[time-nuts] PN sequence generation using GPS
hmurray at megapathdsl.net
Thu Feb 17 23:22:32 UTC 2011
JLeikhim at Leikhim.com said:
> 3) If I block the 1 PPS during PTT, and the receiving end asserts 1 PPS
> reset, will synch be lost? If not how effective will the freewheeling be
> during a 30 second exchange with 10 MHz GPS derived clock reduced to 20 hops/
The receiver has to do the same dance as the transmitter. (Maybe delayed a
bit.) If not, you will be listening on the wrong frequency. Can you use
something like a carrier detect to suppress the reset on the receive end?
(Which might mean that you don't want to start near the end of a second.)
How long you can freewheel depends upon how closely the clocks are
With a good GPSDO and good antenna, both 10 MHz clocks are going to track
within a cycle long term. (that's months, not just seconds) If your antenna
isn't good enough, you have to figure out what happens when it recovers from
holdover. Does it slew the clock to catch up, preserving the cycle count, or
does it jump to the correct frequency and add/delete a few cycles before the
Holdover specs for GPSDOs are on the scale of a few/10 microseconds per day,
so even if one end goes into holdover for an hour, you won't be off much
relative to the time-of-flight delays.
But maybe you are using an inexpensive GPS receiver rather than a GPSDO. How
good is your 10MHz clock? 30 seconds at 1 ppm difference is 30 microseconds.
I picked 1 ppm because the math was simple and it's a reasonable ballpark
for a target. Adjust for your setup.
> 4) 20 hops per second is a 50 ms chip. Two radios 50 miles apart would be
> 270 us delayed. I don't think that should impair the analog FM. So don't
> envision needing any correlation adjustment. Comments?
If you have a 10 MHz clock, it's easy to delay the PPS signal by 2700 clocks.
If you are doing the receiver in software, it may be reasonable to
automatically fine tune the delay.
These are my opinions, not necessarily my employer's. I hate spam.
More information about the time-nuts