Pursuing the (Nearly) Perfect Parasitic Vertical Array for 160 Meters
Part 2: Some Parasitic Possibilities

L. B. Cebik, W4RNL

In Part 1 of this short series, we examined the modeling practices by which analysis of 160-meter vertical arrays are made and designs created. In addition, we took a brief look at phased arrays using a single central support, of which the K8UR system is one of the most outstanding.

In this part, we shall examine the possibilities of using a single driven element and using parasitic wires--guys--in order to create a vertical array. Whether or not practical, the possibilities are most interesting from a design perspective. Throughout, we shall observe the modeling cautions established in Part 1, with all modeling done over radial systems in NEC-4 unless clearly specified otherwise.

A 4-Guy Parasitic Array

It is possible to transform a version of the K8UR array into a parasitic array. However, the guys would extend in the normal fashion. Immediately lost would be the suppression of very high angle radiation from the antenna. In return, the builder would gain the simplicity of feeding a single element without concern for relative current magnitude and phase on the guys. The resulting array would have the appearance of the outline in Fig. 20.

The achieve maximum performance from the array, it is necessary to use a guy angle a bit less than the standard 30 degrees: about 27 degrees appears optimum. As well, as Table 3 shows, the director, the reflector, and the two side guys have different lengths. The table shows these values along with modeled performance data. The feedpoint impedance assumes base feeding of the 1/2 wl central tower, although shunt feeding and other systems are equally applicable to the array.

 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

    Table 3.  A 4-Guy Parasitic Array
Tower:  80-m high, 250-mm diameter, base 1 m above ground; guys: 25-mm diameter,
spaced 0.6 m from tower with virtual anchor point at 78.7 m; no radials.

Soil        Element Length-m        Gain  TO    Fr-Bk  B/W        Feed Z
Quality     Dir   Ref   Sides       dBi   angle ratio  -3 dB      R +/- jX
Very Poor   72.75 79.0  83.5        2.4   26°   29.3   154°       1590-4950
Poor        71.25 79.0  84.0        3.9   22°   32.1   130°       1330-4750
Good        70.5  79.0  84.0        5.0   20°   40.1   121°       1200-4700
Very Good   68.25 79.0  83.5        7.2   15°   40.4   112°        910-4690

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

To illustrate the pattern types produced by the array, Fig. 21 provides the azimuth and elevation patterns over Good soil. Because it is common practice to place this type of array over a radial system, I modeled the system in 3 configurations for each soil type. The table shows the values for the model using no radials under the driven tower. Fig. 22 graphs the gain of this configuration, along with two radial systems, one at ground level (Z=0) and the other with the radials buried 0.001 wl (about 0.164 m or 6.5") below ground. The radial systems consisted of 32 wires, each 2 mm in diameter and 1/2 wl long (81.91 m).

The graph shows that maximum improvement occurs with buried radials over Very Poor soil, an improvement of about 1.2 dB relative to the radial-less system. However, the use of radials does increase the elevation angle of maximum radiation by 2 degrees for Very Poor soil--and by lesser amounts for better soils. The use of radials in models shows virtually no improvement over Very Good soil.

In principle, the 4-guy parasitic array gives the builder the advantage of a system that is relatively easy to feed--something like a set-and-forget array. However, constructing the array is somewhat finicky. Changing the reflector length by as little as 0.5 m drops the front-to-back ratio to the 20 dB region, and slight length variations in the side guys have similarly large effects upon performance. Only the director shows slow changes in performance with changes in length. Hence, for any given installation where the ground conditions are not fully known, the potential for mis-building the 4-guy array is great.

A second disadvantage associated with the 4-guy array lies in the area of switching directions. Changing directions with the array requires altering the parasitic guys so that for each one, three lengths are possible. The complexities of such switching are likely no less than trying to optimally phase a K8UR array. Consequently, it is a strong question mark as to whether this system could live up to the promise of a parasitic system to provide greater simplicity than a fully phased guy-dipole array.

The Basic 7-Element 1/2 Wavelength Parasitic-Guy Array

The following section of this report outlines a potential improvement on the 1/2 wl parasitic guy arrays that is based on an antenna developed for AM BC use. It achieves--at only a slight reduction of array gain--the desired simplicity of switching with full horizon coverage for a steerable array. As well, the array is forgiving of small construction errors--that is, changes in element length yield only small changes of performance for all elements.

The array employs a driven element that is roughly 1/2 wl long. Automatically, a 79.2-m main element places this array outside the structural capabilities of most 160-meter operators, but that is no hindrance to a study of the antenna. In the course of the discussion, a number of modeling issues will arise, such as modeling with and without a radial system. These questions will be addressed along the way.

First, however, we should look at the antenna under study, sketched in Fig. 23. It consists of a central driven element, fed at the base. There is virtually no difference in performance when the antenna base is raised and lowered between 1 m from the ground to about 3.5 m from the ground, so long as everything else in the design remains the same.

The guys are 6 in number, 3 acting as directors and 3 as reflectors. Experiments began with the standard 3-element guyed system and proceeded to a 4-element system, as 2 of three equally spaced guys became reflectors. The double reflector systems showed some improvement. Adding the third reflector guy in line with the director (for a 5-element array) again showed improvement. However, not until I added the two remaining directors to fill out the sloping parasitic system at 60 degree intervals did improvements reach their limit.

The headings of Table 4 and Table 5 provide dimensions for the radials for two radically different models. The first uses a large 250-mm diameter main element (roughly 10"), with 25-mm diameter (about 1") guys. The second employs a more modest 25-mm (1") main element, with 2-mm (between AWG #14 and AWG #12) guys. The differences between the performance levels of the two versions range from less than 0.1 dB over Very Good soil to about 0.2 dB over Very Poor soil for all cases modeled.

The change in main element and guy diameters does require a small shift in parasitic element lengths for optimal performance. The reflector length changes are minimal: 81.5 m for the fat main element, 82.0 m for the thin driver. The changes needed in the director lengths are more extensive: 67.9 m for the fat driver, 71.0 m for the thin main element. All three directors and all three reflectors may each have identical lengths.

The guys may turn out to be, for some installations, pseudo-guy wires. They begin in fairly close proximity to the main elements, spaced only far enough away for safety relative to the high voltages at the ends of both elements. Each parasitic guy begins about 0.6 m away from the main element, about 1.1 m down from the peak height of the main element. The dimensions are partial functions of performance optimization and partial functions of the angle of the guys relative to the main element. The designation "pseudo-guy" arises from the fact that an angle of 27 degrees relative to the driver turns out to be best for the parasitic guys. (The angles away from 27 degrees toward 24 degrees and toward 30 degrees result in a slow rate of performance degradation. Thus, concern over precision--for example, in a worry over catenary angle changes--would be misplaced.) Since the optimal angle is less than normally used for top guys in commercial installations, the parasitic guys may best be implemented as non-structural wires. Of course, to prevent unwanted interactions, the structural guys are presumed to be composed of non-conductive materials. Incidentally, given the element diameters involved, all elements are modeled as copper. Changes in materials make differences only in the 0.0X column of gain figures and less than 1 dB in the front-to-back figures.

Performance and Radials

Table 4 and Table 5 present the modeled performance figures for the two versions of the array. Note that there are three sets of figures for each antenna model. The first group models the antenna over various soils using no ground radial system. The second places a radial system exactly at ground level. Had the driven element been connected to this junction, NEC-4 would have yielded some unusable figures. However, with no source segment attached to the Z=0 junction, the calculations produce normal results. The third set of figures places a radial system 0.001 wl below ground, the same depth as used with earlier models. In the present array, each radial is 81.91 m long, that is, a half wavelength. Both the at-ground and buried radial systems consist of 32 radials each.

 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

                Table 4.  7-element parasitic array: Version A

Driver = 79.2-m vertical, 250 mm diameter; 3 reflectors = sloping 25-mm guys,
81.5 m; 3 directors = sloping 25-mm guys, 67.9 m; single 81.91-m (0.5 wavelength)
radial system, 2 mm diameter, uniform segmentation: 21 segments per 1/2
wavelength (where used); NEC-4.

Soil Type         Gain  TO Angle    Front-to Back     Source Impedance
                  dBi   degrees     Ratio dB          R +/- J X Ohms

No Radials
Very Poor         2.15  25          19.92             3256 + j 3760
Poor              3.70  23          23.84             3329 + j 3806
Good              4.78  20          28.12             3350 + j 3854
Very Good         6.86  15          36.07             3410 + j 3893

32 Radials, 1/2 wavelength long, at ground level
Very Poor         2.64  26          22.38             3234 + j 3818
Poor              4.03  23          26.41             3331 + j 3843
Good              4.99  20          30.16             3362 + j 3878
Very Good         6.90  15          36.41             3415 + j 3897

32 Radials, 1/2 wavelength long, 0.001 wavelength below ground
Very Poor         3.27  27          25.54             3316 + j 4000
Poor              4.28  24          27.67             3364 + j 3937
Good              5.08  20          31.29             3390 + j 3917
Very Good         6.91  15          36.10             3420 + j 3905

 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

                Table 5.  7-element parasitic array:  Version B

Driver = 79.2-m vertical, 25 mm diameter; 3 reflectors = sloping 2-mm guys, 82
m; 3 directors = sloping 2-mm guys, 71 m; single 81.91-m (0.5 wavelength) radial
system, 2 mm diameter, uniform segmentation: 21 segments per 1/2 wavelength
(where used); NEC-4.

Soil Type         Gain  TO Angle    Front-to Back     Source Impedance
                  dBi   degrees     Ratio dB          R +/- J X Ohms

No Radials
Very Poor         1.94  25          22.02             2701 + j 2431
Poor              3.55  23          26.85             2698 + j 2325
Good              4.66  20          31.97             2724 + j 2266
Very Good         6.77  15          33.13             2706 + j 2185

32 Radials, 1/2 wavelength long, at ground level
Very Poor         2.48  26          25.44             2779 + j 2398
Poor              3.90  23          30.75             2732 + j 2294
Good              4.89  21          33.94             2735 + j 2238
Very Good         6.81  15          33.31             2706 + j 2178

32 Radials, 1/2 wavelength long, 0.001 wavelength below ground
Very Poor         3.18  27          30.37             2844 + j 2214
Poor              4.19  24          33.09             2770 + j 2203
Good              5.00  21          35.41             2741 + j 2188
Very Good         6.83  15          32.73             2707 + j 2168

 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

All the of models used the same set of element lengths. There is little change if one extends the towers close to the ground (<1 m). As well, it is possible to develop specific reflector and director lengths, not only for each central driven element diameter, but as well for each soil type. However, the reflector will vary only about 1 m across the range of soils used in this study, while the director may vary as much as 4 m.

The 6-guy parasitic array does not reach the gain levels of the K8UR optimally phased system for any given soil type. The K8UR advantage in gain is a consistent 0.6 dB, largely because the beamwidth of the parasitic array is about 35 degrees wider between -3 dB points than the fully phased array. The beamwidth of the parasitic guy array varies between 128 and 129 degrees for all versions. As well, the 6-guy system does not cancel out the very high angle radiation as well as an optimally phased K8UR array.

Compared to the 4-guy system, the 6-guy array has slightly less gain (0.3 to 0.4 dB), but equal front-to-back potential. Unlike the 4-guy system, the parasitic element lengths are less critical, especially the directors. For instance, over very good ground, one model varied the length of the director by a total of 5 m with a net change of gain of 0.05 dB and a net change of front-to-back ratio of under 2 dB. The reflector lengths are somewhat more critical. However, the key is to keep all directors and all reflectors as close to the same length as feasible. Once this is achieved during construction, one gains the advantage of operational simplicity in the 6-guy array: once set, it requires no operational adjustments and few mechanical adjustments over the lifetime of the system.

In Fig. 24, we can see the modeled gain differences among the different modeled radial systems. As soil quality improves, the differentials in gain disappear. Even with Very Poor soil. there is a maximum differential of about 1.1 dB between the no-radial and the buried radial models for the fat-element version shown in the graph. The thin-element version shows a differential of 1.25 dB.

Fig. 25 shows differentials in front-to-back ratio, using the fat main element. Once again, the greatest differential occurs over Very Poor Soil and amounts to about 5 dB. As the soil quality improves, the differentials decrease. The thin element numbers can be easily garnered from Table 5.

In Fig. 26, we can compare the performance of the radial-less version of the antenna over various soils, using overlaid elevation patterns. Since the far-field performance is not just a function of what is occurring within the near-field region of the antenna, but depends also on the soil quality at a distance in the so-called Fresnel zone, we expect rising performance with improved soil quality. Anyone interested in a study of the antenna in more complex soil conditions can create an inner and outer soil quality differentiation using standard NEC capabilities.

A secondary factor involved with soil quality is the elevation angle of maximum radiation (Take-Off or TO angle). Here, the tables make a good guide to modeling report numbers. Although all three versions of each of the two models show a 15 degree TO angle over Very Good soil, we should also note the increasing TO angle as we move the radials from ground to below-ground level for Very Poor soil. The differences are not large, but they are consistent for both the fat and thin main element models.

As main elements go, both the fat and thin elements are short, relative to a resonant length. However, they are both approaching the region in which resistance and reactance approach their peaks. The thinner element is electrically shorter and thus shows lower resistance and reactance numbers at the source. Interestingly, bringing either main element within 1 m of ground suffices to carry the element beyond resonance so that it shows a capacitive reactance.

The small changes in antenna performance occasioned by the presence or absence of a radial system under that antenna are bound to arouse questions relative to experiential reports of much greater changes for real 1/2 wl vertical systems. The differential of experience and modeling does not represent a fault of the modeling software, but owes to a quite different factor. Because the model places the source directly on the element, it represents a simplified model of the actual antenna system. An actual antenna might look something like the system on the right of Fig. 27, where a network does more than simply transform impedance for the convenience of the feedline to the remote source. The network also completes the antenna system at ground. For maximum system efficiency (or minimal loss), the source ground point and the antenna base ground point must show zero potential differential. At best, we can only approximate this condition in real systems, and a large radial system is one way to achieve this goal.

The simplified model that uses a source placed on the element represents, then, the maximum that such a system can achieve. Any losses in portions of the system not modeled must be separately accounted for, or one must create an adequate model that will accurately show such losses. In the end, one must view even the model using an extensive buried-radial system as a simplified model of the full antenna system. Since each parasitic element--as well as the main element--is coupled to the ground (with or without radials), the results that might emerge from a fully adequate model would be affected in rather complex ways. Even over Very Good soil, where the presence of a radial system appears to have minimal consequences for radiation patterns, overall system effectiveness considerations for physical antennas still make the use of a considerable radial system advisable with 1/2 wl radiators and parasitic elements.

Some Further Possibilities and Concerns

On the assumption that implementing the array is feasible enough to entertain further development thoughts, there are a few cautions to observe. Foremost is lightning protection. The matching network should provide a full and heavy path from deep ground to the main element. For many purposes, such a path will yield relatively good protection for the unterminated parasitic elements in the array. However, if the location of the antenna is in any way sensitive, then added protections should be provided to the lower ends of the parasitic guys in the form of very high impedance RF but very low impedance charge paths to ground. At the very least, each parasitic element should be provided with a static discharge path. An array of this order is not a toy and should be treated with the same care as commercial BC installations.

Having registered due caution, we can turn to happier potentials of the array. One of the more interesting is the fact that with a switch and a half dozen remote relays, we can steer the direction of the array in 60 degree increments. Fig. 28 shows the simplicity of the switching arrangement: a 3-pole, 6-position rotary switch. The object is to convert a director into a reflector with the remote relay adding a length of guy to a director. (Of course, relays with wide terminal separation and high voltage contacts are necessary, and should not be operated with power applied to the antenna.)

Since opposing guys will always be of opposite types (reflector or director), the simple switch suffices to provide relay activation, so long as the contacts are properly staggered to direct the antenna according to operator wishes. The indicated clockwise change in direction can be altered to suit the user's needs. Missing from the figure is the return line from each relay coil.

For most purposes, the broadness of the array pattern will not yield great increases in signal strength for a single 60 degree change in heading. However, one may also think of the switching arrangement as a rear null direction switch. The switch position might easily become a method of increasing the signal-to-noise ratio by deep-nulling a bit of QRM. The directional capabilities of the system give the array a bit of an advantage over shorter phased systems which would require several towers to achieve the same directional capabilities.

As a second potential for the array, consider erecting two of them spaced about 1/2 wl apart. The array might well look like the outline sketch in Fig. 29. For both sub-arrays, the directors are all on one side of the line between the two elements, with all of the reflectors on the other side. If we feed the system in phase, we obtain some interesting improvements in both gain and directivity.

Fig. 30 shows the azimuth pattern of the 250-mm diameter version of the array over good soil at a TO angle of 21 degrees. In exchange for a decrease in the absolute 180 degrees front-to-back ratio, we obtain an entire rear area that is never worse than 18 dB down from the forward lobe. As well, the forward lobe now has a beamwidth of under 64 degrees.

 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

               Table 6.  Two 7-element parasitic arrays in phase

For each array:  Driver = 79.2-m vertical, 250 mm diameter; 3 reflectors =
sloping 25-mm guys, 81.5 m; 3 directors = sloping 25-mm guys, 67.9 m;  array
spacing 81.95 m (1/2 wl); NEC-4.

Soil Type   Gain  TO Angle    Front-to Back     Beamwidth   Source Impedance
            dBi   degrees     Ratio dB          degrees     R +/- J X Ohms

No Radials
Very Poor   4.88  26          15.99             63          3837 + j 3726
Poor        6.46  23          19.06             60          3914 + j 3726
Good        7.58  21          21.54             58          3793 + j 3750
Very Good   9.63  15          26.54             56          4062 + j 3733

 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Table 6 shows the entire set of numbers reported for the large-diameter version of the array. The smaller diameter figures are similar. For every level of soil quality, the gain increases between 2.75 and 2.8 dB, with a worst-case beamwidth (Very Poor soil) of 63 degrees. As well, the front-to-back ratio reflects the performance around the entire rear section of the pattern.

The increase in performance has its price. The dual array is now a single direction affair--or at most, a reversible direction array, if there are two desired directions close to 180 degrees apart. Lost is the ability to swing the array in 60 degree increments at the twist of a switch.


The 7-element parasitic-guy array described here is just a design exercise, not intended for implementation in the amateur radio environment. However, it does demonstrate what may be possible in terms of achieving phased performance from a single parasitic array through the judicious placement and dimensioning of guys. In the process of developing the array, the exercise has also let us set into some perspective the fairly poor state of modeling of previous sloping-guy parasitic arrays and to show some directions that promise better modeling results.

One final note: the array design (for any frequency and variation of materials) is proprietary and described by permission for information purposes only. Commercial applications may not be undertaken without express permission or license, and a patent application may presently be pending. However unlikely such applications may be, this note is a necessary final word on our foray into developing a nearly perfect parasitic vertical array.

Updated 2-1-2002. © L. B. Cebik, W4RNL. The original item appeared in AntenneX for January, 2002. Data may be used for personal purposes, but may not be reproduced for publication in print or any other medium without permission of the author.

Go to Main Index