Emission Current Stabilizer

V. M. Razin

(Presented by the scientific seminar of the Faculty of Physics and Technology)

In many electrovacuum devices a high degree of constancy is required for the electron emission current from a heated cathode. The cause of a change in the emission current may be both a change in the emission properties of the cathode itself under conditions of nonconstant vacuum and the action of external factors: voltage fluctuations, changes in the distance between electrodes under the action of mechanical influences, etc.

Fig. 1

In some devices, such as a VI-3 vacuum gauge, a PTI-4A leak detector, and others, for the purpose of maintaining the constancy of electron emission a stabilization circuit is used, shown in Fig. 1. In this circuit the cathode emission current, under the action of the extracting voltage, passes through the resistance \(R_c\), producing across it a certain voltage drop. This potential difference controls the anode current of lamp \(L_1\) in the circuit of the secondary winding of the regulating transformer \(T_p\) in such a way that, when the emission current increases, the anode current decreases; accordingly, the filament current in the circuit of the filament transformer \(T_n\) also decreases, thereby compensating for this increase in the emission current.

When the emission current decreases, the anode current of lamp \(L_1\) increases; accordingly, the filament current also increases, compensating for the decrease in the emission current. To improve stabilization, in some circuits a direct-voltage amplifier is inserted between the resistance on which the voltage drop is produced by the emission current and the grid of the regulating tube. The features of the method of regulation under consideration can be clar-

... using the equivalent circuit of Fig. 2. Let us note some of the features of this method.

1. In order to be able to regulate the power in the resistance \(R_n\) within sufficiently wide limits, it is necessary for the regulating resistance \(R_p\) to absorb power of the same order as in the load, or even greater. A consequence of this circumstance is the low efficiency of the method under consideration. We further note that if it is required to regulate considerable power, then the regulating electron tube must also be designed for large anode dissipation.

Fig. 2

In the instruments mentioned above, a beam tetrode 6P3 is used as the regulating tube; its permissible anode dissipation is 20 W, which makes it possible to regulate, within the required limits, a load power of the order of 5 W.

Fig. 3

When the load power is increased, it becomes necessary, for regulation purposes, to use a larger number of tubes, as a result of which the circuit becomes bulky.

1. An advantage of the method under consideration is the low inertia of the regulating element.

For the purpose of regulating the pulsed emission current of the electron gun of a betatron, we developed and used the circuit of Fig. 3, in which a choke-type magnetic amplifier \(D_r\) is used as the regulating element. The use of a magnetic amplifier made it possible to carry out regulation over very wide limits with a load power of about 10–15 W, without any significant complication of the circuit.

The circuit (Fig. 3) operates as follows. The pulsed current from the conducting layer of the vacuum chamber of the betatron is fed to terminal 3.

In the circuit of the parallel-connected resistance \(R_5\) and capacitance \(C_3\), the pulse current is averaged (the mean current value is of the order of 50–100 \(\mu\text{A}\)), and the voltage drop across the resistance \(R_5\) due to this current is compared with the value set by means of the potentiometer \(R_4\).

Fig. 4

The difference between these voltages controls the anode current of the 6N7 tube. When the emission current increases for one reason or another, the grid potential of the 6N7 tube becomes more negative, and the anode current of the tube decreases. A decrease in the current in the control winding of the magnetic amplifier reduces the saturation of the core, as a result of which the inductance of the alternating-current windings increases, the voltage drop across them increases, and, correspondingly, the voltage on the filament transformer connected to terminals 2 of the circuit in Fig. 3 decreases. As a result, an increase in the emission current is accompanied by such a decrease in the filament voltage that the resulting change in the emission current proves to be very small. When the emission current decreases, the circuit operates in the opposite direction of changes in the electrical quantities.

Fig. 5

During operation of the emission-current stabilization circuit developed by us, it was found that under certain conditions this regulating circuit loses stability and becomes self-excited, with undamped oscillations of the form shown in the oscillogram of Fig. 4.

Near the stability boundary, the transient process has a practically little-useful character (see the oscillogram of Fig. 5). The character of the undamped oscillations indicates the fact that the system under investigation is substantially nonlinear. Obviously, the element introducing the greatest non-

…nonlinearity in the given system is the filament (here, the dependence of the emission current on the filament current is substantially nonlinear).

As is known, the theoretical analysis of essentially nonlinear systems is associated with serious mathematical difficulties and can be carried out only in a very approximate form. For this reason we undertook experimental investigations of the stability conditions in our circuit.

Fig. 6

The dynamic properties of the stabilization circuit can, to a first approximation, be considered with the aid of the block diagram of Fig. 6, consisting of three consecutively connected relaxation elements, i.e., elementary first-order elements. The parameters of the first element are determined by the properties of the grid circuit of the electronic amplifier; the second element corresponds to the electronic and magnetic amplifiers, and the third to the cathode.

Fig. 7

Fig. 8

The time constant of the first element is determined by the product \(R_5 C_3 \simeq T_1\) (Fig. 3). The time constants of the second and third elements, \(T_2 = 0.026\) sec and \(T_3 = 0.21\) sec, were determined from oscillograms of transient processes in the filament-heating circuits (Fig. 7) and of the magnetic amplifier (Fig. 8). The calibration on these oscillograms, as on all the others, has a frequency of 50 cps.

The critical gain of the entire circuit is determined from the condition [1]

\[ K_{\mathrm{cr}} = \left( \frac{1}{T_1} + \frac{1}{T_2} + \frac{1}{T_3} \right) \left(T_1 + T_2 + T_3\right) - 1. \]

Using this expression, one can calculate the value of \(K_{\mathrm{cr}}\) for different values of \(T_1\), and thus correctly choose the parameters of the stabilizing circuit.

In order to check the correctness of the determination of the constants \(T_2\) and \(T_3\), the critical gain was measured directly in the circuit at different values of the time constant \(T_1\).

The calculated and experimentally determined values of \(K_{\mathrm{cr}}\) for several values of the time constant \(T_1\) are given in the table.

The considerable discrepancy between the experimental and calculated data indicates that the system is substantially nonlinear and that, as a result, exact results cannot be obtained in the linear approximation. The coincidence of the general character of the dependences nevertheless makes it possible to choose, by calculation, the value of the time constant \(T_1\) correctly, if the values of \(T_2\) and \(T_3\) are known, without resorting to experimental determination of the dependence \(K_{\mathrm{cr}} = f(T_1)\).

REFERENCES

1. M. V. Meerov, Fundamentals of Automatic Control of Electrical Machines, GEI, 1952.