So You Want to Sell an Antenna. . .

L. B. Cebik, W4RNL (SK)

The following notes are not what they seem. My tongue was firmly in my cheek as I wrote these notes. In fact, I bit it several times just to bring me back to reality.

You have been working in your garage/basement/barn shop for days/weeks/months. The antenna design on which you have been working is finished. When you install it in the field, it actually works.

At this stage, most hams would simply operate and enjoy the contacts they make, whatever the context: contests/QRP/rag-chewing/etc. A subset of these operators would begin to catalog the things that the antenna will not do and begin research on the next antenna to replace this one. However, a few antenna builders will reach the conclusion that they can make up kits or packages and sell the antenna to other hams. But there are many antennas already on the market.

How do you make your antenna stand out so that people will buy it?

America is not only the the land of the free, it is also the land of marginal advertising. We began the tradition of claiming that toothpaste, deodorant, shaving cream, and cars will attract women to men and that hair dye, perfume, and mascara would attract men to women. Biology, health, intelligence, and personality had nothing to do with the matter. An interesting facet of the history of advertising is that we began to believe such things. Once belief set in, the facts no longer mattered.

So it is with antennas, although I am not aware that any amateur radio device has formed the attractive function between the genders. Still, effective advertising emphasizes the positive. Indeed, it puts the positive in bold-face, surrounds it with a frame, and then lights the frame so that it outshines all other considerations. The merely neutral and certainly the negative lurk silently in the shadows, unexpressed, unmentioned, unacknowledged.

A good engineer will want to tell a complete unvarnished story to anyone who may wish to use a given antenna. The story will be filled with facts and figures, and told without similes or metaphors. That is simply good engineering ethics. It is also why large corporations carefully separate their engineering and their marketing departments. Nothing kills a quick and irreversible sale like the whole plain story.

So the first step in learning how to sell your antenna--your magic antenna--is to come in from the shop, take a shower, put on clean clothes, and become your own marketing department. The second step is learning how to selectively express truths about your antenna to make it appeal to a collection of consumers who are waiting to hand someone their money. The decisive element in whether that money goes to you is not the intrinsic quality of your product. Instead, it is the show that you give them to acquire a belief in your product. Once you have that, even mediocre performance will not convince them that they did not buy the best that there is.

The following notes are designed to show you ways in which you can present some kinds of antenna products so that they engender belief in the product. In no case will any of the techniques falsify a fact. Indeed, for each case that we present, we shall see both the whole unvarnished story and the fact to present. After all, there are laws against false advertising--not to mention a plethora of ethical and religious traditions.

A Wire Antenna Example

Let's start out simply. In our shop, we have developed an off-center-fed 137.5' AWG #14 copper wire antenna. The feedpoint is 42.88' from one end and 94.62' from the other. At the feedpoint, we install a 4:1 balun and use coaxial cable as the main feedline. What can we possibly say about such an antenna that will distinguish the product from all of the other wire antennas on the market? Of course, we can use plastic insulators having exotic names, such as polycarbonate (and the many various trade names for the material). Or we can emphasize the fixture used for connecting the feedline to the wire and extol its virtues for durability. But most antenna buyers want to hear something about performance. They do not want an antenna that merely radiates. They want a "gusher," a "signal pusher," a veritable "RF volcano." How can we give them what they want?

First, we can emphasize the gain on 10 meters--without necessarily saying that we are talking only about 10 meters. Remember that silence is golden in selling. We can say that the gain potential is higher for our OCF than it is for a mere or plain center-fed doublet. We shall not be falsifying in the least. Fig. 1 compares the free-space E-plane (azimuth) patterns of a 135' doublet and our OCF at 28.5 MHz. The maximum gain of the OCF's strongest lobes is indeed numerically greater than the maximum gain of the strongest lobes of the center-fed doublet.

Of course, we shall also be silent about the fact that the OCF has a non-symmetrical pattern, compared to the doublet's symmetrical pattern. We shall also simply fail to mention that the OCF has more lobes and consequently more nulls than the doublet. And we shall certainly not mention that the OCF has many weaker lobes than the doublet on 10 meters, especially broadside to the wire itself (which runs across the page in the pattern).

If we are up on some antenna theory, we can make a stronger claim, again simply being silent about the fact that we are talking only about 10 meters. We may claim with numerical justification that the maximum gain is greater than we can obtain from a 2-element Yagi. Once more, we are falsifying nothing, as shown in the comparative free-space E-plane (azimuth) patterns that appear in Fig. 2. The black line indicates the pattern for a relatively standard 10-meter 2-element Yagi.

We shall not mention that the maximum gain differential is only 0.4 dB. Nor shall we mention that the maximum-gain lobes of the OCF are very narrow compared to the broad main lobe of the Yagi. We shall also be silent concerning the fact that the Yagi has a significant front-to-back ratio (10-11 dB across 10 meters) and is designed to be turned directly toward the desired station. (Pushing a 137.5' length of wire around in a circle for aiming is a daunting task, to be sure.) And we shall also not compare the element lengths (17.5' maximum for the 10-meter Yagi vs. 137.5' for the OCF).

Very likely, we shall also fail to mention that the gain of the OCF decreases as we reduce frequency. At 80 meters, the bi-directional pattern is virtually indistinguishable from a center-fed doublet pattern, although the doublet shows about 1.5-dB more gain. We shall also not mention that with the OCF configuration chosen, the lobes develop twice as fast as in the doublet. So the doublet, when 1 wavelength long, still has only 2 main lobes broadside to the wire, but the OCF shows a non-symmetrical cloverleaf. A symmetrical 4-leaf clover appears for the doublet at 20 meters, but the OCF has 8 lobes on this band. All of this information is likely only to confuse the potential buyer, so we shall refrain from mentioning such unimportant details.

More important to selling our OCF is the fact that we intend to feed it with a length of coax. If we model the antenna for 200 Ohms impedance, we shall obtain a fair picture of the SWR curves at 50 Ohms once we install the balun. Fig. 3 shows these curves.

The initial curves are not promising, although we can cover parts of the harmonically related ham bands with under 2:1 50-Ohm SWR with the balun. In fact, we are likely to measure better performance once we install the balun. First, if we measure the SWR at the far end of RG-58 coax (nearly the lossiest coax that will handle amateur transceiver power levels), we shall obtain better SWR figures. Second, the balun itself may introduce losses that reduce the SWR values. The actual losses will depend upon balun design and its ability to handle higher levels of reactance. As a sample, I inserted an arbitrary 50-Ohm load in series with the feedpoint to simulate possible balun losses. The result is a revised set of SWR curves, shown in Fig. 4.

Note that all levels are considerably lower. The antenna covers the CW portion of 80 meters and just about all of 40 and 20 meters with 2:1 SWR or less. As well, the antenna covers the main activity regions of 10 meters. These are all facts we might wish to stress, and a lossy test coaxial cable, such as RG-58, might even improve the coverage. Of course, we shall not show the actual curves, although we might show a sample if it is below 2:1 across an entire band.

At the same time, we shall not mention that balun and cable losses take their toll on the energy available at the antenna itself. The arbitrary loss insert into the system reduces 10-meter gain on the OCF by well over 1 dB. We might well just omit any reference to losses in the feedline and impedance transformation system and extol the virtues of the coverage we obtain using a coaxial cable. Besides, the tuners we find inside rigs can extend the coverage to all of the listed bands.

Many hams mis-label the type of antenna used here as a Windom. However, that 1929 invention used only a single wire as the feeder, and--of course--it radiated. The antenna that we are manipulating is an off-center-fed doublet, or OCF. Radiation from the feedline is minimal, although there are at least small imbalances in current on the two lines (whether we use parallel line or coax). The main radiation is from the horizontal wire, so we can safely say that the OCF is better than a Windom. If folks wish to believe that it is better than other marketplace OCFs that call themselves Windoms, that is their responsibility.

Every product needs a name. To sound truly professional, the name needs to leave the impression that it is descriptive. Since the antenna covers the HF bands, the letters "HF" are ready to use. Numbers--up to but not over 3--leave a good impression. So we can use 137, which is the rough length of the wire. So we have The All-Band OCF HF-137 as a perfectly plausible name. Note that we have avoided non-professional terms like "gusher" and "volcano." Those names belong to unlicensed users of the RF spectrum, not to relatively newly licensed radio amateurs with FCC-assigned call letters. The upward move in status calls for an antenna that has been engineered.

Our small ad in a magazine might look like Fig. 5. You can always enlarge it with testimonials, once they begin to roll in. Note that the ad seems to appeal not only to the beginning antenna user, but also to the experienced advanced user of complex expensive antennas. That gives the antenna's high quality and performance claims verisimilitude, even if no advanced antenna user ever buys one or even seriously considers using one. The claim is not false, since an all-band wire is in fact a good back-up and emergency antenna for complex stations.

I have not suggested a price for the antenna (actually, the antenna kit with wire, insulators, and balun). Do not make the price too low. First, you will find that the income does not cover the parts you buy plus the time it takes to make up kits and mail them out. Second, too low a price smacks of not being a professional-grade antenna. On the other hand, do not price the antenna too high and put it out of reach of your target market.

Remember that it makes very little difference that the highly positive impression that we leave with our emphases and our silences might just be a misimpression. Wire antenna buyers tend to be newer hams (in contrast to many experienced hams who build wire antennas from scratch) who normally do not have a comparable second antenna for making A-B tests. Hence, they will not be able to determine that the OCF is overall no better than or worse than a center-fed doublet, and they will thank us for not needing a separate antenna tuner. Nor do wire antenna buyers have test equipment capable of measuring system losses either directly or indirectly. They are only interested in what the SWR meter says and what they hear. Going from no antenna to our magical OCF design will let them hear a lot. We can even anticipate receiving testimonials praising the antenna, and we can always add excerpts to our advertising. User judgment counts, especially when highly favorable, regardless of the user's technical qualifications for making such judgments.

A Wire Bi-Directional Array Example

Our second example of selling an antenna involves a very old bi-directional beam called the Sterba curtain. The outline of this venerable array appears in Fig. 6. The core consists of one or more half-wavelength long sections, where each section requires 2 horizontal wires vertical spaced by 1/2 wavelength, with the whole array at least 1/2 wavelength above ground. To complete the array, we have on each end a pair of vertically stacked 1/4-wavelength wires.

At each end of the array, we connect the top and bottom ends of the horizontal runs with a single wire to close the array. However, the sections between the ends are connected with parallel transmission lines that are 1/2-wavelength long with a half-twist to ensure that the top and bottom wires in any section are in phase with each other. The sections should be open-wire (ladder) line for a velocity factor as close to 1 as possible. The actual impedance of these lines is not critical: any high value will do, and the array works fine with lines that are spaced from 3" to a foot. There are 2 possible feedpoints. We can place the feedpoint at a lower extreme corner for convenience. However, the natural losses even of copper wire will bend the bi-directional beams by about a degree or two. The version shown in Fig. 6 uses an odd number of interior sections and places the feedpoint at the exact center of the lower wire. The antenna ideally requires a 600-Ohm feedline to an antenna tuner for effective use with modern amateur equipment. However, since we shall specify the use of a tuner, virtually any high-impedance parallel transmission line will work well.

Fig. 7 shows the free-space E-plane or azimuth pattern for the array at its design frequency. For a 5-section Sterba, 2 wavelengths long at the design frequency, we obtain two main lobes, each with a free-space gain of about 10.4 dBi. The Sterba curtain has more gain than a 5-element Yagi. Of course, a 5-element Yagi has a beamwidth that is twice as wide as the Sterba, and the Yagi is designed to be rotated. However, we shall not let these facts deter us.

Commercial services stopped using Sterbas long ago. They found better large wire arrays, such as the rhombic, for point-to-point and SW broadcast services. Only hams still build occasional Sterbas. As well, virtually all Sterbas in commercial and government service were large and used in the lower HF region. However, you may note from Fig. 7 that our design sample is in the middle of 10 meters.

The Sterba for 10 meters is 69' long and 17.25' high. The minimum height above ground for effective service on 10 meters is 17.25', a half-wavelength. That places the top wire at 34.5' above ground. The interior wires in this model are 1' apart. In the end, the Sterba curtain requires about 1/2 the field space required by a 135' center-fed doublet. The fact that it requires almost 3 times as much wire as the 135' doublet (310' vs. 135') will not slow us down in the least.

The comparison of overall length between the 10-meter Sterba and the 135' doublet is critical to the discovery that will make the antenna a hot seller among wire users. That discovery is a simple one: we can use the feedline and tuner system to load up the antenna on all bands from 80 to 10 meters. We have invented The All-Band Sterba Curtain.

Fig. 8 shows the patterns of the 10-meter Sterba on the remaining HF bands. None of these bands yields a pattern with the high-gain bi-directional beams that we obtain on 10-meters. The Sterba curtain is a high-gain bi-directional beam only on its design frequency, where the half-twist interior connecting lines ensure that the current magnitude and phase are the same for each pair of top and bottom wires. On all other bands, the lines are no longer 1/2-wavelength, and so each pair of top and bottom wires will have different current conditions. As well, adjacent sections of the array have different current conditions. The result is that even the currents in the lines are no longer equal in magnitude and opposite in phase. Hence, the interior vertical wires may radiate to one or another degree of efficiency.

In the end, for all bands from 80 through 10 meters, we obtain the free-space performance levels shown on the left in the following table. The total pattern is composed of vertical and horizontal components, and the second data column list the dominant component.

A Comparison of Free-Space Gain and Feedpoint Impedance Values:  A 10-Meter Sterba Curtain and a 135' Center-Fed Doublet
10-Meter Sterba Curtain 135' Center-Fed Doublet
Frequency Maximum Gain Dominant Feedpoint Impedance Maximum Gain Feedpoint Impedance
MHz dBi Polarization R +/- jX Ohms dBi R +/- jX Ohms
28.5 10.41 horizontal 570 + j 20 5.27 2600 - j 850
24.94 5.43 vertical 930 + j 170 4.70 130 - j 1000
21.225 4.29 vertical 2300 - J 1700 4.58 2900 + j 980
18.118 2.17 mixed 20 + j 80 4.66 130 + j 15
14.175 2.32 vertical 3700 + j 1800 3.83 4100 + j 130
10.125 -1.20 mixed 110 - j 470 3.49 90 - j 320
7.15 2.30 horizontal 250 + j 1800 3.73 5000 - j 2400
5.358 0.52 horizontal 30 - j 360 2.67 430 + j 120
3.75 1.78 horizontal 95 - j 180 2.08 90 + j 100

Of course, if we place the antenna in the vertical position prescribed for lower-HF versions of the array--that is, 1/2 wavelength above ground at the design frequency--the antenna will be much too low for more than NVIS operations below 20 meters. But let's not dwell on that difficulty. The chart tells us that we can load the antenna on all HF bands with not too much strain on the tuner limits with any common parallel transmission line.

The table's right-most columns provides us with the free-space performance of the usual 135' center-fed doublet. The impedance listings show us why some doublet users prefer somewhat shorter or longer lengths for the antenna wire. At 135', the antenna is nearly resonant at the center of 80 meters. Hence, on 40, 20, 15, and 10 meters, it is very close to a perfect even number of wavelengths long, resulting in extremely high values of impedance. If we shorten the antenna to about 125', the antenna is closer to resonance at 4 MHz, with even harmonics that fall outside the ham bands. Hence, the impedances within the ham bands give the tuner a little less pressure on its matching limits. We may also note in passing that the Sterba curtain feedpoint impedances are in almost all cases more amenable to common parallel transmission lines.

Fig. 9 presents the free-space E-plane patterns for the 135' center-fed doublet for comparison with those of the all-band Sterba curtain. Although we cannot prevent the emergence of multiple lobes, note that the lobes evolve in an orderly and predictable progression as we move from 80 to 10 meters. The table tells us that the doublet actually has a higher maximum gain on all but 2 bands. However, newer wire users are more interested in an easy match than they are in the gain numbers (except for 10 meters), so we may simply pass over the gain columns in silence.

We are now ready to advertise our all-band Sterba curtain. In accord with good wire antenna marketing principles, we shall stress the positive aspects of the array and hold our tongues with respect to anything that someone might cite as a negative. The result might be a small ad like the one in Fig. 10.

You will recognize most of the claims from the discussion, including the potential misimpression left by the gain claim. We shall not worry about of oxymoron of calling a Sterba curtain (essentially, a monoband array with a fairly narrow frequency range for maximum performance) an all-band antenna solely on the basis of the ability to effect a match. The ad also stresses the all-copper construction of the antenna, since copper has a good reputation as a conductor. Of course, aluminum would do as well, except for the difficulty of creating junctions. However, we shall package the kit as a single length of wire to start and stop at the feedpoint. We shall throw in a handful of 1' plastic rods to create the vertical phase lines. You can reserve the directions for staking out the lawn to construct the antenna for a paper instruction sheet: words and paper are cheap.

A Yagi Example

In our wire-antenna examples, we have moved from marginal and dubious claims to blatant misimpression. However, when we turn to beam antennas, we change directions and slide into the realm of the downright subtle. In the marketing arena of beam antennas, the name of the game is gain--with a little bit of front-to-back ratio thrown in to cloud issues that might otherwise be clear. The heyday of gain claims for Yagis and other beams occurred in the 1980s, before the emergence of antenna modeling software. Indeed, for decades, some Yagi marketers tried to deny the validity of antenna modeling software in order to sustain overzealous gain claims. Such marketers would argue that models did not reflect adequate range tests, and if their antennas failed to show well in actual range tests, they claimed that the tests were flawed from an engineering perspective. Those were halcyon marketing days, indeed. One common habit was to make a simple sum of each gain advantage. For example, adding a reflector to a dipole added about 4 dB of gain. Adding a director to a dipole added about 4 dB of gain. Therefore, adding both a director and a reflector to a dipole driver must add about 8 dB to the dipole's inherent gain. Note that in the 1980s (and later) many marketers (whether of antenna products or ideas) refused to reference the decibel to any particular comparison or to a standard. Instead, they would claim a gain of XX dB and let the reader try to figure out what that meant. Of course, they hoped that most readers would be so impressed by the number XX that they would forget to ask about the standard of measurement.

In Europe, it is still popular--if not standard--to express the gain of a beam in terms of dBd, that is, gain in decibels over a dipole. However, even dBd contains an ambiguity that we shall explore. One rendering--the most usual--of the term is the dB gain over a theoretically perfect dipole in free space. The perfect dipole has an infinitesimal wire diameter with no wire losses, and hence its value is always 2.15 dBi, where dBi is the gain in dB over an isotropic source. So a gain claim for a 3-element beam over ground might by given as 11 dBd, instead of the US custom of reading the value as 13.15 dBi. Of course, the dBd gain is not the gain over a real dipole at the same height as the tested Yagi. It would show a gain of about 7.65 dBi or 5.5 dBd. Hence, the performance improvement of the Yagi over a real dipole at the same test height would be the same in both dBd and dBi terms: 5.5 dB. However, a claim of 5.5 dB improvement over a dipole does not sell as many beams as a claim of 11 dBd. The dBd purveyors, of course, defend themselves by claiming that a gain of 13.15 dBi is too high and too theoretical a number, conveniently forgetting that the basis of dBd is equally theoretical and equally misleading unless we also cite the gain of a dipole at the same height above ground in the same terms of measurement.

In this part of the exercise, we shall not try to sell a re-named dipole with a gain of 7.65 dBi or 5.5 dBd. Instead, we shall embark upon a more difficult task. First, we shall develop a fairly standard 3-element Yagi for 10 meters. In fact, it will be no better or worse than most of the 10-meter Yagis on the market. However, our task will be to make our Yagi stand out above the crowd.

Fig. 11 shows the general outline of our Yagi. The 3 elements fit on an 11.2' boom (12' with the necessary allowance for boom-to-element hardware). The sample uses 1/2" aluminum elements, which is impractical in actual construction. But the performance of a version with stepped element diameters will not vary significantly from the baseline model. The design frequency gain is about 8.11 dBi in free-space. The gain is about a full dB greater than the standard Yagi in the ARRL Antenna Book. Of course, that Yagi fits on an 8' boom, but we need not emphasize the smaller size of the comparator.

Yagis in this class have resonant feedpoint impedances in the mid-20-Ohm range. Therefore, this design shortens the driver to provide some capacitive reactance in series with the feedpoint resistance. By connecting an inductive reactance of the correct value across the feedpoint, we obtain an L-network matching section that raises the impedance at the terminals to 50 Ohms resistive. The inductive reactance has the form of a shorted transmission line stub, which we also call the hairpin or beta match. There are 3 general ways to elevate an impedance in the 25-Ohm range to 50 Ohms. With a resonant driver, we might use a 1/4-wavelength section of 37-Ohm coax (parallel sections of RG-59). We may use the capacitively reactive driver with a gamma match, with its rod and series capacitor. By comparison with the beta match, the gamma is mechanically more complex with more junctions to weather over time. The beta match introduces the fewest mechanical junctions to suffer from the long-term effects of weather. There is no evidence that any one of the 3 matching systems has any more or less loss when new and solidly constructed than any other system.

The 3-element high-gain Yagi performs in a normal fashion. A completely fair description of the Yagi would show the performance curves of the antenna across the intended operating range, 28.0 to perhaps 29.0 MHz. Fig. 12 shows the gain pattern and the front-to-back pattern. The curve marked "front/back ratio" shows the 180-degree value, while the curve marked "front/side ratio" provides the worst-case values for the front-to-back value.

The curves not only give us the operating characteristics of the antenna on 10 meters, but as well, they provide us with the information we need for the first set of marketing claims about the antenna. For example, the gain climbs in nearly linear fashion as we raise frequency from 7.87 dBi in free space to 8.44 dBi. The average gain is 8.13 dBi. However, in the passband, the Yagi has a gain potential of 8.44 dBi maximum. The curves show an average front-to-back value of nearly 22 dB (180-degree or worst-case) across the band. We may remain silent about the fact that front-to-back ratio drops below 20 dB at both ends of the band, and certainly we shall not talk about the under-15-dB value at 29 MHz. Instead, we shall note that the maximum front-to-back value is over 29 dB at 28.4 MHz. We may also forget to note that the values for maximum gain and for maximum front-to-back ratio do not occur at the same frequency. We shall be in good company with some Yagi makers on the commercial market if we simply cite the maximum values for these parameters.

Fig. 13 shows the free-space E-plane (azimuth) patterns for the Yagi at the design frequency and at the band edges. If we wish to display a pattern, we shall certainly use the mid-band pattern, leaving the band-edge patterns to the users imagination. However, as we continue our marketing research, we may find an even more attractive pattern.

To complete the picture, Fig. 14 graphs the feedpoint resistance and reactance of the Yagi, along with the 50-Ohm SWR curve. The graphs include the beta match. Due to the transformation of complex impedances, fixed-component matching networks tend to reverse the direction of curves that we might obtain of the feedpoint impedance components without a matching section. The resistance would rise with frequency, and the reactance would form a curve with more capacitive reactance at the low end of the band and more inductive reactance at the high end. These details are significant for understanding more fully how the Yagi operates in this configuration. However, they are much too detailed for advertising the beam.

The SWR curve does not cover the entire band at the antenna terminals with a value that is below 2:1. However, the 50-Ohm SWR is below 2:1 across the most active region of 10 meters, which many operators terminate just above 28.8 MHz.

Besides stressing the positive attributes of our beam, we may also wish to compare its performance with a dipole as a more graphic demonstration of its stellar quality. The best way to do this is by testing or modeling the Yagi over ground and comparing it with a dipole at the same height above ground. The best test heights to use for this simple ploy are about 7/8 wavelength or 1.375 wavelengths. Otherwise, we may fail to arrive at the best marketing numbers in the comparison.

The following table shows the basic performance figures for the compared antennas over average ground. On the left are the numbers for the dipole. On the right are the numbers for the Yagi. The Yagi data includes a front-to-back ratio, which is not relevant to the dipole. Like the Yagi, the dipole is made from 1/2"-diameter aluminum. The + and - signs record the maximum and minimum values of selected parameter values within the limits of the height increment used.

A Comparison of Dipole and 3-Element Yagi Performance at Various Heights Aboce Average Ground at 28.5 MHz

Dipole 3-Element Yagi Yagi - Dipole
Height Gain TO Angle Feed Impedance Gain TO Angle Front-Back Feed Impedance Delta Gain
WL dBi degrees R +/- jX Ohms dBi degrees Ratio dB R +/- jX Ohms dB
0.3 5.73 48 85.59 + j5.95 10.47 35 29.17 52.44 + j0.83 4.74
0.4 6.27 35 80.53 - j8.28 11.45 29 30.69 + 50.32 - j1.84 5.18 +
0.5 7.23 28 68.14 - j9.14 12.19 25 25.63 48.21 - j1.98 4.96
0.6 7.76 + 23 63.47 + j0.35 12.69 22 24.74 - 46.78 - j0.59 4.93 -
0.7 7.51 20 69.82 + j7.09 12.95 19 30.72 47.63 + j1.33 5.44
0.8 7.14 - 17 77.27 + j3.78 13.09 17 38.70 + 49.40 + j1.02 5.95
0.9 7.20 15 76.67 - j3.40 13.21 15 28.13 49.65 - j0.38 6.01 +
1.0 7.63 14 70.44 - j5.04 13.34 14 24.77 - 48.54 - j1.25 5.71
1.1 7.91 + 13 67.09 - j0.22 13.41 12 25.17 47.63 - j0.58 5.50 -
1.2 7.77 12 70.37 + j4.17 13.48 11 29.62 47.86 + j0.61 5.71
1.3 7.52 - 11 75.15 + j2.57 13.54 11 32.71 + 48.93 + j0.70 6.02
1.4 7.52 10 75.15 - j2.08 13.60 10 27.75 49.30 - l0.15 6.08 +

A dipole's gain undulates as we raise the dipole height, with peaks at about 5/8 wavelength and every half wavelength above that level (1.125, 1.625, etc.). The peaks become less pronounced with increasing height, and above 2 wavelengths are not especially discernable. Likewise, the dipole shows minimum gain values at about 3/8 wavelength and every half wavelength above that level (0.875, 1.375, etc.). With its inter-element coupling and the projection of the reflected wave forward of the structure, a beam shows less sensitivity to the height above ground, although we can detect differences in the rate of gain change between increments as we elevate the antenna. More sensitive to changes in height is the front-to-back ratio, with peak values (meaning least rearward radiation) corresponding roughly to the dipole heights of minimum gain.

When comparing a Yagi to a dipole, pick a test height that will provide the best marketing numbers while falling within the range that users might place their antennas. At a test height of about 7/8 wavelength, our 3-element Yagi shows over 6-dB gain beyond that of a dipole. Being able to honestly claim more than 6-dB gain advantage will attract buyers more surely than a claim of 5 dB or even of 5.5 dB, both of which are equally honest numbers, but relevant to different test heights. Note that between heights of 0.5 wavelength and 1.0 wavelength, the gain differential varies by more than a full dB. (If we were interested in selling a dipole, we might pick a test height of 0.5 wavelength and then claim that our dipole's gain is less than 5 dB lower than a 3-element Yagi. Or, preferably, we would select a short boom Yagi and claim the gain was less than 4 dB lower. Even better still, we might claim the dipole to by down by only about a half S-unit relative to a 3-element Yagi.)

We may also use a selected test height to obtain a radiation pattern to use in some of our advertising. The selected pattern must use a test height that yields the highest front-to-back ratio. Fig. 15 shows the azimuth pattern at the TO angle listed in the table for a height of 0.8 wavelength. We shall not be concerned that the height is not the same level from which we derived our gain differential number. The azimuth pattern is normalized to the maximum gain within it. Hence, it does not show the gain value associated with it. Some programs print that value to the side, but we can always graphically delete the number. An azimuth patterns with a very high front-to-back ratio almost always leaves the impression of superior performance, even in the gain category.

We now have accumulated enough positive facts to warrant calling our beam "The Superior 3-Element Yagi." So our ad might take the form of Fig. 16. I have omitted the pattern shown in Fig. 15, but you may wish to fit it into the space.

Of course, we shall not in the ad tell anyone that the use of a beta match requires elements that are well insulated and isolated from the boom. Hence, the construction uses one or another means of achieving this goal. We shall also not mention that some methods are better than others, since our ad does not discuss this facet of the beam at all. Performance and tough durability are the hallmarks by which we want to attract customers to this antenna. Even though all antennas deserve preventive maintenance at least once a year--if not more often--most antenna buyers want to believe (without having to say so out loud) that they can install the antenna and have it work to specification indefinitely.


The principles folded into the three examples in this exercise should guide you, whatever the type of antenna that you wish to market from your garage or workshop. They are not fantastical products of a fertile or even a cynical imagination. Every one of these principles has appeared in printed advertisements for antennas sold somewhere in the last 2 decades. The exercise antennas differ from the sources, and I have applied the principles in a concentrated fashion to the samples. Therefore, anyone who has used any of the principles can always say "That is not me."

Nevertheless, the principles sell antennas and yield what marketing and advertising count as being factual and evidentially based claims. Remember that selling an antenna depends as much on what we do not say as it does on the claims we put into print.

Note also that we have not appealed to any untested, unproven, or even controversial theories of operation for our antenna. While such theories appeal to a small market segment, they limit sales among the largest part of the antenna-buying public--our target consumer group. Most folks want to relate to the antenna by using terms that they have heard and that they think they understand. The more positive links between buyer and product, the stronger is the buyer's belief in the product.

Some skeptics--usually found among buyers rather than sellers and engineers rather than marketing experts--might ask whether these principles are fair. A discussion of that question would need another exercise entirely.

Return to Index