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=2aPvHXUSk+vo8Y4zZok9WZqo0SSBOOH+qwMkHl0V4NIcKBnvZY99MzuxAkiFSDCY7bqYSXMkg/cqRL9Ryl5P18/1K0Kt9pTLwkK/P4tt4O787kW/IdZfStntE1+mbGYWeVafbPPOcteD915qpK/lX76MSFdGhWW9ulpBT3AhWds=  ;
Message-ID: <1C7BCD9DEB29472F9BB467C4EE390840@JimPC>
From: "James Moritz" <james.moritz@btopenworld.com>
To: <rsgb_lf_group@blacksheep.org>
References: <008f01caf784$779a11d0$0301a8c0@your91hoehfy9g> <Pine.LNX.4.64.1005192058370.18919@paranoid.lipkowski.org> <4BF4FBF4.2020208@iup.uni-heidelberg.de> <4BF53399.4010403@iup.uni-heidelberg.de>
In-Reply-To: <4BF53399.4010403@iup.uni-heidelberg.de>
Date: Fri, 21 May 2010 02:10:29 +0100
MIME-Version: 1.0
DomainKey-Status: good (testing) 
Subject: LF: Re: Clipping or blanking/9kHz intermods
Content-Type: text/plain;
	format=flowed;
	charset="iso-8859-1";
	reply-type=response
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Content-Transfer-Encoding: quoted-printable


Dear Stefan, LF Group,

----- Original Message -----=20
From: "Stefan Sch=E4fer" <Stefan.Schaefer@iup.uni-heidelberg.de>
...>
In Spectrum Lab there are different methods for noise reduction (e.g.=
 of
lightnings). Those were discussed here in the last weeks. While
Jim/M0BMU preferes the clipper (using 0 dB above AVG if i remember
correctly), Markus/DF6NM prefers the noise blanker.
...>
I don't really have a particular preference; both seem to give a simil=
ar=20
improvement in apparent SNR, and both will introduce some intermodulat=
ion,=20
since they are inherently non-linear techniques. Your spectrograms see=
m to=20
confirm there is not a great difference... BTW, the Spectrum Lab clipp=
er,=20
where the clipping threshold tracks the average signal level, does see=
m to=20
give better results than simple fixed-threshold clipping, especially=
 when=20
other signals or narrow-band noise are present.

...>
I still have to reduce the IM in the first stage (2nd order low pass
filter at the output of the first stage does not better the situation)=
,
either by reducing the wire length or adding a series resistor or...hm=
,
will see (any ideas?).
...>
At my QTH, there are 4 MF broadcasters within 1km, giving combined pea=
k FS=20
of 10s of volts per metre, so I get these problems a lot... I find it=
 is=20
important to keep an open mind about possible sources of this kind of=
 QRM.

The first thing is - are you sure the intermodulation actually occurs=
 in=20
your active whip? Any length of wire acts as an antenna, and will ofte=
n be=20
connected to some non-linear device (e.g. mains cable connected to rec=
tifier=20
diodes), so there are an almost infinite number of places outside the=
=20
antenna circuit that the 9kHz IM product could be generated.

A prime candidate is the earth connection - if your antenna is high up=
 on a=20
building, the earth connection will be carrying considerable=20
broadcast-frequency current - I have found this often results in=20
intermodulation, perhaps due to some kind of electrolytic effects occu=
ring=20
at the actual ground electrodes. This effect makes my TX antenna unusa=
ble=20
for reception at many frequencies. Any mains power supply connected to=
 the=20
antenna preamp can cause a similar effect if the internal diodes,=20
regulators, etc. are not well decoupled at broadcast frequencies.

Another place where intermods can occur is the PC, which will typicall=
y have=20
long cables connected to it (mains, network, sound card etc...) again=
 acting=20
as antennas and carrying MF broadcast currents. Any intermod generated=
 in=20
something connected to the PC can be coupled into the sound card input=
s by=20
the ground loops that tend to exist inside the machine. For example,=
 I found=20
using my 9kHz antenna/preamp at home with a laptop with AC adaptor gav=
e a=20
similar 9kHz line to the one you see. I eventually found this was stil=
l=20
there with the preamp output disconnected from the sound card, but wit=
h the=20
earth connection present between the computer and the preamp. I assume=
 the=20
cables connected to the preamp were acting as an antenna, and flowing=
 to=20
ground through the PC and adaptor. The cure was to wind several turns=
 of the=20
AC adaptor lead through an EMC ferrite core to reduce the MF current=
 flowing=20
in it. For the DK7FC grabber, you could easily test if the 9kHz comes=
 from=20
the antenna or PC end of the system by un-plugging the optical fibre.

Then there are usually many other cables, or large pieces of metal, ne=
ar the=20
antenna, via which intermods can be occuring. The resulting 9kHz=20
voltages/currents will be picked up by your antenna. Moving the antenn=
a is=20
the easiest way to check this, or also to solve the problem, usually.

But then, if the FS is high, it may be that the active antenna is the=
=20
problem after all. In your circuit, it is unlikely that IM is occuring=
 after=20
the input stage due to the presence of the bandpass filter. You could=
 try=20
adding a source degeneration resistor to reduce the gain and improve=
 the=20
linearity of the input stage (say 100R, but you would also probably ne=
ed to=20
change the gate bias voltage to get a sensible bias current). I would=
=20
suggest putting about 47R resistor in series with the gate, to reduce=
 the=20
likelihood of VHF oscillation occuring, which can cause strange proble=
ms,=20
including intermodulation at low frequencies. Probing the drain with=
 an=20
oscilloscope would reveal what large signals are present. Also, the dr=
ain=20
resistor in the second stage needs looking at - 1k is rather too high=
 a=20
value for a 12V supply and a FET with zero gate bias and Idss of 10 -=
 25mA.

BTW, problems with high broadcast station field strength were some of=
 the=20
reasons leading me to favour a loop antenna design, for 9kHz and other=
=20
frequencies as well. It is much easier to include filtering directly=
 at the=20
antenna terminals, since the loop is much lower impedance. Also, groun=
d=20
connections are less of an issue.

Cheers, Jim Moritz
73 de M0BMU