Return to KLUBNL.PL main page

rsgb_lf_group
[Top] [All Lists]

Re: VLF: Software PSK modulation

To: [email protected]
Subject: Re: VLF: Software PSK modulation
From: Wolfgang Büscher <[email protected]>
Date: Sat, 23 Jan 2016 17:36:34 +0100
In-reply-to: <[email protected]>
References: <[email protected]> <[email protected]>
Reply-to: [email protected]
Sender: [email protected]
User-agent: Mozilla/5.0 (Windows NT 6.3; WOW64; rv:38.0) Gecko/20100101 Thunderbird/38.5.1
Hello Markus,

Thanks for the description - very interesting. I didn't take my time to study your configuration yet, but I wonder about the reason why the ones and zeros must be separated by spaces.. is it the lack of a file-access function which can read input character-by-character instead of number-by-number from a text file ? If so, this could be added easily.

Cheers,
  Wolf DL4YHF .


Am 23.01.2016 um 16:52 schrieb Markus Vester:
For yesterday's EbNaut transmission, I have not employed the usual Rubidium oscillator and automatic antenna variometer hardware, but reverted to the software modulation scheme which had been originally used for our first tests in April 2014. This is an extension of the "green line" anti-glitch method, with 1pps-sync in SpecLab and RF feedback of the analog output to the ADC input. To generate PSK modulation, the phase prescription for the software PLL is no longer fixed at 180° (green in the colour spectrogram), but switched between 90° (red) and 270° (sky-blue) by a SpecLab script which periodically reads 1 or 0 symbols from a text file.
 
The primary advantage of this method is that no external hardware modulator is needed. With the feedback picked up from the antenna, it also intrinsically compensates for wind-induced resonance variations and can do without an automatic variometer. Thirdly, the phase switching occurs smoothly by frequency variation, reducing switching transients and far-away sidebands. The main disadvantage is that the phase ramping takes time (about 1 to 3 seconds) and also introduces some jitter, so this method cannot be used effectively for fast signaling with symbols shorter than 5 seconds.  
 
If you would like to try, a SpecLab configuration file can be found at
http://df6nm.bplaced.net/VLF/_vlf_tx_pll_psk.USR
 
To use it, you first need to set up GPS 1pps and audio feedback as described in
http://df6nm.bplaced.net/VLF/VLF_Transmit_phase_correction.txt
This can be done either with a stereo input soundcard (1pps on the right, RF feedback on the left channel), or mono with both signals combined through resistors (RF feedback about 5% of the 1 pps level). Make sure that the 1pps samplerate detector is locking properly (i.e. mostly green), and your PC time is close to the nearest second.
 
Then you need to encode your message using Paul's ebnaut-tx software
http://abelian.org/ebnaut/
and press "Save" to make a text file. For use with the SpecLab script, you then need to insert spaces as separators between the 0 and 1 digits, e.g. converting "1001..." into " 1 0 0 1..." by replacing 0s and 1s in notepad (note that "1 0 0 1 ..." works just as well but may produce a surprising result when reopening the file - search for "bush hid the facts" ;-) As an example, my recent transmission is at
http://df6nm.bplaced.net/VLF/psk_8k19a_73.txt
The file should be saved as psk.txt and placed in your SpecLab home directory.
 
Then you will want to select
- the TX frequency in the second oscillator of the test signal generator,
- the symbol duration as periodic actions interval (eg. 10 seconds),
- the start time, by letting scheduled actions open psk.txt a couple of seconds before. You should also close the file after the transmission,
- and finally tick "send unmodulated test tone at TX frequency" in the digimode terminal to let the signal appear on the left output.
 
A couple of notes regarding frequency selection: For straight carrier experiments, staying within very few mHz of 8270 Hz can show the signal on various grabbers. However EbNaut PSK reception would be susceptible to co-channel interference over a wider bandwidth, eg. +- 0.1 Hz for 10 second symbols. As the PSK signal will anyway be invisible on the grabbers, it might be advisable to choose e.g. 8970.1 or 8969.9 Hz for future PSK transmissions, thereby nulling potential carriers near 8270.000 but still staying inside the decimated IQ bandwidth for EbNaut recordings (+-0.25 Hz here).
 
As SpecLab aligns the absolute phase to time-of-day, the TX frequency for overnight transmissions should be a multiple of 1/86400 Hz to avoid an Alpha-like midnight phase step.
 
Hope this may help to get a few more EbNaut signals on air... Big thanks to Paul and Wolf for making all this possible.
 
All the best,
Markus (DF6NM)
 

<Prev in Thread] Current Thread [Next in Thread>