These pages are concerned with the way audio frequency signals are handled by valve radio technology: in other words, signals after demodulation (detection). The processing requires small signals (in the sub-milliwatt order)to be increased sufficiently to drive the loudspeaker or loudspeaker system at adequate volume (anything from 100milliwatts to 5 watts or higher). It is also essential that control of this amplification is designed in - volume control. Also and optionally, tone control.

Output stages in conventional valve radio fall into one of several categories: 

Class 'A' is where current equal to the complete AF signal flows through a given output valve.

Class AB is where a given valve may be cut-off for part of the input signal. (This is where another valve takes over as in push-pull)

Class B is where each amplifying valve carries current for only half of the signal input.

Each category has its advantages and disadvantages, either in terms of initial cost, power consumption, complexity of circuit or requirement for specialised components or valves. The pages accessed from the top link explore typical vintage radio output stage arrangements. The methods used may not be quite those of the modern valve Hi-Fi amplifier although obviously they have much in common and certainly the best AF sections of quality radio chassis can give a fine account of themselves in that respect.

We start on this page by looking at basic inter-valve coupling methods.

Above, left:

R1 is the anode load for a valve of the double-diode triode type. One of the diodes would be the detector, the other diode commonly being used for AVC (AGC). Assume that an amplified AF signal appears at the anode of the triode. The signal is developed across the resistor R1 and coupled to the control grid of the pentode output valve via C1. This capacitor is usually in the order of 0.01 microfarad, often a paper dielectric and the very often leaky, causing a positive DC voltage to appear on the pentode control grid. This is not a desirable situation as it is likely to result in the early demise of the valve and damage to peripheral components, so the capacitor is a candidate for replacement during restoration of any vintage radio.

R2 is the grid earth return, needed to keep the grid from assuming a DC voltage potential. R3 is the cathode load resistor. Current through this resistor caused by the conduction of the valve creates a positive potential on the cathode, making the grid negative with respect to the cathode: this is called 'biasing'. The capacitor C2 acts as a reservoir for the cathode load. Without it, the potential at the cathode will vary as the signal varies. This is one way of applying negative feedback, however, so some manufacturers omit the capacitor or fit a very small value. C3 is a fixed-value tone 'corrector', bypassing a proportion of the higher audio frequencies to chassis (earth). These are fitted to improve the 'toppy' response of the pentode - at least, that is the claim made for their existence. Usually they are to be found wired across the primary of the output transformer, here shown only as a block labelled 'output'.

Above, right:

The double diode triode is employed again but the output valve is now a beam tetrode - still with the fixed tone corrector C8. This circuit uses more components and is a refinement on the simple one discussed above. Note that the anode load now consists of a split load - R4,5. R4 and C4 are a decoupling circuit designed to prevent AF current from entering the HT supply where they might reach the circuits of other valves in the set, causing instability. C4 shunts AF to chassis. C5 is an additional bypass capacitor to remove any IF current that may have reached the anode of the pentode. Its value is chosen so as to remove the high frequency IF currents without impacting the low frequency AF currents and is in the order of 100-500 pF. Also additional is the resistor R7. This is known as a 'grid stopper' and is there to help maintain stability with the highly efficient output tetrode. Such resistors have a very small value and therefore a negligible effect on AF signals.

All rights reserved. © VRW 2006/2010