Message-ID: <12844122.980174825746.JavaMail.imail@chilly.excite.com>
Date: Mon, 22 Jan 2001 06:46:58 -0800 (PST)
From: "john sexton" <computernetworks@excite.com>
To: rsgb_lf_group@blacksheep.org
Subject: LF: Larries sigs, Easygram, Spectran and Argo
MIME-Version: 1.0
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>


Hi All,

Last night I used Argo and saw Larry's signal from about 02.10 till 04.03.
Could have been longer but Argo crashed soon after, probably because of the
way it fills up the hard disk with photos taken every 1 minute, despite the
fact that at the dot width of 30sec, it takes 15 minutes to complete a
window width. It also seems to ignore the folder that you supply and puts
them whereever it pleases. Unfortunately I didn't actually see his call
sign, because the message faded out at just that point, but it was exactly
on the frequency that Larry stated and was clearly QRSS. The 30sec maximum
is a real problem when assembling the jig-saw like puzzle the following
morning.

After a frustrating week using Easygram to try to see Larry's signal, I
finally understand why I could not see his signal with any certainty and why
my observations did not match those of Andy and Jim.
The best resolution that Easygram can achieve is about the same as
Spectrogram 0.3 Hz and this is simply not enough. It is enough for VE1ZZ's
signals, but I now realise that these are much stronger. Larry is
considerably further to the West, perhaps 500 km or so.

The problem that I have found with Spectran in the past is that it can be
difficult to see the start and end of QRSS dots and dashes, whereas
Spectrogram manages to do this quite well. I mistakenly thought that this
would be good for Larry's signals too.
For really weak signals one needs Spectran or Argo (both from the same
stable and with the same 0.03 Hz resolution) at highest resolution. To avoid
the blurring, 3 sec dots will not do. 30 seconds might be good enough but 90
seconds as at present may be necessary. Longer will probably not help,
because of the QSB on the signal.

Now that Larry is transmitting QRSS(SSSS.....) it is possible to identify
the signal from its content, whereas just sending dashes as he was at the
beginning of the week, makes identification in the presence of QSB only
possible by knowing the exact frequency of the received signal and matching
it to the transmitted frequency, a non-trivial matter.

I have found that there are three main sources of error in determining the
exact frequency of a line on the screen:
Rx accuracy
Sound card/ Software accuracy
Measuring the position of the line

In my case, the Rx accuracy is limited by the 2.655 Hz tuning steps, which
means that one can be out by up to half this amount approx. 1.35 Hz, but it
is possible to calculate this error.
The Sound card/ Software accuracy I have found to add up to a further 0.5 Hz
to this (others have reported 2Hz). Interestingly I found Argo different
from Spectran by 0.4 Hz approx.
None of the packages allow measuring the position of the line to an accuracy
much better than 0.1Hz, although with a accurate scale one can improve this
to about 0.05Hz. Much better isn't really possible because of the thickness
of the lines and the non-linearity of screen images.

My method of calibrating combined Rx/Soundcard/Software package was to match
30 or so Loran lines, using a table of differences.

I make Jim's "two carriers on almost exactly 137.790" to be at:
137789.85 and 137790.05 Hz to an accuracy of 0.05Hz. I found them to be
fading and blurring but present most of the night of 21/22 January.
Jim reported a signal on 137787.5. I saw this too but it is not Loran. The
Loran line is at 137787.8473.
I can see the Loran line as well when the loop is orientated N/S. After
careful checking of frequenciesI am now sure that I am not seeing Loran
lines other than from the Lessay/Sylt chains.

Some of the lines reported by Andy appear to be offset in frequency by a
small amount, when compared with the Loran lines.

I plan to use the new version of Argo, when it settles down and hopefully
the problems with capturing images have been sorted. At the moment this is
the only tool which can do the job.

The map of this part of our band, that I have created will be posted on my
web site.

73, John, G4CNN
e-mail: computernetworks@excite.com
url: www.g4cnn.f2s.com


_______________________________________________________
Send a cool gift with your E-Card
http://www.bluemountain.com/giftcenter/