Regulating accuracy

[bp1000]

Is it the same advice for all mechanical watches?

My new steinhart ocean green is gaining 4 seconds a day it seems. Possibly touching 5. But I have kept it horizontal face up.

I'm trying the vertical crown up to see if it looses a bit to try and regulate it.

Does the horizontal face up to gain and vertical crown up to loose apply?

Thanks

[shortar53]

+4 secs a day is really good for a sub £1000 mechanical watch. Its within the COSC specification for timekeeping that Rolex etc are certified by.

I'd be pretty happy with that.

[R6VED]

shortar53 said:

+4 secs a day is really good for a sub £1000 mechanical watch. Its within the COSC specification for timekeeping that Rolex etc are certified by.

I'd be pretty happy with that.

This.

My Rolex Explorer 1 (214270) runs at about 3.5 secs fast per day. When you consider there are 86,400 seconds in a day that isn't bad :-)

99.99595% accurate :-)

[glazbagun]

In practical terms there's no way to predict positional errors, so what you're doing is the best way to work it out. Take a note of how many hours you leave it overnight in one position, then check the time. Next night try it crown down, etc.

Also, especially with a manual, the watch will have lower amplitude by the first thing in the morning, which has an effect too.

[Riff Raff]

glazbagun said:

In practical terms there's no way to predict positional errors, so what you're doing is the best way to work it out. Take a note of how many hours you leave it overnight in one position, then check the time. Next night try it crown down, etc.

Also, especially with a manual, the watch will have lower amplitude by the first thing in the morning, which has an effect too.

Unless you have access to a timing machine..........

I measure most of my watches - just out of interest really. For most of them a gain is normal when measured face up, and there is normally less of a gain (or even a loss) when measured crown down. But I have a few where that doesn't hold true.

I bet the OP's watch runs a lot more accurately than +4 worn on the wrist. Which is the only measurement that matters, if you think about it.

[bp1000]

I reset it this morning measured against a time.is and so far on the wrist it's lost 0.5 seconds

I was just curious if the method of placement overnight was universal. Ie crown up crown down or horizontal.

Once I figure out the running accuracy I know I could manage it with placement overnight

[ExplorerII]

Horizontal crown down should be best position for your watch.

[Ray Luxury-Yacht]

Ooh, this is interesting, I had no idea that how you place a watch and leave it on a surface makes a difference to the time it keeps! Why dat den?


And also, please explain what this means:

glazbagun said:

Also, especially with a manual, the watch will have lower amplitude by the first thing in the morning, which has an effect too.

I've an automatic Seamaster, which is 15 years old now. I still can't believe how incredibly well it keeps time. I leave it on the bedside table overnight face-up, and wear it all day. It basically never goes 'off' at all. The only time I have to intervene is if I leave it off my wrist for a couple of days and it stops. But I've managed to go months before without it stopping, and I never have to adjust it. I is well impressed with it

[glazbagun]

Basically, (all other things being perfect, which they never are) if the centre of gravity of the balance (the bit that beats really fast when you look at it) is not dead in the centre of it's axis, then the heavy spot will have an effect on the timekeeping (whether the heavy spot is being kicked up or down). This effect changes according to the way the watch is hanging and how far round the balance is being "kicked" with each beat (which is what is meant by "amplitude"- how many degrees of motion the balance swings though).

What I've described would be called poise error, but other things can also affect the timekeeping on the sides, so a broader term would be positional error. Most mechanical watches, almost all in fact, will show positional errors even if everything is perfect, so these can be used as a form of keeping your watch closer to zero average deviation over a day, if you're an OCD nightmare.

This is why old ships Chronometers were sat in gimbals so they'd always stay in one position. The level of wind also has an affect, which is why Chronometers were wound at the same time every day, to keep the amplitude of the watch within limits. Because automatics are winding themselves all day, they only really wind down when they're off the wrist, so the amplitude stays high for much longer.

Edited by glazbagun on Friday 20th June 20:02

[Ray Luxury-Yacht]

Wow, interesting stuff, thanks for explaining that matey. Had no idea about any of that - but it makes sense now you've explained it.

One more question then - I guess the thing you described as going 'up and down' would be going 'side to side' in the watch is laying flat, would it not? So how come it changes if the watch is bezel up or down? I can't see how that would make a difference?

[bp1000]

Very pleased with the accuracy since I reset it and did 20 winds yesterday

Quite surprised steinhart have made a watch relatively cheap and accurate.

It lost 1 second yesterday and placing it crown down last night it gained 3 seconds so is now 2'seconds fast.

Incredible and thanks for the explanation.

Job now is to see how consistent this will be