Almost everyone has heard the effect of switching a light on or off when a radio is in the room; the spark in the switch causes RF radiation which the radio picks up. The spark transmitter did the same thing, though with a modicum of tuning. Not enough, though, to enable the use of spark transmitters now, so don’t try it! They wipe out large areas of spectrum. The UK Amateur Radio Licence used to have a sentence explicitly forbidding the use of spark transmitters, but it no longer does. The regulatory authorities perhaps assume that no-one would be silly enough to try.

The discovery of radio waves is usually credited to Heinrich Hertz; but the American, Joseph Henry (whose name is given to the unit of inductance) noticed that when he was experimenting with large coils and there was a thunderstorm around, he would get unwanted sparks. His coils were picking up the radio waves from the lightning flash - a truly large-scale spark transmitter. (Hertz, incidentally, might have been the discoverer of the photoelectric effect had he not died relatively young. His receiver was a coil which gave a spark across a gap when receiving; Hertz noticed that the spark came a little more easily when the spark-gap was illuminated, but regarded this as an incidental to his main line of enquiry.)

The spark transmitter is very simple, but it generated a large number of technical problems mostly due to very large induced e.m.f.’s when the spark struck, which caused breakdown of the insulation in the primary transformer. To overcome this the construction of even low-power sets was pretty hefty. Low power is a relative term. The wavelengths used were from about 50 to 6000 metres. Wavelengths shorter than 200 m (frequencies above 1.5 MHz, or perhaps I should say 1.5 Mc/s) were all allocated to amateurs even into the 1920’s, with power up to 500 W. ‘Aeroplane sets’ used 200 to 600 m, also with about 500 W; 450 to 800 m for ships with power to 10 kW; 900 to 1500 m for moderate-sized land stations with 5 - 20 kW, and wavelengths over 1500 m (frequencies below 200 kHz) for large land stations with powers up to 100 kW. Some of these transmitters had the key switching the power circuits directly (see below), with perhaps 50 A flowing; no wonder some keys needed cooling. Imagine the Health & Safety Regulators’ views on these!

The essential features of the spark transmitter were:

• An alternator, driven by an electric motor (remember that most mains supplies were d.c. at the time) or maybe a gas or oil engine, producing 120 V a.c. at around 500 Hz;
• The key, which keyed the alternator output which was applied to
• A step-up transformer giving between 10 and 20 kV; the output was connected via r.f. chokes to
• The spark gap
• The condenser (capacitor) which was charged to be discharged through the spark gap;
• An oscillation transformer which transferred energy to the aerial circuit;
• A tuning coil;
• The aerial.

where W = power/W,  N = no of sparks/s,  C = capacity/F,   V = voltage/V. Further, the value of C₁ affected the operating wavelength, so that the frequency at which the transmitter (nominally) radiated depended on the power!  The condensers had to be big and have a high working voltage; Leyden jars were favourite for land-based stations, with mica being the dielectric for ship sets which were necessarily smaller. A unit of capacitance widely used until relatively recent years was the ‘jar’.

The inductance L₁ together with the spark gap and C₁ produces the high-frequency oscillations, which are transferred via L₂ to the ‘open’ circuit. The coupling of L₁ and L₂ was fairly important but affected the tuning, so the function of L₃ is to tune the output of the set. This is also the function of C₃, which became important and higher frequencies.