Topband: K2AV Counterpoise

[Guy Olinger K2AV]

This entire thread has had some really good stuff here and there on
radials. Nicely done, gentlemen and a really good discussion.

:>))  Is there a reason why it will now get buried in the archives with a
searchable subject of "Topband K2AV counterpoise" ?  ROFL. Or was this a
thread hijack back there someplace I haven't found yet.

To the discussion rather than the thread's subject line, I think a blast
from the past is in order. In 1937, they did not have our ceramic
transmitting tubes or transistors of any kind, nor antenna models, nor our
predisposed understandings of how things worked, inherited from beloved
elmers and 50 year old textbooks. But Brown, Lewis and Epstein did have a
curiosity and a sharp commercial purpose with huge potential, and from RCA
very expensive highly accurate thermocouple-based instruments just being
developed on the bleeding edge of radio, and a PAID STAFF with nothing else
on their plate. They had nothing to go on but their measurements. And as
such a few carefully measured and described oddities from the 1937 BL&E
study need to be folded into the discussion:

pp 773-4, Brown, Lewis and Epstein: Ground Systems, [IEEE] Journal June
1937.

[denotes conversion of units, etc. Study conducted at 3 MHz]

"When radial wires were 45 feet long [0.14 wl], the measured resistance was
practically independent of the number of wires. Evidently most of the earth
loss occurred in regions beyond the periphery of the ground system. Fig. 27
is an average curve obtained for this condition.

"When only two radial wires, separated 180 degrees, were used the
resistance was independent of wire length *** since the current vanished
from the wires within a few feet of the antenna." *** [emphasis added]

page 781, ibid.

"For each ground system, the current in the buried wires was measured as
shown in Fig 24. The value measured in a single wire was then multiplied by
the number of buried wires. The current in the wires for an antenna height
of 88 degrees and radial wires 135 feet long [0.41 wl] is shown in Fig 42.
*** We see that the current persists in 113 wires much further [0.10 wl or
32.8 feet] from the antenna than it does for a smaller number of wires."
***  [emphasis added]

As viewed in figure 42, the current remains constant for the first 0.10
wavelength or first 33 feet of the 113 0.41 wl radials.

That means there was no loss in the ground around and beneath from
dielectric loss, or direct or induced current through the ground media. The
BARE, BURIED wire kept all the current in the copper.

page 782, ibid.

"A few tests were made of the action of an earth screen at the base of the
antenna. In the first test, the ground system consisted of 113 radials
wires each 135 feet long [0.41 wl]. The ground screen consisted of a square
copper screen, nine feet on a side. Absolutely no difference in field
strength or antenna resistance could be detected when the screen was
removed and the buried ground system used alone.

[Table embedded in text]  [quote]
----------------------------------------   Resistance  Field intensity

113 buried wires; no earth screen   1.0         1.0
15 buried wires; earth screen         1.62        0.785 [-2.1 dB]
15 buried wires; no earth screen    3.24        0.555 [-5.1 dB]
[unquote]

"The second test was made using 15 buried radial wires and the earth
screen. The relative results are shown above. Thus we see that, with a
[numerically] small ground system, the earth screen furnishes a definite
improvement. However, the results obtained are not nearly as good as those
obtained with the large [113 times 0.41 wl radials] ground system. Further
when the large ground system is used, the earth screen gave no further
improvement."

The study was undertaken by the RCA Corporation, no expense spared, with
the explicit purpose to show that a very expensive tall tower was not
nearly as important as the nature and quality of radials. Secondary
purpose, get more AM transmitters on the air, to help sell more RCA radios
-- capitalism at its best (not sarcastic). There were no computers, no
models, and given the pre-WWII date, such science was in it's infancy, and
they had no reason to bend their measurements. Measurements were all they
had to guide what to do next, with no modeling to make questionable
substitutions for actual measurements.

I spent a lot of time trying to get NEC 4 to generate the radial current
patterns found in Figure 42. Never came close. The assumptions of the
various NEC ground approximation methods cannot generate anything remotely
resembling Mssrs Brown, Lewis and Epstein's ACTUAL RADIAL CURRENT
MEASUREMENTS along their buried bare radial wires.

In figure 42, the first 33 feet carry a constant current, as opposed to a
pair of 0.41 wl radials immediately dumping all current into the dirt
"since the current vanished from the wires within a few feet of the
antenna."

Measured constant current on the 113 0.41 wl radials for the first 33 feet
means that no power was lost in that distance due to dielectric loss, or
direct or induced current loss, in spite of radials being bare and buried.
It was so effective that a 9 foot square of copper screen place above it
made no discernable difference on their highly sensitive thermocouple based
instruments. This brick wall screening effect from the 113 0.41 wl radials
lasted for 33 feet.

Note that using only 15 radials instead of 113 at 0.4 wl cost them 5.1 dB,
measured.

I would also note that the accompanying photographs indicate absolutely
flat, cleared farm land. Also by carefully comparing grey scale in photos
with both the landscape, and the black bakelite meter casings of the
instruments, not only expansive cleared flat land but BLACK dirt. It's in
the photos for permanent evidence. So that's a 5.1 dB loss in GOOD dirt,
not the RF sucking stuff we got here in North Carolina.

The answer to all of this must include unbiased measurements in BL&E, AND
careful, truthful anecdotes, AND careful paradigmic data like N6LF's back
yard. The models may or may not correctly work in all the various paradigms
bantered about here. For sure NEC 4 cannot generate Figure 42's curve for
113 times 0.41 wl buried radials at 3 MHz.

73, Guy K2AV

On Thu, Oct 26, 2017 at 11:47 PM, Dan Maguire via Topband <
topband at contesting.com> wrote:

> K8BHZ wrote:
> >>> The length to avoid is nothing more than a half wavelength, which
> translates the same impedance from end to end
> >>> i.e., the high Z open end translates to a high Z antenna base end.
> This results in minimum radial current.
>
> I'm not so sure I buy that and I don't think N6LF does either.  If you
> look at the section "An Explanation for the Dips in Ga" (Part 1, QEX pg 40)
> in Rudy's document
>
> http://rudys.typepad.com/files/qex-mar-apr-2012.pdf
>
> you'll find this: ["L" is the variable for radial length]
>
> <quote N6LF>
> Why do we see these large dips in Ga for some values of L? We can
> investigate this by looking at the current distributions on the radials and
> the associated E and H-field intensities close to ground under the radials.
> ... For the same current at the feed point, with longer radials the
> currents are much higher as we go out from the base. We would expect these
> higher currents to increase both E and H-field intensities at ground level
> under the radials. ... Since the power dissipation in the soil will vary
> with the square of the field intensity, it’s pretty clear why the
> efficiency takes such a large dip when the radials are too long.
> </quote>
>
> So Rudy seems to be saying that the increased loss is due to higher radial
> currents, not minimum radial currents.
>
> Here's an animation showing how the currents on the vertical and on 1 of 4
> identical radials change as the radial length changes.  Segments 1-15 are
> the vertical with segment 1 being the base (constant 1 amp).  Segments
> 16-37 are the radial with segment 16 being the inner end connected to the
> base of the vertical.  Each radial always has 0.25 amps at the inner end
> but the current distribution across the rest of the radial changes as the
> length changes.
>
> The animation frames start with a radial length 0f 0.10 WL, step by 0.02
> WL, and finish at 0.60 WL.  The frame where the radial current in segment
> #29 peaks at ~1.02A corresponds to the sharp dip in efficiency.  Radial
> length for that frame is 0.36 WL.
>
> https://s1.postimg.org/7a7f6pvu4f/N6_LFani.gif
>
> Above is at 1.85 MHz with 4 radials at 2" above ground.  Below is a
> different animation, this time showing E-field intensity as the radial
> length changes.  For this animation the radial height was 10 ft so the dip
> occurs at ~0.45 WL (frame 9).  That corresponds to the highest radial
> current and the max E-field.
>
> https://s1.postimg.org/45ipxxhp73/MCVert_NF3_D.gif
>
> Dan, AC6LA
> _________________
> Topband Reflector Archives - http://www.contesting.com/_topband
>