My 160m transmit signal with the shunt tower was OK but not great. In contests, I heard many DX stations well but could not work them until the conditions were better. Definitely I was in the middle of the crowd. Still miracles occurred. In 2025 CQ WW 160, special conditions created a duct to GA allowing me to work a pileup of over 200 Europeans, earning #1 assisted US in operator high-power category.
.So with the the new 36 m tower (120′) I set up 4 sq with bent wires as suggested by K9RS and N9NB. .Each element is 12m vertical and 28m bent towards the tower. Since I had little time, I put 8 radials out of 150m spool (500ft) of gauge 14 (1.6 mm) on the ground. . Phasing 1/4 wave lines were made of 35 m (115 ft) RG6 with F connectors, followed by F to PL259 adapters. Phasing was by DXE controller DXE-TFS4-160, and electrical switching was via a Wifi switch by a Green Heron device. Initially bad results as one controls cable was broken and half of F/PL259 adapters were faulty.
The results exceeded the expectations. OK1CF with 599. A small run of EU in the ARRL DX Contest. Even one JA was worked on the first call while a station with 100′ shunt tower and 70 radials could not do it.
I made some RBN comparisons. To EU, the 4 sq is about 10db stranger than my old shunt fed tower. Also about 5 db improvement over the omnidirectional mode. But the old tower is is still OK to the Caribbean. It seems that the new tower north of the old tower works as a reflector to the old tower.
The elements are not yet tuned. Their resonance vary from 1.82 MHz to 1.86, and their resistance from 35 to 48 Ohms. Yet F/B seems to be at least F/B and the dumped power is low: 3% at 1.80, 1% at 1.83, and 10% at 1.92. The omnidirectional mode has a resonance at 1.72 MHz, with SWR going up to 2.0 at 1.84. I plan to raise the resonance of all the element to 1.84 to move the resonance of the omni mode to at least 1.80.
Why 8 radials of 11% wavelength do so well while nearly all recommendations call for 32 25% wavelength radials?
Rudy N6LF has a paper on sparse radial performance. He found that the gain with 8 1/8 wave ground radials is within 0.5 db compared to 32 1/4 wave radials. Few longer radials simply dissipate heat in the ground.
I should have remembered. At one time I had inv L for 160m with 7 20m radials. It worked pretty well. At another time I tried to do better and had inv L with 4 40m radials. The performance was poor, some 6 db below my shunt fed tower. Sometimes bigger and stronger is worse!