Dear
J.B., LF Group,
I did some experiments a while back with different
ground configurations; like several other people, I found that once a
reasonable ground connection was in place, further additions to the ground
system only produce small improvements. Here the soil is clay, so fairly easy
to drive rods into, and probably quite low resistivity. I found that a single
1m ground rod resulted in the antenna system having of the order of 20 ohms
more resistance than multiple ground rods spaced a few metres apart around the
feed point. Adding ground rods one by one had roughly the effect of paralleling
20ohm resistors, and once I reached 6 or so, there was very little improvement
to be had by increasing the number – at this point the overall loss resistance
was still something over 30 ohms (for a 9m high, 40m long inverted L).
When I connected ground rods that were a
greater distance from the feed point (10m or more), very little current flowed
through these – although they might have a low resistance, the overall
impedance of these ground connections is increased due to the distributed
inductance of the connecting wire, so they are not very effective. I expect this
would apply to your well shaft. You can overcome this, and even up the current
distribution, by isolating the antenna feed from the local ground, and equalizing
the impedance of the ground connections, either by bringing them to a common “bus”
point with roughly equal lengths of wire to all the grounds, or by adding
balancing inductance in series with the shorter ground connections (these types
of arrangement seems to have been very popular for LF/VLF antennas in the 20s
and 30s). I tried this in several different configurations, but in spite of
successfully evening up the current distribution over a wide grounding area, it
only reduced the overall antenna losses by a further 10% or thereabouts. I also
tried a counterpoise supported about 2m off the ground, and covering all
available area (about 12m x 50m) – this also only reduced the loss
resistance by a few ohms.
I came to the conclusion that, once this
stage was reached, most of the loss must be due to dielectric losses caused by
the electric field of the antenna penetrating poorly conducting materials
around the antenna (the ground underneath, buildings, trees, etc.). This was
supported by the loss resistance decreasing with increasing frequency (it went
from around 300ohm at 10kHz to 16 ohm at 500kHz as I recall). Increasing
frequency leads to reduced antenna voltage for a given antenna current, so
lower dielectric loss. A similar antenna put up during an expedition in an open
field site, clear of trees and obstructions, had a somewhat lower loss
resistance (maybe 20 ohms), so I think the majority of the loss occurs in the
actual soil under the antenna. As Alan points out, this type of loss can be
reduced by increasing the antenna capacitance (more wire). It can also be
reduced by increasing the height (further from the ground = reduced field at
ground level) and keeping the wire away from trees and buildings. It would seem
that, at least in my circumstances, there is little to be gained by improving
the ground system. Commercial LF antennas are much higher and longer than
amateur ones, so will have much lower losses from this source – for these
antennas, a few ohms of loss in the ground system will make a big difference.
Cheers, Jim Moritz
73 de M0BMU
