Ground resistance varies considerably with frequency, and at HF it can
be very different from the DC or 60 Hz value. (For example, pure water
is virtually lossless at DC and low frequency, but very lossy at HF.)
And in order to get a decent representation of ground, you also have to
know its dielectric constant. The only reasonably accurate method I know
of to make these measurements is a system using a buried transmission
line, called an OWL tester. I've never seen or used one, just read some
papers where it was used.

If your need is for an NEC model, you should realize that the skin depth
of current in soil in the HF range is on the order of 10 - 20 feet,
meaning that significant current exists to well below this depth. Soil
is seldom homogeneous, so to make an accurate representation of real
soil, you'd have to test the soil at various depths down to perhaps 30 -
50 feet or so, then make up some sort of average value for the program
to use. I don't believe any single average value will really be a good
representation of a typical stratified ground.

So the bottom line is that even if you could accurately measure the
surface soil conductivity, you wouldn't have enough information for an
accurate model. And in fact, even if you could measure the conductivity
and dielectric constant to a great depth, the model still wouldn't be
very good because of its assumption that the soil is homogeneous.

About the best you can do with currently available software is to
estimate the best and worst likely soil characteristics which might
occur from the surface to a few tens of feet depth. Run simulations with
both. The antenna performance will probably end up somewhere between the
results of the two. Even that isn't guaranteed, though, since any single
value probably isn't adequate to represent a stratified ground.

Ground modeling is the weak point of all NEC-type programs. Don't put
too much weight on results which are strongly dependent on ground
characteristics.

Roy Lewallen, W7EL

AG4QC wrote:
I was wondering if anyone built some sort of ground resistance tester? I
read that the NEC requires 25 Ohms or less.. But I don't understand how to
measure that. I certainly could use a ohm meter between a ground rod and the
test subject, but then the ground rod I use as a reference could be faulty..
It seems to me it must be measured in some other manner. Maybe low frequency
signal and then measure how much it's absorbed in the earth?.. I know they
sell a clamp on ground resistance tester for big bucks. Considering it's a
clamp on, it certainly don't have any external reference.. Anyway, I was
wondering if anyone had any experience with these testers and maybe building
one?

Joe AG4QC