TX 3

Thermionic Valves

Thermionic Valves, or just valves, and known in the USA and elsewhere as Vacuum Tubes, or just tubes, were a development from the early experiments of Thomas Edison to manufacture a satisfactory light bulb. He noticed an effect which was followed up by Ambrose Fleming in London who eventually produced what we now call a diode, a two electrode device. This could be used to convert ac to dc but could not amplify signals. In the early 20th century, the American Lee de Forest inserted a third electrode to form a triode, however it did not become a practical amplifying device until modified by H. J. Round who worked for Marconi.

The circuit symbols for triode valves are shown below, although these are not representative of the actual construction.The triode comprises a negative electrode termed the cathode. This must be heated to red heat either directly as in the filament of a light bulb or indirectly with a separate heater enclosed in a tube which is itself the cathode. The Anode or positive electrode is usually a further tube which surrounds this cathode at some distance from it. In between is the Control Grid made of fine wire or mesh.

The triode has limitations due to the interaction between the control grid and the anode. This was later overcome by adding a second grid known as the screen grid between the control grid and anode to form a four electrode device or Tetrode. A fifth, sixth and seventh electrode were also added but generally their use was limited to specific applications and were not found in high power transmitters.

In use the anode is connected to a voltage positive with respect to the cathode. The control grid is usually (but not always) negative. When the cathode is heated it emits electrons which are attracted by a positive voltage and so flow towards the anode. This is in fact a flow of current (although the early experimenters got it wrong and established a convention that current flows from positive to negative so the electrons flow in the opposite direction to that which we believe the current to flow). However, the presence of the negative grid limits the flow. Variation of the grid voltage can vary the flow (hence the term “valve”).

CLASS A CLASS B CLASS C

The efficiency of a valve used as an amplifier depends upon the biasing arrangements: I.e. The precise voltages applied to the control grid. Unfortunately the linearity of amplification achieved is in inverse sense to efficiency. A Hi-Fi amplifier would need to be in Class A. In this mode the voltages applied are such that the valve conducts throughout the swing of the signal to be amplified. This is fairly obvious as, if the valve ceased to conduct it could not follow the variations of the input signal and distortion would result. This mode can provide a theoretical efficiency of 50% maximum when operating at full output, which is rarely the case, and in fact practical efficiencies would be less than this. The efficiency falls pro-rata at lower outputs as the valve consumes the same power at all times, even when there is no signal.

In Class B the valve is biased in such a way that with no signal applied no current flows. Only the positive going half cycles of signal are amplified. This is clearly of no use for distortion free amplification. However, in a radio frequency amplifier, the presence of a tuned circuit can provide a flywheel effect which re-inserts the missing half cycle. Linear amplification of RF results. What is more; if two such valves are used, each arranged to amplify a different half cycle and the results combined in a transformer, linear amplification of audio signals can be achieved. This arrangement, known as push-pull is employed in high power modulators. The efficiency can reach a theoretical 78% but a more practical limit would be 65 to 70 % and again varies with signal level. In practice, true class B is rarely employed and some current is allowed to flow even in no signal conditions. This is designated class AB.

For even higher efficiency, class C has the valve biased so that current only flows for less than half a cycle. The smaller the portion of the cycle during which current flows, the higher the efficiency. Figures of over 90% can be achieved. It is highly non-linear and can only be used at RF and even then requires the distortion products to be filtered out.

The following diagrams give some idea of the operation: