My 20M End Fed Half Square Antenna


Dec 9, 2022
First post here so please consider that. I wanted a two-band antenna with some dx possibilities for my new FT-891 here in West Virginia. Reading W4RNL's comments on half squares, I got the impression that they seem to be the way to get the lowest solid wave angle for the least height without going to a beverage antenna. End feeding looked to be a simple way to conveniently get both 40M and 20M capability.

So I bought a Radioman 49:1 unun transformer (eBay) and per AA5TB, used a 0.05λ (7 ft on 40M) bare wire as a counterpoise, running it vertically down the side of the support tree and connecting to a single deep ground rod. Thus I get both a ground and a 7ft counterpoise in one! Gradually sliding a group of 8 ferrite beads over the RG213 coax going away from the transformer, the SWR decreased slightly more at the top of 40M with them positioned at a distance of 7 ft out. So I hedged my bets for my end fed transformer's ground: a 7ft counterpoise, a ground rod, and the beads 7 ft out on the coax to function as a second counterpoise. I used thick rubber insulated AWG#10 wire that happened to be available. The overall configuration: transformer was mounted at 7 ft up on the side of a tree. A 3" alligator clip jumper wire connects the high-Z terminal of the transformer to the tip of a 15" vertical whip antenna mounted above it. The other end of the whip is connected to the main wire, which runs up 10 ft to a crotch in the tree, horizonally north-south across 23 ft, then downward 11.5ft vertically to the end. A piece of synthetic rope attached to a second tree holds up the far corner of the half square. Using an RF Analyst RF-1 antenna analyzer, I trimmed the AWG#10 wire at the far end to 40M resonance. The total path from the transformer high-Z terminal to the end is 46 feet, including the whip (used for resonance adjustments). My SWR values per the FT-891 rig and an old Calrad SWR meter were: 1.3:1 at 7.0MHz, 1.5:1 at 7.21MHz, 1.3:1 at 7.3MHz. The 20M SWRs were worse: 3:1 at 14MHz, 2:1 at 14.150MHz, and 1.6:1 at 14.350MHz. OK on 40M, and usable on 20M SSB. The antenna is thus both a NVIS radiator on 40M and a half square low angle radiator on 20M for dx. I'm somewhat satisfied with it. :)

It is my opinion that the two slightly different counterpoise lengths (one via coax/beads, and the other via the ground wire) are responsible for the two slight SWR dips at the ends of the band. But it troubles me that by extending the whip length out even by an extra 2 feet length doesn't seem to change the 20M SWR response much at all, although extending it does lower the 40M resonant frequency as expected. I'm guessing that the antenna end impedance at 40M is around 3000 ohms, and above 4000 ohms at 20M. And since the transformer output is 2450 ohms, no adjustments to the total resonant length will help the 20M SWR much. I'm really suprised by the low velocity factor of the wire (about 0.66), which by also making up the horizonal section, forces the two vertical sections closer than they should be for a half square. 23 ft apart in air is only 1/3λ on 20M, and not the 1/2λ air phasing traditionally used. That difference is sure to lower the gain by 2dB and decrease the depth of the nulls. One of those nulls is pointed directly at east coast Florida, with their loud signals. :( I considered replacing the horizonal section with bare copper wire wire for the horizonal section to make the spacing problem go away. But for only 2dB more gain? And by fixing the problem the normal bidirectional beamwidth would narrow to place England at its edge, not what I want. So I've decided to live with it the way it is. Although only an 18 ft high NVIS antenna on 40M, I get 40M signals from all over the USA. Low angle vertically polarized 20M gets me Bulgaria, Romania, and Merry Old England. Next I'm going to try to use a spare FT240-43 toroid to make a tapped secondary unun for both 36:1 and 64:1 to see if I can bring down the SWR on 20M. Thoughts and advice?