running rate cal.1570

[Larryscott]

I have an early 60s Oyster Perpetual Date. cal. 1570.

I recently sent it to NY service center. When I got it back, I noticed that the daily running rate was 6-8 sec/day fast, and inconsistent. However, now after more than a month, the rate has increased to an uncomfortable 12.6 sec/day fast. On the upside, however, that rate holds true for +/- 1.2 seconds RMS. So, I am confident that having it regulated, it can have a running rate of 1-2 seconds per day.

Question: over a 2-8 years period, does an older 1570 speed up? or slow down between major servicing?

I want my watch guy to trim the rate by 12 sec/day. But, if the tendency is for the rate to increase with age, do I have it adjusted slightly slow? Or, if they slow down, I should have it set to run slightly fast?

In the past, this watch has run spot on for years at time, but I wasn't tracking it very close. Since it's just been serviced, I'm trying to get a really low running rate with hope for a stable rate for several years.

I would be real happy if I only had to set it monthly just to correct the calendar.

[swish77]

A well-regulated movement shouldn't be doing this, and certainly shouldn't be speeding up and/or slowing down over time. Did you have it serviced by the NYC RSC or an independent? In either case, I'd take it back and have them regulate it properly. Good luck.

[Larryscott]

I sent it to NY. And, it was not a good experience. I had expected the settling down after service, but settling to 12.6"/day is a huge disappointment. I've been tracking it's rate by observation more than once a day and it's 12.6" is rock solid.

I don't know what Rolex does, or how long they let it run before final regulation. However, when it was first delivered it must have been set to the wrong frequency, I guess. It was 15"/minute fast, not day, not hour, but in fact per minute - 6 hours per day fast. They of course addressed that. And, in the past, I had a guy that would trim the rate by what I observed, not by what his machine indicated. Trimming a given number of seconds per day isn't over the top. But getting NY to do it would take weeks, which isn't really an option.

I was looking around on TRF and there are a lot of people reporting greater than 6"/day. The reliability of the 12.6"/d is +/- only a second, so the watch is running well, so to speak.

I was curious what the multi-year trend is at large; do they get faster? or slower with age between service?

[R.W.T.]

RSC seems to set them fast as a rule. I've never gotten a watch back from them that didn't run 6-15 seconds a day fast right off the bat.

Now...you've had this for a while.

I wouldn't expect a settle in of that much.

The main question is...and as always they send a little card with the watches...

You HAVE to wind the watch manually FIRST. It's not a seiko. :-)

It has a stem and crown. The automatic mechanism on a 1570 is very efficient and eventually if you wore the watch 10 hours a day it probably would come in to full wind BUT...it won't start out that way. The origins of the automatic mechanism for WIlsdorf were not really for our convenience but for accuracy by

keeping the mainspring tension constant or more constant than a manual wind. So...fully wound the autowind is to keep it topped up, not to wind it from scratch. :-)

They use formulas to calibrate the watches in quick time. Full wind 3/4 wind 1/2 wind etc.

They figure the "best" overall regulation using that data. To me they ALWAYS run fast.

But to get anything consistent you have to wind it fully first and then wear it consistently.

You have to put it ON the same time everyday and take it OFF the same time every day and leave it resting in the same position for the same basic amount of TIME every day. Now...a 1570 is 19800 bph and the positions are generally very good...so a bit of this is not so critical as say with a 1030 or earlier..

I wouldn't be pleased with 12 seconds on that watch at all. It would drive me crazy.

But fully wind it and wear it exactly the same for a while if you haven't and see what is really up.

It's not going to keep changing or it shouldn't.

I might hit it with a demag first before I did anything else. A slight bit of magnetism could cause a 12 second increase. Maybe not but it's a simple act to try.

[Larryscott]

well, that's why there is TRF.

Winding the watch? When it has run down I just give it a shake, set it, and forget it. No winding necessary, and it runs like the chronometer that it is.

Taking it off at specific times? The watch doesn't dictate what or when I do anything, I just take it off, and put it on as life allows.

And, with that, it runs +/- a second a day. Just like it says: adjusted multiple positions and temperatures.

Now, it's recently back from the RSC, so it's fast. Too fast. But, only varies 1-2 seconds from that rate. so, it's behavior is "certified chronometer" worthy. And that's the only reason that I sent to the RSC. It's over 50 years old and I was hoping to restore it to a 1-2 second per runner.

I never wind it, regardless of when it was last running. That's for cheap, marginal runners, not any Rolex. Give it a shake, put it on. It asks for so little attention. I don't make any concession to it, it works for me - not the other way around, and +/- a second variation (on top of it's very consistent and steady rate) is why Rolex is the one to have.

In the past month, it has strayed only 3 seconds outside of it's rate. Just like H3. For fun, I figure the elapsed time since setting, multiply that by the rate, and voila, the calculated time = true time for just some several seconds. Navigation grade. I will get the 12"/day reduced, and I expect that for the next several years, I will only set it to correct the calendar.

Question still stands: over a several year period, what has the experience been by the members at large: gradual increase in speed, or decrease in speed with age since major service?

[R.W.T.]

Oh...

I really want you to let your watch run all the way down...so that's it's not running. Then I want you to find a watchmaker...with a timing machine and a few minutes to spare. Then I want you to, after you show him that it is NOT running at all, I want you to "give it shake, set it, and forget it"...but first put it on the timing machine and let him show you what's happening...and what kind of time it's keeping through the positions with one shake from a dead stop.

The idea was not to dictate how you wear your watch but you simply can't just look at a 1570 randomly and know what it's doing exactly with 100% certainty.

You can wear it however you want...that's not the issue, but in order to correct the mean variation in time...you kind of need a consistent and KNOWN program to see exactly how much it loses or gains...you need to routinely go through a set of parameters for a few days to see the exact situation and check and record the time at the same time everyday. It's called MEAN variation for a reason...if you stand for 10 hours in a line with the stem down...and check your watch it may very well be...several seconds slower...than it would be if you were resting it dial up for the same time on your desk. Therefore you might check your watch throughout the day with a varied routine and it might be a little fast or a little slow...to the atomic clock. There IS position error to a degree on any watch and the 1570 might have 10 seconds in any position....it might not but it certainly could...so it may loooooooose a little here...and gaaaaaaain a little there...and overall the timing will turn out to be the difference between the two. A consistent wear and recording pattern is absolutely necessary to know what's really going on. If you don't have a full wind and a constant maintenance OF a reasonably full wind...the timing will NOT be any where NEAR AS accurate. I didn't invent physics. It's not my fault.

I was only trying to help you.

[Fredrik]

My experience is that after a service I usually need to have it regulated after a few months up to a year. This is included in the service cost, I can get it regulated free of charge within the warranty period of two years here in Sweden. All three of my 1570 watches gain about 15-30s/month. No increase/decrease in speed thereafter.

I always wind a watch fully before wearing it since they can run a bit fast if the mainspring is wound down.

[Larryscott]

RTW . . .

I have done pretty much what you said re: observation of its rate.

For some 4 weeks, I did observe its indicated time against true time, on some days a few times, on some days only twice. Then I took the hundred or so time checks into Excel. Generated graphs, calculated residual, RMS, (linear equation returned and R^2 of 0.9996). Looked at outliers, only used time.gov and feedback loop NTP services, screened out blunders. I was looking specifically for change of rate, noise and stability of the rate, and found that for the first couple of weeks it was erratic. I was also able to see in the data a gross change in the stability; the settling out period. After discounting the "settling out period", what I found was random variation on the order of 0.05"/hour from the mean of 0.5262"/hour.

With variation in wind, position, activity, when I take it off/on all being random, and short term rate also varying at random, the month long observation (post settling out) returned a "mean" of 12.629"/day over a 24 day period. The data returned a standard error RMS 1.2"/day (1 sigma), maximum residual <2.5", well below the noise level of expectation, and not indicative of any significant systematic behavior, face up, down, on it's side, and small enough deviations to be confident that variations outside of the observed are likely (standard error) to not occur. The observation routine was considered at the outset to satisfy the Nyquist sampling rate, (although not rigorously). So, to examine a finer systematic watch behavior, you are correct. But this 52 year old OP, bought as a wreck on the street in Honduras in 1979, renovated, restored, used and loved. It's not in a science experiment or competing for the Longitude prize (although it could've won it). I am impressed that it is +/- 1"/day against a steady rate. I'll get the 12"/day sorted out.

I did observe the mean, I did respect randomness of conditions. And, I found that given a shake, a set, just put it on, did not effect the rate - not significantly, at least not in a few hours (after which the reserve would be midrange) - contrasted with fully wound. If fully wound v. low reserve had much of a difference that would be grounds for servicing. And on a time graph you get an instantaneous observation, not a day long true behavior.

The excellence of a Rolex is that very little effects its rate, else it's a Chinese generic.

I just found it funny that you said "you have to fully wind it, not like a Seiko". (I guess what, you don't have to fully wind a Seiko but do have to wind a Rolex?) High quality self winding watch is self winding, and low reserve on a chronometer grade 26J watch doesn't effect its rate by more than a factor 0.03 in the short term, and similar with it topped off. Exception: poor condition, second rate service, etc.

which is why there is a forum.

So, any long term behavior pattern between major servicing? do they slow down or speed up? that is the question. Do I have trimmed 11, 12, or 13 seconds per day? Gotta plan ahead, it won't be due for service 2025.