The discussion continues to produce interesting ideas. At the risk of boring
some I would like to throw in some more information.

Ground resistivity is usually measured with a 4 probe technique. ( I used a
similar method for exploring silicon ingot doping, and hence resistivity,
also diffused layers, when I started work on transistors in the early 60s.)
The measurement is based on the classic Kelvin bridge technique for
measurring low value resistors (my first lab experiment at University) low
value 'shunt' resistors with 4 terminals are often refered to as having
Kelvin contacts. The connection is used to avoid the affect of contact
resistance if only two terminals were used. The resistance value seen would
be the required value plus two variable terminal contact resistances. This
is avoided by passing the measuring current through one or a pair of
terminals at each end, and measuring the voltage drop across the other pair.
Provided there is very little current drawn by the "voltmeter" the contact
resistance at the voltage terminals will not significantly corrupt the true
measurement.

The 4 probe ground resistivity measurement works similarly. A potential is
applied to drive a current between the outer probes and the voltage is
measured on the inner pair. Contact resistance is a major problem with this
kind of soil measurement and this technique avoids it nicely. Polarisation
with DC can still be a problem hence the use of either a reversing switch or
more often mains frequency current is common. Strangely enough the
penetration of the probes is not really important although it affects the
calculation if the penetration is of the same order as the spacing of the
probes. (For the archeologists see the probe length and spacing on the
things being heaved around by Time Team's geophys experts.....UK TV)  The
more important factor is the probe spacing as one may consider that the
reading is influenced by the ground to approximately the same depth as the
spacing used (so surface contact is adequate). Some measurements have been
made with spacing of many kilometers. My AVO test leads would be a pushed at
that distance !! Remember that the ground penetration at 136kHz can be
between 5 and 20 metres. John G3PAI and the "cavemen" work through 100s of
metres at 87kHz.

Although ground resistivity obviously correlates in some way with our aerial
loss resistance it is not anything we can do much about. Most of us are
stuck with the ground we choose, usually for other than amateur radio
reasons, to live on. The important factor is to try to reduce the "ground"
loss (which includes environmental factors) as much as possible. Because
there are so wide a variation of situations it really is not possible to
give anything more than rough guidelines for use at any one site. I can only
suggest measuring it and trying different approaches to see which gives the
best returns on effort. I suggest measuring an unresonated aerial since it
avoids a having too many variables. You can also easily measure it over a
range of frequencies. This can be important because occasionally deviations
from the expected resistance frequency curve are seen and sometimes these
can be related to problems unknown to us and outside out control. One might
be a resonant length of mains cable to an outbuilding, of from a supply
transformer. I think you will soon find that there is little to be gained in
most places by using multiple earth stakes, and you can soon reach a optimum
on conterpoise installation. Then the remaining and to my mind most fruitful
area is to increase the aerial top-load capacity.

Cheers de Alan G3NYK
[email protected]