Lately I’ve run into a number of situations where I had to optimize coverage from a less than ideal antenna location. By location, I’m referring to the antenna location on the tower.
(click thumbnail)World Trade Center DTV Antenna
My observations this month apply primarily to UHF antennas, as VHF antennas are usually wrapped around a tower instead of being mounted on a separate support off the side of the tower. With VHF antennas, the spacing between the elements has much more of an impact on the antenna pattern than any reflections from the tower or metal inside the tower.
The only location on the tower where the actual radiated signal can be expected to look like the relative field pattern in the antenna manufacturer’s specifications is at the top of the tower, with no surrounding antennas or supporting structures in the aperture.
For many stations, a single, topmount location is not practical. Many new towers are built as candelabra or starmount designs to provide additional revenue from multiple broadcast tenants. As noted last month in my report on the Dielectric dual VHF-UHF slot antenna, windloading, height restrictions and tower structural limitations may make it uneconomical to stack two antennas.
In areas where tower space is short, a side mount location may be the only one available.
OBSTRUCTING COVERAGE
Most readers are aware that a supporting pipe, tower structure and transmission line in the aperture of the antenna will impact the antenna pattern and coverage in the direction of the obstruction. What may not be commonly understood is that reflected and diffracted energy from these obstructions will impact the pattern of the antenna in the directions that are not obstructed.
Picture a side mounted, omnidirectional antenna. To keep things simple, assume it is mounted a few feet from a single, large pole. Energy from the side of the antenna facing the pole will reflect off the pole and – depending on the spacing of the antenna from the pole – add or subtract from the field radiated from the unobstructed part of the antenna. If the reflected signal is in phase with the signal from the antenna, it will add. Out of phase, it will subtract.
The phase relationship of the main and reflected signals depends on the relative distance the two signals travel from the antenna. The depth of the peak or null created in the pattern will also depend on the relative amplitude of the two signals. For example, two signals will cancel each other out – resulting in a null – if the two signals are out of phase by 180 degrees and equal in amplitude.
It is important to note that the analysis becomes more complex if the pole is very close to the antenna because the pole may have to be considered part of the antenna.
If our hypothetical omnidirectional side mount antenna is mounted on a tower section instead of a single support pole, the analysis becomes even more complex. Each element in the tower has to be considered – the diagonal bracing, each leg of the tower, any transmission lines in the tower, the ladder on the tower, and the like. Any of these elements may block the signal, diffract it or reflect it.
Since the wavelength is shorter at the top of the 6 MHz TV channel than at the lower edge, the impact of each of these elements will be different across the channel. By this time it should be clear that many calculations are needed to determine what the real-world pattern of our omnidirectional, side-mounted antenna will look like.
Fortunately, most broadcast antenna manufacturers and many engineering consultants have the computer software available to analyze this. Plot 1 shows an analysis of an omnidirectional antenna side- mounted on a tower with an 8-foot face.
PATTERN NULLS
How will this affect coverage? Remember that the location of the nulls created by the reflections is frequency-dependent. Where there are deep nulls in the pattern, the frequency response across the channel is likely to be poor. The adaptive equalizer in a DTV tuner may be able to correct for this, but this will extract a penalty in carrier-to-noise ratio.
Dr. Oded Bendov referred to this coverage penalty in a paper he presented at NAB in 1996. (See Coverage Contour Optimization of HDTV and NTSC Antennas by Oded Bendov on the Dielectric Web site, http://www.dielectric.com/broadcast/contour.asp.)
If the receiver is in an urban area, reflections from nearby buildings outside the area in the null are likely to cause severe multipath, extracting an additional noise penalty from the DTV receiver. For analog reception, the multipath and frequency response variations will cause ghosting and possible loss of chroma.
In rural areas, the impact of the pattern nulls may not be as bad, as was indicated in field tests of WRAL-DT, in Raleigh, N.C., using a side-mounted antenna. Tests showed little impact on the ability to receive DTV signals beyond that predicted by the lower radiated power in the nulls.
There are ways to improve the pattern of side-mounted antennas, but some performance trade-offs may be required.
MINIMIZING LOSSES
The obvious way to reduce the impact of reflections from tower members is to reduce the amount of energy hitting them. If the antenna is located outside the main population area – as is the case in cities such as Dallas and San Antonio – a directional antenna can be chosen to minimize energy into the tower. In Los Angeles and Denver there are large mountains behind the main antenna site, so most UHF stations use directional antennas.
Plot 1 shows the impact the tower has on a UHF omnidirectional, side mount antenna on a tower with an 8-foot face. This particular tower is fairly open. The largest transmission lines are two 3-1/8" runs of rigid coax located near a tower leg.
The modeled pattern for the antenna that was finally chosen for this facility is shown in Plot 2. The antenna location relative to the tower is the same in both plots. The antenna used for Plot 2 reduced scalloping in the main population area and reduced the amount of energy into the tower and over sparsely populated areas.
If the antenna is mounted on a self-supporting tower, there is another way to reduce the effect of the tower on the antenna pattern. By mounting the antenna on a sloping tower leg, reflections from the tower will not fall in the main beam of the elevation pattern. This was described in a paper by Carl Eilers and Gary Sgrignoli presented at the IEEE Broadcast Technical Society symposium in September 2000 and reported on in my TV Technology column in October 2000 ("Antennas Garner Attention at IEEE Symposium").
A more rigorous analysis is available in Reradiation (Echo) Analysis of a Tapered Tower Section Supporting a Side Mounted DTV Broadcast Antenna and the Corresponding Azimuth Pattern, by Eilers and Sgrignoli, in the September 2001 IEEE Transactions on Broadcasting.
WHEN ONE ISN’T ENOUGH
When the tower cross section is small or open, a side-mounted antenna will provide some coverage in most directions. This isn’t the case if the antenna has to be mounted on the side of a building or a large supporting structure.
This was the situation facing WNBC and other DTV broadcasters on the World Trade Center. The World Trade Center DTV antenna is shown in the photo. As you can see, the size of the mast made it impossible to side mount an antenna in a location that would provide 360-degree coverage.
I encountered a similar situation with WSNS on the Sears Tower in Chicago.
The answer, in both cases, was to use two antennas on opposite sides of the mast. You can see this in the photo – the Dielectric UHF panels for the DTV antenna (in the middle of the photo) are mounted opposite each other. More panels are used in the direction of the majority of the population. The single column of panels provides some signal on the back of the mast.
In Chicago, two Andrew dual-channel UHF slot antennas were used. One antenna, on the west side of the mast, provides an elliptically polarized signal over Chicago and its suburbs. A second antenna, on the east side, provides a horizontally polarized signal over Indiana.
Deep nulls are created when comparable signals from the two antennas arrive at the receiving antenna out of phase. One of the challenges in designing systems using two antennas is minimizing the antenna pattern overlap and placing the overlap in areas where there is the least population.
In Chicago, Lake Michigan made this task a little easier, although there is one area of overlap over land. To increase the chances viewers in this area could receive a signal, vertical polarization was used on only one of the antennas.
By using a vertically polarized receive antenna or experimenting with the orientation of an indoor antenna, reception should be possible even if the horizontally polarized signal is unusable.
The use of two antennas requires detailed engineering studies and, if possible, modeling of the environment where the two antennas will be mounted. For the Chicago project, Andrew had a scale model of the top of the Sears Tower that we were able to use to analyze the effect of varying the amplitude and phase of the signal to the second antenna.
As it turned out, phase between the two did not have a significant impact on this design, although amplitude did. The physical model also provided a way to verify the accuracy of the computer model.
The use of full-size or scale models can be useful for determining the ideal spot for a side mount antenna on a complicated tower structure. A scale model study for one of the stations I built over ten years ago showed that if the antenna was mounted beyond a certain distance from the tower, a very sharp null developed over a coverage area of interest.
The area impacted by this null would have been difficult to measure in the field because it was so narrow, but it would have had an impact on coverage in some key communities.
CONCLUSION
Side mounting UHF antennas – whether for analog or DTV – can be complicated. I hope I’ve given you an understanding of some of the issues involved and some of the ways to optimize coverage when faced with less-than-ideal options for antenna mounting. Comments are always welcome. Send them to dlung@transmitter.com.
