Then why do my speedometer on the car, truck, van, dads truck all match w/
the gps and never vary over time. If what you say is true, then at one
point it would match the truck and the next point it would have varied and
not match.
All the stuff you talk about adds maybe a .10 variation. Not 3% to 4% which
would be 2.1 to 2.8 mph off at 70 mph. It is not off far enough to tell.
And by the way - it matches exactly w/ the radar speed indicator on the side
of the highway that the state police put up in construction zones.

--
Tony
my boats at http://t.thomas.home.mchsi.com

"Short Wave Sportfishing" wrote in message
...
On Mon, 30 Aug 2004 22:59:06 GMT, "Tony Thomas"
wrote:

My experience has show paddle wheels to be more accurate than the
pressure
pickup but they are usually still off by a couple mph.

As to the accuracy of GPS. Look up any specs on the units at garmin.com.
Accuracy:
a.. Position: 15 meters (49 feet) RMS*
* Subject to accuracy degradation to 100m 2DRMS under the United States
Department of Defense-imposed Selective Availability Program.

a.. Velocity: 0.1 knot RMS steady state
Velocity is very accurate as it is a simple time shift of three or more
satallite signals.

FYI: It's only three sats - the others are considered "spares"

And the specs are correct.

However,

GPS specs are figured as being in "free space" - that is without
environmental considerations such as tropospheric distortion, Doppler
shift, phase shifts, loss of differential data, the geometry of the
satellite constellation at any time, how long the receiver hangs on to
a degrading signal, Horzontal/Vertical Dilution of Precision, clock
drift - oh, hell, a whole bunch of factors. Designers try to account
for the variables, but in free space, it's a perfect world.

In the real world, variables can affect GPS calcuations along the
order of 3 to 4% and even higher depending on one or more combinations
of those variables.

The only true way to determine how accurate the GPS is, is to repeat
the same course, at the same speed, a whole bunch of times, and
average the differences.

And I think you will find I'm right.

Later,

Tom