Content-Transfer-Encoding: 8bit
Message-ID: <3617AC3245C2D1118A840000F805359C0175284F@PDW-MERCURY-1>
From: "Talbot Andrew" <ACTALBOT@dera.gov.uk>
To: rsgb_lf_group@blacksheep.org
Cc: "'PLX'" <Peter.Martinez@btinternet.com>
Subject: LF: RE: TransAtlantic II
Date: Mon, 10 Apr 2000 11:32:04 +0100
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1; format=flowed
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>

It would be unfortunate to limit ourselves to just use of these two
relatively simple pieces of DSP software which at the end of the day are
really just narrowband power detectors in not that narrow a bandwidth.  
Remember, the very first long distance QSO on 73kHz between myself and
G3PLX was made in 0.04 Hz bandwidth using something like 300uW ERP with
plenty of signal in hand - whereas the likes of Spectrogram can only get
down to around 1.3 Hz BW.  The same bandwidth can be achieved with a
filter made from 5 watch crystals at 32 kHz - I've done it.   .

I feel there is a lot more to be gained by coherent integration
techniques.   Use of the VE2IQ software and other slow BPSK modes goes
some way towards this, but all suffer badly from the need for clock
recovery and tracking.  The need to get initial timing information
throws away much of the usefulness of this mode.
The ultimate solution is to use GPS timing.  I know quite a number of
operators in the US are now using GPS locked frequency and time sources,
and they are begining to make an appearance here.   Having the
transition points of the bit intervals known precisely to the
microsecond (if the path length is known) leaves the signal recovery
task to be merely an integration over whatever time is needed for the
bandwidth with no lockup or preamble phase.

TAPR have a range of GPS and frequency locking projects / kits.   Who in
the US Lowfer community have the DSP hardware, even if not the
programming skills, to experiment with these techniques.  Ideally, if a
Motorola DSP 56002EVM module, GPS receiver and carrier frequency known
accurately to milli Hz are availble software can be shared and
experiments made with known equipment capabilities.

Some people are going to spend lots of time/money/effort on antennas and
PAs, why not put a similar effort into signal processing as well.  So
what if it takes 10 hours to send a message, at least it will have been
sent from a station that does not have a farm on which to erect huge
antennas.

Andy  G4JNT


>>From VA3LK

>There is precious little time left to develop computer DSP skills, such
as
>expert use of the software Spectrogram and Gram, .........- we have
begun to understand the issues >involved in bringing TransAtlantic II to
an operational status.


-- 
The Information contained in this E-Mail and any subsequent correspondence
is private and is intended solely for the intended recipient(s).
For those other than the recipient any disclosure, copying, distribution, 
or any action taken or omitted to be taken in reliance on such information is
prohibited and may be unlawful.