RadCom February 2016, Vol. 92, No. 2

February 2016

here is evidence, from

several experimenters, that

elevated radials are often

as effective as many buried radial

wires when driving a vertical

radiator via a coaxial feeder cable

(summarised in ON4UN’s

Low

band DXing

My own experience comparing a 3.5MHz

ground plane (GP) with a ground mounted

/4 radiator confirmed this. The purpose

of either a buried ground system or an

elevated counterpoise, in this situation,

is to provide a return path for antenna

current. The counterpoise may be seen

as predominantly a reactance, between

the feeder braid and ground, whilst the

ground system provides predominantly a

resistance between braid and ground. In

either case the returning current causes a

potential to exist between the coax braid

and ground. Consequently this causes

a current to flow on the braid, causing

wasted power with potential for EMC

problems during transmission and noise

during reception.

To eliminate braid current we must

cancel the impedance between braid and

ground. At the ground connected system,

ideally we would need to introduce

negative resistance! As this is impractical

the other options are; a) to introduce

a feeder choke, but to be effective this

needs to have impedance many times the

typically 350Ω impedance of the braid

and it does nothing to reduce dissipation

in the ground resistance, or b) to make the

ground connection resistance very low, but

this may for example at 1.8MHz require

120 radials occupying half a hectare (over

an acre). For the counterpoise we simply

need to introduce series reactance of

opposite sign. An example is an HF whip

antenna, driven against the metal body

of a vehicle. As it is only a fraction of a

wavelength in any dimension the vehicle

presents capacitive reactance to ground,

which we normally neutralise incidentally

by adding to the inductance that tunes the

whip. A resistive load between inner and

outer of the feeder may alternatively be

achieved by adding the same inductance

between the braid of the cable and the

vehicle. However, in the first case the braid

is at the potential of the radio and therefore

has no current whilst in the second case

the braid is at ground potential and will

therefore carry a current back to the radio.

At our home station we are unlikely to be

sitting with our radio at the feed point of

the antenna and must therefore adopt the

second method by ensuring that the braid

is at ground potential where the inner

conductor connects to the antenna.

Perimeter loaded radials

The radial counterpoise advocated here,

which is probably not original, was

investigated to see if it would reduce the

site area required to accommodate vertical

antennas.

Figure 1

, a top loaded folded

monopole with perimeter loaded radial

(PLR), shows the principle where the

density and screening effect, of elevated

radial wires, are both increased, by joining

their outer ends with perimeter wires. The

technique is applicable to three or more

radials but diminishing returns make

four the most economic quantity. The

particular size of PLR for self resonance, at

which impedance at the centre connection

referred to ground is near zero, requires a

square with sides of approximately

/5.

Although more wire is required, it needs

only a quarter of the land area necessary

for four conventional

/4 radials. Because

precise tuning requires simultaneous

adjustment of all eight wires it is convenient

to make a PLR with sides shorter than

and to rely for resonance on an adjustable

loading inductor at its centre. For example

with the amount of wire that we need to

make four

/4 radials we can make a PLR

occupying just 16% of the conventional

area!

Whilst modelling, for gain, bandwidth

and braid current, I found 0.13

per side

(12.5% area) to be optimum but it was

not critical. It was more important to

have symmetrical horizontal construction

because this minimised the undesirable

tendency for the counterpoise to become

an antenna.

Practical results

Modelling the antenna with PLR was

satisfying but I needed to prove that the

combination would work in practice.

Therefore an antenna was constructed

for 3.75MHz in an area, clear of ground

radials or obstruction, that enabled field

strength measurements from several

wavelengths away.

The test antenna, as illustrated in Figure

1 and which may be linearly scaled for other

bands, was a 48mm (1.8”) diameter metal

mast 12.2m (40’) high, with four sloping

top loading wires of 5m (16’) incorporated

into guys. Initially it had four conventional

horizontal radials of 21m (68’) at 0.03

(2.4m, or 8’) height, supported on poles.

Subsequently the radials were converted

into a PLR with sides of 0.15

(12m, or

40’) in which the total length of wire was

exactly the same as that required for four

/4 radials at 3.75MHz and sufficient for

self resonance at approximately 5MHz.

Finally the PLR was reduced to 0.11

(9.1m or 30’) sides with outer corners

conveniently supported by the guy

ropes. Wires for top loading were 2mm

galvanised steel and wires for PLR were

0.8mm enamelled copper. A roller coaster

inductor provided convenient adjustable

loading of the PLR. A 1m copper pipe of

22mm diameter provided a mast ground

connection for static discharge. Ground

connection resistance, in this case, has

little impact on radiation efficiency because

there is only a small residual current in the

mast below the feed point.

Rather than break the mast with an

insulator, two 5mm shunt feed wires were

connected from the top to insulators at the

feed point and from there via a variable

capacitor and SWR meter to the coax. The

shunt wires, which may be coaxial cable

Reducing an antenna’s

Technical

using perimeter

loaded radials

footprint