Hi Dimitris, ELF,
Well, The day has started here, the carrier is still there, the antenna
current is stable, the tree has still enough solar energy, i have some
coffee here, it looks a bit blue on the sky, so it is a good start into
another good day on that planet :-) Good morning!
Well, the PA for this ELF project is quite different to normal RF
amplifiers. You know, at this frequency (but also als SLF/ULF/VLF, even
LF) the antenna acts as a pure capacitor. Mine has about 470 pF. At a
given frequency and capacity you will get a certain antenna current from
a certain voltage. Since the frequency is so low here, the antenna
current is below 1 mA, despite the nearly 5 kV rms. So the TX power is
very low, that's the first thing.
And how do you generate a sine of 5 kV at 23 Hz? You could use a mains
transformer for 10 kV at 50 Hz and then just go down. But where do you
get that from? There are some old voltage transformers on ebay, taken
from 20 kV power grids.
Like this one:
https://www.ebay.de/itm/MWB-REL20-24-50-2-125kV-90VA-SPANNUNGSWANDLER-POTENTIAL-TRANSFORMER-TRAFO/173246231511?hash=item285647b3d7:g:vqoAAOSwt21av7Vi:rk:1:pf:0
They are heavy and expensive and often do not even handle the required
voltage. The one above is way to expensive and only handles 20 kV/sqrt3,
and just at 50 Hz!
OK that would be fine for getting 6 kV at 23 Hz but if the experiment
works well and i like to try e.g. 9.7 Hz, then the game is over with
that transformer. It would be a large investment for just one experiment.
So i searched for another solution:
I'm using an old TV transformer (think they are called LOPT in english)
to produce a high voltage of about 2 kV. The frequency is about 25 kHz.
The output goes to a homemade Greinacher cascade
(https://en.wikipedia.org/wiki/Voltage_doubler#Greinacher_circuit) to
produce a higher DC voltage. In parallel to the output there is a 10
MOhm HV resistor. This resistor is used to discharge the cascade and to
build a voltage divider to measure the HV. The transformer is driven by
a usual SG2535 switch mode regulator (and 2x IRFZ48Z MOSFETs). So i can
regulate the output voltage (DC) of the cascade. The SG2535 has an
'error amp' input which allows to control the output voltage by an input
voltage. I then use a trim pot between the internal v-ref and gnd to set
an offset in the middle of the regulation range. Then, decoupled by a
4700 uF capacitor i can apply 23 Hz from the soundcard of my PC. This is
decoupled by a 5 VA mains transformer, to avoid EMC problems and to rise
the voltage level a bit.
So actually this is a DC HV supply whose output voltage is quickly
changed, at 23 Hz. So i can tune from 0...15 kV. The DC offset doesn't
matter and the AC component radiates a very very very very small signal,
as we can see!
While trying to keep there series loss resistance as low as possible in
a LF TX system, i switched 2x 10 kOhm in series to the antenna here!
This builds a low pass filter (17 kHz corner frequency, i just see i
could have used 100 kOhm in stead) for the switching frequency. At 325
uA antenna current, it just 'costs' 0.65 V voltage drop! The transmitter
consumes 12V*1.5A.
The parts you need to build such stuff is e.g. available at
https://highvoltageshop.com/ There are very cheap HV diodes and
capacitors. Very useful and rare!
So, the antenna is not tuned. For a tuned antenna i would need a coil
which has 102 kH (kilo Henry) My lifetime is not long enough to wind
such a coil :-)
73, Stefan
Am 07.02.2019 23:15, schrieb Dimitrios Tsifakis:
Hi Stefan,
that's insane! What's the purpose of the DC offset of 7.5 kV? Is the
antenna tuned (unlikley) or you are just applying the volts from a
voltage amplifier?
Hope the antenna is well insulated from people and nature as skin
effect won't save your bacon (unlike the higher frequencies).
73, Dimitris VK2COW
Στις Παρ, 8 Φεβ 2019 στις 3:10 π.μ., ο/η DK7FC<[email protected]> έγραψε:
Hi ELF ;-)
I'm glad to announce that since 15:46 UTC i am transmitting on ELF (the real
ELF, i.e. into the range of 3...30 Hz, ITU radio band 1 ) for the first time
ever.
Here are the parameters:
TX frequency: 22.970 Hz
Wavelength of that frequency: 13060 km
Antenna voltage: 4.8 kV rms (+ a DC component of 7.5 kV)
Antenna current: 325 uA
ERP: 600 aW, or 6E-16 W
Call it QRP!
The goal is to detect the signal on my 3.5 km distant tree grabber, whose E
field is now working again. Since the E field probe is not to far above the
tree, the effective height is not as much as it should ideally be, at this
frequency, so the RX is somewhat deaf. Anyway i can see the Schumann resonances
and the 60 Hz power grid. So i hope it will work.
A spectrogram which is centered on the frequency of interest in now shown at
http://www.iup.uni-heidelberg.de/schaefer_vlf/DK7FC_VLF_Grabber2.html (3rd
window). But maybe 424 uHz FFT bin width is even to optimistic! In a few
minutes we'll see more :-) It takes about 40 minutes until a peak could have
fully developed...
73, Stefan
|
