I use my 40m/20m trap dipole as the radiating element for a "Marconi T" antenna. It is strung between the house and a tree at a height of about 27 feet. The 72ohm balanced feeder is strapped together and used as the vertical. The traps can be ignored at LF. A huge base loading coil is needed to resonate the antenna against ground. I have three ground stakes linked by thick copper wire running the length of the garden, I intend to add more radials to the system, but I want to get something working first !
I use a Siemens D2006 Selective Level Meter as a receiver. My FT847
tunes down to 136kHz but is very insensitive (won't even pick up Radio 4
Long Wave !). The D2006 works very well, and will work down to 10kHz. It
has 80Hz and 3.1kHz IF filters, and will resolve ssb and cw. With its big
meter calibrated in dB, and an attenuator calibrated in 10dB steps, it
makes it very easy to do comparative measurements.
A recent aquisition is a Seimens K2155 SLM - which is more modern,
and has a better filter, but I have yet to put it on the air. Click here for more info
The "Dalek" on the left of the photo is the base loading coil.
I originally wound a coil on the red bucket (120 turns of 7/0.02mm pvc
coated wire) but it was insufficient to bring the antenna to resonance.
The resonant frequency was found to be about 250kHz.
By finding the resonant frequency of the bucket using a
known value capacitor, I could calculate the inductance of the bucket.
This enabled me to estimate the capacitance of the antenna as 150pF.
I could then work out that I needed about 10mH of inductance to resonate
my antenna.
Using a short length of 6 inch dia. drain pipe as a former,
I wound another coil of about 90 turns.
This revolves within the red bucket to make a
variometer, with an inductance which can be varied between 2.3 and
4.2mH. This is most useful for fine tuning the antenna.
Another bucket was obtained and 156 turns of wire wound on it. The
combination of both buckets gives between 9.8 and 12 mH.
This was still not quite enough. A plastic drinks bottle was used as
a former to wind another 170 turns of wire. This light-weight coil was
suspended in a tree above the coil. All three coils together bring the
antenna to resonance nicely.
Both the bucket coils and the pop-bottle coil are based on designs
in the Low Frequency Experimenters Handbook, by Peter Dodd G3LDO, and
published by the RSGB.
A great many taps have been incorporated in the lower part of the
white bucket coil, which allows the receiver to be matched to the antenna.
The Antenna receives the DCF39 transmission from Germany at -55dbm
at the receiver input, during daylight, and the noise floor is at
-111dBm. Some QRSS signals were immediately received, although the
callsigns were not noted. Later in the evening PA0BWL was received using
conventional morse.
To start with I used a TDA2030 Audio amp. This device has a bandwidth of
140kHz and I can squeeze about 6 watts into a 2 ohm load. This is just
on a piece of veroboard - see left.
With great trepidation I connected up a BK electronics 100W audio
amp. Others have found this type of amplifier to blow output FETs and
fuses. However I have had no such problems. I have had 47.5 Watts out of
this amp yeilding an antenna current of 0.8 to 0.9 Amps. This
arrangement allowed a QRSS QSO with Jim M0BMU in Hatfield Herts. My
signal was also picked up in Holland, but no QSO resulted.
The matching network is a simple LC circuit, which transforms the 2
ohm output impedance of the amp to 75 ohms. The output impedance is
placed in series with the inductor and resonated with the capacitor. In
this arrangement the low series resistance becomes equivalent to a much
higher one in parallel to the tuned circuit. The impedance becomes Zout =
X^2/Zamp, where X is the reactance of either the capacitor or the
inductor (they are equal at resonance).
The signal source is a Levell RC oscillator type TG200DMP (which has
a fine frequency control). I hope to be able to use the tracking
generator in the K2155 SLM as an improved signal source.
There is a simple keying arrangement using a 12V relay and this can be linked to a QBasic program running on a PC to send QRSS morse, or to a straight key for conventional CW. For receiving QRSS I use a program called ARGO. QRSS morse is sent at very slow speed, 3 seconds per dot or slower. Using the PC and soundcard, a continuously scrolling spectrum can be displayed and the morse seen as white lines across the screen. It shows up even when completely inaudible in the loudspeaker. Screen shot at left shows transmission by G3XDV received at M0WYE - I think the multiple lines are caused by overload in the crystal IF filter in the D2006. (click the picture for bigger image)
Links
G3XYM - The World of LF: News,
Circuit Ideas
Argo, and other Free software may
be downloaded from I2PHD's site
G3NYK - Alan
Melia's Website, including History of The Rugby Transmitter
David Bowman G0MRF
Circuits and designs for LF equipment
Peter Dodd, G3LDO, Extra
info related to the LF Experimenters Handbook