DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws;
  s=s1024; d=btopenworld.com;
  h=DKIM-Signature:Received:X-Yahoo-SMTP:X-YMail-OSG:X-Yahoo-Newman-Property:Message-ID:From:To:References:In-Reply-To:Subject:Date:MIME-Version:Content-Type:Content-Transfer-Encoding:X-Priority:X-MSMail-Priority:X-Mailer:X-MimeOLE;
  b=zHnDi3jTydItJLNcXxmII7vez6aWeaSEc4teeLzbJaiOahFKTIxfsTmfDCePKtyK0Rj8UhkaifXz6hXhFOdoBItJkrq2S4iCLmrOocUmZ4bOcHPqLSed5aTfEqc7E8lE21oM4UDBvsAsAEK8E36+DkOTEnnpqHH3W7WdrtdI6sw=  ;
Message-ID: <C70B23C4493F48CB8D9A21AB9CA9E69A@JimPC>
From: "James Moritz" <james.moritz@btopenworld.com>
To: <rsgb_lf_group@blacksheep.org>
References: <AANLkTimy87NkJ3VXf4qeQ3NMehA_6HzAe6mHBJ6M_Sts@mail.gmail.com><4D72471D.30904@talktalk.net><AANLkTincU_L7bHAKNmqLWuOmmMv114MY0=FAjKCyP=0m@mail.gmail.com><4D72695E.3090100@talktalk.net><AANLkTikd83tysKokWzpisRYBQuYc2q0-chzhn1B3EfmP@mail.gmail.com><4D7291BD.8080301@talktalk.net> <AANLkTi=GLAyB-fmKj-LvVDERpx5M-qy1P6GriAPOCVij@mail.gmail.com>
In-Reply-To: <AANLkTi=GLAyB-fmKj-LvVDERpx5M-qy1P6GriAPOCVij@mail.gmail.com>
Date: Sat, 5 Mar 2011 22:47:59 -0000
MIME-Version: 1.0
DomainKey-Status: good (testing) 
Subject: Re: LF: Sub 9kHz antenna tests.
Content-Type: text/plain;
	format=flowed;
	charset="iso-8859-1";
	reply-type=original
Content-Transfer-Encoding: 7bit
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org


Dear Eddie, Gary, LF Group,

The loop is tuned, but is very low Q, and has more of a low-pass response 
really. It was done this way partly in order to obtain a wide bandwidth 
around 9kHz, which allows effective use of noise blanking or clipping 
(implemented using the DSP facilities of Spec Lab) to reduce the impact of 
QRN impulses. It also avoids the need for re-tuning, which is very 
convenient. At the same time, it provides a decent amount of attenuation of 
signals above the VLF range, which would otherwise tend to overload either 
the preamp or the sound card input. The frequency response is defined by the 
loop inductance, the shunt capacitor at the input, and the input resistance 
of the preamp stage (about 300ohms, due to the series/shunt negative 
feedback). The low Q means the loop voltage is small, so a fairly low-noise 
preamp is needed. The band noise has always been >10dB above the internal 
preamp noise in practice, so this is satisfactory. The overall gain of the 
preamp is about 60dB at 9kHz, and this was chosen as a compromise - it is 
large enough so that the band noise swamps the internal noise and QRM in the 
sound card input, but small enough so that the utilities and QRN do not 
overload the sound card.

You can get much higher signal voltage out of the same loop antenna by 
changing it to a high Q resonant design. I guess with the relatively high 
impedance load of the sound card input (several kohm), you might get a Q 
between 10 and 100. This would give a voltage gain of 20 - 40dB, more if you 
made a really low-loss loop and optimised a tapping point for the sound card 
input connection. But the bandwidth would also be greatly reduced, probably 
to only a few hundred Hz or less, and this is a bad thing. The narrow 
bandwidth "stretches" the QRN impulses making clipping / blanking less 
effective. Also, accurate tuning is then required.

I suppose one motive for having such a high Q loop would be to eliminate the 
preamp. This might work, depending on how noisy the sound card codec chip 
and the PC it is attached to are. My experience has been that this varies a 
lot between different PCs and frequencies, and which audio input you use. 
E.g. the laptop I usually use has a "Mic" input with considerably higher 
gain than the "line" input. But for some reason, the SNR is worse even 
though the signal level is higher when using this mic input. Also, the input 
noise level may vary depending on the source impedance of the antenna 
connected to the input. Since the preamp components only cost pennies, and 
allow you to get ample SNR as well as several dB improvement through the use 
of noise clipping and "no-tune" operation, this seems a better and simpler 
solution.

When comparing noise levels, it is important to check what the noise 
actually is, especially since PCs and associated stuff are potentially 
powerful noise sources. Listening to the output around 9kHz using the 
SpecLab VLF SDR function is useful - you are hoping to hear the band QRN and 
probably assorted mains-borne noises, rather than a lot of digital hash from 
the PC.

Cheers, Jim Moritz
73 de M0BMU