[TowerTalk] Scaled Antenna

Tom Rauch W8JI@contesting.com
Sat, 22 Apr 2000 00:45:34 -0400

Hi Ron,

First, congratulations on doing some real good work. My 
comments aren't made to discourage you, but rather to point out 
something that most people might not consider. 

> Since my tests were on the 17M band and I wanted everything to be as
> realistic as possible, I began by constructing a scale model (actually two
> of them) of Rohn 20 tower using 1/8" galvanized pipe spaced 1-1/8" on
> center. This section was 6 feet tall. 3 feet of " galvanized pipe was
> attached to the top of the miniature tower and a top hat with a diameter
> of 7-1/2" was used to resonate the model at 18.12 MHz. " fiberglass rod
> was used to insulate the antenna from ground, with a space of 2".

The skin depth on 160 meters is several feet to perhaps 40 feet or 
more deep. On 18 MHz, it is much less. Not only does everything 
work to reduce ground losses on lower frequencies, the ground 
itself has very different characteristics at different depths on 
different frequencies.

Because of that, earth losses are much higher on 18 MHz for a 
given soil conductivity. That's why none of us, except those living 
on saltwater, can work long distances on 18 MHz groundwave and 
all of us can work many many miles on 1.8 MHz groundwave.

Unless the dirt behaves exactly the same (same path losses, 
same conductivity) on 18 Mhz as 1.8 MHz...the data will be flawed. 

> I set the antenna up on a 190' x 170' area near my pond, which stays
> somewhat damp (according to previous owner). This is where I plan to
> install the 160M antennas. 60 radials, each 13 feet long, equally spaced
> at 6 degrees were installed at the test site. The antenna was fed with
> 70-ohm coax through a shunt feed. An MFJ 259B was used to adjust the
> match. One watt of power was fed through a step attenuator to the antenna
> under test and readings were taken with a F/S meter. Antenna pattern was
> clean and quite circular.


> Next, I raised the antenna and used a counterpoise system. This is where
> the surprise came. It only took 8 radials to provide the same F/S as the
> 60 ground-mounted radials! F/S with 60 radials was approximately 135%
> higher than when the radials were ground mounted.

Since you moved the antenna above earth, how did you insure 
propagation to the FSM was identical? What prevented propagation 
losses that are extremely high along the earth on 18 MHz from 
decreasing as the system moved towards spacewave instead of 
groundwave along the earth?

I would expect an increase, unless the distance to the FSM was 
such a very long distance that you didn't decrease path loss by 
moving the antenna.

One thing that clearly points out a flaw in the test is the efficiency 
of a vertical with 60 1/4 wl radials is very near 100%...no matter 
what soil is below the radials. That's because 60 radials pretty 
much look like a solid copper sheet, and the antenna fields are 
almost totally isolated from the dirt below the antenna. 

Unfortunately moving the sixty radials higher caused FS to improve 
to 135% (a 1.3 dB change). That amount of change is absolutely 
impossible due to account for in efficiency changes, since the 
original efficiency is nearly 100%. We certainly can not have 135% 
efficiency, so either the measurement had errors or the propagation 
losses were reduced.
> Is this a surprise to anyone but me? Has anyone else observed the same
> results regarding a counterpoise?

Not me. 

I measured some systems on 160 meters, and it took about 20 or 
so elevated radials to equal 60 radials on the ground. 60 elevated 
radials and 60 ground mounted radials were almost exactly the 
same, well within instrument error.

Since mounting 20-30 elevated radials is a real pain...unsightly...a 
maintenance headache...works on one band mostly....and offers no 
lightning protection I just bit the bullet and buried a bunch of radials.

Since you can't scale the dirt to the much lower 160 meter loss 
values in the test, it doesn't mean much.

By the way, this is why elevated groundplanes work very very well 
on 20 through ten meters...and elevating a vertical above earth 
hardly makes any change at all on 160 or 80 meters.
73, Tom W8JI

FAQ on WWW:               http://www.contesting.com/towertalkfaq.html
Submissions:              towertalk@contesting.com
Administrative requests:  towertalk-REQUEST@contesting.com
Problems:                 owner-towertalk@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm