Hi all,
Now i had a closer look to the scope traces i photographed last weekend.
For the 900m long wire the phase is 42 deg at 8270 Hz. That would mean
that the ground loop has an L of 3 mH if the resistive component is R =
175 Ohm. So i would need 123 nF to series resonate the loop.
The speaker cable should arrive in 2 days so i can do a next test next
weekend, this time with much lower losses. 1 A antenna current is the
goal! Ah and i like to try 2.47 kHz and 1.67 kHz and 1.57 kHz and 0.97
kHz!!!
I even modified our expensive Rohde & Schwarz scope. It's internal power
supply has an output voltage of (guess!) 12.0 V. So i spent a series
diode and now i can run it on batteries, just a normal '12 V' 7 Ah lead
gel accu. It consumes 2.85 A during operation, quite much. With the
scope i can accurately measure the phase and capture the screen. So this
will be another improvement.
Well, 3 mH at 8270 Hz. If this is true, then the loop size can be
estimated. Assuming a simple model of a normal single turn loop with a
rectangular cross section area, where one side is known to be 900m, the
depth of the current into the soil can be calculated. I found the
formula from ON7YDs website
http://www.strobbe.eu/on7yd/136ant/#TransmitLoop and build an Excel
table where i can tune the parameters until they appear reasonable. The
results show something like a loop size of 108000 m^2 and the current
flows back in a depth of 120 m. I'mnot sure whether this is the right
order of magnitude at all :-) The excel sheet also calculates the
radiation resistance (210 uOhm) and efficiency of the antenna (-59 dB)
but this appears to be quite optimistic...
73, Stefan
Am 30.07.2018 um 21:31 schrieb DK7FC:
Hi VLF and ULF friends,
In the attachment you can find 3 spectrograms from transmission of
the last weekend. For the 8270 Hz and 2970 Hz transmission i took
the data of the 900m ground loop, 5170 Hz is from the 450m ground loop.
The 8270 Hz signal was very strong even in 3.8 mHz ('DFCW-600').
Some real QRSS-60 would have been possible!
The spectrograms also show that there was no trace on the frequency
of interest before and after the transmission (no traces from the
right channel ( 1 PPS) ).
A few minutes ago i built the switchable network for resonating the
antenna properly. And i actually ordered 1000 m of that 0.75 mm^2
loudspeaker cable. Thus, in the next experiment i think the Q of the
antenna will be much higher and it will be possible to tune it
accurately. Then, the tuning capacitance will tell us what the
ground loop inductance is! And then, we can calculate the depth
where the current is flowing backwards in the ground, asuming a
simple model in first steps.
Next, we can see whether the inductance or depth stays constant at
lower frequencies (hopefully it will increase!) by resonating the
antenna at 2.97 kHz for example.
BTW i measured the DC resistance of the wire i used last weekend, it
is 116 Ohm, so it agrees well with the calculated 120 Ohm...
The new wire should have 21.4 Ohm only.
The big advantage of this antenna is that no high voltage and large
coils are needed. Usually the voltage is the limiting factor for a
VLF or ULF TX antenna. But not here. Here it is the power of the PA.
The antenna could easily handle 2 kW, which means 5.16 A, no problem
for the wire. And it would be a 20 dB stronger signal than last
weekend...
73, Stefan
Am 29.07.2018 23:11, schrieb DK7FC:
As a next result from today's experiment, here is a spectrum peak
of the 2970.005 Hz transmission. I reached 20 dB SNR in 180 uHz!
This far more than expected, in July anyway.
It is the first ULF transmission (by amateurs) from an earth
electrode TX antenna, detected in the far field.
Spectrograms will be produced soon...
73, Stefan
Am 29.07.2018 16:01, schrieb DK7FC:
Hello Roman, VLF,
I transmitted pure carriers each time. It was just a first test.
Today i've done a second experiment, this time using a 900m long
wire! So the wire should have about 125 Ohm DC resistance. I
measured 220 mA DC at 38.4 V, so the overall loss is 175 Ohm. Very
interesting; the ground loss resistance stays at 50 Ohm although
the distance between the electrodes has doubled!
If i would use the 0.75 mm loudspeaker cable then i can reduce the
losses by 100 Ohm which is more than 50 %, so i will gain 3 dB at
the same output power! Then with a switch mode PA having nearly
100% efficiency i may gain 2 dB more (mmy lonear mode PA was not
well matched today and quite warm). So maybe i can reach 1 A
antenna current with just 75 watts??!!!
Today i tuned to 550 mA antenna current again, at 8270 Hz. With
the scope, i measured the phas, it was more inductive than
yesterday. I actually found a 1 uF MKP-10 cap in my car and
switched it in series to the antenna. This improved the phase
slightly. Then i also found a low pass filter for 137 kHz, the
pypical pi configuration, i.e. it acts as 4*22 nF in parallel here
on this frequency. Switching this in series leads to a low
current, so the C is to small. So for the next experiment i'm
preraring a switchable C network (47n, 100n, 220n, 470n, 1u, 2.2u,
4.7u). It will be even more necessary when i lower the wire
resistance (higher Q).
Well, today the QRN was much lower, especially for the 2970.005 Hz
transmission period. Yesterday this was totally buried in the
noise on my RX on the tree. But today! I transmitted another 90
minutes with much lower QRN background. I already have a clear
spectrum peak but i like to try to improve it a bit more before
presenting it here.
Todays carrier transmissions:
8270.000 Hz : 08:12...09:33 UTC
2970.005 Hz : 09:42...11:15 UTC
Sorry for the confusing email ;-)
73, Stefan
|
