Message-ID: <3C2F2B57.94946C99@netscapeonline.co.uk>
Date: Sun, 30 Dec 2001 14:57:27 +0000
From: "gii3kev" <gii3kev@netscapeonline.co.uk>
Organization: Netscape Online member
MIME-Version: 1.0
To: rsgb_lf_group@blacksheep.org
Subject: Re: LF: Band full ?
References: <122.9e3b3cb.29606ae7@aol.com>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>


MarkusVester@aol.com wrote:

> Hi transatlantic gang,
>
> Band full? No, I believe there is plenty of space, it's just that we may have
> to move a little closer: QRSS or DFCW watched with 10 mHz resolution could
> leasurely be spaced at 100 mHz intervals. As the level differences between
> different Eu sigs seen in America will hardly exceed 15 dB, spectral widening
> during the keying is not a problem. Propagational doppler spread should also
> be less than 10mHz (though there has been one unexplained observation of wide
> doppler on a ZL-P29 LF path, not sure whether this was real).
>
> So it all becomes only a question of frequency accuracy and stability. Some
> stations (e.g. BMU, FTC, YXM and hopefully myself) are within 10mHz of stated
> freq (10^-7), which requires a calibrated OCXO or a very good TCXO. All
> others I have seen lately apparently use standard <1ppm crystal oscillators
> and have drifted less than 50mHz, which still is good enough. Only those rigs
> generating LF by mixing two independent HF xtals are probably not stable
> enough for real narrowband work.
>
> 7fsk must be regarded as a special case. In my opinion, much narrower tone
> spacings (20 to 50 mHz) should be preferred, and could even improve
> legibility because the sequence can be connected visually much easier.
>
> Though I don't particularly favor the thought of "frequency ownership" (still
> the anarchist in me, at age 43?), I agree that allocations will make life
> easier, and the receiving side can associate T-grade traces to their owners.

I do not favour frequency ownership either and random frequency selection for the
mode within a subband is to be preferred. Let the monitoring station identify the
information being sent like call etc
and not be dependent on frequency identification. With my recent first
transatlantic qso I picked an unoccupied frequency and transmitted and was
identified by a number of stations in both the USA and Canada and in Europe. The
identification was on callsign and other info that I sent and not frequency
dependent. With a frequency dependent system, then anything seen on that
frequency would be attributed to a particular person, especially under marginal
conditions, and this could lead to misinformation.
de Mal/G3KEV


>
>
> So some 30 stations could work in a 60s-Argo slot in parallel. (Btw I do
> prefer the "90s slow" setting with equal frequency resolution.Temporal
> resolution is anyhow limited by FFT bandwidth, the slower display simply
> saves screen space.) And if that were not enough, we might eventually have to
> ask Alberto for a higher window.
>
> 73 es have a nice sunny sunday
> de Markus, DF6NM