Re: GPS for satellite obs

Leigh Palmer (
Thu, 23 Jan 1997 10:20:01 -0800

The time shown on my (Garmin 38) handheld GPS unit screen is not ten
seconds different from Universal Time; it is always within a second
of agreement when the unit has acquired at least four satellites.
While it is true that the four atomic beam clocks aboard each of the
satellites keep a continuous time, the clocks are all synchronized
often from the ground, and getting them to send a signal which can
be interpreted in terms of civil time is simply a matter of maintenance.
There is a crystal clock in the unit, however, which continues to
keep time between periods when the unit is in active use. This
crystal clock time is no better than a cheap watch, and considerably
worse than my own cheap wristwatch which only creeps about a second
a week, always faster. The time resets as soon as a better time can
be obtained from the satellite signals. I haven't noticed whether
this occurs on acquisition of the first satellite signal or at some
later time (but that's something I'll check the next time I turn it
on). I am reliably informed that handheld GPS units can now be
purchased in the US for $150 retail. I paid $200 for mine a year ago
and I consider it a great toy. While the unit is cheap, the company
gouges for the interface and external power cables. I can't find a
suitable connector to make up my own cable from a spare Mac printer
cable I've got.