Hi ULF friends,
Last Friday i run another carrier transmission on my 900m long ground
loop antenna. It was the second experiment at 470.1 Hz. The
last experiment failed because just a few 1/10 dB were missed to reach
the 14 dB SNR level which is the criterium for a valid detection.
During the last week i managed to reduce the background QRM on the E
field significantly (at least 10 dB) in that frequency range by
changing the ground connection on the RX system, so it was an EMC issue.
In the recent experiment the DC measurement showed 1 A at 58 V, so the
loop loss was just 58 Ohm, which is one of the best values so far.
Conecting the PA to the LiFePo4 accu produced 620 W RF power and 3 A
antenna current (!) in the first moment. But the DC/DC converter became
a bit to warm and i (thought) had enough time so i reduced the input
voltage. The power dropped to 460 W and so the antenna
current during the experiment was 2.75 A. The carrier started at
14:50 UTC and was on the air for 2.5 hours!
It was a quiet day and so i decided spontaneously to do the experiment.
But i've been at work in the morning and so i was late, a bit to late
actually, since the band already opened and the QRN increased during
the experiment. Who knows what happened to the phase over that path.
The only RX station detecting the signal was my own tree in 57.6 km
distance. Due to the rising QRN the last 30 minutes did not help
to rise the SNR of the postprocessed spectrum peak. But 2 hours were
fine to get a clear copy!
Usually i have to ways to determine the final SNR: One is to read the
output of the EbNaut decoder decoding a '*' message. The ouput, if
using the -f15 -f16 -M'*****' options, then shows the carrier S/N
directly.
The other method is a script that calculates the SNR dB value based on
the spectrum peak and 0 dB noise level. This 0 dB noise level
calculation is done according to the description in SpectumLab's help
sites. Here is the quote:
"The definition of noise levels is not easy. Here is the basic
algorithm of the 'noise' function:
- An array of amplitudes (usually dB values) from the last FFT
calculation is sorted into order of increasing amplitude.
- The amplitude of the lower quartile value (for example bin
number 256 in a sorted set of 1024 points) PLUS 3dB is then returned as
an estimate of the mean noise level.
This technique automatically throws away very high values
(strong signals) that would otherwise affect the result."
This technique is used in my script. I find the result adequate, the 0
dB level is visually in the center of the noise. But sometimes the SNR
statements are a bit different between my script and the result of the
EbNaut program. Maybe it is due to the fact that i am calculating the
mean level based on just 1000 bins arround the spectrum peak, and
EbNaut watches a higher bandwidth, which may contain more QRM peaks
that rise the mean noise baseline?
That may explain why my script whows an SNR of 18.2 dB whereas
the EbNaut decoder says it is an SNR of 16.17 dB. Anyway, well
above 14 dB.
The signal was strong enough, it could have carried a 6 character
EbNaut message.
Attached is a 424 uHz spectrogram showing a short carrier trace just
before the high QRN sets in. Also attached, a single spectrum peak
detection and the result for an EbNaut message (showing the carrier
S/N).
So, it is the first amateur radio signal on the 638 km band
detected over such a distance i think.
73, Stefan
|
470.1Hz_carrier_in_424uHz.png
Description: PNG image
470HzCarrier2hours_dBscale.png
Description: PNG image
470.1Hz_carrier_EbNaut.png
Description: PNG image
|
