Hi Jim, my ideas when I floated this project idea with Wolf was that it
should be possible to do the test with a local good ovened oscillator as a
phase reference. I have three cheap  (old ...nixie display) Racal counters
which all have ovened standards rated at 5 parts in 10^10 /day (also 5 in
10^10 per degree C) , these can easily be set up against an off-air
standard. I was hoping that with the use of a stable source to measure the
phase or timing shift at darkness and dawn. As far as I see it the
'reflection level' probably rises from 50kms up to 100kms in aperiod of
about 45 to 60 minutes as the ionisation due to the sunlight is 'turned
off'.

Incidentally does anyone know a source of 4.9152Mhz TCXO modules at a
sensible amateur price? I have a couple of Synth LF gens that could do with
a 'pacemaker' implant

The problem is that the only source that seems suitable is DCF39 which has
an idle period long enough to make a reasonably accurate phase measurement
between the data bursts. As has been noted after a data burst, the idle
carrier returns with a random phase shift. The idea put to Wolf was to
measure the transmitted phase in the ground wave area as a reference. These
measurements only need to be accurately enough tagged to identify the
particular period, which is usually at least 10 secs long. I was not too
worried about the rate of change of phase but more about identifying the
change in time delay from day to night. I was interested to see if there was
any way we could determine how many hops the signal took on paths of
different length and to try to understand the indifferent signal over
2000kms-2500kms when the 4500km Atlantic path is working well. My problem is
I am really too close to DCF39, getting the majority of the signal by ground
wave.

My though is that main T/A path is a 2-hop path but that single hope signals
can be present. I suspect this last comment maybe contraversial, but there
is
evidence in the CFH plots to support it. I dont have the original
calculation in
front of me but I seem to remember that the delay between ground wave and
one hop over 2000kms is about 30kms or 100usecs. I guess a figure would be
more like 60 to 100 usecs.  I had not thought of it at the time but I
suppose one could also try to measure the time of arrival of the first bit
of a particular burst. This would require very accurate clocks.

It may not be practical or feasible or it may be impossible to extract any
useful information, but the idea has brought up a lot of ideas and has set
people thinking about accurate measurements and references .


I apologise for generating all this technical discussion on the
refector.....but I am enjoying it !!


I am trying hard to keep the propagtion notes up to date, and at the moment
they are updated almost every day, thanks to sterling service from Brian
CT1DRP who sends me the current DCF39 plot.


There is also some new ground loss measurements and a plot of Finbar's
results of loss against top-load capacity for his main aerial after
increasing the top-load to 5 umbrella wires.

Cheers de Alan G3NYK
[email protected]
http://www.alan.melia.btinternet.co.uk