[UK-CONTEST] Vertical radials

[David, G3YYD]

The best book to read on radials is Low Band DXing by ON4UN and read the 
section on verticals.

The key thing to remember about radials is they act to screen the 
underlying lossy soil from the RF current (note current). A radial 
system of any sort shape or whatever should be thought of as a ground 
screen. Therefore the spacing between the radials ends is dependent on 
wavelength of the RF current. The smaller the wavelength the closer the 
ends of the radial need to be. But the shorter the wavelength the 
shorter the radials need to be for the same screening effect.

The other key item is the distribution of the RF current. Assuming the 
radiator is a quarter wave long or less then the current maximum is at 
the start of the base of the radiator and then declines as we move out 
from the base of the radiator. With loss proportional to current squared 
this means that the most gain in terms of reducing loss with the ground 
screen is near the base of the radiator. As the ground screen extends 
further out from the base the reduction in loss per unit length drops 
very quickly (remember the square law). In other words relatively short 
radials will get you rapidly to the point where loss reduction is not 
worth the extra effort and cost. The distance depends on just how much 
the extra dB is valued and the loss of the underlying soil, but 
generally an eight of a wavelength length is sufficient and a quarter 
wave giving another 1dB (poor soil conductivity giving more loss 
improvement with the extra length).

The spacing of radial ends to achieve maximum screening is 2.5m at 160m. 
Doubling this to 5m spacing will increase loss by about 0.5dB (some soil 
and length dependency) but half the effort and half the wire. For higher 
bands reducing the radial end spacing pro-rata, i.e 1.3m on 80m 
(1.8/3.5*2.5).

As for material, avoid using buried galvanised wire. I tested it at my 
location, London clay,  and it dissolves away within about 2 years due 
to either soil chemicals reacting with the zinc or through the battery 
effect with the other metals it is in contact with. I use 1mm 
multi-strand copper wire plastic coated. Give Westlake a call and see 
what he has got in stock - it will not be on his web site. For RF 
performance the wire diameter can be very small as so many are being 
used in parallel. It is the physical strength required to avoid breakage 
that determines wire diameter.

I also suggest the coax feeder is run under the ground screen to 
minimise coupling between the radiator and the coax outer. This needs a 
bit of pre-planning i.e. lay the coax before the radials!

Do not use radiators over a quarter wave long as the peak RF current 
moves out from the base and that means the radial lengths need to 
increase to meet this situation to keep ground losses down (more cost 
and effect for no gain). If you want more signal strength then use more 
radiators to form a vertical beam.

David G3YYD