Team C6APR activated Crooked Island (NA-113) in the Bahamas for the 2008 IOTA contest, employing phased vertical pairs on 80 through 10 meters (see Figure 1). The 80 and 40 meter verticals were full sized and equipped with a switching arrangement for endfire or broadside radiation. For 20 through 10 meters, we installed two pairs of Cushcraft R5 verticals at 90 degrees to each other. One pair provided a northeast/southwest pattern while the other provided a northwest/southeast pattern. We separated these two broadside arrays by 100 feet to avoid any interaction and fed each with separate feed lines switched at the operating positions. The issues with the R5s were obtaining a proper match on each band and optimizing the pattern for 20 meters. The design problem for the 80 and 40 meter arrays was being able to switch the array's pattern from broadside to endfire via remotely switchable phasing lines. Since we were taking all of the equipment to the island on a private plane, weight was a key factor.

The R5s

Feeding a broadside array on a single band is relatively simple, and there are several ways to accomplish this. One common method is to feed each vertical via a quarter-wave (90 deg) 75 ohm coaxial transformer (series section). This raises the impedance of each antenna to 100 ohms, but connecting them with a T adapter provides a 50 ohm load for the transmission line to the shack. For a multiband antenna, however, this becomes a switching nightmare and requires separate quarter-wave sections for each band. In addition, it's only really possible to optimize element spacing for a single band.

To address the matching problem, we placed a 50 to 25 ohm CWS ByteMark, www.cwsbytemark.com, model UN-22-25 unun (unbalanced-to-unbalanced transformer) between the two verticals, which were then connected with equal lengths of 50 ohm coax (see Figure 2). Optimum spacing for 20 meters is about five-eighths wavelength (225 deg), and the patterns on 10 and 15 were not at all optimum. At one point during the development stage, we placed a passive half-wave element for 10 meters in the center of the array. This really cleaned up the pattern on 10 meters, but it was too sharp for a two-array system and left large areas with poor coverage. Consequently, we abandoned that approach. The resulting patterns for 15 and 10 meters, however random, still provided good azimuthal coverage on those bands. We decided to leave the array optimized for 20 meters. Had we anticipated better conditions on the higher bands, though, we might have opted for something closer to half-wave (180 deg) spacing on 20.

Patterns

Figures 3a, 3b and 3c depict array patterns. The 20 meter patterns show a very clear distinction between the northeast/southwest array (gray) and the northwest/southeast array (white). On 15, antenna selection was not too effective in optimizing signals. On 10, it was extremely azimuth dependent and somewhat illogical. We used Antenna Model 2.0.0.660 and EZNEC to analyze the antennas and develop the patterns.

You must take care in tuning the antennas so that each has a nearly identical SWR pass-band response - that is, the same SWR at the same frequencies, not just the same SWR bandwidth. You could have the same bandwidth but different frequency points, and the system would not work well. It's worth noting that staging the antennas inland resulted in performance that did not compare to the much lower radiation angle and significantly higher effective gain we experienced once the antennas were erected on the beach. That said, ground conductivity had little effect on antenna tuning. Once the system was tuned inland, we didn't have to touch it at C6.

All R5s were mounted on eight-foot poles to enhance the radiation angle. Because IOTA is a multi-mode contest, the antennas were tuned to band centers and, on 20 meters, had an SWR bandwidth of 1.8:1 from band edge to band edge. The bandwidth on 15 and 10 was much greater, and the SWR was 1.5:1 and 1.3:1 at band edges. respectively. For a single-mode contest, the SWR bandwidth would be less than 1.3:1 across the mode on all bands.

For the complete version of this article as published in the NCJ, view the PDF version.