Janine Elliot charts the rise and fall and rise again of the Thermionic Tube/valve in this fascinating and informative article. 

1960 was an epic year. Not only was it an epic year for me, but it was the year BBC Television Centre was opened (on 29 June 1960), the first edition of British soap ‘Coronation Street’ and also when Jimmy Hendrix did his first paid gig. It was also the year that old-hat valve-based HiFi reached its peak before transistors would see them parked unscrupulously in the bin until, like aged vinyl, it would begin its second renaissance.

Whilst many manufacturers during this important time would be producing products both valve and transistor based – and even a mixture of the two – some companies, like British manufacturer Lowther were quite  late in disbanding the bulky, inefficient, fragile and hot component that was first constructed in 1904.

By 1970 most Hifi companies had turned to manufacturing transistor based equipment, particularly led by Japanese and American companies, and then by British and other countries. However, a number of companies such as Quad and Leak worked with both at the same time. For example the acclaimed Leak Stereo 20 (and less acclaimed Stereo 60) valve power amp was produced until 1967, but their Stereo 30 transistor amplifier ran from 1963-67. Similarly, the Quad II valve amp first appeared in 1953 but ran until 1969. In America Fisher Radio Corporation ran a TX-300 transistorised amplifier at the same time as the X-100 valve amplifier. The transistorised version was 3 times the price of the valve amplifier. Lowther’s LL18 mono, LL18S stereo and LL26 MkII valve amplifiers ran on until 1975.  Interestingly, they brought out their first transistor amp, the Transistor Amplifier Unit (a preamp available in three versions of high impedance, medium gain, and high gain) way back in 1956. It is true to say that early transistorised amplifiers were worse than their valve replacements or alternatives, hence why early valve amplifiers fetch such good prices in antique audio-fairs. Leak’s Stereo 30 transistor amplifier carried an advert saying that since it was 43% of the size, 48% of the weight and 89% of the price of an equivalent valve amp, plus apparently greater reliability, how could the valve possibly survive?  They worked out it was “500% better” because it didn’t get hot, unlike the valve. It wasn’t their best creation.

Whilst the transistor was first patented as an idea by American physicist and electronic engineer Julius Lilienfeld way back in 1925, with the first produced transistor in 1949, its sheer price prevented it superseding the valve until those prices came down and valve prices increased; supply and demand in its infancy, then. Figures from the RCA Company in the US show that they sold just over a million tubes in 1922, but by 1924 this had risen to over 11 million. With the advent of broadcasting in 1922 (BBC, on 18th October) the needs for valves increased significantly for both broadcasters and listeners alike. Valves and tubes did have their problems, though, preventing their fame possibly having been higher. They were not only quite expensive (even in 1950’s, buying a Leak mono power amp would cost you a month’s wages), but running them wasn’t that simple. For a start they were power thirsty.  Early valves used directly heated cathodes and that meant you needed big batteries to run them. There were no PP9’s then, nor would their 9 volts be sufficient. My 1950’s Eveready “portable” Sky King radio uses a massive B136 battery with 4 terminals (LT+ 1.5v, HT 90V plus two negative terminals), which were not cheap and nor were they lightweight. This radio was actually one of the very first truly portable radios you could buy with 4 B7G valves.  Once indirectly heated valves were developed, this opened up their use considerably and they were more widely used in radios.  The 1950’s and 60’s saw a great rise in valve amplifiers, tape recorders and radios. Most “Hifi” was part of the home décor, so size didn’t matter. Indeed, much of the components came in “chassis form” for you to put inside your teak or mahogany wooden frame, complete with Garrard or Collaro turntable plus loudspeaker, just as originally the motor car came in a chassis for you to contract a ‘coach builder’ to create the bodywork.

The story of the valve (or thermionic tube) goes back to the American Thomas Edison, when he noticed a plate placed in the electric lamp that he invented offered rectification of AC to DC current. He noted the effect but didn’t do anything more. That was left to a Brit’ on 16^th November 1904.

Lancaster born John Ambrose Fleming (he created the famous ‘left-hand-rule’) built the first valve to be used in an electronic circuit; a “Fleming Diode” was a two element (anode and cathode) three wired vacuum tube as an “instrument for converting alternating currents into continuous currents”. To explain how it worked it had two connectors for the cathode which was heated up with a 12V battery, which sent negative charges to the anode. Applying AC current to the cathode and anode terminals and with a resistor in parallel between them created DC between the resistor. This was then used as a detector and rectifier in the then important world of radio telegraphy. The following year the Fleming Diode began being manufactured by The Edison and Swan United Electric Light Company and within a few years was being used in the Marconi-Fleming 2 valve wireless receiver. I say two valves; actually there was one valve with another one in reserve since this new tube technology was so unreliable. The first Triode was made in 1906 by American Lee de Forest, called the ‘Audion’, (pictured) a fish-bowl shaped tube in a low vacuum, which itself led to the “Type A”,  which as well as being designed as a detector of wireless signals could also be used in amplifiers. It was to appear therefore in amplifiers along the newly created long-distant telephone lines. This was followed by the British “R” valve which was a higher vacuum triode, which itself led to a number of other triodes.

The name “thermionic” in relation to valves relates to the use of thermal processing; the use of heaters. In normal life metals have lots of electrons which move about in response to the application of electromagnetic field. Normally the positive charge on the atoms cancel out those negative (electron) charges. However, if heated up those electrons have more energy, and in a vacuum they move about quite freely.  Usually electrons would be fired from a hot cathode into a vacuum known as ‘thermal electron emission’ or the ‘Edison Effect’.

Triodes, have three electrodes inside an evacuated glass envelope: a heated cathode (usually achieved by a separate filament) from which electrons leave it, then a grid, and at the other side another plate which was the anode. The ECC83 is a commonly used “double-triode”; if you look inside there are two separate units in the vacuum tube. The Pentode, invented by Mullard/Philips, by its name has five electrodes in total; namely the addition of two extra grids in the centre. This is to better control the cathode to anode current flow. As electrons flow about they could tend to work their way back to the cathode and so the third “suppressor” grid is there to prevent this from happening. An example is the EL34 tube. A single EL34 creates an amplifier of around 11 Watts.  A Tetrode has just two grids (hence 4 electrodes). The KT66 (pictured) and KT88 is a beam tetrode invented by GEC (the ‘KT’ stands for Kinkless Tetrode)… two of these in “push-pull” can create around 100W.

Commercially manufactured valve amplifiers began to appear in 1913, just before the start of the First World War. World wars seem to create great inventions; the First WW saw U.S. Army Major Edwin Armstrong in 1918 invent the superheterodyne principle, whilst in France, a method of getting better radio reception by moving the signal to a fixed intermediate frequency (IF) was invented. All radios would duly go this way, but it meant more valves were needed, so that meant prices becoming cheaper as more were made to supply demand. Between WW1 and WW2 amplifiers tended to be very heavy, big and ugly, mostly using triodes such as the PX4 and PX25. Two PX4’s in “push-pull” Class A setup would easily give 10W. They also had low distortion, making them popular until the 1950’s … amplifiers such as Decca’s PA/VI used them.  Similarly WW2 was a key point in the development of the thermionic valve, being needed for electronic equipment from radio receivers and transmitters, to electronic warfare equipment and radar systems, and much more. Where early valves were large, after WW2 they began to get smaller. Now, in the 21^st century valve amplifier manufacturers are, in the main, keen to make valves highly visible, whether large or small.

There are many types of valves and their numbering system is highly complicated, and would take too long to explain here. There were a number of different generic codes to describe their design and use, for example the RMA system, RETMA, EIA , etc, etc. Certain letters would describe the manufacturer of origin (eg PL = Philips, SV – Svetlana) and numbers and letters to describe more details of the most common valves used in hi-fi equipment of this period such as ECC83, KT88, KT66,  B7G, EL84, EL34, PX25, ECL86, etc. The standard valve system negotiated between Philips and Telefunken in 1933-34 is the easiest to explain. Using this system the iconic ECC83 means E= +6.3v, C = small signal dual triode (ie effectively two triodes … CC), 83 relates to the chronological order; this being a 9-pin tube (numbers 80-89 were all 9 pin tubes.). ECC83 was the 1950 European version of the 12AX7 (pictured), originally conceived in America in 1947 by RCA (originally having the developmental number A-4522, just to confuse even further and being two 6AV6 triodes in one vacuum tube). The ECC83 is thought to be better than the 12AX7 by some. With the 12AX7 the ‘12’ stands for 12V. Bear in mind this is a double triode so 6.3V+6.3V sort of makes 12. The tube has a centre-tapped heater separated from the cathode (so it can be either one or two heaters). Whichever version, this famous valve is commonly used in first stage/preamps connected one after the other until arriving at the required amplification (each works at an amplification factor of 100). The Leak Stereo 20 has three in total, plus two EL84 pentode output valves per channel. The EL84 had two ‘similar’ relations, the PL84 and UL84, but they were used for smaller television amplifiers (E stands for 6.3V heater, P for 300mA heater and U stands for a 100mA heater).

Today valves are playing an increasing importance in home audio, with rediscovered tubes from the forgotten past reappearing in newer kit as the ‘popular’ valves become too popular for some. Whilst some might say certain valves are better than others, and I agree certain tube manufacturers produced the best of each type of valve, it is more important how the tubes and other components are combined in the single package of amplifier or control unit, that defines how good it really is. Some, like the Leaks, Quads, Radfords, Pye Mozart, Rogers and many others will go down in history as epic, but many others are best forgotten. Now companies, particularly new ones from Eastern countries, are reliving the magic of the thermionic tube, and other favourites  such as McIntosh and Audio Research, and more recently Manley Labs have never stopped recreating valve paradise showing that maybe we haven’t yet quite reached the pinnacle of hi-fidelity valve audio.  Now valve based audio is on the up in the tables, meaning that perhaps we might again reach that pinnacle of 1960 where both valve and transistor products are of an equal par. I hope so!

Janine Elliot