Antenna Tests Revisited
by David Waelder
Reaction to the antenna test article has generally been favorable. Several members wrote to say they appreciated the specific information to help guide choices in antenna selection and deployment.
Some readers with extensive knowledge of the theory and practice of radio transmission had questions about the methodology of the test procedure and the validity of the conclusions. The purpose of the test was always to examine a variety of antennas to determine how they compared in actual use in the field and to what extent variations in deployment translate into advantages or liabilities. To be valuable, a superior system should produce observable results in line-of-sight testing. An advantage that is reliably apparent only in laboratory tests may be genuine but of limited consequence for the user. Additional testing with devices like a spectrum analyzer is an effort to quantify and confirm the results from the field and is really tangential to the effort. Still, measurements and results should be consistent.
Wolf Seeberg and Henry Cohen raised some issues that relate to design advantages that may not be revealed in simple walk tests. Wolf is the proprietor of a video rental company and was, for a long time, a member of Local 695. Henry Cohen is the proprietor of a radio rental and service facility in New York and publishes an online journal on radio performance. They both pointed out that antennas with a circular polarized design, like helicals, are not desirable because they have additional gain but rather because they receive out-of-phase signals with minimal attenuation. In ordinary operation, we strive to maintain consistent antenna orientation for best results. With a belt pack transmitter, the antenna is typically vertical so we align the antenna on the receiver vertically. As the signal bounces off buildings, however, the phase of the signal can be altered just as the spin on a cue ball is shifted as it strikes the cushion. The consequences of phase shifting are usually minor but may have real consequence in a situation where the transmitter and receiver are moving. Doing car-to-car work in city streets, with signals reflected by buildings as the cars pass, is a prime example. Drop-outs may occur as multiple signals, some in phase and some out of phase, arrive at the receiver simultaneously. The circular polarized antenna copes with these phase reversals more effectively than log-periodics that are designed for use in a particular orientation.
This is a valid point; a simple walk test does not reveal a characteristic that might be a significant advantage in a scene with moving cars. And, insert-car scenes are relatively commonplace in our work. Still, this is an advantage that applies only in limited circumstances. Both the Sennheiser CP antenna and the PWS helical design are unwieldy devices to rig and deploy, at least compared with sharkfins, and are conspicuously more expensive. Since they seem to offer little advantage in an ordinary walk-and-talk, I would recommend against purchasing them as part of the regular kit unless you are employed on a cop show where insert-car work is a weekly event. But it would be well worth renting a pair for those days when moving-car work in an urban setting is scheduled.
Wolf also raised the issue of consistency of performance over a range of frequencies. The original tests were performed at 561.800 MHz in Block 21, a popular choice in the Los Angeles area. However, antennas are typically tuned to a particular frequency and may not perform optimally at other frequencies. It was Wolf’s contention that some particular antenna designs offered more consistent performance over a range of frequencies. I took several antennas down to LSC to check performance over their operational range using their spectrum analyzer. We compared signal loss over a range of 450 MHz to 700 MHz at 10 MHz intervals. Some of these designs are rated for performance up to about 900 MHz but the FCC prohibits radio-mike operation above 698 MHz.
Results were interesting. Virtually all of the designs I tested were strong at 450 MHz and exhibited a drop around 500 MHz or 550 MHz. Then they tended to recover and stay nearly flat until 700 MHz. The drop around 500 MHz was typically about 5 dB; none of the previously tested designs exhibited the larger losses that Wolf predicted. I did notice that there was some performance variation from example to example, not just from one design to another. I tested different examples of both PSC and Ramsey LPDAs and found some differences even between two examples of the same design. Observed differences may indicate some variance in manufacturing runs or it may just be a consequence of slightly different hook-up hardware. While there were measurable differences, nothing I observed would alter the basic conclusions of the earlier tests.
In the interest of full disclosure, it should be noted that this set of spectrum analyzer tests was conducted inside a building where reflections would certainly compromise results. But our observations were generally consistent with a previous round of testing done outside at a distance of 500 feet. For tests at multiple frequencies, we needed to use a bench analyzer that can simultaneously transmit and receive radio signals. The portable device used in the previous testing can only be configured to perform one task at a time.
While we were taking measurements, I also took some readings using ordinary whips to investigate the question of how much signal is lost with a mismatched antenna. This is an issue that comes up from time to time as users, with an antenna from an alternate block ready-to-hand, question the importance of an exact match. The answer is that it seems to depend on the frequency of the signal. At Block 21, the use of a Block 27 antenna resulted in a signal impairment of only 2 dB or 3 dB when compared with a properly matched antenna. It didn’t seem to make any difference whether the mismatch was at the transmitter or the receiver end; the loss was the same. But, with a Block 27 signal, the use of a Block 21 whip at either end resulted in a 10 dB loss compared with a properly matched antenna.
These challenges raised some interesting issues but I stand by the broad conclusions of the earlier article with one modification. The circular polarized antennas do seem to offer a genuine performance advantage while moving in an urban environment.
A brief reprise of conclusions is in order:
1. Higher gain antennas offer improved performance but the range advantage is only about 20% or 30%. No antenna doubled the effective range.
2. All of the log-periodic designs seemed to offer a similar performance advantage relative to 1/4 wavelength whips.
3. In an open environment, a good dipole antenna, like the Lectrosonics SNA600, yielded very nearly the performance of the directional sharkfins. But the directional antennas may offer an advantage in a crowded RF environment by restricting unwanted signals.
4. There was a small performance benefit to wide diversity spacing.
5. Performance was improved when the receiving antennas were raised for clear line-of-sight to the transmitter. However, once line-of-sight was achieved, we saw no further benefit from additional altitude.
6. In normal usage, helical and circular polarized antennas offered no identifiable advantage over log-periodic designs. However, circular polarized designs offer an advantage when moving in an environment that reflects radio signals.
As always, I am indebted to Coffey Sound, Professional Sound Corp. and Location Sound for the loan of equipment to test. And special thanks are owed Location Sound for the use of the test bench and to Victor Solis for his operational skills. I should also note that Henry Cohen’s criticisms were in response to my request on the Lectrosonics User Group. Errors and omissions are mine alone.