Consider the 2-element wire 40 meter beam. It is an improvement over the dipole in several ways. 1. It provides forward gain; 2. It provides rear attenuation; and 3. It lowers the elevation angle of maximum radiation by a few degrees. Most of these advantages are captured in the elevation plot below: Gain = 9.5 dBi; Front-to-back ratio = 14 dB; TO angle = 35 degrees; Feedpoint Z = about 50 ohms, all at a height of 50' over average ground, centered at 7.15 MHz.
For reasons that appear below, let me give the performance figures for an elevation angle of 18 degrees, near the lower -3 dB point. Gain = 7.5 dB; F-B = 12.7 dB.
Here are the dimensions that will yield this performance. Driven element = 66'; reflector = 70'; spacing = 20' with #12 or #14 copper wire. However, the 2:1 SWR bandwidth of the antenna covers only about 2/3rds of the 40 meter band. Moreover, especially on the low end of the band, the pattern goes to pot.
Let's redesign the antenna by making one simple change: increase the wire size to 2" in diameter. The dimensions for this fat wire Yagi are these: Driven element = 64'; reflector = 70'; spacing = 20'. Now the 2:1 SWR bandwidth of the beam exceeds the limits of 40 meters, as the following table demonstrates:
Frequency Gain Front-to-Back Feedpoint Impedance SWR MHz dBi dB R +/- jX ohms 7.0 7.9 11.1 36.4 - 13.4 1.6:1 7.1 7.6 12.5 45.3 + 3.3 1.1:1 7.2 7.4 12.6 54.0 + 18.3 1.4:1 7.3 7.3 11.9 62.2 + 32.0 1.8:1
Of course, 2" wire is impractical, and 2" diameter tubing is too heavy for any installation. So the antenna is impractical--unless we remember that we can simulate fat wire with an array of thin wires spaced apart. The simplest scheme to achieve most of the benefits is to use two wires making a flat wire about twice the diameter of the wire used in the model. This 2:1 rule of thumb is not precise, but adequate for most simple design cases. Take two wires and a bunch of spacers (1/2" thin wall CPVC is an adequate substitute for varnished or parafinned wooden dowels) and make lengths of flat 4" wide wire. Not only connect the ends, but as well solder shorts across the wire periodically. Now we have the material for a wide-band 40-meter beam.
Below are outlines of the two beams. Taking the Yagi first, he uses 2 65' long #12 wires, spaced 21' apart. From each, he hangs a length of 50-ohm transmission line (9914 with a velocity factor of 0.78). The lines go to a switch, whose common terminal goes to the coax coming from the shack. Carrol switches in one direction, making the hanging line in that direction an extension of the shack coax and hence, the line to the driven element. The remaining line is not connected and becomes the load for the reflector.
For this Yagi design, to maximize the performance pattern, the reflector needed an inductively reactive load of about 75 ohms. Transmission lines between 0 and 90 degrees long, when shorted, provide inductive reactance. Between 90 and 180 degrees, transmission lines provide inductive reactance when open circuited. Carrol chose open-circuit 146-degree lines (43' 4") to suit his situation. However, you can also use shorted lines of 56.5 degrees (16' 8.7") to do the job. If you need to bring the line near the ground for switching, you can add 180 degrees (53' 2.3") for a total shorted line length of 70'. If you use shorted lines, just be sure that the "unused" switch or relay positions go to ground; if you use open circuit lengths, leave the contacts open.
The next figure provides azimuth and elevation patterns of the beam at its projected 55' height.
Note the gain--a little over 10 dBi--and the front-to-back ratio--a little over 14 dB. Although not astounding when compared to highly elevated many-element 20 meter beams, the antenna will enhance 40 meter operations very nicely--and in two directions.
Searching for a little better front-to-back ratio, AA2NN adapted the Moxon rectangle for reversible operation. The sketch provides the dimensions of Carrol's #12 copper wire model. Because the equalized Moxon rectangle optimized for front-to-back ratio has a slightly higher feedpoint impedance, Carrol used 70-75 ohm cable as his projected feedline. The "hanging" feed-load lines are 75-ohm, 0.83 velocity factor coax. Carrol used 42' 7" lengths of open circuit line for this antenna, although corresponding shorted lines might also have been used. AA2NN does remind us that coax loading lines are not lossless and may be lower in Q than we may initially think, especially when we use them in longer lengths for convenience. The resistive losses will decrease gain by a small amount.
The projected performance of the Moxon version of the reversible beam shows clearly the enhanced front-to-back ratio and the reduced gain relative to the reversible Yagi. Which of these two very usable antennas one might select will depend both on the needs of one's operating situation and on how much high horizontal space one can give to the antenna. The Moxon is almost 20' shorter than the Yagi.
My thanks to AA2NN for letting me add these antennas to this note.
The wire Yagi is the ultimate in simplicity for a directional antenna, but it may not be the best for all types of operating goals. We often forget that we can add parasitical elements to almost any wire antenna. Parasitical extended double Zepps were known back in 1938. More practically, a half square will fit the half wavelength horizontal space of our Yagi, with vertical wires dangling from the 50' high point to about 12 to 14 feet or so above ground. Can we add a reflector about 20' or so behind a half square and change the bidirectional pattern to a monodirectional one? Yes, as the plot below demonstrates. The operating performance of a #12 wire parasitical half square is given by these numbers: Gain = 6.6 dBi; F-B = 23 dB; TO-angle = 18 degrees; Feedpoint impedance = 56.9 + 3.4 ohms. The reason for giving the 18-degree performance figures of the Yagi is now apparent.
The dimensions of this wire parasitical half-square are these: Horizontal length of both elements = 68'; driven element vertical length = 34.8'; reflector vertical length = 35.9'; spacing = 20.4'.
The advantage of the half-square is that at elevation angles below 18 degrees, its gain drops off much more slowly than does the Yagi gain. In addition, it lacks significant gain above 35 degrees, reducing incoming high angle QRM and QRN. These are, of course, advantages to the DX operator; the Field Day and Sweepstakes operator may prefer the Yagi precisely because of its higher angle radiation pattern.
One distinct disadvantage of the wire parasitical half-square is narrow bandwidth--about 100 kHz on 40 meters. To increase the bandwidth both in terms of 2:1 SWR and pattern retention, we must increase the wire size to about 6" in diameter. Then we obtain these dimensions: horizontal length of both elements = 68'; driven element vertical length = 35.2'; reflector vertical length = 37.6'; spacing = 20.4'. Some may find it odd that we increase the element lengths as we fatten the wire of the half-square. However, remember that the half square belongs to the family of 1 wl loop antennas, and like a quad, lengths grows with wire diameter.
With these dimensions, we can achieve the elevation plot below.
Here is a chart of performance checkpoints through the 40-meter band:
Frequency Gain Front-to-Back Feedpoint Impedance SWR MHz dBi dB R +/- jX ohms 7.0 6.9 10.1 37.5 - 21.8 1.8:1 7.1 6.9 21.2 60.5 + 4.6 1.2:1 7.2 6.5 18.6 79.3 + 15.1 1.7:1 7.3 6.1 10.0 85.9 + 23.1 1.9:1
The design center of the fat-wire half square was 7.07 MHz. Selecting this lower frequency was necessary to preserve a directional pattern across the band with a reasonable SWR figure at both band edges.
Like the fat-wire Yagi, the fat-wire parasitical half-square requires construction of the antenna wires using the same principles, but this time with a spacing of about 12".
The figure below compares the shapes of the two antennas and summarizes both #12 and fat-wire dimensions.
Expect both the driven loop and the the reflector to be a bit shorter than a resonated single loop, with the driven element shorter than the reflector.
If you care to scale some numbers from 7.15 MHz, here is a right angle delta loop and its 2-element counterpart. Given are the baseline and height (one is twice the other), and the sides are about 1.414 the height. This model had a maximum height of 60.4' which was held constant for the 2-element version to achieve comparable TO angles (17 degrees for the model)
Antenna Baseline Height Spacing single ra delta 60.8' 30.4' --- 2-el ra delta driv. el. 59.3' 29.65' reflector 60.6' 30.3' 20.5'
When made into a parasitical beam, the deltas also show reduced 2:1 SWR bandwidth (relative to their resonant impedance). At 40 meters, both SWR and pattern begin disintegrating somewhere around +/- 50 kHz from the design point with #12 wire. Widening that bandwidth depends upon using truly fat wires with equivalent diameters of about 6" at 40 meters for full band coverage with reasonable gain and F-B (arbitrarily defined here as 3 dB gain over a single loop, greater than 10 dB F-B, and less than 2:1 SWR).
K1KP uses a simplified version of the ON4UN feed for his 80 meter delta loops, which have their apices up at 70 feet and base legs about 8 feet off the ground. The apices are spaced 20 feet apart, with the bases spread to a distance of about 50 feet. He reports that the feedpoint impedance is close to 100 ohms. Each feedpoint runs to a central switch, roughly as sketched in the drawing. (Not shown in the drawing are baluns at each loop feedpoint to isolate each antenna. Also not shown is the tilt of each loop toward the other.)
Coax sections A and B to the switches are lengths of RG-11/U foam (with a higher velocity factor than non-foam coax) coax, 36 feet long. The switch is a relay that selects one feed as the driven element. 16 more feet of RG-11/U foam coax adds to the 36 feet on the driven element to form a quarter wave matching section, yielding a 50-ohm impedance for the coax to the shack.
The relay also shorts out the end of the other line forming the reflector. The shorted 36 foot coax line functions as a loading inductance to lengthen the electrical size of the loop in use as a reflector.
K1KP reports reasonable flat SWR and detectable gain over the single loop with this system, which is fairly close-spaced (average distance = about 1/8 wl) as parasitical systems go. It represents an ingenious way to switch beam directions and simplify feedline requirements without sacrificing performance from the wire array.
Incidentally, models of the half square and the single loop DMS (otherwise known as a side-fed rectangle) show about a dB gain advantage over the delta loops, and this gain also transfers to parasitical arrays of them. Of the antennas investigated, the half-square has the highest gain and front-to-back potential at more than 6.5 dBi and more than 23 dB respectively. The side-fed rectangle shows nearly comparable figures, but is among the most narrow-banded of the SCV configurations in parasitical application. The side-fed rectangle should be used in a single-loop configuration for parasitical use, since the feedpoint impedance reaches about 40 ohms at a spacing of 25' on 40 meters, while the double loop variety has a feedpoint impedance of over 120 ohms at the same spacing.
Moreover, it is feasible to electrically tune the reflector of any of the SCV parasitical arrays with no significant change of beam performance. This fact makes it possible to design two identical loops/half squares for resonance in the beam configuration and lengthen the reflector with a coaxial stub. The stub can become part of the feed cable when the loop serves as a driven element and can function as an inductive reactance when the loop is a reflector. Using a switching system similar to the one used by K1KP, but designed for direct 50-ohm feed and reflector stub, a reversible beam results with excellent front-to-back ratio and about 3.2 dB greater forward gain than a single loop/half square. Remember that if the required reactance is low, calling for a short stub that will not reach the switch in the center of the two loops, you can use at least two means of getting a longer stub. First, you can add a half-wavelength of coax (remembering velocity factor) to the stub. Second, an open ended stub, which is capacitive at less than 1/4 wl become inductive over 1/4 wl. This latter technique will likely be the most useful for these applications.
As the figure shows, a 2-element wire Yagi may give the same or slightly more gain at 17 degrees elevation when mounted at the apex height of the delta. However, its elevation angle of maximum radiation is about double that of the SCV group, and the SCVs have higher gain below the 17 degree mark, with 1/2 power points ranging from 7-10 degrees elevation, depending on the actual antenna height. Hence, the choice of antenna types depends on the user's operating goals.
The 70' by 21' rectangular area necessary for these antennas is about the
same. If the concept of a fixed position wire beam is useful, then the
decision as to which antenna to build may rest on which performance
characteristics one prefers relative to one's operating goals. However,
whichever you build, it is wise to take the added pains of fattening the
wires to give good performance across the band.
Updated 9-7-97. © L. B. Cebik, W4RNL. 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.