▼ Menu

How the Valve Works

Wireless World December 15, 1938.
    
Extras ▼

 

The Tetrode

In this third article of the series, the screen-grid valve or tetrode is dealt with in detail. It will be followed by a discussion of the pentode and the latest kinkless tetrodes.
Left: the glass envelope has been removed from this AC screen-grid valve to show the arrangement of the electrode assembly more clearly. Right: an anode displaced and one side of the screening grid torn open to show the grid and cathode elements.

Following the triode we come to the tetrode or four-electrode valve. Briefly, a tetrode can be described as a valve with two concentric grids between anode and cathode. There are, however, several different kinds of tetrode. The earliest types were those known as bi-grids, and they were used in two ways. In one the outer grid was the control grid and the inner grid was kept at a fixed positive potential by a battery.

The result of this was a relatively high anode current for a low anode voltage and some types were produced that would function with an anode supply of no more than 6 Volts. The performance, however, was generally inferior to that of the triode and they were not greatly used.

The other use of such valves was as a superheterodyne frequency-changer, the signal being applied to one grid and the other being used in conjunction with the anode for the oscillator. Even for this purpose the valves never had a very wide vogue. See 41MDG.

Tetrodes, as they are understood to-day began with the screen-grid valve, and for some years were very widely used. They were then largely superseded by the pentode, for it was found that by the introduction of a third grid the characteristics could be considerably improved. Since then it has been possible to obtain the desired improvement without the third grid and the result is that the tetrode is now staging a come-back.

The Screen-Grid Tetrode

In view of this it will probably be confusing to adhere strictly to the principle of treating all tetrodes together and then going on to discuss pentodes. It will be simpler to deal with tetrodes and pentodes together, in the order of their actual development.

We start with the screen-grid valve. This was developed primarily to reduce the grid-anode capacity to a negligibly small figure and so to avoid the instability problems inherent with triodes in RF amplifiers. The S625 was the first screened grid valve. It was found, however, to possess, in addition, other desirable characteristics. See M-OV's screened grid valves.

Essentially the valve consists of the cathode, grid and anode of a triode arranged in much the same way as in a triode, but with another close-mesh grid between the grid proper and the anode. It is clear that with such an arrangement if this extra grid is maintained at an unvarying potential with respect to cathode it will act as an electrostatic screen between grid and anode. The screening will not be perfect as the grid is a mesh structure, but it is actually fairly complete and reduces the grid-anode capacity to about 0.001-0.005 pF. only. Naturally, this is obtained at the expense of increased input and output capacities, for the grid and anode each have a capacity to the screen. Notice also that the screening is only effective if the screen grid is maintained at a fixed potential relative to the cathode.

In practice, the anode has a potential of about 200-250 Volts applied to it and the screen is kept, positive with respect to cathode by about 60-100 Volts. The control grid is kept negative, just as with a triode. In use the valve is connected up very much like a triode., the main difference being the necessity for providing the screen potential.

The internal action is rather different, however, and the characteristics obtained are quite different. Owing to the presence of the screen grid the anode voltage now has very little effect on the anode current, for it is very largely screened from the cathode by the screen grid.

The easiest way to understand what goes on is to picture the cathode, control grid and screen grid as the cathode, grid and anode of a triode, and ignore for the moment the anode proper. Electrons emitted from the cathode form the usual cloud around it which we call the cathode space charge, and many of them are attracted to the screen grid by its positive potential. In flying to it they pass through the meshes of the control grid, since this is negative and repels them. The number reaching the screen grid depends upon its potential and upon the control-grid potential, since the force acting on the space charge is a function of the potential of both electrodes.

The Screen-Anode Space

Now comes the difference from the triode. The electrons fly to the screen at high velocity, and since it is not solid, but a mesh structure, many fly through its interstices. Those that do so find themselves between two positive electrodes, the screen and the anode. The anode is the more positive of the two, but whether it exercises the greater attractive force on any individual electron depends on the position of that electron.

Electrons which only just succeed in passing the screen and so do not move far from it are attracted more by the screen than by the anode and so fall back into it. Higher velocity electrons passing through the meshes of the screen to a greater distance from it are attracted by the anode and so carry on their course to it. The anode can collect only those electrons which have sufficient velocity to carry them beyond the main influence of the screen and, within limits, it collects all these irrespective of its actual potential. Consequently, the anode current is not affected to any large extent by anode voltage.

An increase in screen potential, how-ever, increases anode current for it attracts more electrons to itself as a whole and more consequently pass its meshes. Similarly, making the control grid less negative increases both screen and anode currents.

From the above it might be thought that the screen current would be much greater than the anode current. This is not so, however, for many more electrons pass the meshes than land on its wires, and in practice the screen current at the normal operating point is rarely greater than one-quarter of the anode current.

Now the grid-volts-anode-current curves of a tetrode are of much the same shape as those of a triode, but the anode-volts-anode-current curves are quite different. As the anode current is largely independent of anode voltage, the AC resistance is very high and the curves take the form of nearly horizontal lines.

The mutual conductance is of the same order as that of a triode, 1-4 mA/V, so that the amplification factor is also very high. A typical valve might have a mutual conductance of 2.0 mA/V with an AC resistance of 250,000 Ω, giving an amplification factor of 500.

Fig. 10. - Curve A illustrates a typical triode characteristic and curve B that of a tetrode of the screen-grid type. Note the negative resistance portion bc.

Now we know that with a triode the anode current falls gradually to zero as the anode voltage is reduced, in the way illustrated by curve A of Fig. 10, and we should naturally expect the same thing to happen with a tetrode. Usually, however, it does not.

As the anode voltage is lowered the anode current falls only very slowly at first until the anode potential becomes of the same order as the screen-grid voltage. For a further decrease in anode voltage the anode current falls off rapidly. At another critical voltage, however, this falling-off ceases and a further drop in voltage makes the anode current rise! At a still lower voltage the rise ceases, and thereafter the current falls off steadily to zero at zero anode volts.

This state of affairs is depicted by curve B of Fig. 10. The first part considered is the flat top dc over which normal operation takes place. At d the anode and screen potentials are of the same order, and the current falls rapidly over ad as the anode voltage is further lowered. The rising current at still lower voltages is represented by bc, and the final fall in zero by ab.

The kink bcd in the curve represents a defect from most points of view, although on occasion it can be put to good use. The ideal tetrode curve would not have this kink, and can be obtained with pentodes and some of the latest tetrodes.

It is, however, instructive to see the reason for this kink. It is caused by secondary emission. Electron emission can be obtained not only by heating a suitable substance as in the case of the cathode, but also by bombarding a substance with sufficiently high-velocity projectiles. Electrons are such projectiles, and if they strike the anode with sufficient velocity each electron may knock out several others.

The number of electrons knocked out, or secondary electrons, depends on the material of the anode and upon its condition, apart from the primary electron velocity. Although one primary electron may cause several secondary electrons to be emitted from the anode, these secondaries have a lower velocity than the primary.

In the case of a triode with negative grid, the secondary emission from the anode does little harm, for the electrons all fall back again into the anode. The same thing happens in the tetrode as long as the anode voltage is greater than the screen voltage, for then the anode attracts them to itself.

When the anode is at a lower potential than the screen, however, some of the secondaries may be more attracted by the screen than by the anode and so pass to this electrode, thus reducing the anode current.

Secondary Emission

Suppose we start from the point a with zero anode voltage. The anode current here is zero because it does not attract the electrons which pass the screen. These electrons consequently fall back into the screen and in passing through the screen and falling back to it they form a cloud outside the screen. There is consequently a space charge outside the screen and it is from this that the anode draws electrons to form the anode current when the anode potential is raised over the range ab.

Over this small range the action between anode and screen space-charge is similar to that between the anode and the cathode space charge of a diode. As the point b is reached two things are happening. The anode is exercising sufficient attractive force to collect nearly all the electrons passed by the screen, so that the space charge is rapidly disappearing, and many of the electrons arriving at the anode do so with sufficient velocity to knock out secondary electrons.

At first these secondaries fall back to the anode, but as the anode voltage is raised and the primary electron velocity increased, the secondaries have a greater velocity and come within the attractive force of the screen. Over the range bc the anode current falls with increasing voltage because although an increase in voltage attracts a greater number of primary electrons to the anode, still more secondaries leave the anode and go to the screen.

This does not go on indefinitely, however, because the rising anode voltage at length begins to exercise a considerable attractive force on the secondaries, and so tends to prevent their passing to the Screen. At the point c the tide turns and thereafter the anode regains control and secondaries pass to the screen in fewer and fewer numbers. The anode current consequently rises rapidly over cd and for higher voltages the anode not only collects nearly all the electrons passing the screen, but is powerful enough to prevent most of the secondaries from reaching the screen.

Fig. 11. - The tetrode curve may show negative values of anode current, as over the range fcg.

In some cases, the whole curve is lower than is shown in Fig. 10, and the anode current may be negative over a portion of it. Such a curve is given in Fig. 11, and over the range fcg the anode current is negative; that is, instead of flowing from the anode to the HT supply, it flows from HT to anode.

The action is the same as before, but at the points f and g there are as many secondary electrons leaving the anode and passing to the screen as there are primaries arriving at the anode. Consequently, the two balance and the external anode current is zero. Between f and g, more secondaries pass from anode to screen than primaries arrive at the anode and the current is negative. This negative current is not common, but is sometimes found. More often the curve is of the type shown in Fig. 10.

The chief disadvantage of this curiously shaped curve is that the output of the valve is restricted, for the instantaneous anode voltage cannot be allowed to swing below the point d. This is unlikely to happen in any case when the valve is used as an RF amplifier handling only weak signals, and it is for this purpose that the valve was first introduced and proved very successful.

With large signals, however, difficulties arose, and as an AF valve the kink in the curve seriously limited the power output and rendered this type of valve unsatisfactory in the output stage. The screened tetrodes were, of course, in any case unsuitable for power output owing to their low anode current, but types specially designed for output stage work were inferior to triodes on account of the kink in the curve. It should be noted that over the range be where a fall in anode voltage causes an increase in anode current, the AC resistance of the valve is negative, and it will consequently make a tuned circuit connected to the anode oscillate. Used in this way the stage is a dynatron oscillator and has many applications.

While the dynatron oscillator is very useful, it is not very widely used because individual valves vary considerably in their characteristics over this portion of their curves. Secondary emission is rather difficult to control in manufacture, and in any case valve makers try to avoid it rather than to promote it in these valves. Consequently, when a dynatron oscillator is wanted, it is often necessary to pick the valve specially.

Return to Series Contents or back to The Triode or forward to The Pentode

Use browser back button to return.
art-320c