Re: LF: EbNaut test

[Markus Vester <markusvester@aol.com>]

Hi Stefan,

the frequencies of recent detections (e.g. IW4DXW) on our two opds grabbers seem to be within a mHz, so your 10 kHz derived receive LO should be ok. Could it be that the strong LF transmit signal is interfering with either the GPS itself or the 1pps samplerate detector?

My opds image currently shows a zoomed section around on 137517.000 which you might use to check your TX frequency. 

Domenico mentions a possible problem when using synced resampling in the TX instance. I haven't run into that yet as I'm currently not using resampling, neither for TX or RX instances (only 1pps phase-lock but no resampling). The disadvantage on RX is that the frequency offset from the FFT center frequency is not tracking samplerate deviations, but this is only a small effect (e.g. about 30 uHz error 30 Hz offset and 1ppm drift).

One advantage of SpecLab TX with audio feedback is that the phase can recover from audio dropouts on the ADC and DAC side. This might potentially become a problem for EbSynth on a RasPi with USB soundcard.

Best 73,
Markus

-----Ursprüngliche Mitteilung-----
Von: DK7FC <selberdenken@posteo.de>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: Mo, 19. Jun 2017 16:56
Betreff: Re: LF: EbNaut test

That is strange indeed. Maybe something is wrong here? Maybe the old GPS module is not running properly? I had to replace a mains socket recently and had to shut down the complete station for that.
I would like to run a carrier transmission. Could someone of you check it for frequency accuracy? Or we do it the other way. Domenico, your OP32 signal has always been very accurate, i remember 137485.000 Hz. Can you run a transmission so i would see the same offset here maybe? Or maybe Markus, or Riccardo?

73, Stefan

Am 19.06.2017 16:27, schrieb Markus Vester:

>> Where does that -3.4 mHz offset come from?

That's strange. I observed exactly the same frequency offset here (137499.9825 Hz RX LO + 30.0141 Hz offset). On the other hand, your TX signal generation "should" provide a GPS-accurate frequency.

The 0.3 s time lag is expected due to soundcard DAC latency.

Best 73,
Markus


-----Ursprüngliche Mitteilung-----
Von: DK7FC <selberdenken@posteo.de>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: Mo, 19. Jun 2017 15:44
Betreff: Re: LF: EbNaut test

Hi Domenico,

Thanks for the detailed report and the EbNaut decode over 1159 km. It was still in daylight, so not bad for LF in mid June and a 376 second long message...
Where does that -3.4 mHz offset come from?

Well, i realised that i don't even need to build the XOR gate circuit. I simply used the PC that generates EbNaut for VLF, i.e. SpecLab, which has a PPS+NMEA input. I choose a frequency of 12530 Hz. Then i built an LF up-converter (in 2011) that has a LO of 125 kHz. So i come out on 137.53 kHz. That 125 kHz LO frequency (4 MHz xtal down-divided) is PLL locked to the 10 kHz ref output of the old Jupiter GPS, which is running here since 2010. The 10 kHz and the 125 kHz are also used on the RX side. So i can generate stable signals on LF...
It seems it works :-)

73, Stefan


Am 19.06.2017 14:59, schrieb Domenico IZ7SLZ:

Hi Stefan, LF

 as i suspected, i found my linux-pc storage hard disk full. A new 1 TB HD is now in order.

So unfortunatly i can't give you an EbNaut report obtained using full vlfrx-tools programs.

OPDS SL session has recordered a file witch covers your 16:30 EbNaut transmission. I got a nice decode using almost the same offsets found by Markus DF6NM.

I have tried to use also linux ebnaut,  after converting the wave file with following shell script:
------------------
#!/bin/bash
sox 06181637.wav  -t dat  -|         
sed '/;/d' |
ebnaut -v -dp4K19A -F29.9966 -N10 -S1 -r111.11111 -c2 -L200000 -PS -v -T724.5  # |tee dk7fc.txt
-------------------

But it gives,  of course, almost the same results.

At the link http://qsl.net/i/iz7slz//DK7FC/   i have uploaded FFT raw and wave files, in case someone wants to try the decoding.

So Stefan, congratulations for your first LF-Ebnaut transmission. Can you tell us how did you generate the signal ?

There is a very easy and cheap method actually in use by Riccardo IW4DXW  and Tony IK1HGI  for generating a coherent carrier for EbNaut.Congrats to Riccardo for this nice idea.
They are using directly the timing pulse TP5 output of the neo-8 GPS . In fact this timing  port (TP5) can be reprogrammed up to 5 MHz.

The BPSK modulator consist of an XOR port in between the GPS TP5 output and the PA driver. Cool !

73 Domenico, IZ7SLZ

 

On 18 June 2017 at 23:44, DK7FC <selberdenken@posteo.de> wrote:

Thanks Markus, nice. We could exchange 1000 characters in 2 minutes, if the Eb/N0 calculation is correct :-)
Maybe i will try another one with Domenico when he is back. And is RN3AUS reading this email?
Is someone else interested in an LF EnNaut test, if the equipment is available,maybe W1TAG?

73, Stefan

Am 18.06.2017 22:07, schrieb Markus Vester:

Hi Stefan,

a fine and error-free decode, best with 0.3 s extra delay. A phase glitch occured here near the middle, presumably on the receive side.

Best 73,
Markus



-----Ursprüngliche Mitteilung-----
Von: DK7FC <selberdenken@posteo.de>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: So, 18. Jun 2017 18:35
Betreff: Re: LF: EbNaut test

PS: This time 100 Hz higher

Am 18.06.2017 18:28, schrieb DK7FC:

Hello Markus,

Oh, you are right. Same message on 16:30 utc. ..
I hope it works...

73, Stefan

Am 18.06.2017 16:56, schrieb Markus Vester:

Hi Stefan,

I'd like to try. But the QRG 137.43 kHz was below my opds capture range (137.45 to 137.55 kHz).

73, Markus




-----Ursprüngliche Mitteilung-----
Von: DK7FC <selberdenken@posteo.de>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: So, 18. Jun 2017 16:47
Betreff: LF: EbNaut test

Hi LFers,

My first EbNaut message is running at the moment. Not sure if the timing
and keying is OK. Could someone check if it can be decoded? Don't make a
secret out of the message.

Parameters: 4K19A, 1sec, 10 char, CRC16, Start:14:36 UTC (maybe there is
a timing offset between +-1 sec? pse report).

73, Stefan