How accurate is your COSC watch really

[pistonheadforum]

Genuine question and interested in the spread of numbers for folk who wear a COSC watch on a daily basis.

If you can set it for a single day and find out how fast/slow it's running it would help as trying to build a database of manufacturers/models with real world performance.

Thanks in advance.

[julian64]

Not much help but I sent a Breitling off to Switzerland a little while ago. They suggest they will guarantee 30 seconds a day after their overhaul/service. I thought that was setting a reasonable bar, but maybe not?

[hilly10]

I can honestly say in the real world my first generation Planet Ocean is within 2 secs a day.

[Tom1312]

Tudor Pelagos routinely within 2 seconds a day.

[Barchettaman]

pistonheadforum- said:

If you can set it for a single day and find out how fast/slow it's running it would help as trying to build a database of manufacturers/models with real world performance.

Thanks in advance.

Best of luck. What an utterly pointless exercise, however.

[pistonheadforum]

Barchettaman said:

Best of luck. What an utterly pointless exercise, however.

Yes it is rather, but it will help me settle a bet that COSC is a good standard and worth getting. I just need a fair few to prove a point.

There are beers involved in the bet so it's important!

[Nigel_O]

Submariner - 3 seconds slow per day since new eight years ago

[Macneil]

In my experience of ownership, Tudors are very very accurate, Omega not so much, disappointing in fact, still within COSC spec though.

Currently have a Sinn U50 and a Seiko Willard that are non COSC and more accurate than the Omega.

[CardShark]

My Tudor Ranger is -1s per 24h worn on the wrist, my Pelagos is, IIRC, 2s but can't remember if that's + or -.

They're both COSC however 2 of my (non COSC) Sinns run to around a couple of seconds per day, as does my Black Bay 36. I've others that aren't too far from that as well.

My most accurate watches are my non COSC Grand Seikos, one with a (quartz regulated) Spring Drive and the other a 9F quartz. Both are, from memory, around 1.5s out over a 6 month stretch!

Edited by CardShark on Wednesday 29th March 20:54

[Somebody]

My real world experience: the in-house modern Tudor calibres (BB58 & Heritage BB) are the most accurate (0s per day).

Saying that my father handed me a 1997 Rolex 16233 that has never been serviced. It gained 4 seconds over 6 days when I last wore it, which I think is stupendous.

[Barchettaman]

pistonheadforum said:

Yes it is rather, but it will help me settle a bet that COSC is a good standard and worth getting. I just need a fair few to prove a point.

There are beers involved in the bet so it's important!

Aha. If beers are involved then I thoroughly approve

[HTP99]

Put my Christopher Ward COSC forward an hour on Sunday when the clocks changed, just checked and it is 10 seconds fast compared to my phone which has an analogue clock on the display which shows seconds, meaning 2 seconds fast per day.

[Fiedka]

Seamaster 300 heritage with 8912 movement.
Gains 1.0 seconds if left dial up overnight.
Crown down and it then averages 0 seconds per day.
Bought early December and have not reset the thing yet (jumping hour hand took care of daylight saving time change).
Quite remarkable really.

[gregs656]

I think I was 9 days into March before I noticed I hadn't advanced the date on my PO

[Tango13]

My GMT2 is losing about 2~2.5 seconds a day, the only AD in the county claims they sent it back to be re-calibrated after the last service but I have my doubts, I suspect it just sat in their safe for a few weeks.

Past experience suggests it should be good to a couple of seconds a month if carefully calibrated.

[blingybongy]

Never checked.

[ianrb]

Tudor Ranger runs about 2 sec. per day slow when worn.

I also have a Citizen 'Radio Controlled' watch which has nether lost nor gained so much as a second over 15 years.

[julian64]

ianrb said:

Tudor Ranger runs about 2 sec. per day slow when worn.

I also have a Citizen 'Radio Controlled' watch which has nether lost nor gained so much as a second over 15 years.

it has, the radio has just controlled it

[Tango13]

julian64 said:

ianrb said:

Tudor Ranger runs about 2 sec. per day slow when worn.

I also have a Citizen 'Radio Controlled' watch which has nether lost nor gained so much as a second over 15 years.

it has, the radio has just controlled it

My brother has 3 Citizen sliding head lathes, he was told by one of the sales reps about the time they all went to visit the factory in Japan and they were all staring at their radio controlled watches as they got closer to date lines just to watch them update

[gt40steve]

julian64 said:

Not much help but I sent a Breitling off to Switzerland a little while ago. They suggest they will guarantee 30 seconds a day after their overhaul/service. I thought that was setting a reasonable bar, but maybe not?

That sounds a mistake ?

ISO / COSC chronometer accuracy must be within +6 to -4 seconds a day as I understand it.

Several manufacturers are approved to test & certify in house, so no, not every watch goes to Switzerland for calibration.

My Bremont was well within this specification for several years. It eventually started gaining, mainly due to magnetism (it's not one of the Faraday caged movements).

Since I returned it to the factory for a service, it is back to 'as new' accuracy. Stripped, cleaned, demagnetised, reassembled, tested, regulated and certified again with a two year guarantee. (£375 FYI)

I've recently had a visit to the factory & saw the service department, new builds, and testing process first hand. Interesting, fiddly stuff !