I found that the best approach was to use the mininec ground and place
a resistive load at the ground end which can be varied. You may find
that anything in a range behaves mostly the same or completely changes
the outcome, depending on the frequency, and which object you were
changing the load on. This tells you what to worry about.
As to non-resonant sections, in close proximity to an antenna, they
are a MYTH, unless shorter than 3/8 wavelength or so on the frequency
in use.
This is so because the guy section under consideration is excited
NON-SYMMETRICALLY voltage-wise, non-symmetrically phase-wise, and
CONTINUOUSLY excited along its entire length. The current maxima can
MOVE, the current pattern extends and becomes non-sinusoidal, dropping
only moderately as so-called non-resonant lengths are approached. The
non-non-resonant behavior is increased if you add 5-10 pf incidental
capacitance between consecutive sections because of egg-type
insulators.
Your (Pete's) 21' insulators are highly effective, eliminating roughly
80 percent of side/back pattern degradation. What is better is about
1/2 wavelength of phillystrand from the tower at the lowest frequency
used. But bang for the buck, your insulators are great.
If you are talking about a tower-top tribander, and you only care
about gain, you don't mind some moderate swr changes, and you would
rather be able to hear some to the sides and the rear (an oft
expressed view), an insulator at the tower does all that is needed.
FORWARD GAIN is essentially UNAFFECTED by guys. This is at the core of
the variance of contradictory anecdotal material. SWR will vary
somewhat due to guys, depending.
If you want your FRONT-TO-BACK and FRONT-TO-SIDE optimal and
unaffected, then you need to remove any part of anything longer than
3/8 wave on the HIGHEST frequency used from a sphere of radius 1/2
wavelength at the LOWEST frequency used, with the sphere centered on
the radiator midpoint. Pete's 21' insulators are approaching this.
There are more things to check if there are guys above and below a
lower antenna.
If, in addition to the above, you put an insulator at the bottom of a
guy, with turnbuckle, etc to a ground rod, and bridge the insulator
with a 300 ohm non-reactive resistor, you get nearly as good as
phillystrand. This roughly terminates the guy in its characteristic
impedance and dissipates power that would otherwise re-radiate or be
induced into the antenna off-pattern. The 300 ohm value is
non-critical and assumes an RF resistance of 10-100 ohms in the ground
connection.
(Note that elements of 40 meter beams below tribanders, etc, have
their elements symmetrically excited along their length. They behave
in the traditional manner.)
I am SLOWLY developing the workup for this complete with graphs, etc,
but it is hard, painstaking work. It will be seen that the effective
non-resonant section is a SPECIAL CASE that applies in certain
circumstances, and is not the general rule.
Up next, "Do antennas receive the same way they transmit? Not always."
Think about unequal and non-symmetrical excitation of stacked or
phased antennas.
73, y'all, Guy.
On Tue, 21 Sep 1999 08:32:18 -0400, you wrote:
>
>In modeling structures attached to a real (non-miniNEC) ground, EZNEC
>cautions that results may be invalid because no value has been specified
>for the loss resistance in the grounding connection.
>
>Lacking the means to measure such resistance directly, I'm wondering if
>anyone knows of rule-of-thumb resistance values to use for situations such
>as guy anchors embedded in concrete -- Rohn standard style -- and tower
>bases set on or in concrete and grounded through a fairly typical grounding
>system, such as 3 x 8 foot ground rods and #2 copper wire.
>
>Thanks in advance, folks...
>
>73, Pete N4ZR
>Sometimes a tower is just a tower
73, Guy
--. .-..
Guy Olinger, K2AV
k2av@qsl.net
Apex, NC, USA
--
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
|