a bit of a miracle that I’ve been able to move the regulator arm to a position where it so precisely matches the way I wear it and wind it.

The way I understand it, the specifications of a movement doesn’t really specify the accuracy of the watch and not really its precision either but basically just its extreme rates depending on the position of the movement. If we wind a movement consistently and keep it in the same position, most healthy movements will be very precise, although a chronometer grade movement even more so.

The way I think of it, to achieve precision, the flow of energy from the mainspring to the escapement must be constant as possible and the escapement adjusted to perform equally in all positions. This is of course extremely complex with several interacting parameters and it’s a topic far beyond my present level of knowledge.

Anyway, one especially important factor to achieve precision is the poise of the balance wheel. For example, with my mentioned Vostok movement I’ve observed that in pendant up position the rate is +24 seconds per day whereas in pendant down position the rate is -25 seconds per day. This could be the result of many factors, but to make a long story short, in this case I am fairly convinced that it’s a clear sign that the balance is out of poise. I have ordered a poising tool (still waiting for it) and will attempt to improve this imbalance through static poising. If I succeed my watch will (at least momentarily or I'm afraid likely permanently) become less accurate but hopefully more precise and the watch geek in me more satisfied and happy 

Edited December 20, 2020 by VWatchie

[JohnR725]

16 hours ago, spectre6000 said:

I can't find service sheets,

Finding technical information about watches is always a challenge as most watches don't have a service sheet. Then if they have anything at all It would probably be a parts list this is because watch companies have this silly notion that people who repair watches know what they're doing and they don't actually need a detailed service sheet. Then even if you had anything resembling a service sheet seldom does it have any specifications on performance expectations.

The French a very long time ago used to publish books of information on their watch movements. The one I'm currently looking at has a nice section on the French watch industry talks about types of watches how they run but I don't see anything on timekeeping other than a little on the chronometer timekeeping. Then even the movement sheets basically are the caliber size the beats per hour and a rather nice drawing of the movement with some minor technical descriptions but no ratings specifications and strangely enough zero parts list.

To give you an idea of what I'm looking at I googled the title and Amazon came up first interesting price considering mine was free a very long time ago which makes you wonder if maybe there is a PDF out there somewhere?

https://www.amazon.com/French-Watch-Movements-Technical-Characteristics/dp/B00AQKAW34

17 hours ago, spectre6000 said:

My expectation is that a high end mechanical movement should be (manufacturer's spec) in the single digits seconds +/- per day, and can be within a second. What about a more run of the mill fully jeweled movement (thinking your typical A Schild or Eterna)? What about jewel free or single jewel movements? Fully jeweled modern Russian movements are -20/+60 S/d, which is higher than I would expect for a similar Swiss movement. What are the general rules of thumb? 

Another thing before we get the timing specifications is you should look at performance? My definition of this is the amplitude which also translates to the running time of the watch. For instance you start off with the watch it needs to be serviced wind it up put it on the machine you might  perhaps regulate it but with a set mainspring possibly extremely dirty lubrication issues is it going to run for 24 hours? This is one of those misleading things where watch is in pretty bad conditions set mainsprings etc. wound up tight on the timing machine can sometimes look pretty good.

Then another thing that will come up with all of this what exactly do you mean by seconds per day? That seems pretty straightforward your question but watch on machine regulates read off seconds per day. Except that's not what the watch companies do? Usually they time in more than one position the better quality or better timekeeping of the watch more positions. Then they usually do averaging.

Then I will attach a whole bunch a PDFs for your reading enjoyment first up Omega Working Instruction 81: Basic Checks & Settings of a Mechanical Watch. Before he gets what's interesting in here have you noticed how on all the newer Omega stuff they watermark who downloaded it on every single page in the corner? This one came from cousins material house. Makes you wonder why the watch companies are paranoid over who has their documentation. On page 3 are things that you should be aware of the next page is the timing specifications. You'll notice lots of numbers and this is done with the nifty machine that does this all by itself. In case you're wondering how all the numbers come into existence there is another PDF that comments on how to do all those calculations. Even though your watch specification may be for only one position is always nice to look at the watch in multiple positions. Then basically when you look at the bottom line all the concerned about is the X and the Delta.

As I pointed out normally timing specifications aren't usually found in the service information. It's becoming a little more common now but originally it just wasn't there. Except usually were a lot of times it was there was just in another document that you may not have been aware of. So for ETA there the manufacturing information PDFs. I'm attaching Two of them because each of these watches has multiple of grades. Then the 6497 should be recognized by everyone on this group. Scroll through it to get a lot of technical information and finally get to timing. You'll notice that unfortunately if you look at enough of these it doesn't seem to be a universal if it's this grade it will keep this time it varies between the various watches. Or basically each watch has its own timing specifications unless it's chronometer grade and then that's controlled by whoever does  that.

You'll notice that they also  in addition to giving you the specifications farther down they specify how all these specifications came into existence. So basically when you look at the rating is an average of the positions. It also noticed that they're more concerned of amplitude at the end of 24 hours.

Then back to your original question is read all the information you'll discover that your original question has problems. Conceivably we could regulate a one jewel watch to keep really good time. But if you look at the manufacturing information you find out that the better grade watches a keep better time keeping have better materials that the made out of. So on the short term  your one jewel watch can be regulated that have nice timekeeping but temperature changes long-term use etc. probably won't keep that timekeeping.

8643_WI_81 BASIC CHECKS AND SETTINGS OF A MECHANICAL WATCH_EN.pdf witschi X-D-DVH-Di-Im-N_EN.pdf ETA 6497-1 Manufacturing Information.pdf ETA 2892-A2 Manufacturing info.pdf

[JohnR725]

Yes if I didn't give you enough to read, I checked one other source I have so I will attach that and snip out something. If the watch companies didn't print timing specifications sometimes it can be found in the books that came with timing machines. After all you have a timing machine what sort of timing should you expect the watches you're trying to time?

 

1452880703_GreinerMicromatWatchtimingmachine.pdf

[Poljot]

7 hours ago, VWatchie said:

Anyway, one especially important factor to achieve precision is the poise of the balance wheel.

Unless you have done some work on the balance (staff replacement, etc), you probably don't need to poise the balance wheel. I have not yet seen a single balance wheel Vostok, Poljot that would required poising or truing. And if you did replace the Staff, poising shall be done with the roller fitted.

[spectre6000]

You guys think more of my skills than is warranted! I was truly only looking for the rule of thumb type chart that comes with the timing machines JohnR725 posted there. I want to know if, after I service a movement, I screwed it up or not. Ballpark at best, as that's as good as I expect to get. I serviced one of those small French movements, and I was certain I borked it, but I later realized I just didn't know how inaccurate cheaper movements can be. I don't recall how it graphed out, but I'm certain it was within reason for what should have been expected.

[watchweasol]

Hi As an add on to the Griener posted by JohnR725 you may find this handy to have.

Witschi Training Course.pdf

[JohnR725]

1 hour ago, Poljot said:

Unless you have done some work on the balance (staff replacement, etc), you probably don't need to poise the balance wheel.

Yes I'm in agreement unless you've done something to the balance wheel leave it alone. It would only be if for some reason you wanted to make it better than the factory made it. But when you look at the manufacturing information you can see that trying to convert a lower grade watch to a chronometer grade watch isn't going to work because there's other things beyond the balance wheel that makes it a chronometer grade.

 

8 minutes ago, spectre6000 said:

You guys think more of my skills than is warranted! I was truly only looking for the rule of thumb type chart that comes with the timing machines

I had a suspicion about this I gave you an epic novel and you are looking for?

2 hours ago, spectre6000 said:

I want to know if, after I service a movement, I screwed it up or not.

 

Perhaps we should look at the question differently or what is the real question?

 Learning watch repair is interesting how long do you think it takes to learn watch repair? If you answered a number anything less then forever you answered incorrectly. Yes there are people good at watch repair but watch repair is an ongoing learning process that will take forever. Unless you have a very tiny world then maybe it will occur faster.

Real question here is not just watch timing but an evaluation of how you are doing for learning watch repair.

This is why for beginners I always recommend a timing machine and a brand-new preferably Chinese clone of the Swiss pocket watch because it's cheap. This way can put it on the machine see what it's doing. Disassemble reassemble see if it still doing the same thing. If you start with a broken watch and it's still broken whose fault is that which is really what this question is all about is it not?

As rewind to be using the timing machine to evaluate some of this you should establish a timing procedure. Ideally you should establish a repair procedure especially when you're learning. On the watch comes in you evaluated what do you think it needs for repair to run again? Write it down. Establish a timing procedure yes on a dirty watch if it runs time it. Write that down. Then proceed with the repair afterwords they can go on the timing machine again and because you are so eager initially I would've recommended timing again in 24 hours later but you probably didn't but now is the time to do. In other words timing when it's wound up and 25 hours later like the watch companies do.

When you're all through it to repairs look at to do repairs agree with the original valuation? Now you're all done hopefully the watches running keeps time except you're not done?As your collection is probably relatively small this will be easy to do wait a month put it back on the timing machine see how it's doing as you get more and more watches you probably won't do it on a monthly basis but from time to time check your watches how are they doing are they still performing the way they were when you finish to repair. This is where you'll find out if that cheap bottle of used oil on eBay was a worthy investment or not. Or whatever else you're doing good or bad. Always interesting to look at your watch five years from now see if you can even see the lubrication on the watch for instance? Is it even running five years from now.

Then I thought I'd throw in a few more pictures balance out all the text. Yes Omega really does have timing specifications of almost every single watch they ever made in modern history. But what if it's beyond that that's covered also so I snipped  that out.

If you looked at the manufacturing information above or even the timing image you can tell these watches are being timed and three positions so they are a better grade of watch. Then as a reminder there only evaluating for timing purposes which is why they only have one dial position. The watchmaker would've already checked the watch in both dial positions in very likely all the positions.

 

 

[watchweasol]

Hi  There is some sound reasoning and sense In what JohnR725 has said and would be well advised to read and remember, I have been doing clocks/watches for 50+ years and don't know half of it and probably never will . Learning is an ongoing process through life. Just have confidence in your own ability. If you need help ask. no one knows it all' but forums like this cover most bases there is always some one who will have an answer, Shared knowledge, professionals Amateurs, beginners, we all benifit Long may it continue.

[VWatchie]

1 hour ago, JohnR725 said:

Yes I'm in agreement unless you've done something to the balance wheel leave it alone. It would only be if for some reason you wanted to make it better than the factory made it.

And the reason could be as simple as that, wanting to make it better, to improve the precision, motivated by pure interest.

1 hour ago, JohnR725 said:

trying to convert a lower grade watch to a chronometer grade watch isn't going to work because there's other things beyond the balance wheel that makes it a chronometer grade.

That is for sure, but trying to achieve some improvement, rather than aiming for chronometer grade, I personally find interesting enough. I found the below section of Mark's Vostok service video quite inspiring.

 

[spectre6000]

@JohnR725 I hope my response wasn't too disappointing. Like you said, it takes a long time to learn watch repair, and I'm at the beginning of that journey. I recently realized that part of the beginning of the road is learning how to tell if I have succeeded or failed. A few days ago, I serviced a low grade pin lever movement. I thought I botched it, but it turns out I actually did a great job. My expectations just weren't aligned with the reality of cheap movements like that. My intent was to figure out how to tell if I've succeeded or failed, where previously even a fantastic success would have been perceived as a failure.

This evening, I serviced a movement for the first time with a result that I can call an unqualified success! It's well within the range one would expect of an average movement, and by the time it's run in and regulated, may even fall into the range of what the chart up there is calling a precision movement! This is a first for me. Like you said, it's going to be a long journey.

Once I get a few of these under my belt, I'm going to have to come back through this stuff when I can make more sense of it. I'm nowhere near poising balances or adjusting pallet stones or anything. Right now, I just need to figure out how to quit borking up hairsprings consistently.

[watchweasol]

Spectre6000,      You are doing very well, and I an sure the more you do the better it will get. We all have dog days when everything you do turns to s***, but its these days we learn from  again well done  

[VWatchie]

15 hours ago, spectre6000 said:

I'm nowhere near poising balances or adjusting pallet stones

I dreaded adjusting and re-fitting pallet stones for a long time, but once I got around to it, it wasn't as hairy as I expected it to be. I've summed up my experiences in this thread and I thought you might find it useful.

 

Edited December 21, 2020 by VWatchie
misspelled and missing words

[spectre6000]

1 hour ago, watchweasol said:

Spectre6000,      You are doing very well, and I an sure the more you do the better it will get. We all have dog days when everything you do turns to s***, but its these days we learn from  again well done  

Thanks. I'm more than a little stoked. After the however manyth wrecked movement, I was starting to wonder if I hadn't finally discovered the depth of my mechanical aptitudes. This might be the most difficult learning curve I think I've ever attempted. The "mechanical puzzle" part is a cinch, but the slow exactitude of it is a real challenge. Big cast iron cylinder heads don't mind if you bang them around a little, but hairsprings phone it in with as little as a sideward glance.

[spectre6000]

1 hour ago, VWatchie said:

I dreaded adjusting and re-fitting pallet stones for a long time, but once I got around to it, it wasn't has hairy as I expected it to be. I've summed my experiences in this thread and I thought you might find it useful.

 

I've got a Grana that I believe is in a pallet stone adjusting way. I tagged and bagged it for later when I'm feeling confident enough to tackle this.

[JohnR725]

13 hours ago, spectre6000 said:

I hope my response wasn't too disappointing. Like you said, it takes a long time to learn watch repair, and I'm at the beginning of that journey. I recently realized that part of the beginning of the road is learning how to tell if I have succeeded or failed. A few days ago, I serviced a low grade pin lever movement. I thought I botched it, but it turns out I actually did a great job. My expectations just weren't aligned with the reality of cheap movements like that. My intent was to figure out how to tell if I've succeeded or failed, where previously even a fantastic success would have been perceived as a failure.

This discussion and you in this discussion is not disappointing. Disappointing is sometimes posting a response and never seeing it if it was helpful at all.

I was thinking of something one of my instructor said? "Do not please the timing machine" Maybe that might work here. In other words all you want to do is get the watch to run preferably better than it was before. But don't get obsessed with it has to look perfect on the timing machine or it has to be perfect don't worry about that. Worry about's okay don't worry at all try to have fun and see if You can get the watches to run.

A unfortunate problem especially for new people is the assumption that every single watch you work on is going to run perfect. Or for that matter even a run at all.