Hello, TopBanders,
A couple of weeks ago I promised a re-post of data for shunt-feeding Rohn
25 towers. I've spent many hours since then trying many things and
comparing results with the real-world values of my own towers. The
closest thing I keep coming up with is to use EZNEC to model the actual
triangular tower sections, with evenly-spaced cross-braces ("rungs") at
appropriate levels. The following data is for Rohn 25 towers without any
top-loading. Keep in mind that although using this method is in close
agreement with my shunt-fed towers, my aluminum towers taper slightly and
are top-loaded with HF beams. I don't know if this method of modeling is
also valid for non-top-loaded towers.
My last posting on this was based on a loss of 3-1/2" (due to overlap)
per each 10' Rohn 25 section, or a net 9' 8-1/2" per section. It has
occurred to me that this overlap is for N - 1 sections, i.e., there are
only 8 overlaps for a 9-section tower. To correct this error, each of
the towers in this posting is 3-1/2" taller than in the previous posting.
The following tower heights also assume that none of the bottom section
is buried in concrete.
Here are the results as yielded by EZNEC v2.0. Values shown are for a
design frequency of 1830 kHz. In all cases, #8 gauge was used for the
shunt feed wire and was spaced 24" from a tower leg. The capacitor
voltages shown are what to expect for 1 kw of 1830 kHz RF fed to the
tower (for 2 kw pep SSB, multiply the voltage by 1.414). For those who
have high aspirations, I've included info for 3/8-wave, 1/2-wave, and
5/8-wave shunt-fed verticals (20-, 26-, and 33-section Rohn 25 towers,
respectively).
# of Tower 1/4-wave Gamma Gamma Capacitor
Sections Height Resonance Height Capacity Volts
------------- ---------------- ---------------- -------------
------------- --------------
9 87' 8" 2651 kHz 77' 1/2" 80 pF
4839
10 97' 4-1/2" 2390 kHz 65' 9" 109 pF
3553
11 107' 1" 2176 kHz 51' 2-1/2" 163 pF
2391
12 116' 9-1/2" 1997 kHz 33' 7" 287 pF
1355
13 126' 6" 1846 kHz 21' 11" 575 pF
677
14 136' 2-1/2" 1716 kHz 32' 2" 508 pF
765
15 145' 11" 1603 kHz 46' 7" 347 pF
1121
20 194' 5-1/2" 1216 kHz 82' 9" 134 pF
2894
26 252' 8-1/2" 938 kHz 90' 1" 96 pF
4049
33 320' 8" 740 kHz 87' 1/2" 88 pF
4399
The following table shows the performance of the same shunt-fed towers
over various types of earth as described in the ARRL Antenna Book
(chapter 3, table 1, of the 17th edition). Shown are the gain of each
tower compared to the 13-section tower over "good" ground (the one very
close to 1/4-wave). The value just below the gain figure is the angle
above the horizon at which that gain occurs.
TYPE OF FAR-FIELD GROUND:
# of Very Very Fresh Salt
Sections Poor Poor Good Good Water Water Perfect
------------- -------- -------- -------- -------- ----------
--------- -----------
9 -2.154 -0.965 0.091 1.684 0.470 3.179 3.489
29.5 27.0 24.2 18.4 23.4 7.9
0.0
10 -2.204 -0.992 0.073 1.694 0.482 3.217 3.531
29.1 26.7 23.9 18.1 23.2 7.8
0.0
11 -2.264 -1.026 0.048 1.703 0.493 3.260 3.581
28.8 26.3 23.6 17.9 22.9 7.7
0.0
12 -2.324 -1.059 0.025 1.718 0.510 3.313 3.640
28.3 26.0 23.2 17.6 22.5 7.6
0.0
13 -2.384 -1.093 0.000 1.732 0.526 3.368 3.702
27.8 25.5 22.9 17.3 22.1 7.5
0.0
14 -2.439 -1.123 -0.022 1.748 0.544 3.423 3.764
27.4 25.1 22.5 17.0 21.7 7.3
0.0
15 -2.489 -1.150 -0.046 1.769 0.567 3.485 3.834
26.9 24.6 22.1 16.7 21.3 7.2
0.0
20 -2.684 -1.243 -0.105 1.937 0.751 3.906 4.300
24.0 22.0 19.7 15.0 19.1 6.5
0.0
26 -2.566 -1.138 -0.051 2.342 1.229 4.735 5.201
20.4 18.5 16.5 12.6 16.1 5.5
0.0
33 -1.205 +0.525 +0.039 2.793 2.039 5.923 6.551
17.9 15.0 12.9 9.8 12.6 4.3
0.0
It's interesting to note that the above table tends to agree with my 80m
observations with a 1/2-wave vertical over "very good" earth (Houston)
and "good" (?!) earth (Mojave Desert). Close scrutiny of the table
suggests that, unless you have a fresh water far-field ground (/MM on the
Great Lakes?) or better, the tall verticals are not worth the extra money
and effort involved. Also, it's doubtful that ultra-low angles of
radiation are desireable on 160m.
A week or so ago I received a message from Dave, K1FK, who pointed out
that according to Jasik (whom I believe was a broadcast engineer from
more than 35 years ago), had shown that the equivalent radius of a
triangular-shaped tower is .4214 times the radius of a circle which can
just enclose that triangle. For Rohn 25 (12.5" per side), such a circle
would have a radius of 3.041192543", or a diameter of about 6.08". If
you multiply the side dimension of any triangular tower by .4866, you
will find the diameter of the Jasik equivalent circle. I'm forever
grateful to Dave for alerting me of Jasik's finding, however, be aware
that Jasik's formula is based on sinusoidal current distribution whereas
NEC-based programs use the "method of moments" method of calculation.
Following is a table of Rohn 25 shunt-fed towers using a Jasik equivalent
cylindrical tower 6.08" in diameter. Again, the shunt-feed wire is #8
gauge and spaced 24" from the outer radius of the Jasik equivalent tower
(27.04" from the tower center).
# of Gamma Gamma Capacitor
Sections Height Capacitor Voltage
------------- ------------ --------------
---------------
9 83' 0-1/4" 77 pF 5034
10 72' 0-1/2" 105 pF 3709
11 57' 8-3/4" 154 pF 2528
12 40' 0" 261 pF 1487
13 26' 0" 515 pF 755
14 34' 7" 492 pF 790
15 50' 1-1/2" 332 pF 1170
I lend you these observations, based on modeling my top-loaded shunt-fed
towers:
1) Modeling the tower and beam structures in EZNEC yields accurate
(within less than 1%) results for both towers for all four shunt-feed
parameters -- gamma wire size, height and spacing, and required
capacitance.
2) Modeling the Jasik equivalent tower/beam structures in ELNEC requires
that the gamma height be 2 to 3% lower than the real world (using real
world wire size and spacing). The capacitance required in the models is
about 25% greater than in the real world.
3) Modeling my tower/beam structures using a cylindrical tower equivalent
0.55 times the width of the triangle (a bit larger than the Jasik
equivalent) yields accurate results for wire size, height and spacing,
but the required capacitance is still about 17% high.
4) Observation #2 above is contrary to the results posted in this
e-mail, i.e., Jasik equivalent results require higher gamma height and
slightly less gamma capacitance than their respective triangular tower
models.
5) The conclusion is that, for top-loaded towers, modeling the triangular
tower structure in EZNEC with the HF beam atop is an accurate method of
finding the shunt-feed parameters required. The method of using
cylindrical tower equivalents for top-loaded towers will give a "ball
park" figure, at best.
6) The accuracy of any of the models presented here for non-top-loaded
towers has yet to be verified. Anyone shunt-feeding a non-top-loaded
tower that can provide their real-world tower and shunt-feed parameters
is most welcome to present them to see how they agree with these methods
of shunt-feed modeling.
73, GL, de Earl, K6SE
--
FAQ on WWW: http://www.contesting.com/topband.html
Submissions: topband@contesting.com
Administrative requests: topband-REQUEST@contesting.com
Problems: owner-topband@contesting.com
|