High-voltage stabilizer

I. A. GABRUSENKO
Submitted 1962 | SovietRxiv: ru-196201.13722 | Translated from Russian

Abstract Generated abstract

This paper describes a high-voltage stabilizer designed for electrostatic deflection in studies of a 250 kV electron beam. The stabilizer uses a parallel regulating circuit with a GMI-83 pulse tetrode and a differential three-stage DC amplifier, allowing the deflecting voltage to be set between 9 and 16 kV while addressing pulsed plate currents and transient recovery through output shunting and circuit analysis. The design considerations include limits on tube dissipation, recovery time after beam-induced pulses, and reduction of amplifier grid-current errors. Measurements indicate that the deflecting voltage instability over eight hours does not exceed 5 x 10^-5.

Full Text

HIGH-VOLTAGE STABILIZER

I. A. GABRUSENKO

The electrostatic method of deflecting charged particles is widely used in accelerator technology and in various physical installations. The high-voltage stabilizer described here, intended for studying an electron beam with an energy of 250 kV, makes it possible to set the deflecting voltage \(E_i\) within the range \(9 \div 16\) kV and maintains it with an accuracy of \(5 \cdot 10^{-5}\).

The average current consumed by the deflecting plates does not exceed 100 μA. With a small load current it is expedient to use a circuit with a parallel connection of the regulating tube, which substantially simplifies the design of the stabilizer. As the regulating tube, a GMI-83 pulse tetrode is used, whose characteristics in the region of small currents are shown in Fig. 1. As analysis of the characteristics shows, the GMI-83 has quite satisfactory parameters in this operating mode: the slope of the characteristic is \(S = 0.4\) mA/V, and the permeability is \(D = 0.005\).

Fig. 1.

It should be noted that the tube is used in a mode for which it was not intended. The presence of a large space charge at a negative voltage of the first grid close to cutoff leads to the fact that a slight error in the geometry of the tube causes a nonuniform distribution of current over the anode area and, as a result, local overheating of the anode.

This circumstance makes it necessary to reduce the power dissipated by the anode by a factor of 2–3 relative to the permissible value.

The schematic circuit of the stabilizer is shown in Fig. 2. The voltage from the divider and the reference voltage are applied to the input of a differential three-stage DC amplifier with a gain coefficient \(K = 10000\). The latter is regulated within the range \(35 \div 60\) V, which makes it possible to set the specified value of the deflecting voltage. The “zero” drift of the amplifier does not exceed 1 mV; the stability of the supply voltages (\(+250\) V, \(-250\) V) is no worse than 0.5%.

Part of the scattered electron beam reaches the deflecting plates. The current pulse of the deflecting plates reaches 100 mA with a duration of 5 μsec. The decrease in the deflecting voltage under the action of the current pulse must not exceed \(2 \cdot 10^{-5} E_i\), for which purpose the stabilizer output is shunted by a capacitance \(C_{\mathrm{bl}} = 2\ \mu\mathrm{F}\).

Fig. 2.

For calculating the transient processes occurring after the termination of the current of the deflecting plates, the following obvious equality may be written:

\[ i_c + i_a = \frac{E_0 - e_i}{R_a}, \tag{1} \]

where \(i_a\) is the anode current of the regulating tube,
\(i_c\) is the transient current of the capacitance \(C_{\mathrm{bl}}\),
\(e_i\) is the instantaneous value of the deflecting voltage.

Substituting into (1) the expressions for the currents

\[ i_c = C_{\mathrm{bl}} \cdot \dot e_i, \qquad i_a = e_i \cdot n \cdot K \cdot S, \]

we obtain the equation of the transient processes

\[ \dot e_i + e_i \frac{S \cdot n \cdot K}{C_{\mathrm{bl}}} - \frac{E_0 - e_i}{C_{\mathrm{bl}} \cdot R_a} = 0; \tag{2} \]

here \(n\) is the gain coefficient of the high-voltage divider.

Fig. 3.

Using a linear approximation of the characteristics of the regulating tube, the transient processes can be represented as two successive processes. During the first, the tube is closed, and the voltage rises at the rate

\[ \frac{E_0 - E_1}{C_{\mathrm{bl}} \cdot R_a}. \]

The regulating tube opens—

takes place when \(E_i - e_i < \dfrac{E_o - E_i}{n \cdot K \cdot S \cdot R_a}\); voltage recovery occurs with the time constant \(\tau = \dfrac{C_{\mathrm{b.a}}}{S \cdot n \cdot K}\). The duration of the transient processes in this case does not exceed \(3 \cdot 10^{-3}\) sec.

One of the sources of instability of the deflecting voltage is the grid currents of the first stage of the amplifier. If an error introduced by the grid currents equal to \(10^{-5}\) is allowed, then, with an output resistance of the high-voltage divider of \(100\ \mathrm{k}\Omega\), the currents must not exceed \(10^{-8}\) A. Figure 3 gives the characteristics of the grid currents of the 6N9S tube, which make it possible to select the operating mode of the first stage. In this circuit the bias of the first stage of the amplifier was brought to \(-0.85\) V at a filament voltage of 5 V, which made it possible to reduce the grid currents to \(5 \cdot 10^{-9}\) A and, correspondingly, the instability introduced by them to \(0.5 \cdot 10^{-5}\). Measurements showed that the instability of the deflecting voltage over 8 hours of operation does not exceed \(5 \cdot 10^{-5}\).

Submission history

High-voltage stabilizer