Let me add that if done off line and the 10 Mhz Reflock II based
oscillator is
used, then you can use a forward backward algorithm tracking the
oscillator
to resample the other channels to remove all systematic biases/errors up
to the
accuracy and the lock of the 10 Mhz oscillator. Not much can be done
with the
phase noise of course. If you want, you can multiply that 10 Mhz to
60Mhz and
use it as the external clocking (rather than the 64 Mhz oscillator now
on board).
The Reflock II could also be used to lock a 64 Mhz oscillator as well,
given an external
reference such as GPS, and then all sample rates will be as needed.
This implies
having a VCXO that runs at 64 Mhz. The phase noise would be low since
you are
taming a crystal oscillator. Thus the accuracy can be improved in
either of these ways.
A TC-VCXO would be ideal. Barring that, put a thermal tab on it. DB6NT
http://*www.kuhne-electronic.de*
makes a slip over microprocessor controlled "dress" for keeping a
crystal at nearly 40C. It is called the QH40A.
Go here:
http://www.kuhne-electronic.de/english/special.htm
and click on crystal heater.
Bob
n4hy wrote:
None of this work needs to be done realtime. It can be done after
the fact on
recorded digital signals. May I suggest that you take that Reflock
II tamed
10 Mhz oscillator and record it in parallel.
Bob
_______________________________________________
Discuss-gnuradio mailing list
address@hidden
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
_______________________________________________
Discuss-gnuradio mailing list
address@hidden
http://lists.gnu.org/mailman/listinfo/discuss-gnuradio