|
Ok Marcus
Thanks for the info on the no-go zones , the
decoder is quite robust in terms of carrier in the
pass band , i think 64 decodes are possible
, may be less on the narrow windows , but still
only some 5 or 10 hz separation needed
I was relating to the correlation <> Op signal ,
where normally only the time is normally
selected by the operator , the software
looking for the lowest s/n to place the carrier , a
good idea in tx/rx but not for tx only ,
especially if there is a 30 min wait to see
if anything happens .. but better than
testing the 4 hour version ..
73 -G.
Sent: Thursday, June 06, 2013 11:47 PM
Subject: Re: LF: VO1NA - 137.7770 kHz
Hi Graham,
yes it's true that the opds correlation detector
does not attempt to directly decode any of the Opera 28-bit message
contents (ie. the callsign). But you do get more than a single yes-or-no
information. Opds tells you fairly reliably which entry from the predefined
callsign list actually produced the hit. The list currently contains 28
candidates which makes up about 5 bits worth. Also it measures
the TX carrier frequency down to a milli-Hz out of a 100 Hz range,
which in itself could be used to encode another 17 bits... So I'd say
it gives you less than 28 bits but certainly more than one bit.
Best wishes,
Markus (DF6NM)
Sent: Thursday, June 06, 2013 7:13 PM
Subject: Re: LF: VO1NA - 137.7770 kHz
...
Where as Op is designed to use standard tx/rx
and soundcards with
afc routines , Marcus is
running a correlation process , looking for
the pattern match , this in effect is
decoding a single data 'bit'
and stability
will affect his decode system , but with another 10 dB
s/n it may extend the range considerably
!
...
|
