"MESSAGE In A Bottle", by The Police, might well have been inspired after gazing into the insides of a valve. Just what that message could be, is the subject of this article.

The electron tube valve is named after its mechanical counterpart found on the wheel rims of motor cars. As the tyre valve lets air pass one way, so the diode (a simple type of electrical valve), allows the passage of current in one direction only.

The diode has two internal metal plates called electrodes and from these it gets its name. Both are made of nickel, but one is coated with a layer of strontium and barium oxides. The coated electrode is known as the cathode and is fitted with a heater coil, and this is the part you can see glowing inside an amplifier. The purpose of the heater is to cause the oxide-coated cathode to give off electrons — negatively charged particles normally bound within atoms at room temperatures.

In manufacture, a vacuum is formed by removing all the air from the glass tube in which the electrodes are assembled. This is necessary for "thermionic emission" to occur and allows the electrons to travel as far as their energy will take them without collision with air molecules.

A tiny fraction of the thermally emitted electrons have enough energy to reach the adjacent collecting electrode, but most fall back to the cathode to be reunited with their grieving families.

If the collecting electrode (or anode as it is more correctly called), is connected to a high voltage source (200-400V) with respect to the cathode, then electrostatic attraction being similar to magnetic, the electrons liberated from the cathode will be lured to the anode, or more technically, a current will flow. Therefore the diode will pass current only when the anode is more positive than the cathode.

Well, this may have explained why valve amplifiers take a time to warm up, but how do valves amplify?

The next stage in the evolution of the valve was the development of the triode, which was equipped with three electrodes (no points for guessing that one). The addition to the basic diode was a fine wire mesh called the control grid which was positioned near to the cathode.

In construction the form of the triode was two concentric cylinders, the anode and grid, surrounding a central cathode. It was found that applying a negative voltage to the grid caused the anode current to fall and a certain value would cut off the flow of electrons completely.

If a resistor was placed between the anode and the high voltage supply, then any change in anode current would be reflected in a change in voltage across this resistor (Ohm's law). In this way amplification had been discovered because a small change in grid voltage resulted in a much larger change across the anode load resistor.

To amplify a low level signal, the procedure was to feed the signal into the grid along with a fixed negative voltage called a bias, so that even with the signal at its most positive the grid would remain negative to the cathode. All that then remained was to calculate the anode resistor... (next four pages omitted on editorial and medical advice).

The triode, apart from shaping present day telecommunications, computers, sound recording and the world as we know it, was a pretty crummy device, especially its high frequency performance. This was due to inter-electrode capacitances, and to combat this a second grid was introduced.

Although the resulting tetrode had much lower parasitic capacitance, the phenomenon of secondary emission gave rise to a very distorted grid voltage—anode current characteristic, making the valve practically useless.

The cure was to introduce yet another grid, which gave rise to the pentode — a type widely used in amplifier output stages. The tetrode has now been re-designed since the early days by making the two grids from fine wire wound around formers in perfect alignment. This arrangement overcomes the previous problems, and gives the beam-tetrode used in the 6L6 and KT66 output valves.

For a perfect distortion-free amplifying valve, the relationship between the grid voltage and the anode current would be linear, that is if expressed graphically, a straight line.

In practice, this is not so, and a pure sine waveform after amplification is found to have its positive peaks sharpened and its negative troughs shallowed. This is the same effect as adding a signal at exactly twice the frequency and so the effect is called second harmonic distortion.

Distortion at all the higher harmonics is also present at lower levels..

For small signal amplification at the front end of an amplifier the valve is working over a small part of its grid voltage—anode current curve and the distortion produced is quite negligible. At the output stages however, where large voltage swings are required to drive the loudspeaker, a method is required to eliminate the considerable distortion this would introduce.

The technique used is the push-pull output stage which uses two valves working in opposition. One valve handles positive transitions of the signal, while the other deals with the negative. In this way only one valve is "working" at a time and the dynamic grid voltage—anode current characteristic for the two valves in tandem becomes a straight line. Many pages of sums can show that this arrangement cancels all the even harmonic distortion, but leaves the odd harmonics untouched.

Another form of distortion arises in a valve amplifying stage if the grid becomes more positive than the cathode. This may occur if a very large signal is applied to the grid as with overdrive. This causes current expected at the anode to flow in the grid and so the anode current will fall creating distortion on each positive half cycle of the signal. In addition the input impedance falls to a low value which loads the previous stage creating further distortion.

If manufacturers knew exactly what qualities made a valve amplifier so attractive to many guitarists they would have synthesised them with solid state techniques long ago. Certainly overloading occurs more gradually, generating predominantly second harmonic distortion at first as the valves become non-linear and then going into hard clipping with the production of largely odd harmonics as the available voltage swing becomes restricted.

Vibrational movement between the electrodes of a valve gives rise to microphonic effects which although not significant below 1kHz may provide mechanical feedback and add yet another "undefinable" to the sound.

The voltages found within a valve amplifier are typically three to four times higher than in a transistor amp of similar rating and this puts the components under greater strain.

When replacing output pentodes or beam tetrodes, it's wise to fit them in pairs. The push-pull circuit of which they will be part requires the two valves to be reasonably matched in order to enjoy the cancellation of nasties like mains hum in addition to the distortion reduction already discussed.

A typical valve line-up for a British 30-watt amplifier would be an EF86 low noise pentode as the input valve, an ECC83 double triode as the output stage driver, two EL34 power output pentodes and a GZ34 rectifier to provide the high voltage DC.

The Roland Bolt range proves there is a demand for valve amplifiers, and with modern components such hybrids give the best of both worlds.