APPARATUS FOR INVESTIGATING THE TRANSMISSION OF BREMSSTRAHLUNG IN PULSED MODE

V. A. VOROB’EV, L. N. DROZDOV

(Presented by the scientific seminar of the Research Institute of Electronic Introscopy)

In most induction electron accelerators—betatrons—a continuous pulsed mode of operation is used, with a supply frequency for the accelerator electromagnet of 50 Hz or higher [1, 2, 3]. In some cases, especially when betatrons are used in nuclear-physics experiments, and also in recording rapidly occurring processes, it is desirable, in addition to the continuous pulsed operating mode, to have a lower-frequency mode, down to the production of single radiation pulses.

Below we describe circuits developed at the Research Institute of Electronic Introscopy of the TPI, one of which provides the generation of single radiation pulses and also regulates the frequency of radiation pulses in the range 0.05–50 Hz from a betatron with a maximum radiation energy of up to 30 MeV, operating in a continuous pulsed mode at a frequency of 50 Hz; the other provides the recording and measurement of single radiation pulses and of pulses with a repetition rate of up to 50 Hz.

The mode of single shots and of an adjustable radiation-pulse frequency is created by introducing a parametric generator into the betatron synchronization circuit used to form the firing pulses of the injection and bias thyratrons. The usual synchronization circuit for betatron control operates as follows: the voltage pulse of the peak transformer—the zero-field sensor—is fed to an amplifier, triggering the delay multivibrator. The trailing edge of the differentiated rectangular pulse triggers a shaper that generates pulses of duration on the order of 5–6 μsec, with a leading-edge steepness of 0.7–0.8 μsec and an amplitude sufficient to trigger the thyratrons of the injection and bias circuits; it is applied to their grids through a cathode follower.

The parametric generator (of the reactive-trigger type), implemented on a low-power thyratron, operates in the mode of a frequency divider for the pulses arriving at its input. The generator is switched into those parts of the synchronization circuit where the necessary conditions exist for its precise triggering with the least phase spread. The variants for switching the generator into the synchronization circuit will be described below. The generator proper (Fig. 1) is implemented according to the thyratron circuit \(L_1\), which makes it possible to considerably reduce the duration of the generated pulse and also to increase the amplitude of the voltage oscillation on the capacitors

\(C_2 C_3\), which increases the stability of the frequency. The initial state of the circuit is as follows: thyratron \(L_1\) is extinguished; for this purpose the voltage on its control grid is chosen below the ignition potential at an anode voltage \(E_a\). The resistance \(R_1\) is chosen so large that a self-sustained discharge in the tube is impossible, even if the grid potential exceeds the ignition potential. In the initial state the capacitor is charged through \(R_1\) to the voltage \(E_a\), and capacitors \(C_2\) or \(C_3\) through \(R_1\) and \(R_2\). The action of a starting pulse with amplitude

Fig. 1. Schematic circuit diagram of the single-pulse generator

\(U_{\mathrm{st}} \gg E_a - U_z\) leads to the firing of the thyratron and to the discharge through it and the resistance \(R_2\) of capacitors \(C_1, C_2, C_3\). Capacitor \(C_1\) discharges faster than \(C_2\) or \(C_3\), and at the output a pulse is produced with \(\tau_{\mathrm{front}} = 0.05\ \mu\mathrm{sec}\) and \(\tau_{\mathrm{tail}} = 1.5\ \mu\mathrm{sec}\). After the discharge the thyratron does not go out, since \(C_2\), which determines the frequency of the relaxation oscillations, is still discharging. As the capacitor discharges, the anode voltage of the tube falls and the discharge ceases. After this, capacitors \(C_2\) and \(C_1\) are again charged from the voltage source \(E_a\) with the time constant \(\tau = R_1 C_2\), which is chosen to be of the order of the deionization time, i.e., \(10^{-4}\ \mathrm{sec}\), so that a repeated discharge does not occur in the tube. After a time \(t \simeq 5 R_2 C_2\), the initial electrical state of the circuit is restored. The shaped sharp pulse, through the cathode follower \(L_2\), is fed to the betatron-control synchronization circuit. The firing threshold of thyratron \(L_1\) is adjusted by the resistance \(R_5\). The frequency of the generator is varied by the resistance \(R_1\) and switch \(P_1\).

The results of testing showed that, if the thyratron generator is connected between the peaker and the input of the synchronization circuit, then, because of the long leading edge of the peaker pulse (30–40 \(\mu\mathrm{sec}\)), the generator has a large phase scatter in triggering, amounting to \(\pm 1.5\)–\(2\ \mu\mathrm{sec}\). This creates certain difficulties in synchronizing the injection phase for optimal capture of electrons. The radiation intensity in the pulse is extremely unstable. Good results are obtained in the case when the injection channel operates in continuous mode at a frequency of 50 Hz at the optimal phase, and single triggering is produced

in the mixing channel. The generator is then connected between the shaper and the final stage. Here the triggering pulse has a low steepness, of the order of \(0.01\ \mu\mathrm{s}\), ensuring clear ignition of the thyratron. Good results were obtained by using, for forming the injection synchronizing pulse, a computing device that measures the voltage on the magnet and the injection voltage, calculating the optimal injection phase from the known formula

\[ t_i=\frac{10^4\sqrt{U_{\mathrm{in}}(U_{\mathrm{in}}+1.02)}}{3R_0\omega H}. \]

In this case the most complete capture of the electrons injected into the chamber is ensured. The mixer, however, operates in a single-shot mode or at an adjustable frequency, being synchronized from the circuit with stabilization of the energy level [4]. With such triggering, the spread in intensity does not exceed \(1\text{--}2\%\). The described generator circuit, which provides low-frequency and single-pulse generation of radiation, can be used with equal success in various accelerators.

For recording and measuring the intensity of pulsed radiation fluxes, the integral method [5] is widely used, in which the charge collected on a charging capacitance is usually measured by a dc amplifier or an electrical circuit. This method is used when measuring average beam intensities over some period of time. Since the repetition frequency of sources of pulsed radiation fluxes may reach \(0.05\ \mathrm{Hz}\) and single shots, the integral method gives large errors. In addition, at low repetition frequencies of radiation pulses (\(10\text{--}1\ \mathrm{Hz}\)) the requirements on the insulation of the collecting electrode increase sharply.

The amplitude-time conversion method makes it possible to avoid these difficulties, but has a disadvantage with respect to calibration of the measuring setup, although the error introduced by the electronic apparatus is no more than \(2\%\) [6]. The oscillographic method provides low measurement accuracy (\(\pm 10\%\)). In the course of the work, a scintillation method was chosen for measuring the radiation pulse, with subsequent indication on digital devices. This method makes it possible to measure the absolute value of the instantaneous maximum amplitude of a current or voltage pulse, ensuring high linearity and an accuracy corresponding to \(0.1\%\). The method is as follows. In the circuit of the radiation receiver—the photomultiplier—there arises a current pulse that is fed to a memory device (pulse stretcher), consisting of a capacitor and a valve with a large reverse resistance, which preserves the amplitude value of the voltage pulse to which the capacitor is charged for a time considerably exceeding the duration of the pulse itself. The output pulse from the stretcher is an exponentially decreasing voltage with a time constant of about \(0.6\ \mathrm{s}\). The maximum value of this pulse is measured by an electronic digital voltmeter of type V7-8. Let us consider in more detail the operation of the pulse stretcher shown in Fig. 2.

To stretch pulses while preserving the ratio of their amplitudes, pulse storage by means of a capacitor is used [7]. The principle of such a stretching device is illustrated by the circuit in Fig. 2. The pulse whose duration is to be increased, from the output of cathode follower 1, which has a low output resistance, is fed to capacitor \(C\). The charging time constant of this capacitor is

\[ \tau=(R_{\mathrm{out}}+r_d)C, \]

where \(r\) is the resistance of the diode in the conducting direction; it is chosen to be much less than the pulse duration and less than, or at least of the order of, the duration of its leading edge. Then the capacitor \(C\) is charged to a voltage equal to the pulse amplitude \(U_{\mathrm{pk}}\). The drop in voltage at the input of the circuit is accompanied by cutoff of the diode, since the voltage at its cathode becomes higher than at the anode. If the reverse resistance of the diode is infinitely large, no resistance is connected in parallel with the capacitor \(C\), and it has no leakage, then the voltage at the output of the circuit will remain equal to \(U_{\mathrm{pk}}\). Thus, the capacitor \(C\) stores the signal amplitude. The output voltage must be maintained at a constant or approximately constant level for the time required to record the signal. After this (at the instant \(t_2\)) the capacitor must be discharged, and only after that will the circuit be ready for the arrival of the next pulse at its input. Discharge of the capacitor occurs through some finite resistance \(R\), which is connected in parallel with the capacitor; in this case the output voltage falls with the time constant \(\tau = RC\). The latter must be much greater than the time interval \(t^1 = t_2 - t_1\), if it is desired to obtain a pulse with a flat top. In the simplest case the circuit may contain no special circuit for rapid discharge of the capacitor \(C\), and by choosing the value of the resistance \(R\) one establishes a particular duration of the output pulse, which has a steep leading edge and an exponential decay. The block diagram of the measurement-recording channel is shown in Fig. 3. Here the voltage pulse

Fig. 2. Memory cell

Fig. 3. Block diagram of a two-channel digital intensimeter: 1—radiation source, 2—collimator, 3—absorber, 4—collimators, 5—detectors, 6—photomultipliers, 7—inverters, 8—amplifiers, 9—expanders, 10—digital voltmeters

from the photomultiplier is fed to the inverter, then to a four-stage amplifier with an overall gain equal to 1000. The amplified pulse, with an amplitude equal to 50 V (an amplitude of 50 V ensures operation of the charging diodes on the linear portion of the current-voltage character-

characteristic), is fed to an expander consisting of three memory cells, at the output of which, for matching with a V7-8 digital voltmeter, a cathode follower is installed. The cathode followers of the expander are made according to a cascade circuit with zero potential at the output, which ensures complete locking of the charging diodes at the moment when information is read from the memory cells. The connection between the expander stages is direct. The parameters of the digital intensimeter are: frequency of recorded pulses, \(10\) MHz–\(0.01\) Hz; minimum pulse duration, \(10^{-8}\) s; range of measured voltage amplitudes from \(0.05\) V to \(100\) V. Preliminary tests showed that the measurement error of the digital intensimeter is no more than \(0.5\%\). The use of different detectors makes it possible, with the aid of a two-channel intensimeter (for comparing pulses before and after the absorber), to measure the attenuation coefficients of bremsstrahlung and the accumulation of secondary electrons behind an absorber of various thicknesses and materials during a single radiation pulse (single shots), to study the spatial distribution of the beam density over its cross section, and to carry out a number of other physical experiments.

REFERENCES

1. L. M. Anan’ev, A. A. Vorob’ev, V. I. Gorbunov. An induction electron accelerator—a betatron. Atomizdat, Moscow, 1961.

2. A. P. Grinberg. Methods of acceleration of charged particles. Moscow, Gostekhizdat, 1950.

3. W. J. Bosley, Scient. Instrum., 28, 277, 1949.

4. V. I. Gorbunov, O. I. Nedavnii. Proceedings of the Fourth Interuniversity Conference on Electron Accelerators, 305–308. Moscow, 1964.

5. C. Hamina Westcott, Y. Rev. Jei Instz. 20, 181, 1949.

6. G. R. Kartashov, PTE, No. 1, 73, 1958.

7. A. M. Bonch-Bruevich. Radio electronics in experimental physics. Moscow, “Nauka,” 1966.