[time-nuts] PN sequence generation using GPS

[jimlux]

On 2/16/11 4:58 PM, Joe Leikhim wrote:
> Thanks Hal;
>
> Transceivers A and B (C etc) would extract clock and 1 PPS from their
> own GPS. The idea is to use 1 PPS (or derivative) to reset periodically.
> Yes the propagation delay A-B would limit the hopping/spreading rate
> unless some mechanism to correct the offset at the receiver end. (I
> don't see this as big problem for the experiment).
>

The traditional approach for tinkering is to start in sync, and then 
slow one down until you have the right delay.  After all, you know that 
there's a positive time delay between the two stations.

If the distances are short, you can use some sort of 
pushbutton/flip-flop scheme to drop one clock pulse at a time from the 
generator.

At a 1 Mchip/sec rate, a chip is about 300 meters long.

As others have commented, the usual scheme is to rig up some sort of 
tracking loop.  In the analog world, a popular scheme is an "early/late" 
comparator, since it's easy to get two outputs from your shift register 
that are one chip different. You run both of them into a pair of 
correlators/mixers, and then take the two outputs and subtract one from 
the other to generate your error signal to drive the VCO.