The HV triodes are a good idea (PD500?).
Maybe a GPS-synchronized rotary spark gap?
VY 73
Jacek / SQ5BPF
On Sun, 10 Mar 2019, Wolfgang Büscher wrote:
Date: Sun, 10 Mar 2019 11:05:20 +0100
From: Wolfgang Büscher <[email protected]>
Reply-To: [email protected]
To: [email protected]
Subject: Re: LF: Re: Re Loomis? & ... 12.47 Hz
Greetings,
How about using an ancient high-voltage "ballast" triode, as used in colour
TVs ? They may emit some X-rays when fed with anode voltages above 30 kV (no
risk, no fun) but they are very rugged. I have rescued some of them from the
junk because of the nice large anode gap. They always sat in a large metal
cage near the flyback transformer, along with the HV rectifier tube.
73,
Wolf .
On 10.03.2019 09:58, Jacek Lipkowski wrote:
I haven't tried this with 300m kites, but even ordinary low-band dipoles
can charge quite quickly if they are high enough, so 100uA seems to be a
good approximation. This is already comparable with the 170uA you're
getting at 12.71Hz now (and which will be lower at lower frequencies).
The only problem is the switch, but a high voltage vacuum relay (or a few
in series with a piece of fiberoptic for controlling each of them) should
be sufficient upto a few Hz.
Also note that the more charge in the atmosphere, the more ERP you get. But
it also gets more dangerous. No risk no fun on the "loomis band" :)
VY 73
Jacek / SQ5BPF
On Sat, 9 Mar 2019, DK7FC wrote:
Date: Sat, 09 Mar 2019 22:31:41 +0100
From: DK7FC <[email protected]>
Reply-To: [email protected]
To: "[email protected]" <[email protected]>
Subject: LF: Re: Re Loomis? & ... 12.47 Hz
Hi Jim,
I hope you don't mind that i'd like to share the email with the reflector,
because i've a thought that might be interesting.
In my view, the Loomis experiment it is rather the detection of a changing
current (charge per time) on the RX site. The changing current is
coming from a change in the static electric field, caused by the
shortcutted 'TX' antenna. Something like a current divider.
In 2010/2011 i've done VLF transmissions on my own, using a 300m vertical
kite antenna (having a special licence for that altitude). The
antenna capacity was about 1.5 nF. During an experiment in the summer time
there was a short moment when the vertical wire was floating. It
quickly charged up to some kV, which was quite noticable when i catched
and touched the wire then!! Since that time i carefully kept the wire
grounded during such experiments.
So, it means that the wire charged up, so there must be some continuous
charge flowing onto the wire and, if the wire would be grounded
permanently, you could probably measure a more or less stable current, i
guess it would be some 100 uA.
Now imagine someone else would rise a grounded kite in a few meters
distance. This would certainly affect the current flowing in my kite wire.
The farer both 'antennas', the lower expressed the effect will be and the
higher the antennas, the stronger it will be expressed.
I think the effect would be much better expressed by measuring the voltage
across a 1 MOhm resistor instead, which could be done by using a
scope and some overvoltage protection!
Actually an interesting question: In the summer time, which DC voltage
could be measured over a 1 MOhm resistor when connecting to a large E
field antenna and ground?
And, a next step: If i let my antenna charge up (floating) and then
discharge it in exact time intervals, say each second, then i should see
something at 1 Hz on a suitable receiver. This would already come close to
the experiment i'v done. I'm just replacing the 'natural charge
source' by a high voltage power supply and modulate that voltage (with a
sine wave, not rectangular).
So, to answer your question, i think that Loomis experiment was not
dedicated ELF, it was rather a broad-band spectrum that was radiated,
since
the charged antenna was discharged immediately. For a real ELF
transmission i would say that the carrier frequency has to be at ELF, not
the
modulating frequency. OK here you might say the carrier frequency is 0 and
it is AM modulated...
Try to repeat the experiment! Use smaller antennas and shorter distances.
Could be interesting :-) Rise two 10m high wires in 10m distance in
an open field. Connect one of them to a scope (1 MOhm input resistance),
protect the input with a glow lamp. Keep the other wire floating.
Select 1 second/div. If there is a thunderstorm coming and you can see a
rising DC level on the scope, then do a shortcircuit on the other
wire. I bet you will see the voltage dropping on the scope.
73, Stefan
Am 09.03.2019 19:10, schrieb James Hollander:
Hi Jacek and Stefan, I?d like to suggest that while I can?t say
for sure there weren't ELF frequencies received in the Loomis
experiment of 1866, I?m hesitant to reach the conclusion ELF was
used by Loomis because of the following questions.
1) If the transient current that flowed when Loomis? transmitter
circuit was closed probably lasted only a few milliseconds, wouldn?t
the modulation frequency content exceed at least the upper ELF boundary
30Hz as impressed on the ?carrier??
2) With a 600? long TX antenna and only a galvanometer fed by
similar height RX antenna, wouldn?t any radio waves that might have been
received be shorter than 10x the wavelength for which a 600? TX antenna is
a quarter wavelength? 10x(600?x4)=24000? or about 8km. If the
wavelength is less than about 8km, wouldn?t the ?carrier? frequency
content exceed about 37 KHz?
3) Nevertheless, one might say, if galvanometer deflected
temporarily in Loomis? system, it must have detected some near-DC content
unless some nonlinear element were in the receiving circuit. If I
Fourier Transform a damped DC transient, what is the frequency
content?
4) If there were DC transfer, wouldn't we say it's in the nature of
a current charging an atmosphere-ground capacitance through the
ground resistance, not radio in near field ELF? Or should we say the
meaning of ?frequency? in this case becomes so fuzzy that Loomis both
did and didn?t use ELF?
5) If indeed Loomis communicated any ELF, can?t one still radically
distinguish the 12.67 Hz experiment at DK7FC as involving a
very narrow band continuous wave with 227 hours integration of this
continuous wave to detect it and make it separable from other waves
that could be generated in the ELF band?
I?m new to the subject of ELF, and would appreciate any words of
wisdom you?d like to give.
Vy 73, Jim Hollander W5EST
-----Original Message-----
From: Jacek Lipkowski <[email protected]>
To: rsgb_lf_group <[email protected]>
Sent: Sat, Mar 9, 2019 4:28 am
Subject: Re: LF: RE: RE: Almost touching the ground... | 12.47 Hz
Actually a similar experiment to Stefan's has been done already, and at
much lower frequencies (almost 0Hz :):
http://aerohistory.org/Wireless/loomis.html
In this case the power supply is from the cloud electric field and
probably had quite a few more kV than Stefan's.
Please note the DX distance.
VY 73
Jacek / SQ5BPF
From: DK7FC <[email protected]> To: rsgb_lf_group
<[email protected]>
Sent: Tue, Mar 5, 2019 12:50 pm Subject: ELF: Almost touching the
ground... | 12.47 Hz
Hi ELF friends, During the last 2 weeks i've done another
experiment on ELF, this time
on 12.47 Hz, the 24 Mm band (wavelength 24057 km). Again i've crossed the
local distance of 3.5 km. That's the lowest frequency i've ever
been and it feels like i can see the ground already :-) The dimensions
of everything down there are extreme. I've integrated 227 hours
of a carrier transmission into one spectrum peak, it is shown in the
attachment in 1.25 uHz. This carrier could have transferred an
EbNaut message of nearly 100 characters.
The ERP was 50 attowatt or -163 dBW and the antenna current was 170 uA
only, despite about 5 kV antenna voltage.
I'm now trying to put a step below 10 Hz but the RX antenna becomes less
efficient with each Hz. 73, Stefan
|