Accuracy

[volvos60s60]

I recently bought an Omega 2254.50 here on SC - my first quality watch. It was around 10 seconds a day out so, as it was 5 years old & never been serviced, I thought it was time it was.

Off it went to Steve at Ryte Time & I got it back last week.

I thought I do an accuracy check from the off & it's been on my wrist for 7 days now. Over that entire week it has been +1 second!!!! I'm amazed that this little machine might only need resetting, say, once or twice a year.

So chuffed, just thought I'd tell you & also congratulate Ryte Time on their skills.

[marcus fenix]

Nice. Although I think you'll find there has been an element of luck there, and in the longer term it may not perform quite so well. Most accurate I ever owned was this one:

viewtopic.php?f=1&t=132806&hilit=+later

But it didn't always perform like that. Within a second a day though :) All my current mechanicals are around +3-4 seconds a day.

[Daddelvirks]

Don't worry, it will keep on running fine, it's very possible with that movement.
My Ebel Chrono is now back from service a couple of weeks, and running at less than +1 sec/ a day!

Daddel.

[seikomatic]

I let go the issue once I have more than 3

[Omegamanic]

Nice. Although I think you'll find there has been an element of luck there, and in the longer term it may not perform quite so well. Most accurate I ever owned was this one:

viewtopic.php?f=1&t=132806&hilit=+later

But it didn't always perform like that. Within a second a day though :) All my current mechanicals are around +3-4 seconds a day.

3-4 is very lucky.

1 second per week on the 2254 will not be a consistent thing imo, you are likely looking at a number of variables on gaining time one day and losing time the next.

In order to judge it properly maybe keep a graph and jot down the time every 4hours and see if it gives an indication of timekeeping in use or rest.

[volvos60s60]

Nice. Although I think you'll find there has been an element of luck there, and in the longer term it may not perform quite so well. Most accurate I ever owned was this one:

viewtopic.php?f=1&t=132806&hilit=+later

But it didn't always perform like that. Within a second a day though :) All my current mechanicals are around +3-4 seconds a day.

3-4 is very lucky.

1 second per week on the 2254 will not be a consistent thing imo, you are likely looking at a number of variables on gaining time one day and losing time the next.

In order to judge it properly maybe keep a graph and jot down the time every 4hours and see if it gives an indication of timekeeping in use or rest.

I understand that, & a good point. However, given the swings & roundabouts effects over a 7 day preriod, it's more than I could have hoped for. If's it 3 seconds fast over the daytime, & 2.9 seconds slow at night, I can happily accept that. After all, it's not going to make me late for an appointment. I had been expecting to have to reset set it every fortnight

[kraftar]

OK, a stupid question... how do you guys measure the accuracy of your watches?

[gray]

OK, a stupid question... how do you guys measure the accuracy of your watches?

Not a silly question at all

Personally, I check the time on my watch and see if it's right or not :wink:

Usually against another time source I think might be right.

[volvos60s60]

I use my Seiko quartz which is set against the talking clock. I verify my auto aginst both time sources as any gains losses should be the same against my quartz & the talking clock. I reckon the whole thing is accurage to 1 second

[kraftar]

I use my Seiko quartz which is set against the talking clock. I verify my auto aginst both time sources as any gains losses should be the same against my quartz & the talking clock. I reckon the whole thing is accurage to 1 second

ok, thanks... but that leads to my next stupid question: what is the 'talking clock'?

[marcus fenix]

I use my Seiko quartz which is set against the talking clock. I verify my auto aginst both time sources as any gains losses should be the same against my quartz & the talking clock. I reckon the whole thing is accurage to 1 second

ok, thanks... but that leads to my next stupid question: what is the 'talking clock'?

Dial 123 in the UK. I just look at either a Junghans mantle clock or a Casio 5600 atomic.

[horrovac]

I thought I do an accuracy check from the off & it's been on my wrist for 7 days now. Over that entire week it has been +1 second!!!! I'm amazed that this little machine might only need resetting, say, once or twice a year.

You are being a bit optimistic there, it's not going to be that accurate all the time. However, it is amazing, isn't it? A bit of oscillating metal, being shaken this or that way, being smacked against doorframes and whatnot, and accurate to within a couple of seconds... It is nothing new to me, but it never ceases to amaze me.

[kraftar]

I use my Seiko quartz which is set against the talking clock. I verify my auto aginst both time sources as any gains losses should be the same against my quartz & the talking clock. I reckon the whole thing is accurage to 1 second

ok, thanks... but that leads to my next stupid question: what is the 'talking clock'?

Dial 123 in the UK. I just look at either a Junghans mantle clock or a Casio 5600 atomic.

great, thanks! I learn something new here every day.

[horrovac]

I use my Seiko quartz which is set against the talking clock. I verify my auto aginst both time sources as any gains losses should be the same against my quartz & the talking clock. I reckon the whole thing is accurage to 1 second

ok, thanks... but that leads to my next stupid question: what is the 'talking clock'?

Dial 123 in the UK. I just look at either a Junghans mantle clock or a Casio 5600 atomic.

Since you are probably typing this using a computer, you don't need even that. Just have your computer network-synchronised using NTP (Network time protocol). On Linux, BSD and other unices you will need to configure and run the ntpd daemon, it will then not only synchronise constantly with time servers, but also adjust the operating system's clock skew to make it more accurate even between the updates. Your computer clock will be accurate to about 0.01 second at the very least (and probably irrelevantly far more accurate than radio synchronisation), then you can use the Dreadnought at the top of the forum to set the watch or check against it. Or do as I did and write a simple program to beep at second changes, it is easier to check your watch with this audio feedback.

Android smartphones have ntp support, you just need to activate it (and then get the timefactors android app to set your watch in style :).

Other operating systems also have a level of time synchronisation available.

[EdRonax]

Thats very good indeed, to be honest I'm not really sure how accurate mine are as I usually change every day or so.

[horrovac]

<blah> Or do as I did and write a simple program to beep at second changes, it is easier to check your watch with this audio feedback.

I just thought I might have as well posted the code, in case people can't be arsed typing it themselves:

#include <stdio.h>
#include <sys/time.h>

int main (void)
{
	struct timeval tv;
	time_t second=0;

	gettimeofday ( &tv, NULL );
	second = tv.tv_sec;

	while ( 1 )
	{
		gettimeofday ( &tv, NULL );
		if ( second != tv.tv_sec )
		{
			fprintf ( stderr, "\a" );
			second = tv.tv_sec;
		}
		usleep ( 100 );
	}
return 0;
}

...put it in, say, secbeep.c and compile with gcc secbeep.c -o secbeep.

[Umbongo]

I thought I do an accuracy check from the off & it's been on my wrist for 7 days now. Over that entire week it has been +1 second!!!! I'm amazed that this little machine might only need resetting, say, once or twice a year.

You are being a bit optimistic there, it's not going to be that accurate all the time. However, it is amazing, isn't it? A bit of oscillating metal, being shaken this or that way, being smacked against doorframes and whatnot, and accurate to within a couple of seconds... It is nothing new to me, but it never ceases to amaze me.

The thing that amazes me the most is my humble Planet Ocean. Scuffed, dropped (a few times!), worn at the gym one minute and then just left sitting around for a couple of days at a time when I get a vintage mood on me or, recently, get re-obsessed with my zenith. And it struggles to gain or lose a second A FORTNIGHT! Im dreading the day its up for service - something tells me its going to come back worse!!!

[monogroover]

Android smartphones have ntp support, you just need to activate it (and then get the timefactors android app to set your watch in style :)

There's a timefactors Android app? Thanks!

Hmm, I like it although nicer if it was a widget.

[kraftar]

thanks for the assistance and the program!

[Harrison]

I thought I do an accuracy check from the off & it's been on my wrist for 7 days now. Over that entire week it has been +1 second!!!! I'm amazed that this little machine might only need resetting, say, once or twice a year.

You are being a bit optimistic there, it's not going to be that accurate all the time. However, it is amazing, isn't it? A bit of oscillating metal, being shaken this or that way, being smacked against doorframes and whatnot, and accurate to within a couple of seconds... It is nothing new to me, but it never ceases to amaze me.

The thing that amazes me the most is my humble Planet Ocean. Scuffed, dropped (a few times!), worn at the gym one minute and then just left sitting around for a couple of days at a time when I get a vintage mood on me or, recently, get re-obsessed with my zenith. And it struggles to gain or lose a second A FORTNIGHT! Im dreading the day its up for service - something tells me its going to come back worse!!!

Ok we've got a second a fortnight. Sounds like a new record.

A second a fortnight

Going once....twice...

Can Anyone better that?