Thanks Paul, most interesting again.
Good to know now that the normalisation helps a lot.
It was interesting to see that adding the not normalised signal of the
15th reduced the SNR by 0.1 dB whereas it rose the SNR by 0.2 dB when
beeing normalised.
Also it is very interesting to see the night phase. The day phase is so
stable at about -137.5 deg. an there were easy decodes of each message
whereas the night transmission failed to decode. I can imagine that the
phase will vary much more from day to day. Maybe we will get a decode
when combining two nights... It would be interesting to see the results
from the night transmissions in Italy (Renato), with 2 or 3 nights
combined. Maybe the night transmissions could have an advantage over
certain paths? A few more things to try out.
But things are quite clear now it seems, at least on this path of 882
km. In the past we tried to decode longer messages by using longer
symbol rates but this only increased the transmission time and made
things (phase variations) even worse or at least more critical. With the
new technique i have no doubt that even a 100 character message can be
passed between us. That will happen next week.
Am 16.11.2016 19:48, schrieb Paul Nicholson:
> Later I'll fire up LWPC and plot some examples of long
> distance paths.
For 1uW ERP, DK7FC to Todmorden
http://abelian.org/vlf/tmp/1479317903_9356.png
Daytime signal -34dBuV/m corresponds to about 0.03 fT
I'm not 100% sure that i fully understand what the plot shows. Is it a
propagation model for our path and frequency? Does it show that my idea
of TXing from 8:00...15:44 UTC each day was just the perfect choice?
DK7FC to vlf35 Virginia
http://abelian.org/vlf/tmp/1479318501_18872.png
-54dBuV/m, about 20dB weaker than Todmorden reception,
would need 100 days summation to build up the same
signal.
That means a night transmission would be the better choice for that
path?!? ...because we have 7 hours of a stable path and stronger
amplitudes. But that seems to be a rough (simple) model because the
curves are quite flat. Anyway, it is interesting and we can check the
model by our own experiments.
100 days. Hmm, but the Eb/N0 on your side is about 3 dB above the decode
threshold, so 50 days could be fine as well! And this is not even 2
months!!! :-)
And it would be well worth to try something in 2000...3000 km.
DK7FC to Tasmania:
http://abelian.org/vlf/tmp/1479318832_10010.png
A crazy looking diurnal. At -145dBuV/m, even a century
of summing wouldn't come anywhere close to this.
Now you know why i'm not a friend of simulations ;-)
...but how does it look for 8.27 kHz?
73, Stefan
| 
