Guide to Radio Valves PDF Print E-mail
Written by Bryce Ringwood   
Thursday, 28 October 2010 08:41

Radio valves are the active components of old wireless receivers. These are the devices that amplified the signal, converted it and drove the loudspeaker. In the transmitter, they generated and amplified the transmitted carrier wave and impressed speech on to it, so that a broadcast could be transmitted. This article describes some of their history, what they are and how they work.

How the Valve got its name

In 1905, Ambrose Fleming took out a patent for the thermionic diode. He compared its action to that of a mechanical valve that would only allow the flow of a liquid in one direction. The term 'thermionic' refers to the action of emitting electrons by way of heat. This first valve was called a 'diode' because it had two elements.

A diode valve resembles a light bulb into which someone has inserted a steel plate and as can be seen, some early diode valves look exactly like that.Philips DCG1250   Looks like a lght bulb The steel plate is referred to as the anode and the filament is called the cathode. When the filament is heated, it gives off electrons – negatively charged subatomic particles. If the anode has a positive voltage applied to it, then the electrons will be attracted to the anode and current will flow. If, on the other hand, there is a negative voltage on the anode, then the electrons will be repelled from the anode, but can't return to the filament, so no current will flow.

diode_valve double_diode
Symbol For a Diode Double Diode






The Triode Valve

In 1907, Lee de Forest introduced a third electrode consisting of a grid or mesh of fine wire between the cathode and anode of the Diode valve. This new valve was called a triode (you can see the pattern di=2, tri=3 after the Greek system of numbering).

Small UHF Triode B9a   Base UHF Disk Seal Triode A Group of Valves   and an early transistor  
A2521 Triode DET23 Triode. The disk at the top is the anode A group of Radio Valves. The one on the far right is a transmitting triode. An early transistor is the smallest object, with a nuvistor beneath  

Assuming that a steady current is flowing between the cathode and the anode of the valve, then a fairly small negative voltage applied to the grid could cut the current flow completely. It is this ability of the triode valve to amplify small voltages that makes valve amplifiers and valve radios possible.

Triode Schematic double_triode
Symbol for Triode Symbol for Double Triode

Multi Electrode Valves

Later developments introduced a further grid (the screen grid) to produce the tetrode, capable of even higher amplification than the triode, and then the introduction of a suppressor grid to prevent some unfortunate secondary emission effects caused by high speed electrons bouncing off the anode and back on to the suppressor.

Symbol for Tetrode Symbol for Pentode ech81
Tetrode Pentode Triode-Hexode

In some power amplifier valves, the screen grid is replaced with beam forming plates, as in the beam power tetrode.

In the years immediately prior to World War 2, other multi electrode valves were introduced. These had additional control grids so that they could be used a mixing stages in superheterodyne receivers (for example). These might be hexodes or heptodes, depending on their construction. I'm not sure I've heard of an octode, but there certainly was a nonode valve (type EQ80) designed for use as an FM discriminator.

YL1030 Transmitting ValveAlthough double triode valves were in use in the late 30's, the idea of putting more than one valve in the same glass envelope really took off with the introduction of a tax on radios according to how many valves it contained. I have some double-pentode valves (12AL11), but can't see myself ever using them. The valve opposite is a transmitting valve - a double tetrode.

A better explanation for putting more than one valve in the same envelope is that of simple economics – it is just a lot cheaper. You may see these being referred to as 'compactrons'.

Special Purpose Valves

If you service communications receivers, you will encounter voltage stabilisers. These are a type of diode valve with an inert gas filling. They are designed to hold a high-tension voltage steady at (normally 85, 105 or 150) volts to prevent frequency drift in the oscillator stages of the set. Other gas-filled valves are the thyratron (used in the trigger circuit of some oscilloscopes and the dekatron decade counter tube. Finally, there is the numicator used in early digital voltmeters.

There are also tuning indicators of various types, the favourite being the “magic-eye”. These are small cathode ray tubes designed to throw a shadow on a green background which opens and closes depending on the strength of the received station. Cathode ray tubes are still made in ever dwindling numbers as LCD and LED technology displace them.

Occasionally, you will encounter barreters or current stabilisers. The commonest example is perhaps the 3TF7 current stabiliser made by Amperex and used in the R-390 radio receiver. A year or so ago, Amperex were still making these devices. I can't say these are too reliable. Next one to blow gets replaced by a big resistor.

During the second world war, numerous special tubes were developed for radar and microwave communications. Most of these were oscillators, such as the klystron, the backward-wave oscillator and so on. The traveling wave tube or TWT was an amplifier, and of course, everyone's favourite, the magnetron used today in the microwave oven. You may encounter these in old microwave radio sets.

Finally, there are photo cells, photomultiplier tubes and scintillation counters, as well as Geiger-Muller tubes. Photomultipliers are still made and used today.

Sub-Miniature Valves

The term “Miniature” seems to be reserved for the smallish all glass valves with a B7G base. Some extremely small valves were produced by Hivac among others. They were used in hearing aids and in some radio control receivers. They also turn up in military equipment. Very often Russian military equipment used sub-miniature valves, presumably because they thought they would be immune from the electromagnetic pulse produced by nuclear explosions.

A little more common in the Western world, the nuvistor is only a little larger than an early transistor. These were used in some strategic radios, but were also used in ham radio equipment and in Tektronix oscilloscopes. They were (or are) used in certain microphone preamplifiers, so they are now somewhat scarce and expensive.

Valve construction

The earliest valves used a tungsten filament as the cathode. This would run at a very high temperature and electronic equipment often had a rheostat to control the filament temperature. Thes “bright emitters” soon gave way to a filament coated with an oxide. The elements of the valve were supported on wire supports within the glass envelope., the same way as an electric lamp. This form of construction rapidly gave way to a form where the elements were placed concentrically around the cathode and held in place by mica plates. The filament became an element inside an oxide-coated nickel tube. The filament, now more correctly the heater, was electrically isolated from the nickel cathode.

In order to maintain a high vacuum within the valve envelope, a device called a getter is used. After the valve has been manufactured, the getter is fired using an induction heater. This releases a metallic vapour which deposits itself on the inside of the envelope. The metal may be (for example) sodium, which combines rapidly with any oxygen molecules remaining inside the valve envelope. It gives the inside of the valve a silvery appearance. If the valve envelope has a leak, this turns white.

Not all valves have a glass envelope. Many American valves used a metal envelope – I assume there was no glass envelope inside, since some of my metal valves have quite deep bumps in the surface and work perfectly. Other valves, such as the EF50 and CV327 types (used in the B40 receiver) have a metal exterior covering a normal glass construction.

Valve Reliability

Valves may look like light bulbs, but they don't often fail because the heater goes open circuit, the same way a light bulb filament fails. The corollary is that just because a valve is lit up, it doesn't mean its working. Valves fail because of various emission problems, caused by the cathode failing to emit electrons, or by too high a pressure within the tube envelope making it impossible for the valve to operate.1940s Mullard Advert

Compared to transistors, valves are very robust and almost immune to catastrophic failure of the sort we see in transistor and and integrated circuits. (however, you can drop transistors on a stone floor – not recommended with most valves.)

The first sign that a valve may not be working is that it will feel comparatively cool to touch. This is because the anode current contributes to the heat. This isn't a definitive test – the American's seem to run their valves at lower anode current than their British colleagues. Usually, there will be a cathode resistor in the circuit – if you find that there is no current flowing through the resistor, it could be you have a defective valve.

Also, unlike transistors, valves can be half-bad. If a valve is down on emission, you may find the performance of the equipment you are working on is disappointing. This might be a good argument for purchasing a valve-tester, but they are expensive. Tektronix in their manuals suggest that valve substitution with known good valves is the practical way to check valve performance.

When all is said and done, solid-state devices have a far greater lifetime expectancy than that of valves. A valve might be expected to last, say,10 000 hours. mean time between failures (MTBF). This computer uses the equivalent of about 50 million or so transistors. If it was made using 50 million (very small) valves, it would run for about .7 of a second. Not even enough time to boot.

Substituting Valves

Philips Miniwatt AdvertMany valves have quite similar characteristics and it is very tempting to substitute when you have a near-equivalent to hand. Usually, its just a botch and unless there is a really compelling reason, I always stick to using the original specified valve.

The only excuse might be because there are simply no more of the original valves left, or the cost is prohibitive. For example, the EM34 magic-eye is almost unobtainable now – expect to pay R1500-00 or so. It might make sense to replace it with the similar looking 6E5. (The 6E5 will last much longer – BUT it doesn't have dual sensitivity). Another hard to get item is the PX4 audio valve unless you are OK with about R3300-00 a pair. (I suggest you fly to the UK to collect them personally.) By the way there's a web-site dedicated entirely to this 1920's valve.

Theory of operation

If you have got this far, you are probably thinking or asking yourself “what does it do?”. As mentioned way back at the beginning, a small variation in grid voltage will produce a variation in anode current. What we would like to know is how much the valve will amplify voltage. First of all we need to know the internal resistance of the valve, also called the anode a.c. Resistance or “plate resistance” in American data books. If we fix the grid voltage of the valve to a given value, we can increase the anode voltage and note the change in anode current. Then:plate_resistance

The internal resistance is then change in anode voltage / change in anode current – from Ohm's Law.

Next we hold the anode voltage steady and vary the grid voltage. As we do so, the anode current changes. This allows us to calculate a value:mutual_conductance

change in anode current (mA) / change in grid volts

This is called the mutual conductance of the valve, and is expressed in mA/Volt. If we mulltiply the two figures together, we get a dimensionless value called the amplification factor. amplification_factor1.. the change in anode volts for a change in grid voltage.

Note that when the grid voltage becomes positive, everything becomes different. The grid current (normally a few micro amps) suddenly becomes much larger and the valve conducts heavily. This form of operation is desirable in radio transmitters and computer circuits, but is not always what you want in a radio receiver.

Usually, all these values are plotted out as graphs and presented as characteristic curves. 6am6_curves


In most cases, the valve needs to be biased so that there is an equal output voltage swing for an equal input voltage swing about the bias point. This implies that you need to place a resistor in the anode circuit, and that the valve cathode must be positive with respect to the grid. This was done using a grid-bias battery (see batteries), but later, a cathode-bias resistor was used. 

In radio receivers, it is essential to have a means to reduce the overall gain of the set in response to strong signals. This is accomplished through the use of valves that have a varying amplification factor (µ) with grid bias. Usually, these are pentodes, and are referred to as remote cut-off pentodes or variable-mu. The 6AM6 (EF91) above is a conventional pentode, but there is a variable-mu version EF92 (rarely seen). More about this in the article on superheterodyne receivers. See also the Eddystone 940 receiver, which uses a variable-mu double triode.

Valve Data

Valve data books occasionally turn up in old bookshops. If you see one, buy it. These are absolutely vital to finding out what the pin connections are, if nothing else. There is nothing more frustrating than squinting through the valve envelope in order to work out what connects to what. Just be aware that the data is not always correct and you may have to cross check between different data books. For the most part there shouldn't be a problem.

There are a number of “tube databases” on the Internet. (Googling “Valve Database” doesn't work.).

Valves today

Valves are still made today for audio amplifiers. Many people believe that valves provide a far more pure sound than transistor or MOSFET amplifiers. Exactly why a valve should sound better than anything else is a bit of a conundrum to me. Admittedly, early transistor amplifiers suffered from crossover distortion, which was horrible, but those days are long gone (but don't expect great audio from early transistor radios). I hesitate to pour scorn on the notion, since at a recent audio exhibition I was asked to compare oxygen-free copper cables against ordinary speaker cables and was taken down a notch when I realised the oxygen-free cables sounded crisper and clearer. The amplifier was a Quad using KT88 valves and my hearing is not good, so the difference must be quite appreciable to someone with good hearing.

The audio and guitar amplifier fraternity like to use valves with stable characteristics, free from hum or microphony and they are willing to pay for what they want. Try to get inexpensive valves described as new-old-stock or N.O.S. rather than recently manufactured types with a specification far greater than required-unless you are repairing a modern guitar amp, of course.

Valve Bases

Provided you substitute the correct valves, there should be no problems. I believe the American B7G base has slightly different dimensions from the British equivalent, but I have never had any problems putting an American 6BA6 into a British EF93 socket (They are the same valve). Do beware of the Mazda Octal base – it looks nearly the same s the American “International Octal” base, but you won't get a Mazda VR65 to fit. Also beware the Philips side-contact base, the valves wobble and don't make good contact. Finally, beware the Sylvania 6C9 – it looks like a B9A base valve, but closer inspection will reveal a tenth pin in the middle of the circle. One of these pitched up half-dead in a Sansui FM tuner/amplifier. To make matters worse there is a Mazda 6C9 – a rare valve from an earlier era.



Last Updated on Wednesday, 07 August 2013 13:48
Joomla template by a4joomla