[time-nuts] New precision watch

[Attila Kinali]

On Tue, 20 Dec 2011 13:12:13 +1100
Jim Palfreyman <jim77742 at gmail.com> wrote:

> Why don't they build a watch that measures the temperature and every time
> you accurately set it, it adds to a small database of time change v
> temperature and then adjusts itself internally.
> 
> Over time it would become quite accurate I would think.

Because it doesn't work that way. The temperature curve is not linear.
So you would need to have a lot of reference points to get an acceptable
accuracy. Ie you would need at least a dozen if not more measurements
points, then calculate from that the correction factors (which isn't easy
if you have no control over the temperature).

It's far more economic to put the watch into a climate chamber for half
a day and do one or two cycles trough the expected range. This gives you
much more accurate data and the watch can then do simple lookup in a table
instead of doing complicated calculations.

If you do the calibration right, you can quite easily get below 1ppm 
(not accounting for aging).

				Attila Kinali
-- 
It is upon moral qualities that a society is ultimately founded. All 
the prosperity and technological sophistication in the world is of no 
use without that foundation.
                 -- Miss Matheson, The Diamond Age, Neil Stephenson