The daytime signal peaks at a different frequency, averaging
8270.00709300 Hz +/- 0.5 uHz
between 08:00 and 17:00. I'm also seeing the same shift at
Forest, VA. The frequency is now close to the DDS nominal
frequency. Perhaps the shack has warmed up to the temperature
the OCXO was calibrated at.
Here is the daylight signal at Todmorden, plot centered
on 8270.007093:
http://abelian.org/vlf/tmp/170508f.gif
S/N 12.1 dB in 23.148 uHz and I think we're losing a bit
of amplitude on a long integration because the energy is
not all in one bin due to drift.
These frequency shifts could be a problem for EbNaut. It won't
work if the frequency shifts by more than a few uHz during a
long transmission. The night signal might just about manage
5 chars with rate 1/16 symbols about 22 seconds, if frequency
stable. Day might just do 3 chars with 45 second symbols.
A little more ERP would help greatly here.
Signal is invisible at Forest during daylight path in 12 hour
integration.
Today I'll try to optimise beam heading and blanker settings and
plot some nice spectra.
> Good to learn that I put the decimal point in the right place
> for the ERP!
I wouldn't take these measurements too seriously. The agreement
between LWPC and measured flux is mostly luck. We can say that
signal strengths are consistent with your 10uW ERP estimate - we
are in the same ball park. I'm happy with the arithmetic - no
anomalies.
I've added a couple of entries to
http://abelian.org/vlf/amateur-radio/
Seems I have to update this file almost daily at the moment. VLF
is certainly an interesting band to work with. Everywhere you
turn there's something new to try.
--
Paul Nicholson
--
|
