Antennas Part V -- Special Antenna Techniques (Jan. 1979)

Home | Audio Magazine | Stereo Review magazine | Good Sound | Troubleshooting

The four-part series on FM antenna systems run last year elicited unusually high reader response. Many wrote in to say how much they liked the series.

Many wrote in to say they liked it, but it didn't cover this or that problem they have. Apparently a goodly chunk (or at least a very vocal chunk) of the readers have problems that fall into one or more of the following categories: On-channel interference, severe multipath, or a desire to hear a distant station that is very close in frequency to a local station. These problems have a common characteristic, they are all forms of on-channel interference, hence they cannot be filtered out. The only effective method of rejecting the undesired signal is by extreme antenna directionality, and this is achieved by antenna systems using multiple antennas (arrays). Thus, this final article will deal with esoteric antenna techniques yielding extreme directionality and related techniques for achieving maximum antenna output.

Stacked Arrays

Stacked arrays are combinations of two or more antennas whose outputs are combined in a way that provides nearly double the signal power and an array pattern having characteristics unlike that of any single antenna. The patterns may be characterized by a very-narrow main lobe with deep nulls alongside or extremely high F/B ratio.

The individual antenna outputs are combined by using a two-way signal splitter backwards. The baluns of the two antennas are connected with equal lengths of coax to the "out" ports of the splitter; the "in" port is connected to the long run of coaxial cable that goes to your tuner. The exact amount of gain increase actually achieved depends mainly on the resistive losses of the splitter; 3 dB is the maximum-possible gain (perfect splitter). The splitter must also be waterproof, since it is generally located on the mast near the antennas. The RMS Electronics MA-21.1V or CA-1002/ SM described in Part Il are ideal for this purpose; the 1110 MHz combining gain of these miniature units with diecast housings was measured at 2.6 to 2.7 dB. Do not attempt stacking at 300 ohms impedance unless gain is of secondary importance; aside from the difficulty of installing the combining lines, the best 300-ohm splitters have a combining gain of only 2.2 dB. The coax sections from the antennas to the splitter must be equal within 2 inches. The baluns must be connected so the two antennas are phased properly (yes, just like speakers). If the phasing is reversed, the array output will be less than that of a single antenna, and the pattern will be haywire. So make the balun connections to each antennas as alike as possible. This includes having the same side up, and the same transmission line approach.

To check the phasing, point the antenna array toward a moderately weak station whose direction is accurately known.

This is very important for horizontally stacked arrays. Add attenuation between the tuner input and transmission line until the signal-level meter indication is a little less than half of full scale. Then reverse one balun connection and check the meter. If the signal level has dropped, the original connection was correct and should be restored. However, if the signal level improved, leave the new connection as is. Do this procedure with the AFC off and at a frequency that has no other stations on it or on the adjacent channels.

The type of array needed depends on the direction (or bearing) of the undesired signal relative to the desired signal. If the undesired station or interference is in the opposite direction (interfering signal A in Fig. 1), a single antenna with very-high F/B ratio may do the job (as was mentioned in Part III, Reception Problems). However, for severe cases the technique known as stagger stacking must be used. If the interference is conning from the side (interfering signal B), reorienting a single antenna with deep side nulls may do the job (again, as described in Part III). however, if the angle is very small (as interfering signal C), no single antenna has narrow enough beamwidth to null out the interfering signal and provide maximum output of desired signal. A pair of horizontally stacked antennas is needed.

Fig. 1 Interfering signals.

When the interference problem is not one of bearing but of elevation, vertical stacking vs indicated.

Vertical Stacking. In vertical stacking two identical antennas are mounted one above the other. This is very easy to accomplish mechanically, since both antennas mount on the same mast (Fig. 2). Vertical stacking has no effect on the horizontal-plane beamwidth, it is the same as that of a single antenna. However, the vertical plane beamwidth (or should we say, beam height) is greatly decreased.

Deep nulls appear in the pattern above and below what is now a very narrow front (and rear ) lobe. This means a big decrease in the pickup of interfering signals originating above and below the zero elevation mark.

This reduces or entirely eliminates pickup of auto ignition and ground and airplane-reflected multipath. The null angle varies inversely with the spacing between the antennas, so it is possible (but obviously tedious) to adjust the antenna spacing for maximum rejection of a fixed interference source at your favorite FM frequency. For general (broadband) use the spacing is not critical, but six feet is about the minimum amount recommended for good gain and pattern characteristics all the way to the bottom of the FM band. If large, high-gain antennas are used, the minimum recommended spacing is greater, about eight feet.

The wider the spacing, the narrower the vertical-plane pickup pattern.

However, very wide spacing is impractical because the lower antenna will be too close to the roof unless the array is mounted atop a tower. To achieve the proper vertical-plane pattern, both antennas must receive equal-strength signals, so the difference in distance from each antenna to the roof should be as small as possible.

Never use vertical stacking unless the lowest antenna is at least 20 feet above the roof. This means the 10-foot mast section holding the two antennas should be mounted on at least an 18 foot well-guyed mast, but preferably a tower. The Oct.-Dec., 1977, issue of the Winegard Dealer News uses captioned photographs showing step-by-step how to vertically stack a pair of their CH-6065 FM antennas on a tower. If you are seriously considering this venture, contact Winegard for a copy.

Stagger Stacking. Stagger stacking is a technique that produces an array with extremely high F/B ratio. The antennas are mounted one above the other, but one antenna is a quarter wavelength closer to the signal source, and its cable section is an electrical quarter wavelength longer than that of the other antenna (Fig. 3). Since this technique is frequency sensitive, the antenna displacement and cable lengths should be optimized for the frequency at which you are having your problem. For example, if fourth harmonic CB radiation is being picked up from a transmitter behind your antenna, use 108 MHz for your calculations. If you want to reject a local station on 89.1 MHz so you can listen to a distant station in the opposite direction, use 89.1 for the displacement and line-length calculations.

If the antennas are mounted so each antenna receives exactly the same signal level, the F/B ratio will approach infinity at the design frequency. In practice, this is impossible (for the same reasons discussed above in Vertical Stacking), but the F/B ratio will be very high at the design frequency and higher than that of a single antenna over most of the FM band.

Fig. 2 Vertically stacked antennas. (Photo courtesy Winegard Co.)

The vertical spacing of the stagger stacked array should be the same as if it was a vertically stacked array. The amount of antenna displacement for the frequency of interest is calculated from the formula in Fig. 3. Note that a diagonal brace is required to support the displaced antenna. Naturally, the mounting bracket and clamp must also be relocated. Before running out and buying a pair of antennas, make sure the antenna's mounting clamp can be moved back by an amount' equal to the desired displacement and you do not end up with the mast touching an antenna element.

The difference in cable length can also be calculated from the formula in Fig. 3. This formula is correct only for foam-dielectric coax, either RG59sized with No. 20 center conductor or RG6-sized No. 14 coax.

Fig. 3 Stagger-stacking technique and formulae.

Fig. 4 Horizontally stacked antennas.

Fig. 5 Narrow-beam main lobe and nulls produced by horizontally stacking two Jerrold QFM-9s one wavelength apart.

Fig. 6 Mechanics of horizontal stacking.

Horizontal Stacking. Mounting two antennas side-by-side (Fig. 4) produces an array with the very useful characteristic of being able to reject an undesired signal that lies in almost the same direction as the desired station.

Moreover, it also provides nearly three dB more gain on the desired station than a single antenna. As Fig. 5 shows, the horizontal-plane polar pattern has a main lobe much narrower than that of the antennas used in it, with deep nulls alongside the main lobe. The angle between the nulls and the beam width of the main lobe decreases with increasing spacing between the two antennas. This means that if the interfering signal and the desired station are extremely close together (only a 15-20" difference in bearing), very wide spacing can be used to drop the interfering signal into a null while providing very high output of the desired station. Similarly, if the two signals have a fairly large difference in bearing (50-60°), narrow spacing is indicated.

To determine the spacing (in inches) between antennas needed to knock out an interfering signal, proceed as follows:

1. Measure the difference in bearing between the station you want and the one you wish to reject, using a local map and a protractor.

2. Determine the spacing in inches at the frequency of interest by the following formula:

S_in = 5905 / [sin x f_MHz]

This technique is applicable to rejecting CB harmonic radiation and severe multi-path arriving from the front, although the direction of the interference must be fairly well known for this technique to be effective. Its greatest application is for long-distance reception of a station on the same or adjacent channel as a local station, when the desired and the local stations are in almost the same direction. Moreover, this technique can be used as an alternative to stagger stacking if the desired station is almost (but not quite) opposite to the interfering local.

This is possible because the nulls also appear in the back lobe! Horizontal stacking is much harder to accomplish mechanically than vertical or stagger stacking. Two short masts are needed to mount the antennas, and these must be attached to a crosspiece, which is in turn fastened to the tall main mast. This is complicated by the requirement that the crosspiece be non-metallic. Since the crosspiece is parallel to the antenna elements, a metallic crosspiece will interfere with the operation of the antenna. Additionally, the crosspiece may be as much as 20 feet long (2 wavelengths) to achieve the spacing necessary for small null angles. Redwood crosspieces are often used in commercial applications, but the resultant array will be too heavy for a residence. A lighter and cheaper technique is shown in Fig. 6. Plastic pipe (around 11-inch outside diameter) is used for the crosspiece. The short mast sections can be slid along the plastic pipe to allow relatively easy readjustment of the spacing. However, the center of the plastic pipe must be strengthened 'with braces made by slicing a 12-inch length of steel pipe lengthwise.

The cable sections connecting the antennas to the splitter cannot be run horizontally either. Run the cable down the individual mast sections and from there at a 45° angle to the splitter on the main mast.

Quad Stacks. Two arrays of horizontally stacked antennas can be stacked vertically to produce a quad array or quad stack (Fig. 7). An array of this type has narrow beamwidth and beam height. This is close to the ultimate in directionality and interference rejection. However, an array of this type is obviously expensive, very difficult for the audiophile to build and erect, and even more difficult to keep aloft! Still, if money is no object (yes, there really are people who can afford it), professionally built and installed quad arrays are available from CATV antenna manufacturers such as SITCO.

Increased Output

An additional requirement for low noise, long-distance reception is antenna output that is as high as possible. Preamplifiers are of little use unless the transmission-line run is very long and/or several tuners must be served (see Part IV). This is because modern high-quality FM tuners have noise figures as good as, if not better than, broadband FM preamplifiers.

However, if the line run is very long (150 feet or over), a low-noise, mast mounted preamp will make a useful contribution to perceived S/N ratio by an amount equal to the transmission line attenuation. However, in most cases higher antenna output must be achieved by "noiseless" antenna techniques. Stacking increases output up to 3 dB, but increased antenna height or larger antenna(s) may be needed in extreme cases or if stacking is not used. Each technique has its advantages and disadvantages.

Increased Antenna Heights. The distance between the roof and an antenna mounted on the popular 10-foot mast is small enough in terms of wavelength for the roof (and surrounding objects) to affect the antenna operation to some degree. Experiments conducted by the author on TV Channel 5 (which is not far in frequency from the FM band) indicate that a gain of three to four dB is likely when a medium-sized FM antenna is moved from 10 feet above roof level to 18 feet. The additional spacing removes much of the proximity effect of the roof and allows the antenna to provide maximum output from the signal level available. The only hitch is that an 18 foot mast must be properly installed and guyed or down it comes!

Larger Antenna. If you are using an omnidirectional antenna, changing to a fairly big yagi will yield a healthy increase in signal strength. Changing from a small (3or 4-element) yagi to a really big one (10 or 11 elements) will yield a useful increase in output. However, changing to an antenna with just two or three more elements more than your old one will yield just a fraction of a dB increase in average gain, so is hardly worth the trouble. To get about 3 dB more gain, the replacement antenna must be about 21 times the boom length and element count of the one it is replacing.

Indoor Antenna Techniques

At the beginning of this series, I mentioned that indoor antennas have no place in a discussion of high-quality FM antenna systems. Though I still hold this opinion, many readers point out that they have no choice but to use an indoor antenna, so ask that I please do something to help them.

With some misgiving therefore, I will end this article with a few suggestions.

On the premise that residences where outdoor antennas are prohibited are usually located in strong-signal areas, a high-sensitivity tuner can compensate for the low signal levels obtainable from small indoor antennas.

The problems therefore are multipath and S/N ratio. Each of these can be alleviated somewhat by antenna directionality. Although the simple dips' (be it the ribbon dipole packed with the receiver or a TV rabbit ears) has essentially zero front-to-back ratio, it does have decent side nulls under the right circumstances. The idea is to orient the antenna (while watching the multipath indicator or listening to the interference) so the undesired signal falls into a side null and thus produces little output from the antenna.

Orientability means that the ribbon dipole cannot be taped on the wall or laid under the rug. The dipole should be taped to a wooden or plastic strip about 6 feet long, and mounted on some sort of device that will allow it to remain in the position you leave it. A clamp-on reflector lamp (which sells for about $2) is a good device to adapt; remove the reflector and lamp socket, and use the clamp with universal joint.

The result will not look pretty in your home, but is about the best that can be done under the circumstances. A better-looking alternative is TV rabbit ears, the simple kind having just two telescoping elements. These rods can be adjusted in length to peak the signal, the angle of the two rods can be varied, and the whole antenna rotated.

Just make sure the one you buy has rods each that extend to at least 34 inches.

Every so often a very expensive indoor FM antenna is marketed. Some of them do offer performance superior to that of a ribbon dipole or rabbit ears, but none can outperform even a small outdoor antenna in terms of directionality and low-noise output level. The amplifiers in the more elaborate models can boost the signal level as high as that from an outdoor antenna, but the output necessarily has relatively poor S/N ratio. Still, if you can't put a good antenna up outdoors, the newer indoor "mini antennas" are worth a try.

Fig. 7 SITCO quad-stack array. (Photo courtesy SITCO.)