Message-ID: <22624.200101211519@gemini>
From: "James Moritz" <j.r.moritz@herts.ac.uk>
Organization: University of Hertfordshire
To: rsgb_lf_group@blacksheep.org
Date: Sun, 21 Jan 2001 15:22:57 +0000
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII; format=flowed
Content-Transfer-Encoding: 8bit
Subject: LF: Receiving VA3LK
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>

Dear LF Group,

I monitored VA3LK's frequency again using the same set-up as 
before, and got rather similar results as in the previous two nights. 
The QRN level seemed rather high, adding about 5-10dB to the 
noise floor, however, Larry's signal also seemed to be a few dB 
stronger, so copy was still fairly good. As before, there were 
several periods of good copy, interspersed with fades where only 
odd dots or weak traces of signal were visible. The periods of good 
copy were 2220-2240, 0040-0150, 0225-0320, 0345-0400, 0715-
0745, 0825-0840. The signals seen around dawn are getting quite 
weak, but the noise level also drops off at this time. I have yet to 
see anything before about 2200 in the evening.

>From the other reports that are coming in, it would seem that most 
people equipped for QRSS operation should be able to detect 
Larry's signals; the important thing compared to the "faster" QRSS 
is to obtain high resolution and accurate frequency calibration of 
the receiver, so you know where to look for the signal. Here are 
some details of the methods I am using:

I am currently using a W&G SPM19 level meter with external 
homebrew preselector and demodulator, and DL4YHF's "Spectrum 
Lab" spectrogram software. This setup has the advantage of very 
good frequency stability, but I think most synthesised receivers 
would be able to maintain the +/- a couple of Hz per night which is 
probably required. The frequency error of 2Hz on the soundcard 
seems to be something of a standard.... the easiest way to 
calibrate the frequency within a fraction of a Hz is to look at the 
Loran lines; in my QTH, (SE England), there are lines at 
137.78785, 137.79171, 137.79528 kHz. G4CNN has produced 
lists of frequencies from other Loran chains. I believe Loran uses a 
caesium clock for a reference, so they should be pretty accurate. 
They had to come in handy for something! The transmit frequency 
does not seem to have drifted measurably.

The spectrogram software needs to have higher resolution than the 
0.3Hz that seems to be optimum for 3s/dot QRSS. Spectrum lab 
goes down to 0.042Hz (11k sample rate, divided by 4, 65k point 
FFT, average of 2 FFTs selected in the "spectrum" menu). It also 
has an input monitor 'scope display, so you can check the A/D 
converter is not being overloaded by the QRN. The brightness and 
contrast controls can be set for high contrast (about 20dB for the 
full colour spectrum), which is good for readability, but less contrast 
is better for monitoring the general signal level. The spectrum width 
is set to 5Hz or 10Hz - another reason to calibrate the frequency 
carefully. I find that the receiver generates a weak spurious signal 
at it's center frequency; it is worth checking for this kind of thing 
during the day. I set the timebase for about 35 mins per screen 
(can't remember actual figure), and setting the image capture to 
once every 30 minutes means I can go and get some sleep 
sometimes!

I have been using a large single turn loop antenna, which I 
mentioned before on the reflector, which seems to work quite well 
and gets rid of most of the Loran lines. But I think the TX inverted L 
would work well too.

Cheers, Jim Moritz
73 de M0BMU