It is fairly easy to measure soil conductivity at DC and power
frequencies. See program EARTHRES. Such measurement values apply up to
a few hundred kHz.

A few years back I became curious about what happens to conductivity
at HF. And at what frequencies does soil permittivity begin to matter.
Most amateur activity is at HF and above. Yet, without even thimking
about it, we persist in plugging in power frequency values into
formulae and computer programs.

A volume of soil between a pair of electrodes behaves as a resistance
in parallel with a capacitor.

So this is what I did -

1. Obtain a 16-inch length of galvanised steel tube, 5 inches in
diameter. (Mine was industrial ventilation duct)

2. Block off one end of the tube with a rigid disk of plastic
insulating material.

3. Obtain a 17-inch length of copper water pipe, diameter = 0.6
inches.

4. Locate the water pipe in the centre of the tube, resting on the
plastic disk at the bottom.

5. You now have a coaxial structure of accurately known dimensions.
When empty, Zo = 128 ohms. 1/4-wave at 191 MHz

6. Obtain a mixed sample of soil from various places in your garden
under your antenna.

7. Fill the galvanised tube in easy stages with garden soil. At each
stage compress and pack-down the soil to about the same density as it
was in your garden. Make sure the soil is in contact with the inside
surface of the tube.

8. Cover the top of the soil in the tube with a flat disk, with a
hole in the middle, to discourage evaporation of moisture and drying
out of the soil.

9. You now have 16-inch length of transmission line on which you can
make HF impedance measurements using instruments as simple as
hand-held antenna analysers. Resistance measurements at 50 or 60 Hz
can be obtained from volts/milliamps. Although connecting leads can be
kept very short it is advisable to correct measurements for
lead-length above 10 MHz. Measurements were made up to VHF.

10. Using classical transmission line formulae in reverse, the values
of line conductance G, capacitance C and hence permittivity K of the
"insulating" material, i.e., the soil, can be calculated.

11. Measurements are of input impedance of the line with the other
end open cicuit. The basic equation is Zin = Zo*Coth(A + jB) where A
is line attenuation and B is line phase shift. At the lower
frequencies the line is very short and G and Capacitance and then K,
can be calculated directly from measurements and line dimsnsions.

12. A clinical thermometer can be inserted deep in the soil. If the
test cylinder is too large to fit in the domestic fridge, by leaving
the test sample out of doors overnight in winter the effects of
temperature can be observed as the sample slowly warms up from
freezing. Soil has a high negative temperature coefficient of
resistance. Resistance increases as temperature decreases. My garden
soil is roughly -2 percent per degree C at 20 degrees C.

13. I have made HF measurements in other shaped containers, usually
smaller and plastic, with copper sheets for electrodes. Also in the
garden itself between radials and arrays of relatively short rods.
Any sort of measurements are more useful than none.

Some people say the only way to deternine soil characteristics is to
construct a 1/4-wave vertical antenna, feed it with 50 Kwatt at 500
KHz and measure field strength at 1 mile intervals for 100 miles. And
then do some calculations. Don't you believe it!
----
Reg G4FGQ

Hi Reg,

You've presented a very interesting way of measuring soil characteristics. When
I return to Florida in November I'm going to use your method of measuring the
soil underneath the dipole whose impedances I measured over the frequency range
14 to 15 MHz at various heights above ground, including one set of measurements
with the dipole lying on the ground.

One of the reasons I offered to distribute the data from my measurements is to
see whether anyone can deduce any soil characteristics from the changes in
impedance with height. The changes are significant. For example, the terminal
impedance with the dipole on the ground runs from 470 + j250 at 14 MHz to 570 +
j132 at 15 MHz. The inductive reactance doesn't become capacitive until the
dipole is 2 ft off the ground. In addition, except at zero height, the
resistance component decreases with height, but for every height the resistance
increases with frequency. Do you think any of the soil characteristics could be
determined by such data?

Would you like a copy of my data, just fer the helovit?

Walt, W2DU

"Walter Maxwell" wrote in message
...
Hi Reg,

You've presented a very interesting way of measuring soil characteristics.
When I return to Florida in November I'm going to use your method of
measuring the soil underneath the dipole whose impedances I measured over
the frequency range 14 to 15 MHz at various heights above ground,
including one set of measurements with the dipole lying on the ground.

One of the reasons I offered to distribute the data from my measurements
is to see whether anyone can deduce any soil characteristics from the
changes in impedance with height. The changes are significant. For
example, the terminal impedance with the dipole on the ground runs from
470 + j250 at 14 MHz to 570 + j132 at 15 MHz. The inductive reactance
doesn't become capacitive until the dipole is 2 ft off the ground. In
addition, except at zero height, the resistance component decreases with
height, but for every height the resistance increases with frequency. Do
you think any of the soil characteristics could be determined by such
data?

Would you like a copy of my data, just fer the helovit?

Walt, W2DU

I was working at a company a few years ago, and they built a capacitor of
two plates, about 6" on a side, and 0.25" separation. They measured the
thing on a network analyzer, and then packed the dielectric with potting
soil. Again measuring the results on the network analyzer they were able to
deduce the conductivity and permittivity of the soil. I did not think
potting soil was typical, but still an interesting experiment.

Frank

PS, Walt, I would very much like to receive your experimental dipole pdf.

Walter Maxwell wrote:
. . .
One of the reasons I offered to distribute the data from my
measurements is to see whether anyone can deduce any soil
characteristics from the changes in impedance with height. The
changes are significant. For example, the terminal impedance with the
dipole on the ground runs from 470 + j250 at 14 MHz to 570 + j132 at
15 MHz. The inductive reactance doesn't become capacitive until the
dipole is 2 ft off the ground. In addition, except at zero height,
the resistance component decreases with height, but for every height
the resistance increases with frequency. Do you think any of the soil
characteristics could be determined by such data?
. . .

I haven't looked into this carefully, but one person I know who was very
involved in NVIS operation (where ground characteristics are important)
tried it some years ago. He concluded that it wasn't possible to
set the antenna height and make measurements with sufficient accuracy to
infer the ground characteristics with any confidence.

Roy Lewallen, W7EL

Roy, W7EL wrote:
I haven't looked into this carefully, but one person I know who was very
involved in NVIS operation (where ground characteristics are important)
tried it some years ago. He concluded that it wasn't possible to
set the antenna height and make measurements with sufficient accuracy to
infer the ground characteristics with any confidence.

Roy Lewallen, W7EL

Understood, Roy, but was this person saying that with just one height it
wouldn't give sufficient accuracy, or is he saying that with impedance knowledge
at many different heights there would still be no determination of any of the
ground characteristics?

Would you not like to see my data before concluding it couldn't reveal any
ground characteristics.?

Walt

On Mon, 20 Jun 2005 16:21:58 -0400, "Walter Maxwell"
wrote:
Understood, Roy, but was this person saying that with just one height it
wouldn't give sufficient accuracy, or is he saying that with impedance knowledge
at many different heights there would still be no determination of any of the
ground characteristics?

Hi Walt,

And per my several critiques into this matter, all such broad
proclamations lack the fundamental of drawing a validation through
correlating work in the subject. Let's examine the one point offered:
He concluded that it wasn't possible to
set the antenna height and make measurements with sufficient accuracy to
infer the ground characteristics with any confidence.
This, of course, presumes that this source has any actual
authoritative data. Something that is prohibitively beyond the scope
of an individual to determine (when it is already rejected through
correlations of antenna characteristics and measurements) in the first
place suggests there is none.

Roy has already pointed out the futility of a piece-wise measurement
throughout the bulk of earth soaked by RF to its skin depth. I have
pointed out that these several treatments offered only go to the thin
veneer of soil. Some conclusions drawn were preposterous on the face
of the data offered. Further, to suggest the four lead measurement be
stretched to employing wavelength sized leads is fraught with error
through the denial of those leads becoming what every Amateur already
has, an antenna.

Reg has dismissed the use of an antenna to measure the earth's
contribution of loss, or to distinguish its characteristics by
perturbing the known characteristic of an antenna. Such dismissal is
not an argument - it is a conceit.

Walt, your data is comprehensive enough to build a soil model for the
band you studied. I seriously doubt anyone could challenge your
results if they were internally consistent.

73's
Richard Clark, KB7QHC

Walt, your data is comprehensive enough to build a soil model for the
band you studied. I seriously doubt anyone could challenge your
results if they were internally consistent.

73's
Richard Clark, KB7QHC

Thanks Richard, you've made my day!

Walt

Walter Maxwell wrote:
Roy, W7EL wrote:

I haven't looked into this carefully, but one person I know who was very
involved in NVIS operation (where ground characteristics are important)
tried it some years ago. He concluded that it wasn't possible to
set the antenna height and make measurements with sufficient accuracy to
infer the ground characteristics with any confidence.

Roy Lewallen, W7EL


Understood, Roy, but was this person saying that with just one height it
wouldn't give sufficient accuracy, or is he saying that with impedance knowledge
at many different heights there would still be no determination of any of the
ground characteristics?

I interpreted what he said as meaning that he looked into the method and
determined it wasn't practical. Surely he thought of making numerous
measurements. He's a very capable engineer, so I took what he said at
face value. On the other hand, I don't think he's highly skilled in
making precision antenna measurements, so he might have assumed that a
level of accuracy wasn't achievable which in fact might be.

You might spend a little while with EZNEC looking at how much a change
in ground conductivity or permittivity changes the antenna input Z at
various heights, and how much the height changes the Z with a given set
of ground characteristics. Then consider whether you'd be able to set
the height and make the impedance measurements accurately enough to
infer the ground characteristics with any degree of confidence.

Would you not like to see my data before concluding it couldn't reveal any
ground characteristics.?

I'd like to see your data, but it wouldn't be enough information to
conclude whether the method would be practical or not.

Roy Lewallen, W7EL

It would be interesting to recreate the measurements at other locations. My
location has 500 feet of sand below me. It would be a great improvement just
to have poor soil.


"Walter Maxwell" wrote in message
...
Hi Reg,

You've presented a very interesting way of measuring soil characteristics.
When
I return to Florida in November I'm going to use your method of measuring
the
soil underneath the dipole whose impedances I measured over the frequency
range
14 to 15 MHz at various heights above ground, including one set of
measurements
with the dipole lying on the ground.

One of the reasons I offered to distribute the data from my measurements
is to
see whether anyone can deduce any soil characteristics from the changes in
impedance with height. The changes are significant. For example, the
terminal
impedance with the dipole on the ground runs from 470 + j250 at 14 MHz to
570 +
j132 at 15 MHz. The inductive reactance doesn't become capacitive until
the
dipole is 2 ft off the ground. In addition, except at zero height, the
resistance component decreases with height, but for every height the
resistance
increases with frequency. Do you think any of the soil characteristics
could be
determined by such data?

Would you like a copy of my data, just fer the helovit?

Walt, W2DU

Fred W4JLE wrote:
It would be interesting to recreate the measurements at other locations. My
location has 500 feet of sand below me. It would be a great improvement just
to have poor soil.

Depends on your objective. For NVIS operation with a horizontal antenna,
where you need the reflection, that's probably true. But for a vertical
or for DX with a horizontal antenna, you're better off with the sand.
Perfect ground has no loss; free space has no loss. There's an
intermediate quality of ground at which the loss is maximum at a given
frequency. Unfortunately, this happens to be in the range of ordinary
ground characteristics in the HF range. Your ground should be very low
loss. And your pattern should resemble free space, with a very strong
field at very low radiation angles.

Roy Lewallen, W7EL