Today, especially for DTV facilities, directional antennas primarily provide protection to the service areas of other stations. Measurements of outgoing and incoming interference are based on the patterns of all stations involved, as contained in the FCC's database. Those patterns are normally derived from the operation of the antenna in free space.
To confirm the accuracy of antenna patterns, stations can perform measurements using either a full-scale antenna on a range or using a model in an anechoic chamber. Full-scale ranges have almost disappeared from use. For years, RCA, Harris, Dielectric and RFS each operated a range for full-scale measurements of horizontal and vertical antenna patterns. But today, those ranges no longer exist. To my knowledge, only Jampro still operates a full-scale range.
Measuring antenna patterns
It is beyond the scope of this article to argue the relative accuracies of model measurements vs. full-scale range testing. The main point is to recognize that the commission will accept either method.
Normally, patterns filed with the commission are based solely on calculated values. Television stations do not have to file measured pattern values with their license applications, whereas FM stations must provide documentation of pattern measurements.
The patterns submitted to the commission are based on the TV antenna operating in free space, without respect to the mounting structure or other items in the area. When dealing with a single antenna top-mounted on a tower placed far from other towers or antennas, the assumption of free space is pretty realistic.
The problems occur when other conditions exist, such as when an antenna is side-mounted on a tower. The omnidirectional antenna will no longer be omnidirectional, and the exact shape of a directional pattern will be changed.
The amount of such change can be determined but it isn't easy. One good method routinely used in FM service is to perform the measurements with the antenna mounted on a section of the tower.
Another method is to use models with transmission lines passing through the aperture. Ladders, conduits, climbing safety devices and other hardware are included in the model. If all nearby conditions are accurately considered, both the modeling measurements and the calculations will produce a fairly accurate version of the antenna pattern.
Consider an antenna mounted on a solid surface, such as the cylinders on the Sears Tower. It is obvious that the cylinders will significantly affect the radiation on the opposite side of the antenna. Moving vertically, antennas are mounted on the individual towers in the aperture of antennas on the other tower. A number of parameters will affect the antenna pattern, including the frequency of the radiated signal and the size of the tower members.
The most affected area is where the tower members approach one-half wavelength at the operating frequency. In FM service, that point has long been considered. And 4ft to 5ft face towers are avoided in the antenna aperture. In such cases, an omnidirectional pattern can be changed by as much as 3dB. The same type of effect can occur to TV signals.
Defining a true pattern
So what is the real pattern? Is it the theoretical calculated pattern, the measured free space pattern or what actually occurs when the antenna is mounted on a large structure or in the vicinity of other antennas?
Other types of antennas can also affect directional antenna patterns. Slot antennas normally look like a big piece of pipe at frequencies significantly removed from the antenna of concern. On the other hand, panel antennas are wideband in their operation and are much more significant re-radiation sources.
Years ago, licensees were concerned about the effect of towers mounted on the same structure. An early example was the John Hancock Tower in Boston, where two towers supported multiple antennas. A study done by Andrew Alford determined that the antennas' placement created a saw-tooth pattern in the radiation pattern. The size and number of the teeth in that saw tooth was a function of the frequency involved and the distance between the towers. This effect relates to the multiple antennas located on a T-bar or candelabra structure.
Reasonable calculations can be performed to determine how to best locate antennas on a structure or a tower. In the case of multiple antenna structures, those calculations should be performed to determine the best mounting configuration so that all stations will obtain the best possible performance from their antennas.
These calculations are essentially beyond the normal function or abilities of most consulting engineers. Such an undertaking usually requires a handful of engineers directly involved in the design and construction of transmitting antennas. The calculations are extremely complex and should be performed by computer programs that are proprietary to the manufacturers.
You can analyze the impact of reflecting surfaces in the vicinity of a transmitting antenna by using various scattering programs. Those programs use the transmitting antenna pattern, the frequency involved and the physical characteristics of the structures as inputs. This includes whether the structure is solid, as in a cylinder or microwave dish, or broken, such as a tower. The result will usually be a pattern. Variations of more than 10dB over small azimuth angles are normal. Fortunately, when all the re-radiated signals are combined and incorporated into the actual station service calculations, they tend to average out.
The purpose of performing these calculations is to show that two separate antenna patterns can exist. The first pattern is the calculated pattern the commission wants in the application. In a lot of cases, this pattern might best be described as far from reality. However, the commission doesn't have the means to investigate it and usually accepts the pattern data offered.
The second pattern is the real one. It considers the actual operation of the antenna and the physical environment surrounding the antenna, rather than free space. This pattern realistically describes how the station will perform. It is not simple to obtain this pattern, but it can be determined by providing antenna manufacturers with data about the real world conditions where the antenna will be mounted.
Most stations don't want to spend the money to determine the realistic pattern and will evaluate the performance of the transmitting system by the TV set in the manager's family room, which is operating on cable fed by a fiber-optic system. If the manager is happy, all is well in the world.
Don Markley is president of D.L. Markley and Associates.
Send questions and comments to:email@example.com