Page Content - skip navigation
A number of devices were made during the 19th.Century that developed the technologies that enabled and, to some extent, prompted, the invention of thermionic valves.
Experimental incandescent lamps with platinum filaments in a vacuum were developed by 1840. Frederic Guthrie discovered the loss of charge from heated bodies in 1873. Swan and Edison both patented commercial lamps in 1878. On February 13, 1880 Edison rediscovered thermionic emission and made experiments to show conduction between the filament and a second electrode in the envelope of an incandescent lamp. He patented the use of this device as a voltage regulator.
Other electron devices that preceded valves included the Crookes tube (1875), used to study conduction in low pressure gases, and X-ray tubes. Crookes tubes operate at high voltages (at least several kV) and ionisation of the gas produces free electrons which strike the anode with sufficient energy to produce X-ray photons. These were first systematically investigated by Wilhelm Röntgen in 1895.
Fleming discovered the rectification of the Edison device and patented its use to detect radio waves in 1904. This was a simple diode with either a carbon or tungsten filament. The one shown at left was a commercial version, probably made before 1910.
On February 18th., 1908, Lee DeForest patented the three electrode Audion. De Forest himself admitted that he had no idea how it worked, but did specify that it should contain gas at low pressure. This was the first device that used a third electrode between the filament and the anode to control the current in the valve. This was not a true vacuum tube. Early Audions were very unrelaible due the the gas filling being adsorbed onto the internal walls and metal components.
Many other investigators, including H.J. Round, of Marconi, G.W. White, at the Cavendish Laboratory,and Lieben, Reisz and Strauss at Telefunken developed these devices and improved the reliability, mostly by adding facilities to allow the gas pressure in the valve to be restored to the correct working level. Round's valves had a tube attached, filled with asbestos fibres. When this was heated, with a match, gas was evolved and the function of the tube restored. White's 1916 valve contained mercury vapour and had an oxide cathode. It was put into service with the Royal Navy but proved to be erratic in service.
Dr. Irving Langmuir, working for the General Electric Company, developed a hard vacuum triode, which he patented in April 1915. He had earlier developed improved diffusion pumps, to produce harder vacuums and from 1913 developed his Pliotron, along with a hard vacuum diode, the Kenotron.
The French TM valve was developed from the pliotron. It was the mainstay of radio communication during the first World War and the basis of all the later bright emitter triodes until the early twenties.Most French examples had ungettered spherical envelopes with tubular anodes, spiral grids and straight pure tungsten filaments. British made 'R' types were more varied and included tubular envelopes and various gettering.
German valves of this period used a very primitive electrode structure, which resulted in valves which could only be used for low frequency amplification. In late 1917 they began to make copies of captured TM or R type valves.
Here is a timeline of the development of valve manufacture in the U.K. up to 1930.
"Brights" used pure tungsten filaments and were operated at 2500-2600 K. This is somewhat brighter and whiter than a modern incandescent bulb. All these designs were more or less the same as the 'R' types that were used during the first World War. Filament power was typically 2-3 W (4V and 0.75A). Cold filament resistances of less than 0.5Ω are diagnostic of bright emitters. Gain (μ) was between 6 and 9 when operated at 30 - 80 anode volts.
Dull emitters use a thoriated tungsten filament operated at 1700 - 1900 Kelvins. These gradually replaced bright emitters in domestic receivers in the period from 1922-5.The thoriated tungsten filament was much more efficient, consuming less than a watt. This reduced 'A' battery drain to one or a few hundred milliamps, extending accumulator life between charges by a factor of 5 to 10.
A.C. Cossor in the U.K. and Western Electric in the U.S. both solved the problem of producing a filament coated with alkali metal oxides that was robust enough for prolonged everyday use, before 1925. These were operated at 900-1000 Kelvins, a dull red heat.
Data for these valves was apparently
not published
but I have found some heater
data which
may be of use to those wishing to "light up" one of these valves. There
is a characteristic curve on the B5
page.
Philips had invented the 'Azide' process which increased the emission of dull emitter cathodes beyond that obtainable from thoriated tungsten. This process produced an oxide covered cathode which was more robust and long lasting than other competing methods, for example the A.C. Cossor 'Wuncell' type. Before assembly the tungsten filament was directly oxidised or plated with copper and oxidised. The anode was coated with barium nitride (Azide). After evacuation the anode was heated (probably by induction) and the evaporated barium nitride reduced the oxide on the filament forming barium oxide, metallic tungsten or copper and nitrogen which was adsorbed by the getter.
Mullard commenced production of these P.M. (Philips Mullard) valves in 1925 with the PM3 and PM4.
Cossor had earlier perfected a process for oxide coating filaments. A new process enable them to reduce the filament current to 100mA for their Point One series of valves.
Triotron manufactured valves in Austria, the valves were imported into the U.K, and achieved significant market share as they were cheaper than those made by the B.V.A. cartel. The earlier ones appear to be 'Azide process'like the Mullard 'PM' series above, judging by the dark speckled deposits on the envelopes.
Tungsram valves were manufactured in Hungary, by a long established electric lamp maker, who had patented the tungsten filament in 1903. The name is a contraction of TUNGsten and wolfRAM. Large numbers were imported into the U.K. Like Triotron they achieved significant market share by undecutting the B.V.A. cartel. Again these were 'Barium' valves and may have been produced by the 'Azide process'.
The PX230 and PX4 have quite heavy filament currents and although directly heated they are often used in mains equipment.