These are my logs of working JPC-12 antenna settings from my time in the field. The JPC-12 is a coil-loaded 1/4 wavelength vertical antenna that can be tuned for 40/30/20/17/15/12/10/6m (and maybe 60?)

The most direct way to assess ground system quality is to measure the feed point impedance of a known antenna with a VNA or antenna analyser. For a 1/4-wave vertical, the theoretical feed point resistance over a perfect ground is approximately 36ohm. Measured values higher than this — say, 50–80ohm — indicate significant ground loss resistance. The excess resistance (measured R minus 36ohm) is your ground loss Rg. Improving the radial system and re-measuring shows the direct benefit of each improvement.

Additional ground plane thoughts. The upgraded radials using pairs of wires with a crimped terminal connector worked really well. I currently have 4 wires (2 pairs) cut around 17' and 2 wires (1 pair) cut around 9'. The shorter ones provide good matching for 10/12/15m especially when slightly elevated. Theory says it's desirable to have radials be roughly odd multiples of 1/4 wavelength (1/4, 3/4, 5/4), which makes sense. A perfect ground would effectively be infinite length, but also a perfect ground would give an antenna impedance at the feed point of 36 ohms instead of 50 ohms.

A 1:1 common mode choke at the antenna's feed point to decouple the coax shielding from the radials is also a good idea. Mix 31 or 43 toroid wrapped with coax can provide a 5000ohm impedance at HF frequencies to block CM current on the coax shield. Ferrite cored choke chart shows impedance by frequency for different turns and mixes, black on the bottom half means frequency/reactance doesn't play much role in the impedance calculation, it's more than half resistant (which is desirable so the impedance doesn't vary with transmitting frequency). I ordered one that is 9 turns 350w nickel-zinc "green" probably mix 43? But "green" and other colored cores typically indicated iron rather than ferrite so could be iron mix 41? I hope to test out its performance using the VNA, following the process outlined by Lance. During field testing, my not-very-scientific assessment is that touching the coax shield without a choke made a big difference on the VNA's smith chart on 15m and touching with the choke did not. On 20/30m far less difference was observed, possibly due to the better radial resonance plus the feed line being a more advantageous fraction of the resonant frequency. In any case, I don't see much harm in using it for all bands. The one I ordered is inexpensive though, the end already came off once and I had to solder it back together. And I still want to measure how much common mode impedance it's actually producing.

10m

Theory: 8.4' element 8.1' radial

12m

Theory: 9.7' element 9.4' radial

15m

Theory: 11.4' element 11' radial

Ground spike, radials, feed point, four aluminum pole segments, telescope with 1.8 collapsed. Elevated the ends of two radials using traffic cones to improve SWR.

17m

Ground spike, counterpoise, feed line, four aluminum pole segments, coil (four higher than center of gold), telescope with 0.5 collapsed. Positioning of counterpoise has significant impact on SWR.

Ground spike, radials, feed line, four aluminum pole segments, telescope fully extended minus 1 inch. SWR 1.06.

Theoretical radial length: 13'

20m

Ground spike, counterpoise (13ft 6in), feed line, four aluminum pole segments, coil (one click below red marker near top), telescope with 0.5 collapsed. Had to lengthen during operation.

Similar setup but with elevated counterpoise gave me 1.01 according to CAT. Hard to replicate consistently, likely due to grounding differences.

Theoretical radial length: 16.7' or 18.7' depending on who you ask.

30m

Theory: 23.8' element 23.1' radial

40m

Theory: 33.7' element 32.7' radial

Pages that link here: