First I attempted to measure the efficiency of a small test loop, 6
turns and 66 cm diameter. It was made from a 12.1 meter piece of 729-strand
Decca litz wire (same that Stefan used for his famous loading
coil). Resonating with 23 nF styroflex gave a Q of 400, ie. 0.12
ohm series resistance. Theoretical radiation resistance (above
ground) would be about 12 nanoohms, predicting an efficiency of about
-70 dB. Then I tuned to 138.83 kHz and measured -61.0 dBm available receive
power from DCF39. Assuming a fieldstrength of 3 mV/m here, an idealized lossless
loop or Marconi should deliver +8.2 dBm. Thus efficiency of the test
loop was -69.2 dB, in good agreement to theory. This shows that there is
little attenuation of magnetic fields in the walls and the roof,
despite the thin aluminium vapor barrier above me.
To improve efficiency and power capability,
I then dismantled the little loop and reused the litz wire for
a large single-turn loop. This is about 3.5 m wide wall-to wall, 2.5 m
high floor to ceiling, oriented east-west for
best signal towards Heidelberg. The loop was resonated with
83.3nF, made of 10 pieces 33nF / 2kV FKP-1 capacitors, in 2s-5p
configuration. At 137 kHz, each capacitor is thermally limited
to about 6 A rms (30 A total). As
expected, both the inductance and the Q factor were lower (Q=72, 0.19
ohm loss). Presumably the extra 70 milliohms were contributed by
absorption in steel reinforcement in the floor and walls. The radiation resistance of an 8.8 sqm loop should be
about 0.23 microohms, thus predicting 1.2 ppm (-59 dB)
efficiency.
Before the on air tests, I
had calibrated the DK7FC grabber receiver (180 km) using my normal LF
Marconi, which has a known -30 dB efficiency. A 136.17 kHz
carrier from a signal generator (13 dBm, 20 µW radiated power) produced a
level of about -89 dB in Stefan's Eu window. During the morning hours,
noise was a low -105 dB in Heidelberg.
The big loop was then matched to 50 ohms using 16:1
turns on a ferrite ring (AL 4 uH). Fine tuning to 136.17 kHz was achieved
by 0.44 uF in series to one of the cap branches, and by slightly deforming the
loop geometry. Then I carefully applied power, expecting fireworks at any
moment, or all sorts of unwanted EMC effects due to the magnetic nearfield. To
my surprise, none of this happened. Going up to full 200 watts silently
produced 32 A in the loop, with no adverse effects on electonics or radios
in the vicinity. Only the small ferrite
transformer became hot after a short while. Recalculating core
magnetisation (6V rms, 0.5 cm^2) gave 0.2 tesla peak
which is way too much. This was easily mitigated by going to 32 turns
primary and 2 turns secondary.
For the on air-tests, I had inserted 100
kohm into the Marconi uplead to eliminate possible coupling and reradiation.
Despite the higher afternoon QRN, signals from the loop were immediately
received in Heidelberg, at -78 dB and 16 dB SNR. Thus radiated
power was about 0.25 mW, and efficiency -59 dB, just
as predicted. To double check, I temporarily reconnected the Marconi and transmitted a
dash with full power. Due to slightly lower load impedance (40 ohms) the TX
now produced 250 W (ie. 0.25 W EMRP). As expected, the signal in
Heidelberg indeed went up by 30 dB.
During the following hour, SNR was actually
good enough to marginally read a slow Hell transmission from the indoor antenna.
More pics are at
Well, so what is this good for? Transmit during a
thunderstorm? Maybe there's not much one can do with this kind of power,
but the experiment was certainly instructive, and good
fun.
Best 73,
Markus (DF6NM)
