A replica of an early experimental Fleming Diode.
The front cover of the November 1979 edition of Wireless World is a reference to the fact that in that year is the 75th anniversary of the invention of the thermionic diode, for it was on November 16, 1904, that Fleming filed his patent for 'a two-electrode valve for the rectification of high-frequency alternating currents'. It would be rash to attribute the birth of electronics to any particular device, but there is no doubt that Fleming's diode ushered in the thermionic valve era and, as distinct from earlier scientific work on electrical discharges through gases and vacua, was invented for a practical purpose in communications technology. The following account by Professor Fleming himself, extracted from his book 'Fifty years of electricity' published by Wireless World, shows how he saw the possibility of using the Edison effect for the particular requirement of detecting oscillations in wireless telegraphy receivers.
The article below was written a good while after the events described, in about 1921. Fleming does not mention that at the time of making his invention he wrote to Guglielmo Marconi telling him about it and adding, as an afterthought, 'I have not mentioned this to anyone yet as it may become very useful'.
There is a fourth method for creating continuous waves of steadily increasing importance, called the valve method, which has developed out of an invention made by the author in 1904 of the oscilliation valve for the detection of electric waves.
We have already explained that in the spark system of wireless telegraphy the electric vibrations in the aerial wire are created by the discharge of a capacitor across a spark gap. These oscillations, therefore, come in groups or trains corresponding to each spark, and as there may be from 50 to 500 sparks per second there are 50 to 500 trains of oscillations and, therefore, of radiated waves, each of which may contain 20-100 oscillations or waves. The interval of time between two successive movements of electricity or waves may be of the order of a millionth or a half a millionth of a second. These vibrations are too quick to affect a Bell telephone or even the human ear. If we convert the oscillatory movements of electricity in each train into a single gush or flow of electricity in one direction, then we change the trains into short flows of electricity all in one direction, these gushes coming at the spark frequency viz., 50-100 per second. For such intermittent currents the telephone is very sensitive. Accordingly, it appeared to the author in 1904 that if we could find some kind of conductor which would act like a valve for high frequency currents and let currents in one direction pass, but stop currents in the opposite directions, we should be able to rectify the trains of high frequency oscillations set up in a receiving aerial and detect them by a telephone or any equivalent direct-current instrument. Meditating on this problem the author found the solution by making use of an incandescent electric lamp with a plate of metal sealed into the bulb.
The author had carefully studied in 1883 and 1896, the so-called 'Edlson Effect' in glow lamps discovered by Edison in 1883, and by 1904, as a consequence of the researches of Sir J J Thomson, it was well known that an incandescent filament of carbon in a high vacuum was giving off torrents of electrons or particles of negative electricity. Also, it had been found by the author that the space in a high vacuum between an incandescent cathode and a cold anode could conduct negative electricity from the hot to the cold electrode, but not in the reverse direction. It was not at all obvious, however, that a carbon filament incandescent lamp with a plate sealed into the bulb could be used to rectify high-frequency alternating currents; that is to convert them into continuous or direct currents. Mr Edison had made no such use of his 'Edison effect' lamps, nor had it occurred to anyone, until the author pointed it out, that such a lamp, having a metal cylinder surrounding the filament and carried on a wire sealed through the bulb, could be used to rectify high frequency currents and, therefore, as a detector of electric waves in wireless telegraphy.
The author, however, constructed in 1904 some carbon filament incandescent lamps in which the filament was surrounded by a metal cylinder carried on a platinum wire sealed through the bulb. These lamps had their filaments made incandescent by a six-cell storage battery, and they were connected with the receiving circuit of a wireless telegraph apparatus.
The electric waves striking the aerial wire set up in it rapid electric oscillations or electric currents running up and down the wire. These created, by induction, other electric currents in the capacitor circuit connected to the aerial wire. To one terminal of the capacitor the metal cylinder if the lamp was joined, and the end of the carbon filament in connection with the negative terminal of the battery of cells was connected through a galvanometer or a telephone with the second terminal of the receiving capacitor.
Hence, as the electric oscillations took place in the capacitor, electric currents would flow through the telephone and through the vacuous space, but, as already stated, negative electrons are being given out by the hot filament, and, therefore, negative electricity only can pass from the filament to the cylinder in the bulb, but not in the opposite direction. Hence such a bulb operates to stop all current flow in one direction, but permits it in the opposite, in other words, it acts like a valve for electricity. The author, therefore, called it an oscillation valve and it has generally been named a Fleming valve or thermionic valve. The result is to convert the trains of rapid oscillations produced in the capacitor circuit into gushes of electricity all in the same direction through the telephone. These gushes come at intervals corresponding to the spark frequency, viz., 50-500 per second, and, therefore, produce in the telephone a uniform sound. This is cut up into short or long periods corresponding to the dot and dash of the Morse Code, when the signalling key in the transmitter is manipulated properly.
It was at once found that this thermionic valve gave us a very simple, easily managed detector of electric waves in radiotelegraphy.