It was a few hours before I got to the rx to retrieve the
battery so it continued on an almost dead battery and recorded
two more messages.

VO1NA 2015-12-05/06 16K21A 5S 5C QRO

 start  Eb/N0  Freq

 13:30  13.0   -222 uHz
 15:00   3.1   -256 uHz
 16:30   3.9   -482 uHz
 18:00  15.1   -558 uHz
 19:30  17.3   -228 uHz
 21:00  14.2   -330 uHz
 22:30  18.7   -262 uHz
 00:00  14.8   -253 uHz
 01:30  17.0   -293 uHz
 03:00  22.4    -41 uHz
 04:30  20.8   -208 uHz
 06:00  13.0   -216 uHz
 07:30  17.0   -276 uHz
 09:00   3.1    +20 uHz

Partial reception at 10:30 gave enough to measure +3 uHz.

I was able to measure the phase over the entire 21 hour
recording in 5 minute averages:

 http://abelian.org/vlf/tmp/151206a.gif

This looks to me like propagation, not drift of the tx
oscillator.  It's a pretty typical diurnal.  I'm plotting
phase measured at a frequency offset of -255 uHz.

I think that Joe's oscillator is actually doing very well and
we are seeing mostly the effect of propagation.  The phase ends
up nearly where it started and the total day/night change is
about -1250 degrees or about -3.5 cycles.
--
Paul Nicholson
--