Content-Transfer-Encoding: 8bit
Message-ID: <3617AC3245C2D1118A840000F805359C01752896@pdw-mercury-1.dera.gov.uk>
From: "Talbot Andrew" <ACTALBOT@dera.gov.uk>
To: rsgb_lf_group@blacksheep.org
Subject: LF: Transatlantic  beacon tests
Date: Wed, 26 Apr 2000 12:45:14 +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>

I was receiving Marco's carrier during the night tests at 20 - 30dB
above noise in a 2mHz bandwidth at a distance of 927km.  Daytime was 10
- 15dB lower.

In a Spectran bandwidth of  0.03 Hz that will correspond to about 12db
less S/N  ie 8 - 18dB - a good visible signal on a waterfall,  so
confirms we are receiving consistent levels between us.

Transatlantic test beacons need to be very stable so that the listeners
over there that have the equipment can use very narrow bandwidths.
Spectran may be popular but it will only go down to 0.03 Hz and other
listeners over there do have other DSP equipment that can do a lot
better.

I suggest we don't use 2 Hz FSK (DFCW) unless this is generated by a DDS
or similar source which can absolutely guarantee the tone frequencies.
This sort of frequency shift requires a crystal to be pulled a long way
and results in a rather hit or miss frequency setting accuracy.
Marco's first test showed this when I got a wonderful wobbly trace.

Instead, I suggest a simple on / off keyed carrier in traditional SLOWCW
or whatever it is called now generated from a source with a bit of care
taken over its stability.  Normal CW may take longer to send a callsign
but so what, if someone actually hears it and can measure S/N.  

A simple temperature control, such as the clip on crystal heaters based
on a positive temperature thermistor (as used by all microwave
operators) would help towards getting the sort of accuracy needed.  On
my 10 GHz transceiver kept in the shack, a normal crystal with one of
these devices generally stays within 1ppm    A small packaged TCXO
source, now available from Farnell and other suppliers for a few pounds,
is specified at 1 or 2 ppm accuracy, but my experience of packaged TCXOs
is that they generally do a lot better than this if kept in a normal
reasonably constant indoor environment.   0.2ppm means 27mHz accuracy
which is comfortably on scale for an FFT span of 0.061 Hz (my slowest is
0.03 Hz) giving a resolution of 1mHz.  A heater on one of these could
probably do even better.

For a serious transatlantic beacon project, I have a 5 MHz source
available.  This is a laboratory standard oscillator but MUST be kept
running.  After a few days stabilising, it will stay within a few parts
in 10^-10 (but if switched off and on again, hysteresis reduces the
accuracy and it needs to be reset. Needless to say there is no hope
whatsoever of FSKing it !    I also have a DDS source controlled by the
5 MHz which can give a frequency setting resolution of 1mHz.    

These can be made available to anyone serious who wants to set up and
keep going a suitable high power beacon. 
I can also offer off air frequency measurement to anyone who needs a
beacon measured.

Andy G4JNT


>
>So, as Alan implies, the same QRSS techniques used with amateur QRO 
>(circa 1w erp) stations could be used to bridge the 'pond' at any 
>time provided conditions were reasonable. The obvious solution would 
>be to monitor a station running in beacon mode with QRSS keying.
>>From our experience this requires a lot of very close liaison with 
>the those involved on both sides of the Atlantic and up to now I have 
>not seen much evidence of this, although I would be the first to 
>agree that absence of evidence is not evidence of absence.


-- 
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.