Re: LF: MF: EbNaut

[Domenico IZ7SLZ <iz7slz.domenico@gmail.com>]

Dear friends,

in the Paul's 21.15 decode (screenshot of my previous mail)  i made a mistake in entering the time offsets. I wrongly used comma instead of dot to separate decimal value.

So decode for 21.15 gives now better results. BTW seems there was about 20 milliseconds difference between my and Paul's clock as shown in the second screenshot.

73, Domenico iz7slz

On 1 March 2016 at 09:02, Domenico IZ7SLZ <iz7slz.domenico@gmail.com> wrote:

Hi Paul,

for your analisys, in this folder there are .wav files recordered during the EbNaut test of yesterday night:

https://www.dropbox.com/sh/9c3nnw40nyb5pqc/AABZ6oPxpAY4IwuWFmebMSfba?dl=0

These are available also to whom wants to try EbNaut-rx program playing with offsets as shown in the attached screenshot. So everybody can download and use them.

73 all

Domenico, iz7slz

On 1 March 2016 at 00:43, Domenico IZ7SLZ <iz7slz.domenico@gmail.com> wrote:

Hi Paul,

again congratulations becouse i got some more better decodes.

Here the  results by playing only with nominals time & frequency offsets:

UT   Eb/N0 sym  Eb/N0 carr  Rank

2045  NIL
2100  5.1  7.2   0
2115   -1.3 -1.14  47
2145  4.0  5.8  0

2200  NIL
2215 5.4   7.1  0
2230  2.4  3.3  0

2245 NIL

2300 NIL

So,  seems that fading can interrupt completly the EbNaut signal but, i'm shure that , by making some changes in the freq. offsets,  it will be possible to get  also the  failed decodes.

For further analisys, you can download my FFT's  files from dropbox
https://www.dropbox.com/sh/zu0k37z95ft1xfw/AAAgwsWDdfHVxTByuT8CyY_2a?dl=0

Receiving conditions: 7m vertical whip, RX JRC NRD-92M with GPS disciplinated OCXO, soundcard calibrated with 1_pps from same GPS.

All the best

Domenico, iz7slz

On 1 March 2016 at 00:07, Paul Nicholson <vlf0403@abelian.org> wrote:

Many thanks Markus and Domenico.  I'm really pleased
with that.

I make that about 2022 km.   A distance record for me!
Maybe that is the farthest EbNaut on MF so far?

My callsign decodes with Eb/N0 = -0.4 dB determined
from the symbol errors and +1.0 dB from the carrier
strength.  Still some discrepancy there.

Data from Markus shows invariably a lower Eb/N0 (higher
BER) than suggested by the carrier S/N.  The mode is
not performing as good as it should, and the difference
is expressed in the dB between the two columns

> start  car. sym.
> 23:15 10.0  3.9

missing out on over 6dB there.   The difference is caused
by phase jitter if the carrier Eb/N0 is greater than the
BER Eb/N0.   The other way round would indicate some other
nearby signal in the decoder input band.

The question, is that jitter from my tx or is it propagation?

The signal from EbSynth is good. I'm suspicious of the multiplier.

I think I will have to make up another divider and divide
the signal back down to audio frequency and compare with
the reference.   I have a nifty soundcard program for accurately
comparing the phase of two signals of different frequencies.

Now for a celebratory cup of tea.   The signal stays
on air until 00:00.

--
Paul Nicholson
--

Screenshot 2016-03-01 12.25.03.png
Description: PNG image

Screenshot 2016-03-01 12.29.11.png
Description: PNG image