Re: VLF: Tonite...

["Max IK0VVE" <ik0vve@ik0vve.net>]

DR friends,
 
there's a weak trace on my grabber... quite strong sirocco wind at the moment is shaking my loop antenna and noise is increased.
EbNaut SL session is running, too. 

73, Max IK0VVE 

______________________________

 

...here some of my HAM Radio projects...

SkyGeM Rotator

AntennaDinamica

RCR-DX

www.ik0vve.net

Da: owner-rsgb_lf_group@blacksheep.org

A: rsgb_lf_group@blacksheep.org

Cc:

Data: Fri, 03 Mar 2017 20:17:33 +0100

Oggetto: Re: VLF: Tonite...

Thanks Markus, Jacek and Eddie for the decodes.

I'm planning a new message for tonite but still need some time to build up a new PC into the VLF TX system. The old one has 512 MB RAM, the new one has 2 GB which should make a real difference regarding soundcard glitches. Additionally i'm moving from PPS+NTP to PPS+NMEA.

Announcement follows...

73, Stefan


Am 03.03.2017 14:10, schrieb Markus Vester:

Good decode here as well.

This transmission came just in time for me to test a different setup, using samplerate lock to DCF77 rather than 1pps. I'm using a cheap Delock USB sound dongle with VIA VT1630A chipset, which happens to have only moderate antialias filtering. With 48 kHz samplerate, DCF77 can be picked up in the 4th Nyqist band with only about 30 dB attenuation. SpecLab's samplerate detector initially warns about the too-low samplerate, but anyway processes the subsampled reference correctly. At my distance from DCF77 (160 km), there is a bit of nighttime phase variation, but fortunately this is reduced by almost an order of magnitude when scaling from 77.5 to 8.27 kHz.

After having decoded Stefan's message, I undid the PSK modulation and plotted the received phase in 10 minute averages. There seems to be some minor long-term phase drift, but apparently no phase-jumps from dropouts have occurred.

The same trick could probably also be used with other time signals, e.g. 60, 66.667, 162 or 198 kHz. With appropriate filtering, it is even possible to receive 137 kHz directly by subsampling. This might be a useful way to steer the phase of an LF EbNaut transmitter by GPS-phaselock, removing the need for an accurate local oscillator.

Best 73,
Markus (DF6NM)

 

ebnaut-rx V0.5
input file 03030123a.wav
sample rate 0.514405 per second
rx frequency 8270
file start time 2017-03-02 20:58:21.300
polynomial 16K21A
crc size 28
number of chars 2
block size 40
symbol period 30.000
number of symbols 960
list length 200000
start offset 407.000
freq offset 0.100000
cores 2
skipped 406.294730 seconds to start
initial reference phase 40.0
---------------------------------------------
found XX
list rank 0
reference phase  180  180  180  180
carrier RMS 2.675e+004 noise RMS 1.578e+003
carrier S/N 24.58 dB in 34.7 uHz
carrier Es/N0 -5.24 dB
carrier Eb/N0 14.01 dB
info bit period 2526.32 seconds
symbol error rate 192/960 = 20.000 %
Es/N0 from symbol errors -4.5 dB
Eb/N0 from symbol errors 14.8 dB
Shannon capacity 45.3 bits/hour
Shannon efficiency 3.1 %
symbols file 03030123a-symbols.csv
elapsed 10 seconds
---------------------------------------------
found XX
list rank 0
reference phase  150  150  180  180
carrier RMS 2.672e+004 noise RMS 1.578e+003
carrier S/N 24.57 dB in 34.7 uHz
carrier Es/N0 -5.25 dB
carrier Eb/N0 14.00 dB
info bit period 2526.32 seconds
symbol error rate 186/960 = 19.375 %
Es/N0 from symbol errors -4.2 dB
Eb/N0 from symbol errors 15.1 dB
Shannon capacity 45.2 bits/hour
Shannon efficiency 3.2 %
symbols file 03030123a-symbols.csv
elapsed 126 seconds

-----Ursprüngliche Mitteilung-----
Von: DK7FC <selberdenken@posteo.de>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: Do, 2. Mrz 2017 22:15
Betreff: VLF: Tonite...

Hi all,

It's time for a new EbNaut transmission. Another attempt to get a decode from IK0VVE or IZ7SLZ and of course also for anyone else who is interested :-)

f = 8270.100000 Hz
Start time: Thursday 02.Mar.2017   21:05:00 UTC (already running!)
Symbol period: 30 s
Characters: 2
CRC bits: 28
Coding 16K21A
Duration: 8h, 0m, 0s
Antenna current: 700 mA

If no decode appears, then it may help to repeat the process and increase the list length to very large values, like 2000000.

73, Stefan