A single ground rod, unless in sea water, has a resistance to mother earth
between 50 and 200 ohms. Let's take it to be 100 ohms.

Efficiency of a 1/4-wave vertical, feedpoint resistance = 37 ohms, is 27
percent.

Efficiency of a 5/8-wave vertical, feedpoint resistance = 50 ohms, is 33
percent.

Efficiency of a 1/2-wave vertical, feedpoint resistance = 2500 ohms, is 96
percent.

The difference in radiation pattern in a typical back yard, in the vertical
plane, is neither here nor there.

The 1/2-wave antenna also needs the most simple L and C matching network.

But I'd never recommend a ground rod anyway. Not worth the time, trouble
and expense unless extremely short of real estate at ground level.

Roy, the problem of choice lies in over-complication by too 'clever',
'knowledgeable' old-wives and gurus rather than under-complication.
----
Reg, G4FGQ

"Roy Lewallen" wrote in message
...
In the inverted L or any antenna with a horizontal wire, there's
coupling between the wire and ground. The field from the horizontal wire
induces current in the ground under it. If the wire is low, the loss
produced by this current can be substantial. By putting an elevated wire
under the horizontal wire, you've changed this coupling to the ground,
plus you've introduced a new conductor into the antenna. Mutual coupling
between this conductor and the other wires will change the impedance.

Modeling will give a lot of insight into what all is going on.

Roy Lewallen, W7EL

Thanks, don't know why I hadn't considered "the rest of the half wave". Not
many options other than what I have up, so will pass on the modeling. It's
REALLY a very decent performer on 160, 80 and 40, and unobtrusive in the
summer.

W4ZCB

On Thu, 20 May 2004 17:15:01 GMT, Richard Clark
wrote:

I made the same comparison at 40M. The difference between 120
(quarterwave) radials and 1 amounts to 0.1dB Clip that one down by a
tenth and the difference climbs to an astronomical 0.3dB.

Maybe the difference is the length of the radials. I used ½-wavelength
radials as the peak ground current is at 0.35-wavelength from the base
of the monopole - ¼-wavelength radials would be too short to reach
that area.

Danny

Be careful about making generalizations about this. The position of the
peak current depends on frequency and the ground characteristics. I
believe it's also a function of the height of the vertical. In some
cases there's no real peak at all, but an exponential-looking decay of
current from the base of the vertical outward. This, incidentally, was
experimentally measured and documented by Brown, Lewis, and Epstein in 1937.

Roy Lewallen, W7EL

Dan Richardson wrote:

Maybe the difference is the length of the radials. I used ½-wavelength
radials as the peak ground current is at 0.35-wavelength from the base
of the monopole - ¼-wavelength radials would be too short to reach
that area.

Danny

On Thu, 20 May 2004 19:49:28 -0700, Roy Lewallen
wrote:

Be careful about making generalizations about this. The position of the
peak current depends on frequency and the ground characteristics. I
believe it's also a function of the height of the vertical. In some
cases there's no real peak at all, but an exponential-looking decay of
current from the base of the vertical outward. This, incidentally, was
experimentally measured and documented by Brown, Lewis, and Epstein in 1937.

Roy Lewallen, W7EL


Interesting. I went back to the model and took a look at the current
in the radials.

My model was a 1/2-wave monopole using 120 1/2-wavelength buried
radials. The frequency was 3.6 MHz.

EZNEC (Version 4) reported the peak radial current at about
0.41-wavelength from the base of the antenna. I made two runs. One
using poor ground and one using average ground. The peak current
location was the same in both.

This still leads me to believe that the difference in gain reported
between what Richard had modeled and I found (0.1 dB vs 1.0 dB) is due
to the length of the radials.

In ether case adding that much wire (15,840 feet) for so little gain
sure doesn't seem worthwhile.

73
Danny, K6MHE

In ether case adding that much wire (15,840 feet) for so little gain
sure doesn't seem worthwhile.

73
Danny, K6MHE

========================

Danny, I quite agree.

The current-carrying cross-sectional area of the Earth is enormous at
distances from the antenna base of 1/4-wavelength and greater. Regardless
even of very poor soil resistivity, loss in the soil is sensibly zero.

Furthermore, propagation velocity in the soil is MUCH less than the free
space velocity and I am of the opinion that computer models give a very
distorted picture of what actually happens.

At distances of the order of 1/8 free-space wavelength practically all of
the current flows in the soil. Shallow-buried radials might just as well not
be there. The copper is better used to increase the number of short
radials.

But an increase in the number of short radials is a waste of copper anyway
when the number of radials is already very large.

What B,L&E were doing with 120 radials at MF in 1937 is hardly relevant. I
understand they forgot to determine ground conductivity - an indication they
didn't fully appreciate what they were about. As they were the first in the
field to make such measurements this omission is understandable. But at HF,
soil characteristics are considerably different - factors which computer
model users do not feed into their models.

Computerised antenna model users are inclined to suffer from delusions of
accuracy - drowning, unaware, in a sea of uncertainties.

But there's no real harm done! ;o)
----
Reg.