The IEEE Annual Broadcast Technical Symposium (IEEE-BTS), held this year in Tyson’s Corner, Va., is the best conference for learning about new broadcast RF technology. The overall quality of the papers this year, as usual, was excellent, and antennas received a lot of attention.
TV antenna designs have had their share of controversy. In the past, some papers presented at NAB and IEEE sessions claimed center-fed or branch-fed antennas were superior to end-fed antennas for DTV, while others discounted the difference.
At the 2000 IEEE-BTS, IEEE Fellow Geza Dienes countered the criticisms of end-fed slot antennas in his paper "Computer-Aided Television Transmitting Antenna Design Techniques." He started with a summary of the three common UHF TV antenna types: panel, center-fed and end-fed.
Panel antennas use a corporate, or branch, feed network. Each individual panel usually contains four elements, so phase and amplitude control is limited to groups of four. This type of antenna is capable of broadband performance.
Geza explained that center-fed antennas are actually the composite of two end-fed antennas – one pointing up and the other pointing down. In this "old man" of design approaches, RF is injected near the center of the antenna array. Generally, only two phase and amplitude controls are available – the entire top or bottom half of the array must be shifted. The result is a shifted phase front.
With end-fed antennas, as the name implies, RF is injected at the end of the antenna array and, as Geza explained, exits at a controlled rate and timing from the radiating elements. Phase and amplitude for each radiating element can be adjusted individually.
A properly designed end-fed antenna will produce a flat phase front. However, achieving that proper design requires work. It is necessary to determine the location, distance from the receiver, amplitude and phase for each element in the antenna. Geza said this takes more than 1,800 calculations!
Computers have been used to design end-fed antennas. In the past, with less-powerful computers, it was necessary to make many simplifications and assumptions. With the powerful computers available now, it is possible to vary the parameters of any element and see the result.
Geza demonstrated the software he had developed for use at Andrew Corp., including an azimuth pattern calculator and an array calculator for developing elevation patterns for end-fed antennas. The computer programs make it possible to visualize antenna performance, making it easier to determine what compromises need to be made and to see how stable the design is – it’s possible see how inaccuracies affect antenna performances.
One example showed the variety of pattern options available with a simple array of four panels around a radius. Another demonstrated how the elevation pattern of a 35-bay end-fed slot antenna could be modified to minimize "beam-steering" or differential gain by varying the phase and amplitude of each element in the antenna. He also showed that if more than 2 degrees of electrical beam-tilt is designed into a center-fed antenna, the main lobe will actually split.
Someone asked Geza if it was possible to design a near-perfect elevation pattern for an end-fed antenna – why couldn’t the same technique be used to optimize a center-fed antenna, because it consists of two end-fed antennas. He replied this was because the center-fed antenna appears as two sources a certain distance apart and for this reason you can’t use more than two degrees of electrical beam-tilt in a center-fed antenna.
Another question concerned "beam-steering," a problem often associated with end-fed antennas. Geza answered that this can be reduced to 1 dB over a 6 MHz channel for any viewer.
SIDEMOUNTED DTV ANTENNAS
A common belief is that sidemount antennas won’t work well for DTV. While tower effects have to be considered, Carl Eilers presented a paper he wrote with Gary Sgrignoli that showed if the antenna is sidemounted on a tapered tower section, pattern distortion will not be as bad as it would be if mounted on a straight-leg tower section.
To summarize the results of the analysis, the reflections from the legs of a tapered tower section will fall either above the horizon or below the main beam of the antenna. The tower leg closest to the antenna is angled upward, creating reflections above the horizon. The tower legs farthest from the antenna will be angled down, so reflections from them will fall below the main beam. As a result, pattern ripple is less than it would be with a straight tower section and, correspondingly, the echo pattern improves.
Carl Eilers said there should be no fear of sidemounting an antenna on a tapered tower structure. He did note, however, that the reflection from a tapered tower would slightly modify the vertically and horizontally polarized components of the E-M (Electric-Magnetic) vector.
WIDEBAND SLOT ANTENNAS
Slot antennas are not usually seen as wideband antennas. Gordon Sinclair, in his paper "Adding Channels without Adding Tower Loading: UHF Wideband Slot Antenna Technology," described features of the Technology for Communications International (TCI) Model 888 wideband slot antenna. He focused on the disadvantages present in other antenna designs – single-channel slot antennas, turnstiles and panel arrays – and showed how the TCI design avoided them.
Some disadvantages of these common designs included narrowband operation, high windloading and difficult-to-support mechanical structures. The TCI Model 888 avoids them by using wideband "slots" as the antenna element in a unique antenna configuration.
Antenna elements in the TCI Model 888 consist of a wideband cavity (approximately one wavelength long – 15 inches) with a T-bar feed. Sixteen elements are assembled in a vertical column and branch-fed.
While it wasn’t discussed in the paper, it appears the phase of each element could be adjusted individually, allowing more flexibility in elevation patterns than would be obtainable using the common panel configuration with four elements fed together.
CREATING DIRECTIONAL PATTERNS
The columns are tightly clustered around a small radius, which contains and protects the feed lines. The diameter of the cylinder of columns is about 24 inches. This tight spacing minimizes pattern scalloping and has less windloading than a panel array with comparable gain. While the standard design is omnidirectional, elements can be removed (or disconnected) in one or more columns to create directional patterns.
The paper showed the results of field tests of an omnidirectional Model 888 installed at WRLK. VSWR plots indicated the broadband nature of the antenna and field-strength plots showed the signal from the antenna was comparable to the original RCA TFU-30, with better field strengths closer to the antenna and results within 2 dB of the RCA antenna farther out, except along one radial.
Investigation showed there was a problem with a feeder on that side. The large number of feeders and the difficulty of getting to them for repairs when the antenna is on top the tower is a disadvantage with the design, although repair after installation is likely to be difficult for any type of antenna.
NEW TOWER RATING STANDARD
In the past, antennas have not been considered part of the tower structure when it was analyzed for structural integrity. Adam Jones, mechanical engineer at Dielectric Communications, explained the ramifications of the new TIA/EIA-222-G tower rating standard. This standard is scheduled for release this year.
Many broadcasters will be adding antennas to existing towers or building new towers for their DTV antennas. A thorough understanding of EIA-222-G now could avoid possible problems when the tower or antenna is installed.
EIA-222-G has several significant differences from the previous standard, TIA/EIA-222-F. One change is that antennas are now included in the tower rating standard. For the structure to be studied under EIA-222-G, the mechanical design of the antenna will have to be evaluated. Another is that EIA-222-F was based on the Allowable Stress Design (ASD) methodology, while EIA-222-G uses the more widely recognized Load Resistance Factored Design (LRFD) methodology.
Under EIA-222-F, basic wind speed was based on the "fastest mile averaging period." With EIA-222-G, a "3-second gust measurement" is used. This brings the standard in line with American Society of Civil Engineers wind speed map (ASCE 7-95). Design parameters for icing and wind speed are now defined, based on local conditions.
The burden of these changes will mostly fall on the tower and antenna manufacturer. However, EIA-222-G contains several tables that define exposure category (urban/suburban, open terrain with scattered obstructions, and flat unobstructed areas exposed to wind over open water), a new topographic factor and a reliability class. If there is a choice of locations for a new tower, these factors should be considered when selecting a site.
There are three reliability classes based on probabilities that the design wind speed of the structure will be exceeded during a 25-year period. Most broadcast structures will be considered "Class II" structures, defined as "structures that due to height, use or location represent a substantial hazard to human life and/or property in the event of failure."
The topographic factor is used for structures located on escarpments, hills or ridges that could cause a wind speed increase. For example, structures in the upper half of a ridge must consider wind speedup in all directions (topographic category 4). This wind speedup can increase the wind pressure on a structure by a factor of 2 to 2.5 over that calculated without the factor.
From this brief introduction to EIA-222-G, it should be apparent that tower customers will have to supply more information to their structural engineer to ensure the new tower or addition will meet the new standard.
Contact Doug Lung at [email protected].