[time-nuts] M12T and M12M

[Poul-Henning Kamp]

In message <20060615.101811.786000956.cfmd at bredband.net>, Magnus Danielson writ
es:

>> Is there logic (that I'm missing) to what seems like such an odd frequency?
>
>Hint: 16 x 1023000 = 16368000 Hz
>
>1,023 MHz match the chipping rate of C/A code. Perfect match is not the end of
>the world since you need to compensate for doppler and deviations.
>
>The actual frequency 16367000 is however a bit strange since it is
>2*2*2*5*5*5*13*1259.

When you discipline signals and clocks the way it's done in a GPS
receiver you can save yourself quite a bit of code for keeping
proper track of the sign by choosing a clock which is always on the
low (or if you prefer: high) side of the ideal.

Given that the spikes in the (corrected) PPS on the Oncore UT were
related to just such a sign bug, I'm not surprised Motorola did
this for the M12 family.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.