[TowerTalk] Tuning raised radial verticals

[Gary Johnson]

You might like my article on a recent rebuild of the 80m wire 4sq at N6RO in the July/Aug 2017 issue of NCJ, or download it from: http://www.wb9jps.com/Gary_Johnson/Amateur_Radio_files/Four_Square_Rebuild_at_N6RO.pdf

Each antenna had a vertical element about 66 ft long and two radials about 58 ft long, elevated at 10 ft, optimized for 3750 (SSB) and with a switched loading coil on the radial side for CW operation. Phasing lines were 1/4 wave at 3650. We have a Comtek system.

First comment: The commercial phasing boxes (Comtek and DXE) are optimized for exactly one antenna system, based on feedpoint impedance and mutual coupling: Full-sized tubular verticals, ground-mounted over a substantial radial field. As we deviate from that, the impedances change and the magnitude and phase at each element is no longer exactly as desired. The result is pattern degradation, mostly in the null off the back. But practically speaking, it’s not a disaster and we all get reasonable performance. 

Single-vertical tuning is easily done by disconnecting the bottom of the vertical element of the other three elements. We trimmed the radials for resonance (tune for X=0; SWR is meaningless) at the offset freq, which turned out to be about 70 kHz, so 3680 was the target. At resonance, they looked like 40-45 ohms, not too far from the holy grail ground-mounted vertical.

Common-mode chokes at the feedpoints are extremely important and most of my article is devoted to that subject. The handful of beads recommended by Comtek isn’t really enough but it’s better than nothing. Optimum solutions include: 70 Fair-Rite 2631102002 beads on each feedline; Transition to RG302 (75-ohm) and use many smaller beads (ON4UN used 100 type 73 beads in this configuration), or wind it around a toroid (12T on 2.4 inch type 31); Or transition to a bifilar choke as I did.

More radials will reduce ground losses. We went with two 1/4-wave radials as a tradeoff between performance and the practical problem of routing many elevated wires. Symmetrical arrangements can improve the final pattern. 

With the system fully configured, pay more attention to the dump power from the hybrid than SWR. Bandwidth is best defined by setting a limit on dump power, say 10% of applied power (0.5 dB). Our system had a 275 kHz bandwidth by this criterion. SWR should change slowly and actually is more useful as an indicator that something has failed.

I can also recommend spending time with at least a basic EZNEC simulation to see what happens as you vary the geometry etc. 

Gary NA6O