STABILIZATION OF PHOTOMULTIPLIER GAIN

When working with photomultipliers, one often encounters the question of stabilizing their gain. Since in multipliers the gain coefficient depends strongly on the voltage of the power supply, even small changes in the supply voltage lead to significant fluctuations in the amplitudes of the pulses at the output of the multiplier.

The supply voltage of a photomultiplier is usually \(1000\ \text{V}\) or more, and the creation of a power supply with electronic stabilization for such a voltage, possessing high and long-term stability, is associated with certain difficulties[^1]. The developed circuits with electronic stabilization[^1] require for their manufacture precision resistors and high-quality stabilovolts.

Fig. 1.

Fig. 2.

As power sources for photomultipliers, anode batteries are also used. This method gives good results, but requires a considerable number of batteries.

In a number of cases it is desirable that the high-voltage supply circuit not contain a large number of batteries or a complex circuit,

electronic stabilization, but at the same time would ensure satisfactory stability of the multiplier gain.

In paper 2 a method is proposed for stabilizing the gain of photomultipliers with electrostatic focusing. The principal circuit of a divider with a stabilizing chain is shown in Fig. 1. The stabilizing properties of the indicated chain were studied in paper 3. The method of stabilizing the gain of a photomultiplier consists in fixing the potential of one of the dynodes relative to the neighboring, preceding one, by means of batteries. The variable resistance \(R\) connects two dynodes located before and after dynode 8.

With a fixed voltage of the photomultiplier power supply, the amplitude of the pulses at the output changes with the change in the value of \(R\). Fig. 2 gives a curve of the change in pulse amplitude as a function of \(R\) for the case when stabilization was applied to the external (8th) electrode of the photomultiplier RCA-5819.

Fig. 3.

Fig. 4.

Since multipliers of the RCA-5819 and RCA-C7151 types were used in the work, in which the dynodes are arranged not linearly but along a circumference, the authors considered the operation of the stabilization circuit when fixing the potential of both the external and the internal dynodes.

For stabilizing the gain of the multiplier, the operating point \(A\) is chosen on the fall of the curve given above. In this case fluctuations of the voltage of the photomultiplier power supply will be compensated by the stabilizing cell. A typical curve illustrating the presence of a plateau in the photomultiplier is shown in Fig. 3. The plateau has considerable length (\(\sim 200\)—\(300\) V) and a slight slope. The width of the plateau and the initial potential depend on \(R_A\). When the operating point is placed at the beginning of the plateau, the gain of the photomultiplier falls only by 20% in comparison with the gain given by the multiplier at the same supply voltage but without the stabilization circuit.

The battery voltage necessary for operation of the stabilization circuit was from 90 to 135 V, depending on the individual properties

photomultipliers. As the measurements showed, the statistical distribution of pulses from the scintillating crystal remains unchanged.

The operation of the stabilization according to the circuit of Fig. 1 and with the potential of the internal dynode fixed was considered. In this case the curve of variation of the multiplier gain as a function of the value \(\bar R\) had two maxima. The operating point could be chosen on either the first or the second descending branch of this curve. When the operating point was set on the first descending branch, stabilization could be obtained at lower voltages on the multiplier (\(\sim 600\) V), but the plateau had a smaller extent (\(\sim 30\) V) and was less flat.

In those cases where it is undesirable to use a battery, and where stabilovolts with good parameters (type 5691) are available, the circuit of Fig. 4, described in paper 4, may be used. In principle the circuit differs in no way from the circuit shown by us in Fig. 1. The voltage between dynodes 5 and 6, for example, is fixed by the 5691 stabilovolt. In this circuit the fixed potential of the internal dynode and the plateau amounted to 80 V.

The described method of stabilizing the gain of photomultipliers with electrostatic focusing is attractive because of its simplicity and can be easily implemented without special difficulties. No current is consumed from the stabilizing battery, and it can operate for a considerable time. Some loss of photomultiplier gain (10–20%) can in some cases be compensated by increasing the voltage of the power supply.

Yu. Sh.

References

1. W. A. Higinbotham, Rev. Sci. Instr. 22, 429 (1951).
2. G. A. Morton, RCA Rev. 10, 525 (1949).
3. R. Sherr and J. B. Gerhart, Rev. Sci. Instr. 23, 770 (1952).
4. G. L. Guernsey, G. R. Mott, B. K. Nelson and A. Roberts, Rev. Sci. Instr. 23, 476 (1952).