INJECTOR AND INJECTION LINE OF A 300 MeV SYNCHROTRON

L. G. Kositsyn

In the 300 MeV synchrotron, a pulsed accelerating tube at 260 kV is used as the injector. Electrons are introduced into the accelerating chamber by means of an electrostatic inflector. The beam deflection angle in the inflector is \(47^\circ\).

The design of the accelerating tube is shown in Fig. 1. The body of the tube consists of porcelain and steel cylinders bonded with epoxy adhesive. A duralumin cathode flange is bonded to the other end of the porcelain cylinder. The length of the porcelain cylinder is 30 cm. The body of the tube, provided with electrostatic shields, was tested at a voltage above 300 kV.

The accelerating gap, which is a Pierce gun [1], is placed in a steel cylinder at ground potential. The shape of the gun electrodes is shown in Fig. 2. The cathode of the accelerating tube is a lanthanum-boride pellet 5.6 mm in diameter. With a filament power of 120 W, the accelerating tube delivers an output current of 4.4 A. The current oscillogram is shown in Fig. 3. The beam divergence is \(5 \cdot 10^{-2}\) rad; the beam diameter at a distance of 24 cm from the anode does not exceed 2 cm.

From the accelerating tube, the electrons enter the vacuum chamber through the injection line. The focusing system of the injection line must provide, in the vertical plane, either a parallel or a converging beam \((\gamma_z \le 5 \cdot 10^{-3}\ \text{rad})\). To ensure beam parallelism at the exit from the inflector in the horizontal plane, the beam must have a crossover at the focal point of the inflector. The dimensions of the beam cross section are determined by the gap of the inflector. To avoid particle losses on the inflector plates, the beam height must not exceed 25 mm, and the width must not exceed 8 mm. The system of two quadrupole lenses described above satisfies these requirements.

The focusing scheme of the beam by quadrupole lenses is shown in Fig. 4. In the horizontal plane \((X)\), the first lens focuses the beam and the second defocuses it. In the vertical plane \((Y)\), the order of focusing and defocusing is reversed.

The optical power of the lenses is calculated by a trial method using the two-term formula:

\[ \begin{aligned} q &= \frac{q'_0}{\sqrt{K}} \sin \sqrt{K}l + q_0 \cos \sqrt{K}l,\\ q' &= q'_0 \cos \sqrt{K}l - \sqrt{K}q_0 \sin \sqrt{K}l, \end{aligned} \qquad \text{for a focusing lens;} \]

Fig. 1. Design of the accelerating tube: 1—fluoroplastic cylinder; 2—steel cylinder; 3—cathode flange; 4—anode; 5—cathode block; 6—tube; 7—rods; 8 and 9—screens; 10—cathode flange electrode.

Fig. 2. Shape of the beam electrodes: 1—cathode; 2—cathode electrode; 3—anode; 4—steel cylinder; 5—beam boundary

Fig. 3. Oscillogram of the accelerating-tube current

Visible labels in the diagram: plane \(Y\); plane \(X\); quadrupole lenses; \(l\), \(d\), \(l\), \(L\).

Fig. 4. Diagram of beam focusing by quadrupole lenses

\[ \begin{aligned} q&=\frac{q_0'}{\sqrt{K}}\operatorname{sh}\sqrt{K}\,l+q_0\operatorname{ch}\sqrt{K}\,l,\\ q'&=q_0'\operatorname{ch}\sqrt{K}\,l+\sqrt{K}q_0\operatorname{sh}\sqrt{K}\,l, \end{aligned} \]
for a defocusing lens;

where \(q_0\) and \(q_0'\) are the initial values of the coordinates and angles of inclination of the trajectories at the entrance to the lens;

\(q\) and \(q'\) are the coordinates and angles of inclination of the trajectories at the exit from the lens;

\(K\) is the specific optical strength of the lens:
\[ K=\frac{1}{BR}\frac{\partial B_x}{\partial Y} =\frac{1}{BR}\frac{\partial B_y}{\partial X}; \]

\(l\) is the length of the lens along the axis.

For the specified values \(l=2\ \text{cm}\), \(L=6\ \text{cm}\), \(q_0=1\ \text{cm}\), and \(q_0'=0.025\ \text{rad}\), values were found: \(K_1=0.04\), \(K_2=0.035\), and \(d=5\ \text{cm}\), which provide the best approximation to the required beam parameters at the exit from the focusing system. The magnetic-field gradient
\[ \frac{\partial B_x}{\partial Y}=\frac{\partial B_y}{\partial X} \]
for \(K=0.04\) and \(BR=1380\ \text{G}\cdot\text{cm}\) is \(75\ \text{G}/\text{cm}\).

The pole profile of the lenses is made in the form of arcs of circles with a radius of \(33\ \text{mm}\). The magnetomotive force is produced by four coils located on the pole cores. Each coil contains 1250 turns. The lens power-supply circuit provides adjustment of the current in the coils from 0 to 150 mA.

The inflector is a cylindrical electrostatic capacitor made in the form of a bent duralumin box of rectangular cross section. The high-voltage electrode located inside the box is divided lengthwise into three equal parts. Each of the parts is mounted on a separate feedthrough insulator and is powered from a separate adjustable high-voltage source. This makes it possible to correct distortions of the beam trajectory in the inflector caused by stray magnetic fields in the straight section. The grounded plate of the inflector is insulated from the box and is likewise divided into three parts. Each section has an independent lead to a coaxial connector mounted on the end section of the injection line. The plates are grounded through a resistance of \(100\ \Omega\). Observation of the signals from the plates, caused by electrons striking them as they pass through the inflector, makes it possible to perform fine adjustment of the voltage on the high-voltage electrodes.

With a mean radius of curvature of the inflector of \(35\ \text{cm}\) and a gap of \(1\ \text{cm}\), the calculated voltage on the plates is \(12.35\ \text{kV}\). The stabilized sources for feeding the inflector plates are designed for voltages up to \(15\ \text{kV}\). The transverse section of the inflector is shown in Fig. 5. The profile of the high-voltage electrode was selected in an electrolytic bath from the condition of obtaining the greatest extent of the working region. With a high-voltage-electrode width of \(40\ \text{mm}\), the working region is \(35\ \text{mm}\).

For beam alignment, a system of iron-free concentrated deflecting coils is used, mounted on the end section of the injection line. The vertical and horizontal deflections are combined in one and the same region. In the horizontal plane, the alignment system must provide compensation of the beam deflection

\(\gamma = 0.1\ \mathrm{rad}\), caused by the stray magnetic field. In the vertical plane the requirements on the alignment system are lower by a factor of 2–3.

The relation between the deflection angle of the electron beam and the magnetic field has the form:

\[ \gamma = 5.86 \cdot 10^{-2}\frac{\sqrt{1-\beta^2}}{\beta} H_0 l_0, \]

where

\[ H_0 \]
is the magnetic-field strength in oersteds;

\[ l_0 = \frac{\displaystyle\int_{-\infty}^{+\infty} H\,dl}{H_0} \]
is the effective length of the coil in meters;

\[ \beta \]
is the relative velocity of the electrons.

The required magnetic-field strength for horizontal deflection is produced by coils containing 1000 turns each. The vertical-deflection coils have half as many turns. Stabilized power supplies provide smooth regulation of the current in the coils from 0 to 150 mA.

In addition to the electrical alignment of the electron beam, mechanical adjustment is provided for the accelerating tube, the lenses, the inflector, and the injection line as a whole.

Fig. 5. Cross section of the inflector

In the final section of the injection line, clipping plates are installed. Their purpose is to introduce particles into the accelerating chamber during a time shorter than the duration of the injection-voltage pulse. The clipping plates are powered from a special pulse generator [2]. The duration of beam clipping is adjustable from 0.1 to 6 μsec. In the latter case, the duration of electron injection and their energy spectrum are determined entirely by the accelerating-voltage pulse. Short clipping durations (1.1–0.2 μsec) are necessary for a detailed study of the beam motion during the first revolutions [3].

The current of the accelerating tube can be measured in two ways: with interruption and without interruption of the beam. The first measurement method is carried out with a movable Faraday cylinder, the second with an induction magnetic sensor. The magnetic sensor is a coil of 180 turns wound on a ring of F-250 ferrite. The sensitivity of the sensor is 3.85 mV/mA.

Control of the beam shape and of its centering relative to the axis of the line is carried out by means of a grid and a screen coated with a phosphor. The grid is located at the beginning of the injection line and can be set at an angle of \(45^\circ\) to the beam axis. The screen is installed at the entrance to the inflector. To allow the beam to pass through the screen, there is a slit 36 mm high and 9 mm wide, located exactly opposite the working gap of the inflector. The beam is observed through viewing windows. Limitation of the dimensions of the beam introduced into the inflector is achieved by a collimator with adjustable slit dimensions in two planes. Collimation of the beam protects the inflector from bombardment-

...of tuning with unwanted electrons and is necessary in studying the first revolutions of the particles in the accelerating chamber. During the adjustment of the focusing system, the collimator was used to study the distribution of electron density in the beam in two directions by the methods of a movable target and a point aperture, and also to study the angular characteristics of the beam. Measurements of the angular characteristics of the beam at the inflector exit showed that the beam diverges in the horizontal plane at an angle of \(5 \cdot 10^{-2}\ \mathrm{rad}\), while in the vertical plane it converges at an angle of \(3 \cdot 10^{-3}\ \mathrm{rad}\). The above values of the angles are smaller than the permissible ones.

REFERENCES

1. J. Pierce. Theory and Design of Electron Beams. 1956.

2. L. I. Minenko. Power supply of the deflecting plates of the beam injection system of the 300 MeV electron synchrotron. This collection.

3. V. P. Anokhin, V. N. Eponeshnikov, L. G. Kositsyn. Correction of the magnetic field with respect to the beam and the quasibetatron mode in the 300 MeV synchrotron. This collection.