WWVB controlled clocks

Frank Reed (f-reed@metabien.com)
Sat, 21 Nov 1998 14:53:15 -0700

Several months ago there was some discussion here of how to get a UTC time
reference to a precision of 0.1 second for timing of observed satellite
events.  At the time, I concluded that a short wave radio receive tuned to
a WWV frequency was the least costly for me.  Other means, including low
cost radio clocks which receive the 60 kHz signal from WWVB, consumer
oriented GPS receivers, and time signals via the Internet, had typical
timing errors of a few tenths of a second.

Recently, I came across a post in the sci.geo.satellite-nav new group that
changed my mind about my Radio Shack WWVB radio controlled clock.
Previously, I had thought that the frequency drift of the clock oscillator
could cause time errors of a few tenths of a second to accumulate between
self calibrations by received WWVB signals.  Here is the post by Michael
Rathbun that changed my mind:

----------------------------------------------------------------------------
---

I have measured an approximate 150 millisecond per day drift on a freshly
reset clock that has been synchronized once.  However, when the same clock has
performed regular syncs over a couple of days, it appears that it steers its
timebase based on the corrections.  

Dropping the antenna in a steel box after the clock has been running for a few
days causes the "no sync" indicator to appear after about 96 hours, but there
is no detectable drift at that time when I compare the clock display to a WWV
broadcast.  Disabling sync using the button press apparently results in
resetting the timebase to its default calibration, which is the reason for the
steel box in this test.

I haven't tested how well it compensates for temperature variations, however.

mdr

--------------------------------------------------------

My own experience several months ago tends to confirm this: after showing
"out-of-calibration" continuously for about two weeks, my RS clock did
*not* show an expected drift of several seconds.

I now believe I this clock will is reliable to within 0.1 second or so.

I think I paid around $40 when it was on sale at Radio Shack last spring.
Similar clocks are available from Oregon Scientific:


http://oregonscientific.com/timefram.html

I once saw a post from a man who paid only $30 for one of theirs, but he
didn't say where.

- Frank


--
Frank Reed
Metabien Software

http://www.metabien.com/maps/   Topo maps on CD-ROM