Do watches gain and lose

[TJGladeRaider]

I have been wearing a vintage Rollie DRSD since I got it back with Ziggy Lume. Since I have been wearing it, I have not had to set it since the date flipped to March 1 - three weeks ago.

I just went to the US Official Time site to set the watch, and I was amazed to find it dead on - I mean to the exact second.

Now I know that nobody's mechanical watch keeps perfect time - including this one as I check them all when I first get them and I have never had one be spot on after a couple of days. My MBWs all seem to keep time within +/- 5 secs per day, so dead on after three weeks isn't plausible.

So my question is, how is this possible?

One possibility that occurs to me is that I could have set it to the wrong time to begin with - say 2:13 when it was actially 2:14. That would account for a steady loss of about 2.5 secs per day which is believable, if it just happened that the second hand was spot on when rechecked as a coincidence.

Another possibility is, I may have reset the watch since I got it back from Ziggy - I suspect that most of us reset our watches as a matter of habit, so the fact that I don't remember doing so doesn't really mean much. If I reset it a week and a half ago but set it to the wrong minute, that would account for a steadly loss of 5 secs per day which is very normal for these things. Again, allowing for the seconds hand being did on as mere coincidence.

Otherwise, I see only two possibilities:

1. The watch gains time sometimes and and loses time other times, in this case having preceisely balanced out.

2. The watch neither gains nor loses at all - and nobody needs to tell me how unlikely that is.

Any other possibilities?

Bill

[longshot]

if you take it off at night there is a fairly common answer

mechanical watches are positionally sensitive. i have played around with this some time ago with my seikos and found that some gain time crown up at night, and lose crown down. some seem to be the other way. there are a couple scf forum guys who have set up spreadsheets and calculated gain and loss in the 6 positions and tracked the results over time. if your watch is gaining 5/6 seconds a day and you set it on the nightstand, say crown up, and it loses 5/6 seconds overnight it is possible for it to be dead on after a few weeks. or really close anyway!

that's why you see that engraving on really fine movements about "adjusted to X positions and temperature"

[TJGladeRaider]

if you take it off at night there is a fairly common answer

mechanical watches are positionally sensitive. i have played around with this some time ago with my seikos and found that some gain time crown up at night, and lose crown down. some seem to be the other way. there are a couple scf forum guys who have set up spreadsheets and calculated gain and loss in the 6 positions and tracked the results over time. if your watch is gaining 5/6 seconds a day and you set it on the nightstand, say crown up, and it loses 5/6 seconds overnight it is possible for it to be dead on after a few weeks. or really close anyway!

that's why you see that engraving on really fine movements about "adjusted to X positions and temperature"

That could be it. I do take the watch off when I go to bed.

I'm going to pay close attention and check it again after a few days. It sure was a shock to go to set it and find it was dead on.

Bill

[david-aussie]

u can actually record the position of the watch with crwon up, crown down , face up etc and

see whether it gain or loose.

that way, u can use the postion if rest to compensate the wearing of the watch during the day.

[spa]

My latest rep runs -1sec after 55 days on the wrist. (checked with reliable source).

I didn't touch the inside up to now, (and i am not going to as long as it keeps like that).

No other rep of mine did so well so far...they all run from +7 to +15 secs per month...

[crystalcranium]

Adjusted to 5 positions. That means the watch is adjusted and is evaluated in its' accuracy in 5 positions. This evaluation best represents the position of the watch in daily use. Yes, watches adjusted to near perfect isochronism will fall on either side of the knife's edge of perfection due to influences other than balance spring length, most often influences such as temperature and friction in the balance wheel pivots due to position. Watches tend to gain dial up, the position they are in on your nite stand 8 hours a day and loose crown down, the postition they are in most of the day on your wrist. If your watch is a fine machanical marvel, these variations will tend to even out over a period of time. If it is a very fine timepiece, the positional variations will be small and even. A Rolex 3135 is a beautifully balanced movement that is usually +- a second, perfectly compensating, in a variety of positions. I don't know of any movements that swing broadly and compensate. Usually 3-5 seconds difference in positions plus and minus is about the maximum for a fully compensating positional movement.

This is the rub that some mechanical watch enthusiasts have against quartz watches. An accurately adjusted quartz watch is a marvel of accurate time keeping but a slightly out of adjustment quartz movement will not positionally compensate. In other words, a quartz watch that is half a second fast out of adjustment will gain a half a second every day additively where a mechanical watch that varies a half a second a day will tend to even out the variations in time over time.

I have a beautifully adjusted Omega SMP Chrono that may be off by two or three seconds on any given day but usually at week's end, it's spot on. This was the kind of accuracy initally insisted upon by the railroad standards over 100 years ago. They weren't particularily interested if your pocket watch was 10 seconds off in the middle of the week as long as the 7 day variation was not outside of 30 seconds.