[time-nuts] S/A, Receiver Glitches, Y2K etc
Tom Clark, K3IO (ex W3IWI)
K3IO at verizon.net
Sun Mar 12 18:10:19 EST 2006
Don did not mention which basic receiver is inside his AUTEC box. Some
receivers do exhibit glitches. Some of these are documented in my 2000
ION paper Low-cost, High Accuracy GPS Timing available on
http://gpstime.com. You might find some of the Timing for VLBI
tutorials to be of some interest also. The ONCORE v.3 firmware is
known to have glitches 50 nsec glitches every 19.5 nsec.
Also, S/A has been turned off since May, 2000 and the US Gov't has
made a commitment that it will stay off.
The GPS "Y2K" problem occurred in August 1999. The timing broadcast is
in terms of elapsed seconds since Jan.6 1980. The seconds are sent as
a 20-bit word in order to accommodate 604800 seconds per week, and a
10-bit (range=0-1023) week counter from 1980. The 10-bit counter
overflowed 1023 weeks after 1980 in Aug, 2000. Some manufacturers
started their conversion algorithm later (like when their first
receiver was delivered), and their week ambiguity occurred later. And
there may be some manufacturers that have receivers that will glitch
at a later day -- who knows? In general these glitches happen at the
start of a new GPS week late Saturday nite.
More information about the time-nuts