[time-nuts] GPS fadeout
murray at rakon.co.nz
Wed Aug 22 18:02:44 EDT 2007
Of course that's why GPSDO units have holdover capability, to help them
through the not-infrequent periods when a reliable 1pps is not
available. Even with the best antenna in the best location, there will
be times when your usable constellation of satellites is as low as 5,
simply because of where they all are at the time.
If any of them are in less than optimum locations, it doesn't take much
interference to stop you getting a fix.
I've also noticed that different receivers have quite different
strategies when it comes to changing from a setting satellite
(approaching the horizon or antenna mask angle) to a rising one. The old
6-ch Motorola receiver in the Z3801A seems to use a very nice strategy,
where it abandons a setting satellite when a rising one reaches the same
elevation, and switches over seamlessly, but others (such as the 12-ch
Furuno one in the HP Z3815A) seems simply to chase the satellites down
to the mask angle and then go searching for another. Obviously with only
6 channels you need a very careful approach to satellite choice!
This is a long way of saying that knowing that the errors on the
satellite position increase at low elevations, the satellite selection
algorithm plays an important part in the ability of the GPS unit to
maintain a good fix with a limited number of available satellites.
Murray Greenman ZL1BPU
More information about the time-nuts