Joseph T. Semak

President, Tele-Images, Inc., San Diego, California


Recently, the advantages of certain new FM antenna designs have become recognized. As a result, there has been substantial interest in the use of special antenna designs. The principle catalyst for these efforts stems from new Federal regulations limiting the permissible levels of human exposure to non-ionizing RF radiation in areas near broadcast towers, a problem that can usually be solved by application of the correct antenna. Unfortunately, it is generally unknown to broadcasters that many other problems occurring at the transmitter site such as RFI/EMI, antenna coupling, and pattern distortion can be controlled, reduced, or eliminated through the appropriate use of a specially designed antenna.

This report will familiarize the FM broadcaster with the various operational concepts, design, and construction applicable to all FM broadcast antennas; this is necessary to fully understand the specialty antenna. The application techniques and performance characteristics of the specialty antenna are thoroughly covered, with emphasis on practical situations.


For many years now, the FM transmitting antenna has seen relatively few changes in its design although many different styles of antenna bays (radiating elements) have appeared, they are almost always integrated into an array using a standard vertical separation of one wavelength (one lambda). This applies to both the side mounted ring type antennas as well as (reflector) panel types.

All FM antenna systems (except single bays) utilize the broadside array principles to develop gain over a single bay of the same type. This is true regardless of the type of bay used: side mounted rings, directionalized rings, or panels. A standard broadside array for broadcast applications consists of a number of identical bays stacked vertically along the same axis at some fixed separation, all of the bays being supplied the same amount of RF power at an identical phase angle. (In the cases of null fill or beam tilt, there are small variations of the amplitude and phase, respectively, in the RF currents fed to the different bays.) This arrangement will produce a certain amount of gain over a single bay due to the concentration of power in the plane normal to the axis of the antenna array (coverage area). This concentration of power is called the main beam (beam maximum). As the overall length of the antenna is increased, the width of the main beam narrows proportionately, thus increasing the concentration of power and producing more gain.

Antenna gain

Definition.–Antenna gain is simply described as the amount of power concentrated in the main beam of an antenna as compared to the amount of power concentrated in the main beam of the reference antenna, assuming that both have equal input power. For FM broadcast, the FCC has adopted the reference constant as a one-half wavelength linear dipole with a one kilowatt input power. For convenience, gain can be computed given the equivalent E field (field strength) values at a fixed reference distance, such as one mile.

Controlling parameters.–It is important to understand that the gain of a broadside array is not in direct proportion to the number of bays (on the antenna) as is commonly believed. Gain is primarily a function of the effective length or aperture of the antenna in wavelengths as measured from the centers of the two outermost bays; assuming that the bays are spaced at least every 1.5 wavelengths along this length. The number of bays contained within any given antenna aperture will vary with different interbay spacings; however, the resulting changes in beamwidth and gain will be rather small as long as the antenna’s effective length (aperture) remains unchanged.

It is interesting to note that the maximum amount of gain for a given length of antenna aperture generally occurs when the bays are separated by five-eighths to three-fourths of one wavelength; however, other considerations overshadow this slight advantage, dictating bay separations based on the economy of one wavelength spacing, or the necessity of radiation suppression; calling for one-half wavelength bay spacing.

The directivity possessed by each of the bays in the antenna array is the final contributing factor to the gain of the array. The influence that the -individual bay has on the overall system, and the relationships involved are discussed in detail. later in the text.

Determination of gain.–The broadcast industry is used to describing (incorrectly) the gain of an antenna simply by the number of bays that it has. Presumably, this is because it is convenient with antenna systems using one wavelength interbay spacings. However, this will undoubtedly result in considerable confusion with the introduction of specially modified antennas.

It is suggested that it is appropriate to characterize an antenna by the effective length of the array–the first bay through the last bay–in wavelengths. This is not difficult, and by doing so a person can estimate an antenna’s gain regardless of the number of bays or the separation usedFor example: a three wavelength (effective length) antenna that is omni-directional and circularly polarized will have a power gain of approximately two regardless of its bay separation. Note that in this equation the antenna’s effective length is essentially divided by two; this takes into account the power division that takes place in a circularly polarized antenna.

Exact antenna gain can be determined through various mathematical and graphic procedures that are available in the antenna design textbooks. One method, known as the Poynting vector method, models the antenna array in the center of a large, imaginary sphere; thus allowing the power flowing out of the sphere to be determined in terms of density per unit of area. From this, the total power equivalent density and peak (main beam) power density can be determined. Relating the two power density figures will yield the concentration of power in the main beam relative to the total input power, and therefore, the gain of an isotropic radiator. This figure has to be corrected if a different gain reference is desired. When referencing to a half-wave dipole, the gain of the dipole–2.15 dB–has to be subtracted from any gain figure that has been referenced to an isotropic source.

Another method, algebraically derived, projects gain based on the knowledge of the RF currents flowing in the array. When using this method, the mutual impedances of every bay into every other bay must be known in order to establish the radiation resistances within the array. This is not practical though, as it is very difficult to accurately determine the mutual impedances through any type of analysis.

Unfortunately, no simple formula for gain exists that will give exact results.

Measurement of gain. –Gain can be determined through controlled measurements of an actual antenna. If performed with care, measuring the elevation pattern (and azimuth pattern, if necessary) of an assembled antenna on a good antenna test range is considered an acceptable method of determining gain, and is often done on television antennas. As an offshoot method, one could measure the elevation pattern of a single. bay (of the type to used on the antenna) and multiple the results by a calculated array pattern. This will yield an accurate figure if properly done, and can be. accommodated on a smaller test range.

Radiating Elements

Most common FM antenna bays (radiating elements) are based on the use of one or more. half-wave dipoles configured in such a way so as to produce the desired polarization, power handling, and bandwidth characteristics. Recently, quite a bit of attention has been centered around the design of the bay, (radiating element) and if it’s a panel antenna, the reflector as well. Different types of bays can produce substantially different radiation patterns; thus, the selection of a bay design is critical and plays a key role in the way the over-all antenna will perform in virtually all aspects.

Radiation Characteristics. -Ideally, an antenna bay for all side mounted ring type arrays and most panel types should possess the same relative field distribution for the elevation pattern as a perfect half-wave linear dipole and/or short horizontal loop located in free space. The radiator used as the reference depends upon the polarization(s) used. If the mode is vertical polarization, a vertically oriented linear dipole is the reference; for horizontal polarization the comparison is made against a horizontal loop. A perfect linear half-wave dipole radiates an electric field that is proportional to the cosine of the angle of radiation as compared to any point normal to the axis of the dipole. The short horizontal (end-loaded) loop has a donut-shaped field pattern (distribution) that is substantially the same as a linear dipole’s, but it is horizontally polarized. The horizontal loop is formed from a linear dipole to effectively provide the omni-directional, Horizontally polarized coverage that cannot be obtained from a linear dipole located with its axis parallel to the earth’s surface.

Operating Impedance.–Antenna bays must be designed to present correct and equal impedances within any array. This is generally accomplished through gamma matching and/or transmission line type matching principles applied on or inside the bay’s feed stem, respectively. This adjusts the dipole’s characteristic (low) impedance up to the proper level for correct RF distribution. Additionally, any reactance (undesirable) must be removed from the antenna in order to provide a purely resistive load at the operating frequency. This is done in several waysthe most common method is trimming the element length for proper resonance during manufacture. The bays used in modified antenna designs are usually identical to those used in the standard model antennas; however, there are sometimes differences in the way the operating impedance is determined and adjusted.

Mutual impedance.–The mutual impedances that result from the coupling of antenna bays is an area of particular concern when special antenna designs are approached. This is because the bays in most modified antennas are separated by less than a full wavelength. At large (one wavelength or greater) bay separations there is an insignificant amount of coupling among the elements; However, this situation rapidly changes as the bay separation is reduced to one-half wavelength. At this point, significant amounts of coupling will. occur causing changes in the bays’ radiation resistances that are related to both the magnitude. and phase of the coupled energy. Changes in a bay’s radiation resistance will produce similar changes in its input impedance. This will, in turn, affect the amount of power the various bays will radiate. Because mutual impedances result from the close physical presence of other bays, it is a non-uniform effect over the length of the antenna. The bays assigned to the center of the array will be affected to a greater extent than those at the ends of the array. Obtaining even power distribution throughout the entire array is the primary obstacle to overcome when large amounts of mutual impedances are present. To this end, the manufacturer must determine to what extent an array is (or will be) impacted by mutual impedances and correct the power distribution by setting the individual arrays’ input impedances so that each bay assumes the correct level when operating in the array. For one-half wavelength bay separations, the center bays will tend to have a lower radiation resistance than the outer bays; therefore, the center bays will receive more power. Power distribution within a specially modified antenna must be carefully controlled, otherwise the desired effect can be ruined.

Feed systems

Standard applications.–When employing a one wavelength antenna configuration the manufacturer simply taps the. main feed line at each bay location, resulting in an array with all of the elements fed in phase. This allows antennas to be produced rather simply and economically. The amplitude of the RF input to each of the bays shunted across the main feedline is dependent upon the impedance that each bay presents to the feedline. The bays are typically the same impedance in any one array, and thus the power for an eight bay antenna will divide, across the bays in the same way it would across eight resistors, of equal value connected in parallel.

Special Applications.–There are unique variations to the feed systems of antennas that have bay separations other than one wavelength. Uniform, in phase operation (RF excitation, not spacing) of the antenna’s bays is requisite in any broadside array. In order to achieve this, several variations upon the standard feed schemes can be used.

In the case of one-half wavelength separations, a sub-feeder system can be used or the elements can be shunted directly across the main feedline at half-wavelength intervals by inverting every other bay. The latter method corrects the phase inversion inherent at every other feedline tap point by reversing the side of the dipole receiving the RF excitation; effectively restoring the bay spacing to in-phase operation.

It is recommended that the sub-feeder type of system be. used to avoid compromise in antenna performance due to the asymmetry present in most bay designs. For the cases of bay separation greater than one-half wavelength, the common branch feed system is recommended as the most practical means of achieving the proper RF distribution.

Branch feeds.–When it is inconvenient to shunt bays or sub-feeders directly across the main feedline, as is the case with most panel antennas and certain modified antennas, a branch feed is used. A branch feed consists of a power divider at the input to the antenna from which individual feedlines–branches–run to their respective bays. This allows the proper phasing to be maintained simply by keeping the lengths of the branches a multiple of one wavelength.

Impedance matching.–There are different philosophies among the various manufacturers as to what the optimum impedance distribution within an antenna’s feed system should be. Questions arise as to where the transformations should take place in the antenna in order to achieve proper impedance matching and power distribution, while obtaining a standard 50 ohm input. Several different systems have been used to meet these requirements in the many different available antennas, and although each has its advantages, they will not be reviewed here since the subject is well covered in other publications addressing broadcast antennas. It is mentioned so that the reader is reminded of these differences, and the ways they may impact the methods described in this text.

Basis for alternative antenna designs

In practice, even the best antennas never quite perform the way they theoretically should–the way they really need to. This is best illustrated by recalling the discussion of the ideal antenna bay and the dough-nut shaped field pattern that it should have. Now imagine that the nulls in the fields above and below the antenna are not very deep. Lets say there is about 10 dB less radiation (approximately 0.3 relative E field) in the general directions of 0° and 180° (the zenith and nadir of the elevation pattern, respectively) as compared to the beam maximums at 90° and 270°.

Unfortunately, this depicts what the author has typically field measured from the “good” antenna bays. This is the primary source of the excessive radiation that is present below many FM broadcast antennas.

At this point, it should be apparent that if practical antenna bays could radiate power as hypothetically they do in their ideal equivalent form, there would be no downward radiation. There would be no need for any unusual bay separation; one wavelength bay spacing would work fine. This is because the overall field distribution characteristics, i.e., the azimuth and elevation patterns, from a broadside array are the product of the same patterns for both the individual radiators and the array.

Interbay spacing is an important parameter of the broadside array, and as a variable, is incorporated into all of the modified antenna designs considered in this report. In theory, interbay separation can (and may) be any value from infinity close to over a wavelength apart; all values included will allow gain from the array. Changing the number of bays within a fixed aperture will. change the arraypattern. The array pattern is defined as the full elevation pattern of a broadside array that substitutes (imaginary) isotropic radiators (equal radiation in all directions) in place of the bays actually used. The array pattern is calculated. It is the array pattern that generally determines what useful applications a modified antenna will have. This will be considered in further detail elsewhere in the text.


Modified antennas fall into two general categories: The first group consists of those antennas designed to reduce the extraneous RF fields, which represents the majority of current applications. Group two is made up of antennas designed for a specific application other than RF radiation suppression specifically. There are significant differences among these groups in both construction and end result.

Antennas for RF radiation suppression

The requirement for antennas falling into this group is primarily the reduction or elimination of undesired downward RF radiation from the antenna. Referring back to the discussion on antenna bays, the principle cause of radiation along the axis of an antenna is due to the imperfect characteristics of a practical transmitting antenna bay. In an antenna with bays spaced at one wavelength the array pattern is such that the radiation from all bays will add in phase along the axis of the array, forming an end-fire effect. Consequently, this augments the undesirable aspects of each bay. Since it is unlikely that anyone is going to devise and implement the perfect radiator, any approach taken to control unwanted downward radiation must take into account and compensate for the shortcomings of the individual bays.

Technique.–A suitable approach to the elimination of excess downward radiation involves modifying the spacing of the bays along the length of the array in such a manner that the location of the radiators causes a phase cancellation of signals above and below the array. This method can be referred to as space-phasing (of the bays) and does nothing more than achieve the desired array pattern. Locating bays every one-half wavelength along the length of the antenna will satisfy the above requirement provided the antenna’s total number of bays is an even number. (Assumes correct phasing of input.) An antenna with one-half wavelength separations, it can be seen that waves traveling from the antenna and normal to its axis add in-phase as in any other broadside array; however, as the waves from a bay radiate along the axis (vertically) of the antenna they traverse the adjacent bays where a virtually complete cancellation takes place as a result of the 180° electrical distance between the bays.

Other space-phasing schemes are possible, and depending upon your requirements, may be worthy of investigation. These alternatives can offer similar characteristics to the half wavelength system at reduced cost. For example, three-fourths wavelength bay separation could be utilized on antennas with four, or multiples of four bays. This type of antenna will have similar RF suppression as the one half wavelength case., although not to the same degree. Expect the downward RF suppression of a three-fourths wavelength antenna to be at least 3 dB less than that of a one-half wavelength type of equal gain. Cancellation within an array using three-fourths wavelength separations takes place across every other bay. In a similar fashion, other antenna system designs are possible if they too are configured to leave every bay on the array out of phase with another bay along the array. An example of this would be the use of ten bays in conjunction with 0.9 wavelength interbay spacings.

The primary reason for considering a special antenna system with other than one-half wavelength bay separations would be either lower required performance and/or lower cost. (Applies to group one antennas only.) A half-wavelength type configuration will typically use about twice as many bays for the same amount of gain as a standard model antenna, and the feed system is complex; making this antenna expensive. It is, however, the best choice for a difficult transmitter site situation.

When contemplating specially designed antenna systems with other than one-half wavelength style, careful planning is necessary because of the extensive feed systems, particularly on the larger antennas. A branch feed is probably the only practical method of feeding these antennas, and with a large branch feed installation, special attention must be directed towards integrating the feed system onto the tower. This could partially offset any cost savings provided by these latter antenna systems. Don’t forget to take tower windloading into consideration anytime an antenna with increased windload is contemplated.

Application.–An antenna used for RF radiation reduction must be properly applied to yield the best results. The gain of the array as well as the arrays location on the tower are important considerations that must be chosen to provide the desired cone of silence. The cone of silence as defined by this author, encompasses the area below (and above) the antenna that has RF radiation suppressed by at least 20 dB relative to the main beam, measured at an equal distant point. The depression angle that defines the cone of silence can be found by referring to the proper elevation pattern for the antenna in question, and locating on it the point for which the relative field falls below 0.1 (-20 dB) and remains below this level. through -90°; the nadir of the elevation pattern. (This applies to the complementary side of the pattern, the zenith, as well.)

Obviously, the area contained within the cone of silence is directly related to the antenna height and antenna gain. It is strongly recommended that the broadcaster take whatever steps necessary to obtain as much antenna height above ground level as possible and use it in conjunction with as much antenna gain as is practical. In this way it is possible to minimize the downward radiation from the system. The additional height increases the space losses and in many cases requires a reduction in station ERP. Both effects benefit the overall effort.

The cone of silence is not the only area that requires careful attention when there is concern about excessive downward radiation. The minor lobes radiated from the antenna can contain a substantial amount of energy even when using a modified antenna. The amplitude and location of these lobes can be determined from the elevation pattern in much the same way as the cone of silence was determined. The antenna height and gain are, again, the controlling factors. Fortunately, when maximizing the area within the cone of silence, the probability of an objectionable amount of radiation from the minor lobes will be reduced. Exercise caution though, in instances where the ground elevation rises substantially in the vicinity of the transmitter site, and/or low gain antennas are used.

Once the excess downward radiation has been reduced to an acceptable level, the RF fields from the antenna’s minor lobes that strike the earth near the transmitter site will then, most likely, be the predominate source of strong ground reflections. Keep this in mind when estimating probable field strengths near ground level at any site. These reflections can add 6 dB to a calculated E field value at a point of a 100 percent amplitude, in phase reflection.

When either null fill or beam tilt is used on any antenna, the downward radiation situation is further complicated by the effects that these options have on the power distribution in the minor lobes. When null fill is used, there will be additional radiation directed into the first (and sometimes second) null, but there will also be more total power radiated towards the ground at virtually all angles below the main beam. The use of beam tilt, on the other hand, results in a redistribution of energy among the minor lobes. Although the total energy (in the minor lobes) does not significantly increase when beam tilt is used, the existing energy can be redirected into a undesired location. When null fill and beam tilt are used together, expect combined effects exceeding those illustrated for either option independently.

All of this is not to say that null fill and beam tilt are bad options that should never be usedmuch to the contrary, they are very useful tools when properly applied. What is stressed, howeveris that null fill and beam tilt should not be indiscriminately applied.

There is an interesting point to note about the way an antennas calculated nulls occur in practice: When in an array near field, i.e., at the transmitter site, the nulls actually encountered are generally quite shallow (if any exist at all) compared to those suggested by the (far field) elevation pattern. For planning purposes, it is practical to ignore the nulls altogether and consider the minor lobes as one broad lobe, defined by the maximas of all the minor lobes.

Improvements resulting from use of a special antenna.–Within the cone of silence, a 10 dB to 20 dB reduction of field strengths over those transmitted from a conventional antenna (of the same gain, ERP, and location) are generally obtained. The author has observed real nulls (as opposed to ground reflection induced nulls) exceeding 30 dB directly below a modified antenna. All of these figures represent a very substantial reduction.

Because of the presence of tower re-radiation and ground reflections, it is unlikely that the improvements theoretically possible beyond the levels given above can be realized. The elevation patterns suggest that complete suppression is available at certain depression angles, but don’t count on this phenomenon in practice, i.e., at your site!

There are secondary mechanisms that further assist the modified antenna in achieving radiation reduction: When the tower is excited by a conventional antenna, the tower’s structural members will radiate a substantial amount of energy towards the ground via the same phase additive process that occurs on the antenna itself. If the tower is instead excited every one-half wavelength, the re-radiation from the tower members will tend to cancel itself as it would on the antenna; although the affect will not be as complete.

Likewise, when parasitic elements are mounted near an antenna bay for pattern correction purposes, they too can radiate a (very) substantial amount of energy towards the ground. This is particularly true if they are horizontally polarized. And once again, a very desirable reduction in the electric fields present below the antenna can be realized when the parasitics’ undesired radiation is subject to cancellation through space-phasing.

With all of that power no longer wasted on the ground and up into space, it can be concentrated in the main beam; where it belongs. In fact, most antennas never realize their rated (generally calculated) gain due to various losses; those discussed in this text included. The chances of getting closer to ideal are definitely improved through the employment of a modified antenna design. The typical antenna can lose 0.5 dB to 1.0 dB of gain in all polarization’s due to radiation in unwanted directions (excluding radiation normally present in the minor lobes), relative to a one-half wavelength type modified design. With an extremely poor bay, loss of gain due to excessive undesired radiation can substantially exceed these levels.

Additional applications for group one antennas.–The inherit characteristic of an antenna designed to suppress unwanted RF radiation makes it an equally poor receiving antenna in the same directions it was designed to protect, under most circumstances. (The law of reciprocity, as applied to antennas.) Because of this, a special antenna is an ideal tool to increase the coupling loss (i.e., reduce the coupling) into other antennas at many transmitter sites. This, of course, will reduce the likelihood of spurious emission products.

The reduction in antenna coupling that can be achieved for a given situation is not easily defined. Many aspects influence the antenna coupling equation, including; antenna gain, antenna separation, the geometry of the separations, the presence of the tower(s), the polarizations used, frequencies involved, etc. As a general guideline, when one of the coupled antennas is a one-half wavelength type of design, a nominal 15 dB improvement in coupling losses can be expected compared to those figures attainable with a regular antenna under the same set of circumstances (assumed to be somewhat ideal). The range of improvement to be found under most reasonable circumstances will probably be 10 dB to 20 dB. The use of two similar special antennas will further improve this figure.

The above guidelines regarding the use of a special antenna must be qualified: The maximum amount of loss will occur when the antennas are (1) stacked vertically, e.g., one directly above the other; (2) leave a relatively large distance (more than a wavelength) between the two closest bays, and (3) are close in frequency. Note that this last parameter does not apply when two conventional antennas are used.

An undesired increase in coupling will occur whenever one of the parameters described in the above paragraph is compromised. Individual stations’ antenna systems should not interleave or partially overlap one another on the same tower. This can lead to incurable coupling problems (among other things) regardless of the type of antenna employed, standard or special. When antennas are adjacent to each other on separate towers located a few hundred feet apart, the antennas with large apertures will generally exhibit less coupling than their smaller counterparts. Antenna coupling can be further reduced by employing a reverse sense of polarization (cross polarization) for one of the transmitting antennas. This will yield a small to medium improvement (3 dB to 10 dB) in coupling loss.

Maximizing antenna coupling losses from the beginning can result in substantial benefits for a station that is located physically close to another station(s) that is close in frequency; e.g., 800kHz. When situations like these arise, along with the spurious emissions, the filtering requirements can get very tough–if not impossible. Not convinced? Ask someone who has really battled one of these cases and you will probably change your mind. In cases where every dB of coupling loss counts, a properly applied modified antenna is an excellent place to start.

Finally, the last application for a category one special antenna, to be mentioned here, is so obvious that it is often overlooked. The reduction or elimination of RFI (radio frequency interference), and EMI (electromagnetic interference) at the transmitter site can be readily accomplished with a special antenna. Needless to say, this is a direct result of the reduction of downward radiation. Remember, best results are obtained when the transmitting equipment is located near the base of the tower, the antenna is located on a high tower, and has a relatively large amount of gain.


General.–There are applications for special antennas that are not bound by the same configurations that radiation suppression requires. This allows us to build an antenna with modified interbay spacings, but without regard to exacting requirements necessary for proper space-phasing cancellation. These are the group two antennas referred to previously.

The fact that these antennas are not specifically designed for RF suppression does not suggest that the antenna will have a large amount of downward radiation. Whenever an antenna is modified so its interbay separation is a value other than one wavelength, a certain amount of space-phasing cancellation will always take place. In some cases, this advantage can be significant even though it is not a design goal.

There are potentially many cases where an antenna could be custom built to suit a particular requirement. The one application that will be discussed is a situation that will probably prove to be popular. This case involves the use of a special antenna to substantially improve the coverage of a station through improved ability to control the azimuth pattern of the antenna.

The problems of pattern distortion, and the corresponding methods of correction are quite complex. It is beyond the scope of this paper to cover the intricacies of antenna patterning; hence, it is assumed that the reader is already familiar with the subject, or can gain access to material that treats antenna patterning in-depth. Therefore, this discussion is limited to background needed to understand the application of the antenna design.

Problem.–Whenever a ring type FM antenna is side mounted on a tower, the azimuth pattern (horizontal plane pattern) actually transmitted from the from the antenna is usually severely distorted. Variations in the relative field strengths at different directions of azimuth will typically exceed 10 dB, and will leave different patterns for the horizontal and vertical polarization’s. This type of problem can vary from moderate to intolerable depending upon tower size, structural design, antenna mounting location, stand-off from tower, operating frequency, etc.

Several manufacturers have pattern development programs that allow the customer to purchase an antenna that has been modeled on a tower replicating that used by the station. The engineers develop the pattern as best they can according to the requirements, e.g., omni-directionaldirectional, etc.. Occasionally, parasitic elements are used to aid in obtaining the desired pattern. Testing is done either full scale or with scale models. A single bay is usually employed and is generally tested at several points of elevation along the sample tower section to determine what effects the tower’s structural members’ locations will have on the azimuth pattern, relative to the antenna bay. The manufacturer then takes these patterns and attempts to arrive at an average pattern representative of how the antenna is expected to perform once installed. Since an antenna’s bays rarely fall at the same regular interval as the tower’s repetitive structural members, the technique of pattern averaging to arrive at a final pattern has been considered a reasonable and proper approach.

Research on the part of at least one manufacturer has revealed that this pattern averaging procedure can introduce large discrepancies between the projected pattern and the actual installed pattern. This suggests that the final pattern is generally not the simple average of several patterns given practical circumstances. And there is no known mathematical function or relationship that can be applied in each case to arrive at the correct result. Obviously, the use of pattern averaging to predict the final pattern is inherently flawed, and cannot be relied upon unless the various patterns involved are quite similar to begin with. This is usually not the case when pattern distortion is sufficient to warrant correction. It should be evident that it would be difficult to quantify the final pattern for an antenna with n number of bays, and the array arbitrarily attached to the tower.

Solution.–A unique approach can be taken to eliminate the uncertainty of averaging the patterns for the various bay elevations, allowing arrival at the correct pattern with considerable confidence. The technique requires an antenna with altered interbay separation, and is very simple: Just place each bay at a position along the tower that concurs with the repetitive nature of the structure. In this manner, it is possible for each of the bays to illuminate the tower in the exact same way, and in a relative bay location that yields the best overall pattern. Obviously, each bay will radiate the same azimuth pattern, and as a result the final predicted pattern will be equivalent to the bay pattern.

The exact bay separation required to achieve the above set of circumstances is dependent upon the tower. Virtually every tower has girders and braces located every few feet, and it should be possible to locate the bays anywhere from one-half wavelength to one full wavelength apart. Only in rare cases should it be necessary to exceed this recommendation. All bays must be fed in phase, as required by a broadside array, and this can most conveniently be done with a branch feed under most circumstances.

The gain of the antenna is determined after the bay separation and total aperture have been selected. The necessary figures can be derived from the total number of bays to be used, and knowing their respective locations on the tower. Recalling the earlier discussion on gain, it can readily be seen that this is all that is necessary to determine the array pattern, and thus, the gain over a single bay of the same type. How this is translated into the ultimate gain figure is dependent upon the exact antenna configuration (directional, non-directional, etc.) and the methods preferred by the antenna manufacturer.

Resulting benefits.–An antenna employed in the above manner will provide the. best overall pattern control possible on a medium or large size towerwithout resorting to the use of a costly panel antenna system. These special antennas will produce beamwidth and gain that is similar to that of a conventional antenna of the same approximate length, and will yield some suppression of downward RF radiation. The exact amount of which can be determined from the antennas specifications.


The author has had the opportunity to research, design, and have built, a specially modified antenna which made an important transmitter site available to a station that was previously precluded from the site due to operational difficulties.

In this particular instance, a special antenna designed and used as outlined under the group one antenna guidelines in this text, solved several “insurmountable” problems and turned a virtually hopeless situation into one of the best signals in the market. This serves to illustrate one of the many cases where a special antenna could provide the means for solving a difficult problem, a solution that might not be available in any other form.

A modified antenna is best thought of as a tool that can be used to solve a specific problem (or problems), should it be in an existing installation or for one that is on the drawing board.

It is urged that any station interested in utilizing a special antenna first carefully evaluate the exact needs for the particular installation in question. A modified antenna should not be arbitrarily installed, as the results may be both disappointing and costly.

In order to obtain the best results with an antenna installation, the utmost attention to detail is required during all phases of planning, manufacture, installation, operation, and maintenance. This cannot be overstated.

Finally, on a topic as intricate, as that presented in this text, it is difficult, if not impossible to cover every potential aspect. In order to keep this paper at a reasonable length, it was not possible to address every situation; such as the use of a modified antenna for multiplexed stations, etc. The author wishes to apologize for any omissions, and hopes that any errors discovered in the text will be brought to his attention so that they may be corrected.


The author wishes to thank the following persons for their contributions that assisted in the preparation of this paper:
Bob Gonsett
Don Hobson
Chris Holt
Jim Levitt
Special thanks goes to Bill DeCormier at Dielectric Communications for his many valuable suggestions, and to Dielectric Communications for supplying several of the illustrations used in this text.


P. C. Gailey and R. A. Tell, “An Engineering Assessment of the Potential Impact of Federal Radiation Protection Guidance on the AM, FM, and TV Broadcast Services.” U.S. Environmental. Protection Agency, Non-ionizing Radiation Branch, Las Vegas, Nevada.

J. D. Kraus, Antennas, McGraw-Hill Book Company, Copyright 1950.

F. E. Terman, Radio Engineers’ Handbook, pp. 776-800, McGraw-Hill Book Company, Copyright 1943.