[TowerTalk] 160 radials

[jimlux]

Hal Kennedy wrote:
> For those who believe they are putting 1/4 wave radials on the ground -
> it might be important to know/remember that the velocity factor of wire
> on the ground is approx. 0.5.  Quarter wave radials are actually approx
> a half wave electrically -  which is why it takes so many of them to get
> a monopole down to 35 ohms - each radial presents a high impedance if
> its 135 ft long and on the ground. 

Not really.  I'd contend that a wire laying on a lossy substrate (soil) 
isn't going to have a real well defined resonance.  A propagation speed, 
certainly, but the loss from one end to the other will be so high that 
by the time the reflected wave gets back to you, it will be 
infinitesimally small.

if the wire is perfectly straddling the perfectly flat interface between 
the soil and air, the propagation velocity will be, to a first order, 
about 1/sqrt(epsilonsoil/2).. If you use the usual 13/0.005 sort of 
model for soil properties, then you're looking at around 1/sqrt(6.5) or 
around 0.5.  If the dielectric constant is a bit lower, or the wire is a 
bit higher, then the oft observed factor of 2 will be the case.

Though, the loss is so high, that for all intents and purposes, radials 
laying on the ground are non-resonant.  There are good sounding 
geometric reasons for picking a length comparable in height to the 
radiator, but even those don't always stand up to rigorous scrutiny. 
It's convenient, though...

N6LF has a lot of good analysis on radial fields (and tediously 
collected experimental data that is consistent with the analysis) on his 
webpage.



  BC stations use quarter wavelength
> (mechanical quarter wave) radials because 120 of them will provide a low
> impedance when placed in parallel and current share well since each are
> a high impedance. 


One wants to be careful about reading too much into broadcast practices.
a) it's a different frequency
b) they're interested in having a very radiated ground wave for 
regulatory reasons
c) they have different cost/benefit analysis assumptions than the usual ham.
d) they have various preferred implementations from a regulatory standpoint.



>  
> You can easily prove this to yourself.  Lay an 80 meter (or higher in
> QRG) dipole on the ground and check it with an MFJ.  It will be resonant
> near 160 meters.  Do it quick with a 10M dipole - it will be resonant
> around 20M.  You can't do that test with a 160 dipole as it will be
> resonant around 900 KHz and the MFJ won't go that low.

Actually, this is one (of several) approaches for measuring soil 
properties. While your demo is a nice way to demonstrate that the 
dielectric properties of the soil affect the resonance of a dipole, I 
don't know that I would use it to infer that ALL properties of a dipole 
scale. In particular, I'd be willing to bet a six pack of your favorite 
frosty cold beverage that the feedpoint Z of the dipole on the ground at 
twice the observed resonant frequency isn't all that high, compared to 
the observed feedpoint Z at "resonance".  Without actually running the 
numbers, I'll go for within a factor of 4.

Measuring resonance of a dipole on the ground was studied by G. Hagn at 
SRI back in the 70s as a means to determine soil properties at HF, in 
connection with work on developing good field expedient antennas for use 
in jungles.  He later pioneered the use of the OWL (open wire line) 
technique for soil property measurement (it's in the ARRL antenna book).

The dipole on the ground technique has the problems that a)it takes a 
lot of room and b)it's very sensitive to layout and distance of the wire 
from the soil.  Sticking a chunk of transmission line into the ground is 
a lot more reliable.