Message-ID: <5.1.0.14.0.20030120171137.00b158e0@gemini.herts.ac.uk>
Date: Mon, 20 Jan 2003 19:14:42 +0000
To: rsgb_lf_group@blacksheep.org
From: "James Moritz" <j.r.moritz@herts.ac.uk>
In-reply-to: <004401c2c0a2$25622860$9801a8c0@gigacom.com.br>
References: <23.2aadfb3f.2b5c5522@aol.com>
MIME-Version: 1.0
Subject: Re: LF: Loops x Q
Content-Type: text/plain; charset=iso-8859-1; format=flowed
Content-Transfer-Encoding: 8bit
Sender: <owner-rsgb_lf_group@blacksheep.org>
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org

Dear Marcus, LF Group,

It seems to me this is quite a large loop, and will generate a large output:

I think across an untuned loop V = 2.1e-8.f.N.A.E

where N = turns, f is frequency in Hz, A is area in m^2, E is f/s in V/m

For your loop, and a f/s of 1uV/m, this gives 0.21uV @ 400k, 0.85uV @ 1600k.,
With a Q of 200, the voltage at resonance will be increased to 42uV @ 400k, 
170uV @ 1600k

The loop noise depends on the RX bandwidth, but in 10kHz BW it will be 
roughly 6uV @ 400k, 12uV @ 1600k. So for 1uV/m, SNR would be 17dB @ 400k, 
23dB @ 1600k. The JFET noise will be negligible. I think most of the time 
the atmospheric noise level in 10k BW at MF will be well above 1uV/m, so it 
should be more than sensitive enough. Reducing the Q to 100 due to 
resistors or varicap will make the output voltage decrease by 6dB and loop 
noise decrease by 3dB - still probably OK. Reducing the RX bandwidth will 
improve the SNR, of course.

The tuning is a problem - 4:1 tuning range needs 16:1 capacitance 
variation, including the stray capacitance of the loop. If L is about 446uH 
from the table, this means 350p - 22p. The stray capacitance of the loop 
may be greater than 22p, so you may not get the full tuning range. Using a 
varicap diode, it is unlikely the minimum- to maximum capacitance ratio 
will be high enough. it may be necessary to reduce the number of turns and 
use a larger tuning capacitor.

The other problem is that local stations will give very large signals - eg. 
10mV/m from an 1MHz MF broadcast station will result in about 1V across the 
loop at resonance, which will probably lead to severe intermods from both 
the preamp and the varicap. You could connect the preamp across only 1 turn 
instead of all 9 - this would still give several uV out for 1uV/m f/s, more 
than enough for a decent receiver. In fact, since the loop output is so 
large, you could probably eliminate the preamp by using a single turn 
coupling loop to a 50ohm RX input. The air-variable tuning capacitor would 
also be a good idea for this reason, unless you live a very long way from 
any broadcast stations!

Hope this is helpful,
Cheers, Jim Moritz
73 de M0BMU


At 13:35 20/01/2003 -0300, you wrote:
>Hi ...
>
>After read an article last year (Wireless World) about loop antennas X DCF39
>calibration, I constructed a octogonal RX loop, #16 awg wire, to cover from
>400KHz to 1.6MHz.
>
>There´s 9 turns, 5mm spacing, 1.9 meter diameter. About 55 meter of wire.
>
>This is the first version. Later I´ll build another one to 136KHz.
>
>Acording with this article, I should expect a Q=200 ... 220 from such loop.
>
>I considered the NTE618 to tune it and a push-pull pre-amp (J310) with gates
>conected to the loop ends. No resistors there. The article consider that for
>such loop at this frequency will have something like .20uV of thermal noise.
>The resistors will add some extra noise there so I removed them.
>
>But the Q=210 ... and the NTE618 Q is 200.
>
>Is there any gain If I use a dual-gang air variable driven by a step motor +
>reduction in place of the NTE618?