Dear Scott, Roger, LF Group,
I'm afraid the gyrator is a non-starter - there are different types of
gyrator, but the concept is a feedback circuit that, when you apply a
voltage to an input node, drive a current into that node with appropriate
magnitude and phase to simulate the impedance you want (more usually an
inductance). Usually these are small-signal circuits, used in active filters
and similar applications, but no doubt you could make a high power version.
But the out-of-phase current in the simulated reactance is being supplied by
an amplifier - so you are getting rid of the reactive load current in one PA
by adding another, much bigger, PA that has to deliver the (much bigger)
reactive current instead - might as well design the original PA to handle
the reactive load in the first place...
I don't think the tuning problems of a loop at 9kHz would be as severe as
Scott expects. The frequency and so the loop reactance is about 15 times
lower than at 136k, so also for a given loop current the capacitor voltage
is also 15 times lower - for the 10m x 10m loop, the voltage would be about
74V RMS at 9k, compared to 1.1kV at 136k. The reactive power rating
requirement of the capacitors is thus also 15 times lower. I have recently
been experimenting with a 136kHz PA design that has about 100V, 30A
circulating in a 0.4uF capacitor made up of 8 small polypropylene units in
parallel without any great drama - if anything, 9kHz would put less stress
on the capacitors. As for Q requirements, the loop resistance will be
reduced, but probably not by a factor of 15 - the skin effect makes R
proportional to sqrt(f) at high frequencies, so falls off slower than the
reactance as f decreases. Also, R will level off at VLF as the AC resistance
approaches the DC resistance. Going with the previous calculation I did,
loop reactance is about 2.3 ohm, loss resistance about 0.1ohm, so Q is only
23 - quite modest compared to 136k antennas.
I don't think a small loop of the type discussed is viable for the kind of
relatively long-range tests that have been going on lately is concerned. But
a vertical of a similar size is a difficult thing to drive at 9kHz, with
loading coils in the range of 1 Henry or more, and loss resistances that
seem to be creeping into the kilohm range. So if you are attempting to
radiate powers of the order of a uW, using only a few 10s of metres of wire
in total (as per G3XBM's original enquiry), the loop might actually be a
better bet. To find out if it really would, some 9kHz impedance measurements
would be a good start.
Cheers, Jim Moritz
73 de M0BMU
|
