Quite late in response, partly done much earlier. Sometimes I have time,
and sometimes I have a house and a spouse.
Peter's original question in this thread was why a radial and mesh mix in a
counterpoise should have the radials and mesh bonded together. That is
quite a different question than why mesh at all, or how good is mesh by
itself, or mesh versus radials, or just how does "ground" behave at the
seashore.
In the 1937 Brown, Lewis and Epstein (BL&E) study, field strength was **
MEASURED ** at one mile at the frequency of 3.0 MHz over 15 times 0.4
wavelength radials, as well as over 113 times 0.4 wavelength radials, with
and without a 9 foot square mesh ground screen *over the radials*. {BL&E:
Ground Systems, Proceedings of the IEEE, June, 1937, p. 782}
Over the 113 radials ZERO change, "with" vs. "without" ground screen,
relative field strength 1.0 (0 dB)
Over the 15 radials,
"With" the screen: relative field strength 0.785 (-2.1 dB)
"Without" the screen: relative field strength 0.555 (-5.1 dB)
"With" was a *measured* improvement of 3.0 dB over "without", but still
significantly less than the dense 113 radial field.
Assuming that the mechanism is the mesh covering loss that would otherwise
be beneath it...isn't it interesting that over 15 ground radials 3 dB was
lost in the first 9 feet away from the center of the vertical at the
ground.
We should understand the why's of losing 2.1 dB from the dense radials and
then 3.0 dB more without the screen. Blowing an S-unit would certainly seem
significant. AND that was 15 times 0.4 wavelength evenly spaced radials,
not 15 miscellaneous length, shape and position radials.
In 1937 the only calculation device available for calculation from formulas
was the slide rule. Computers, running at GHz frequencies solving antenna
problems with quadrillions of intermediate steps in intricate and advanced
algorithmic methods, were six decades in the future. As such BL&E must be
given credence, as a number of very interesting graphs of RF fields and
current could not possibly have been anticipated, or "modeled" by any
process or means available. They HAD to MEASURE to have anything.
BL&E were being paid to go through this exercise and had a paid staff for
assistance, and funds to purchase the best test equipment of the time. They
could expend the resources to actually measure everything. They overcame
the obstacles so often defeating our modern personally-funded attempts with
the deep dollar pockets of the RCA corporation. RCA in turn had a huge
vested interest in completion of the research, building up commercial AM
broadcasting.
BL&E were explorers in a field mostly still unknown. They have
extraordinary graphs that put the lie to some corners of today’s thinking
where models’ approximations have been accepted as fact. We *should* have a
science of the limits of modeling and issues with “boundary situations”,
but seemingly do not.
BL&E DID have VERY sensitive thermocouple field strength and power
measuring devices with exceptional accuracy. When I was employed by AT&T
Long Lines in 1963, way back when I still had hair, variants of these were
still employed for standard maintenance practices. They were significantly
better than anything I myself have owned since. Just BULKY :>))
Sherwood describes a comparison where *only* mesh screens are used as an
*alternative* to radials. His study does *not* measure the *interaction* of
a mesh over radials in any configuration, as queried by the originator of
this thread.
Good luck, all and may your 160m wire "smoke 'em" in tomorrow's Pre Stew.
73, Guy K2AV
On Mon, Jun 4, 2018 at 3:24 PM, Grant Saviers <grants2@pacbell.net> wrote:
> For simulating a solid conducting plate in NEC with wires, Roy Lewallen
> (EZNEC author) advises 0.1 wavelength on a side squares of wires. That
> would be 16m for topband. Since adding the mesh is a search for tenths
> more db's, be conservative and use half his recommendation, 8m or 25'.
>
> If a reasonable radial field of 36 wires x 125' were installed, then
> pi*D/36 = 21' tip separation (or do the trig for 10 deg angles). However,
> buried radials don't need to be 125' long due to the effects of ground,
> thus the tips will be closer together.
>
> So IMO, based on that guidance, matting at topband wavelengths is not
> worth it, especially considering how close together wires are near the base
> of a vertical.
>
> If a research answer is required, with NEC4.2 (EZNEC Pro/4) you can bury a
> mesh in the ground, connect it to the radials, and analyze it. NEC2 will
> give close answers with radials and the mesh an fraction of an inch above
> ground at a fraction of NEC4 cost. It would be a real PITA to build the
> wire model for radials and a square mesh since wires have to connect at
> segment junctions.
>
> Grant KZ1W
>
>
> On 6/4/2018 0:04 AM, Raymond Benny wrote:
>
>> Peter:
>>
>> Is this a question or what you are saying is a know fact?
>>
>> I am very interested in this outcome since I will a installing a TX 4SQ
>> system where some existing ground radials. Over time, I have heard both
>> pros and cons on this subject. I would like to read some research or
>> documentation on this subject.
>>
>> Tnx,
>>
>> Ray,
>> N6VR
>> Near Prescott, AZ
>>
>> On Sun, Jun 3, 2018 at 1:14 PM, Peter Bertini <radioconnection@gmail.com>
>> wrote:
>>
>> Why would bonding the added matting be required if it is laid over or
>>> beneath an existing radial field? It reduces ground losses regardless.
>>>
>>> Peter
>>> _________________
>>> Topband Reflector Archives - http://www.contesting.com/_topband
>>>
>>>
>>
>>
> _________________
> Topband Reflector Archives - http://www.contesting.com/_topband
>
_________________
Searchable Archives: http://www.contesting.com/_topband - Topband Reflector
|