Re: LF: MF Ebnaut transmissions from Riccardo IW4DXW

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

Hi,

Yes, Paul thanks for your comments.

Riccardo IW4DXW made some tests also today.

First of all he solved a small bug in the math of the PIC leading his frequency out of 0.01 Hz.

Then he made some transmissions on 478460.00 Hz  using down to 0.4 W (abt 1 mW ERP) and i got his messages several times  (QRB = 650 km).

This is the last:

ebnaut-rx V0.3g
input file C:\EbNaut\01062233.wav
sample rate 125.008900 per second
rx frequency 478500
file start time 2016-01-06 22:29:33.770
polynomial 8K19A
number of chars 16
block size 112
symbol period 0.500
number of symbols 1040
list length 20000
start offset 26.230
freq offset -40.000000
cores 1
padded 21.974 seconds at end
initial reference phase -44.0
---------------------------------------------
found IZ7SLZ GN IW4DXW
list rank 0
reference phase  180  180  180  180
carrier RMS 6.491e+005 noise RMS 4.288e+004
carrier S/N 23.60 dB in 1923.1 uHz
carrier Es/N0 -6.57 dB
carrier Eb/N0 4.00 dB
info bit period 5.70 seconds
symbol error rate 296/1040 = 28.462 %
Es/N0 from symbol errors -7.9 dB
Eb/N0 from symbol errors 2.4 dB
Shannon capacity 2068.8 bits/hour
Shannon efficiency 30.5 %
symbols file C:\EbNaut\01062233-symbols.csv
elapsed 10 seconds

EbNaut looks very promising also in MF !

IW4DXW will complete his receiver in few time. Maybe he will be ready for a QSO in the week-end.Hope that Markus and other operators can join the tests on MF.

All the best.

Domenico, iz7slz

On 6 January 2016 at 22:16, Paul Nicholson <vlf0403@abelian.org> wrote:

Nice work!  Well done to Riccardo for getting the coherent
transmission working.

The timing looks good and with the signal available, it'll
probably do 0.1 second symbols easily.  I don't have
anything available to receive coherently at MF so cannot
join in the fun.  Perhaps someone can recommend a suitable
SDR that I can clock from a 10Mhz GPSDO?

Here I'm still working on CRCs and ploughing through the
math to calculate code performance.  Even with knowledge of
the complete weight spectrum of the code, it is proving to
be immensely difficult but I'm making some progress.

The version of EbNaut now on the website has a selectable
CRC length and some improvements that I mentioned a week
ago, but I recommend staying with 16 bit CRC for now.

--
Paul Nicholson
--