Re: VLF: 300 mA on 5170 Hz

[Markus Vester <markusvester@aol.com>]

Hi Stefan,

attached is the result of stacking three days at DL0AO (also appended to table below).

Indeed the daytime carrier was well received, with no indication of attenuation towards the afternoon. Daytime SNR was around 15 dB in 1.5 * 424 uHz, equivalent to -17 dBHz. Nighttime Eb/N0 was ~ 6 dB for 5.7 bits in 8.9 hours which is 6+7.5-45 = -31.5 dBHz, or 14.5 dB less SNR than during the day. On average, nocturnal noise was approximately 12 dB stronger, so we end up with a 2.5 dB stronger daytime signal (albeit this small difference is probably within uncertainties). Anyhow this doesn't look like a daytime 1.5 lambda minimum at all.

Now
https://en.wikipedia.org/wiki/Earth%E2%80%93ionosphere_waveguide
https://de.wikipedia.org/wiki/Ionosph%C3%A4rischer_Wellenleiter
left me confused. The article basically says that for small ranges the simple one-hop ray theory is adequate. But it also states that at VLF freequencies, the bottom of the ionosphere should act as a magnetic (i.e. high-impedance) wall, inducing a reflection with 180° phase shift. This is the opposite of what I had thought - shouldn't the conductivity from the free electrons rather resemble a metal-like electric boundary, like it undisputedly does at ELF frequencies? Reversing boundaries would swap minima and maxima, but in this case neither seems to fit our observation.

Best 73,
Markus (DF6NM)


-----Ursprüngliche Mitteilung-----
Von: DK7FC <selberdenken@posteo.de>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: Di, 24. Okt 2017 23:41
Betreff: Re: VLF: 300 mA on 5170 Hz

Hi Markus,

Thanks for the results. So the night/day SNR ratio looks much better than on 6.47 kHz for us.

Am 18.10.2017 11:12, schrieb DK7FC:

Am 17.10.2017 22:55, schrieb Markus Vester:

As a crude first Approximation to the model, we might look only at groundwave and single-hop skywave. For 230 km range, the geometrical path difference would be 63.3 km (1.37 lambda) for 90 km reflection height (night), and 40.3 km (0.87 lambda) for 70 km (day).

For 5170 Hz it is 1.09 lambda during the nights and 1.44 lambda during the days. That's even closer to 1 lambda or 1.5 lambda, so here we would expect an even stronger experessed day/night difference, if this propagation model is reasonable over that distance.
We will check that!

...now, can you say something about the daytime / nighttime levels (not SNR).  Actually the daytime signal should be quite invisible when the geometrical path difference explains the propagation in that range. But i saw the signal at up to 15 dB in 424 uHz, actually the best daytime SNR on all bands tested between us so far. So again, the oppsite to the expected?!?

58 km is an ideal band for us. Also your recent carrier was 'O copy' even in 424 uHz.
Don't you like to send your call or so in EbNaut during daytime? I might catch it in less than 5 hours!, e.g. http://abelian.org/ebnaut/calc.php?sndb=10&snbws=0.000424&snmps=&code=16K21&sp=15&crc=16&nc=5&submit=Calculate

73, Stefan


Am 24.10.2017 21:37, schrieb Markus Vester:

Good nighttime results at DL0AO on this frequency, using only 8.9 hours of data from the 31.4 uHz western cardioid instance. Decoder output was equal from the two available nights, so surely correct. Summing both with equal weights produced slightly less than the expected 3 dB gain.

Carrier-based Eb/N0:

================
20/21  (no data recorded)
22/23   5.5 dB
23/24   5.8 dB
sum     8.3 dB

24/25   6.3 dB
sum(3) 10.2 dB

Best 73,
Markus (DF6NM)


-----Ursprüngliche Mitteilung-----
Von: DK7FC <selberdenken@posteo.de>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: Di, 24. Okt 2017 18:58
Betreff: Re: VLF: 300 mA on 5170 Hz

...the message was running on the 20th, 22th and 23th.

Today i made a mistake in the SpecLab settings which unfortunetely affectet the transmission.
I watched the alternating colours (due to the PSK) and saw that it doesn't switch the phase as expected. Then i quickly restartet SpecLab, it was bit chaotic but now it should work again. Since the symbol length is 3 minutes it was easy to delete the first 11 symbols and start this special file at 16:33:00 UTC. All in all just the first 8 symbols are affected. 2 of them were transmitted correctly (productive), 2 of them have opposite phase (destructive) and 4 were not transmitted (not really neutral but not so dramatic).
Fortunately the transmission takes 12 hours, so 24 minutes is just 3.3 % or -0.15 dB.
The transmimission should now continue as expected.
Just a note for those who follow it...

73, Stefan


Am 20.10.2017 16:11, schrieb DK7FC:

f = 5170.1000 Hz
Start time: 20.October.2017  16:00:00 UTC (daily)
Symbol period: 180 s
Characters: 1
CRC bits: 6
Coding 8K19A
Duration: 12h
Antenna current: 300 mA

sum_W_22-24_result.png
Description: PNG image

sum_W_22-24_sym.png
Description: PNG image