First computer with real-time clock?

From: Paul Koning <pkoning_at_equallogic.com>
Date: Tue Aug 3 08:12:04 2004

>>>>> "David" == David V Corbin <dvcorbin_at_optonline.net> writes:

 David> Regarding a message that appeared on this thread with the
 David> approximate quote:

>>> Crysital Oscliattors Quite Bad with a tolerance of 0.01& lead to
>>> an error
 David> of over a minute per dat

 David> [sorry I it the delete key too quickly on the message to post
 David> the EXACT text]

 David> 1 Day = 86400 seconds * 0.01% [100 ppm] = 8.64 SECONDS per
 David> day.

Oops.

 David> Not good enough for long term time keeping but much better
 David> than the previous poster indicated.

Yup. Right conclusion, wrong numbers.

 David> I was involved in the development of som Military systems
 David> [1979-1983] that used a tempe0rature stabilized crystal with
 David> 0.5ppm stability. To the best of my knowledge this was a
 David> "state of the art" implementation of automomous time keeping
 David> for 1979.

0.5 ppm for boxes that get carried around in military trucks and
bounced around in the field -- that's quite good. For something
that's sitting in a reasonably controlled environment, that's not
so great; I think that 10^-8 would be considered state of the art for
OXCOs (1970s or not).

For autonomous timekeeping independent of technology, the state of the
art was a second per year or so (that's 10^-8, roughly) around the
early 1900s. First with pendulum clocks (Shortt clock), then around
the 1940s or so crystal clocks came in that could match this. And not
too long after that there came the rubidium (10^-10) and cesium
(10^-14) clocks. Some of that would be found in military gear, I
think (Rb at least, Cs somewhat less likely). Consider GPS
satellites, which have either or both built-in.

            paul
Received on Tue Aug 03 2004 - 08:12:04 BST