How do you determine accuracy?

[camhabib]

I was reading a recent post of a vintage Rolex manual that actually gave instructions on how to store the watch in order to gain or lose seconds every day. Granted I am not sure how relevant this is today with the technological advancements in todays movements, but I'm sure it still has some bearing, as do several other factors I'd imagine.

When you report that you're watch is "+2 seconds / day" or so on, what exactly are you doing to come up with that number? For example, after 48 hours, wearing the watch for about 16 hours, and storing dial up for the remainder of the time, my watch has gained around 4-5 seconds (or 2-2.5 seconds per day if you prefer). Does anyone really take into account things such as wrist time, storage position, etc?

[Tools]

All of those things you mention will affect time keeping.. Yours sounds very accurate, although you should take the same reading with the same source standard for about a week and then divide by the number of days.

Active wear will shorten the amplitude, so an active watch will be faster than an idle watch... temperature affects the movement to a large degree, storage position to a degree, although as you point out, the faster beat movements do not exhibit as much change... and on and on..


No, the numbers mentioned here are mostly a quick look for bragging rights..... I'll bet that even some fudging goes on...you know...just because...

......................

[Watchdog]

To determine the accuracy for my GMT IIC (my daily wearer), I set the watch to the current US time and then checked it every day or so. By the way--without fudging or bragging--my V serial that I wear every minute of the day has been DEAD ON. When I say DEAD ON, I mean DEAD ON. I have never seen anything like it in a mechanical watch. In the last month, it has not lost or gained even half a second.

I live in SoCal and the temperature is obviously relatively constant but this movement is extraordinary.

[camhabib]

Larry, thanks a lot for the reply. I figured most of the numbers thrown around are more just general estimates then exact figures.

How does a watch maker go about adjusting a watch if so many factors go into play? Is there a temperature or position they equilibrate to (or for that matter, should) in order to obtain an optimal timing? What position should I store the watch in to try and make it run a bit slower so it doesn't gain the few seconds a day?

[Tools]

Originally Posted by camhabib

Larry, thanks a lot for the reply. I figured most of the numbers thrown around are more just general estimates then exact figures.

How does a watch maker go about adjusting a watch if so many factors go into play? Is there a temperature or position they equilibrate to (or for that matter, should) in order to obtain an optimal timing? What position should I store the watch in to try and make it run a bit slower so it doesn't gain the few seconds a day?

Now you are talking a different animal.......... Timing a watch properly takes a couple of days and the watch is attached to a vibrometer that actually listens to the escapement lock and unlock giving a calculated daily variance read-out. The watch is then rotated to the various positions and the watchmaker times it to the best overall average performance.

The laws of physics always affect a mechanical watch to some degree, so let's just say that laying flat on it's back, it has minimal friction on the pivots, and minimal drag caused by gravity, so that will allow for the "fastest" performance.

If you want "slower" performance, then lay it on it's side (winding crown pointing up or down). This sort of hangs the balance on it's side for maximum friction on the sides of the pivots, and gravity needs to be overcome to swing the hairspring up on each rotation.........

So, with a modern fast-beat movement, you might not notice a change overnight, but it will most certainly be affecting it's function by some small percentage, and will help maintain accuracy over time........

[flyback]

Originally Posted by Tools

Now you are talking a different animal.......... Timing a watch properly takes a couple of days and the watch is attached to a vibrometer that actually listens to the escapement lock and unlock giving a calculated daily variance read-out. The watch is then rotated to the various positions and the watchmaker times it to the best overall average performance.

This is how watchmakers regulate watches, but I disagree that it is the proper method.

If the owner wears his or her watch daily, the systematic error is usually extremely consistent. The proper method is for the watchmaker to adjust to that. This requires that the owner take time measurements and may require several adjustments.

Some people claim to ask that their watches be regulated to run fast, which is bizarre.