Early attempts at wireless communication involved sending audio frequency signals through the ground via two electrodes buried a hundred metres or so apart from each other. The signals could be detected up to several kilometres away by two ground electrodes similarly spaced or by a loop laid on the surface [1 and 2]. It was originally known as the 'conduction' method, although it clearly involves induction as well. Transmission and reception are thus via near-field effects. The system is now commonly referred to as an 'earth base'.

The author had experimented with this system for years, more recently using modern solid-state amplifiers for transmit and receive. When the 73 kHz allocation became available, it was decided to test the efficacy of an earth base as a radiator at that frequency.

Each end of the base consisted of a pair of 1.2m earth rods driven into the ground about 4m apart and linked together. The ground around the rods was well salted and watered. The distance between the base ends was 230m and 2.5mm stranded wire was used for connection. Transmitting equipment (fig1) was the CW - modulated 73 kHz output of a signal generator, amplified by a 1000W BK Electronics module [3]. The module requires an output impedance of 2W for maximum performance. This was accomplished by interposing a purpose-built matching transformer [4] with a tapped secondary connected to the ground electrodes. To the author's initial surprise a good signal report was soon received from an LF station (G4GVC) 100 kms distant. Far-field effects were clearly responsible. The length of the base was short in comparison with the wavelength of the transmission frequency and the evident success of the earth base as an antenna was puzzling.

BLOCK DIAGRAM OF AMP AND TRANSFORMER/BASE (DOTTED LINE TO DISTANT RODS) (fig1)

To match the amplifier output to the earth base system it was necessary to know the earth base impedance. Initially, measurements made at frequencies up to 73 kHz indicated that the impedance of the ground increased with frequency. The release of the 136 kHz band in the UK and the subsequent successful use of the earth base system at that frequency led to measurements of the earth base impedance at higher frequencies. The results showed that ground impedance rose and then fell as frequency increased, later rising again. (fig 2). No two days measurements were exactly the same. Readings taken over a period of time showed that in wet weather the initial peak and trough occurred at lower frequencies. There was a rough correlation between the amount of rainfall and the displacement of the peak and trough frequencies.

fig 2

A possible explanation of this phenomenon is as follows.

The velocity of electromagnetic radiation in a medium is slower than in a vacuum. The relationship is given as u = c/Ö d where u = the velocity of the radiation in the medium, c = velocity in vacuo and d = the dielectric constant of the medium. Typical values for d are between 29 (moist soil) and 81 (water). This relationship means that the radiation wavelength in a particular medium is shortened by an amount which depends upon the value of the dielectric constant of that medium. When d = 35, a frequency of 55 kHz will have a wavelength of approximately 920 metres. When the ground is saturated with water the velocity factor of the ground will become even lower. Thus it is easily possible for a 230m earth base to be of the order of l /4 or l /2 at the UK LF frequency allocations.

The behaviour of a long line on the ground resembles that of a coaxial feeder with an eccentric dielectric. The earth base system corresponds to a long feeder line terminated at its far end. A shorted feeder will have standing waves on it at a particular frequency. There are two forms of feeder resonance - parallel and series. At 'parallel resonance' which occurs when the shorted feeder length is a multiple of a quarter wavelength, the impedance at the driven end is high. At 'series resonance' when the shorted feeder length is a multiple of a half-wavelength, this impedance becomes low.

No doubt there are other effects that complicate the issue such as current loops in the ground whose lengths may also vary with frequency.

A method of studying the behaviour of the earth base system over a range of frequencies was devised by G3HMO (fig3). The amplified output of a signal generator is applied as before to the earth base via a matching transformer. The Y plates of an oscilloscope are connected across the earth base to measure the output voltage. An 8W 20 watt resistor is put in series with the transformer's output to the base and connected to the oscilloscope's X plates. Signals are sent through the earth base over a range of frequencies. The resultant Lissajous figure gives a visual clue as to what is happening. A Watson Super Hunter frequency counter [5] provides a constant frequency readout. As the signal frequency fed into the ground is altered, an ellipse on the oscilloscope screen flattens and becomes a diagonal line at two different frequencies.

The gradient of the line indicates whether the base has high or low impedance at that point. These may be the occasions when the system possesses parallel or series resonance conditions. The effect of adding inductance or capacitance can be studied.

DIAGRAM (fig3)

Theory indicates that the antenna will transmit best if the length of the earth base is a multiple of l /2 in order to obtain maximum current flow at a particular frequency. It is a matter for further experimentation to see whether this is so and what other factors, such as adding inductance or capacitance or varying the length of the base, affect the efficiency of this 'dielectric antenna' for VLF. One advantage of this antenna is that it presents little visual impact - it can even be buried.

[1] Stanley, Rupert ‘Textbook on Wireless Telegraphy’, Vol. 2. Wireless Press 1919.

[2] Meulstee, Louis ‘Earth Current Signalling. The History of the Power Buzzer’.

Journal of the Royal Signals, Spring 1988.

[3] BK Electronics

Unit 1

Comet Way

Southend on Sea SS2 6TR

England

[4] Electro-Inductors Ltd. / Aluminium Inductors Ltd.

29 Lower Coombe Street

Croydon

Surrey CR0 1AA

England

[5] Counter from Waters and Stanton

22 Main Road

Hockley

Essex SS5 4QS

England