Changes in timekeeping accuracy

[Junior]

Four weeks ago I took delivery of a new 500m water resistant watch with an ETA 2824-2 movement, adjusted in five positions (it says here!) and apart from immediately changing the strap, wore it straight from the box. The accuracy was very impressive from the word go, with minimal daily gain, and a total gain of 5 seconds in 10 days; I was very pleased, to say the least, and wore the watch 24 hours a day, checking the accuracy each night when I'd finished work (as it was a new watch). Everything was going well until five days ago when I checked it and ...... it had gained 12 seconds in 16 hours. I reset it and it did more or less the same the next day - this time gaining 20 seconds in 24 hours.

I then checked it, unworn, in various positions and over a three day period it performed as follows:

Day 1: Dial up: +6 seconds in 8 hours

Day 1: Dial down: +8 seconds in 11 hours

Day 2: Crown down: +6 seconds in 7 hours

Day 3: Crown up: +11 seconds in 19 hours

The gain seems to be between 14 seconds and 20 seconds a day whether worn or at rest in any of the 4 more usual 'testing' positions, and as I have never experienced anything like this before with a new watch, I wonder if anyone has any idea why this might be - particularly bearing in mind the initial extremely good results? The watch has not been dropped, knocked, subjected to any sudden changes in temperature, nor has it been shouted at or verbally abused, so I don't think I have hurt its' feelings! I'd be interested to hear views on this (to me, at least) odd phenomenon.

[spluurfg]

Well, gaining 14-20 means that it's still got very good positional variance. I.e., it's average daily rate is about 17 seconds, and positional variance is +3/-3. This shows that it is still a well adjusted watch, but the overall rate (regulation) needs improvement. Positional variance (adjustment) is definitely the harder thing to fix versus a simple average rate (regulation)

You could probably fix it yourself easily using the rate adjustment, or send it to a repairer you trust, or if it's still under warranty just send it in. But basically the (solid) positional variance shows that it probably doesn't need adjustment to the balance wheel or any disassembly or re-lubrication etc.

[seffrican]

As spluurfg says, the positional behaviour is good, so all it needs is a gentle tweak of the regulator screw in the (-) direction.

If you have a decent case opener and the right size screwdriver, you're ready to go

[Junior]

Thanks folks. Have you any idea why everything was prefect for three weeks and then the wheels fell off? Nothing about the treatment of the watch changed in any way whatsoever.

[seffrican]

Magnetic fields. Cosmic rays. Hair loss. Earth tremor. Weight gain. Moon phase. Global warming. Hairspring grit. Oil overdose.

Could be anything.

[Junior]

You're right - it was probably the Boks almost losing to Ireland at the weekend!