Dr. Henri Wilson wrote:
On Wed, 20 Feb 2008 11:58:55 +0100, "Paul B. Andersen"
wrote:

This table show how the clock error for the satellite with PRN 2
has evolved during 30 days. Note that the clock runs
a bit too fast, so the error increases day by day.
The error has increased 564 ns during 28 days, which is
an average of 20.1 ns/day, or +2.3E-13 too fast.
This is as expected (or better), since it is less than
the precision of the frequency standard.

So the clock which was GR corrected by -4.4647E-10
when it was manufactured, is running +2.3E-13 too fast.
That is, the error in clock rate is 0.05% of the GR correction.

This proves conclusively that the GR-correction is correct
within the precision of the clock.

All you have done is prove that the 'GR correction' is NOT the right one.
The clocks drift way off compared with the Earth clocks.

Thanks for the new contribution to my collection of
Wilsonian thinking:
"If the prediction is in accorance with the measurement within
the precision of the measurement, then the prediction is proven
wrong."


--
Paul

http://home.c2i.net/pb_andersen/