Many moons ago I remember having discussions
about the performance of 32k768 benders. We
used them in pacemakers, mother natures 37.0
deg crystal ovens. For the watch chip designers
the temp curves were a part of their lives. The
daily usage was a part of the plan for where to set
the rock. The wrist temperature is a part of it and
the number of hours each day on the wrist. For
watches that sat unworn the times drifted a bunch.
I believe that that was the origin of the 4M194304
references which started showing up on the fancy
translate spensive, watches.
Imagine drift numbers for a watch sitting on a
dresser in an unheated cabin over winter months
and one closer to the Equator. Not an easy way
to compensate the drift without a temp tracker
and a varactor in the watch. I guess that that is
now old news.
Greg
On 4/6/2011 12:44 PM, Tom Van Baak wrote:
Although we normally deal with nanoseconds and picoseconds
here on this list, the essence of being a "time nut" and the allure
of precision timekeeping is there in the micro- and milli-second
world as well. I can personally tell you that the Allan deviation of
a pendulum clock is equal or more interesting than an H-maser.
As Bill just mentioned, for those of you interested in wristwatch
or pendulum clock timing, please see Bryan's site:
http://www.bmumford.com/microset.html
Antonio, I used one of Bryan's Microset timers to measure my
(WWVB radio controlled) wristwatch:
http://www.leapsecond.com/pages/Junghans/
There's no need to wait a month before you see a 1 second
error! -- If you can measure to 1/10 second you only have
to wait a few days, right?
If you can measure to 1/1000 second you can see the error
within an hour. If you use a Microset measure to 1/1000000
second you can time and rate errors within seconds or minutes!
But at this level you will also be able to measure variations due
to temperature or humidity or pressure or orientation, or phase
of the moon, etc. You can monitor time, rate over time; rate drift
over time; rate over shock, etc. The Microset is really nice.
Note that just because your watch is accurate to 1 second a
month does not mean it will be accurate to 12 second a year
or 2 minutes a decade or 1/4 second a week or 1.4 ms an hour.
The extreme non-linearity in performance is part of the great
fun of analyzing the performance of any clock, in the short- and
long-term, from mechanical to quartz to atomic clocks.
/tvb
Dear Time-Nuts,
I am afraid of being off topic with the following.
If so, I sincerely apologize.
I would like to know the precision to be expected
on time keeping from a Tissot mod. J378/478 S wrist
watch and how could be verified the fulfilment of
that specification without waiting for a long time
(probably more than one month) to observe an error
of one second.
Instructions on how to build a test basket or similar
layout would be most appreciated.
Thanks in advance,
Antonio
CT1TE
_______________________________________________
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
