Magnetic Field Correction by the Beam and the Quasi-Betatron Regime in a 300 MeV Synchrotron

V. P. Anokhin, V. N. Eponeshnikov, L. G. Kositsyn

Commissioning an accelerator is an independent and complex task. Preparation for commissioning consists of magnetic measurements and adjustment of the magnetic field, obtaining the required parameters of the injected beam and of the radio-frequency system, achieving reliable operation of the electromagnet power-supply circuit, the vacuum system, synchronization, and other accelerator units.

The commissioning process consists in selecting the initial injection conditions, in the final correction of the magnetic field, and in adjusting the switching-on time and the parameters of the radio-frequency system.

Fig. 1. Signal from the first current flag

The initial position of the mechanical adjustments of the synchrotron injection line corresponded to the geodetic settings. The initial values of the electrical parameters of the injection system corresponded to the data obtained when tuning the injection line without switching on the electromagnet.

As was to be expected, injection of the beam into the accelerating chamber with the electromagnet operating was achieved after compensation of the stray magnetic field by the system of alignment coils on the injection line and by selecting the voltage on the inflector plates. Indica-

…tion of the optimal passage of the beam through the inflector gap was carried out from the signal of the first current flag (Fig. 1).

The currents in the magnetic-field correction circuits were set on the basis of processing the results of magnetic measurements. Obtaining the first turn of the beam with the given correction-current settings usually presented no difficulty. However, in order to obtain the quasi-betatron regime in the synchrotron, additional adjustment of the magnetic-field correction was required. However accurately the magnetic measurements were made, and however carefully the currents in the correction circuits were selected, at that stage of the work it was impossible to take into account many local field perturbations that accompanied the installation of the vacuum chamber with numerous metal flanges, beam-indication devices, the inflector, the high-frequency resonator, etc.

Fig. 2. Signal from a current flag located at the end of the first turn

The final correction of the magnetic field was performed using the beam, on the basis of an analysis of distortions of the instantaneous orbit.

The principal means for indicating the shape and position of the beam on the first turns were movable fluorescent screens and shadow screens [1]. The beam was observed visually. The study of beam motion in the accelerating chamber was preceded by adjustment of the focusing system and of the injection line to obtain an optimum signal from the current flag located at the end of the first turn (Fig. 2). The beam was then collimated, and the cutoff system was switched on. For reliable indication of individual beam turns, it was necessary to set the collimator slit dimensions to 2 mm horizontally and 4 mm vertically. The cutoff duration was 0.2 μsec. Under these conditions the beam dimensions on the screens did not exceed 10–15 mm.

The distortions of the orbit over the first two or three turns were estimated on the basis of the method of eliminating free oscillations [2]. The components of the free radial oscillations of the beam in the synchrotron on the first two turns are shown in Fig. 3. The instantaneous orbit is parallel to the horizontal axis $\Theta$. The discrepancy between the first and second beam turns at azimuth $a$ is caused by an error in the injection angle; at azimuth $b$, by a mismatch between the injection instant and the instant at which the instantaneous orbit passes through the radial coordinate of the inflector. These two azimuths were the principal points for observing the beam at this stage of accelerator tuning. By selecting the injection angle am—

Fig. 3. Components of free radial oscillations:
1 — caused by the difference between the injection angle and the optimum angle;
2 — due to the mismatch between the moment of injection and the moment at which the instantaneous orbit passes through the radial coordinate of the inflector.

the amplitudes of the components of the free oscillations of the beam were reduced to values close to zero. In this case the first two or three revolutions of the beam actually merged, and the electron trajectories were close to the instantaneous orbit. The effect of orbit contraction (\(\sim 0.7\ \mathrm{mm/rev}\)) was not taken into account.

Analysis of the distortion of the instantaneous orbit indicated the presence of a first-harmonic component of the magnetic-field inhomogeneity. The amplitude of the distortion of the instantaneous orbit reached \(2\ \mathrm{cm}\). By adjusting the current in the circuit of the corresponding field correction, it was possible to reduce the radial distortions of the orbit by a factor of four.

Eliminating the vertical oscillations of the beam required a more detailed study of the regularities of its motion. The initial injection conditions for the beam in the vertical plane were also selected by the method of eliminating free oscillations. In this case, mechanical adjustment of the injection channel in the vertical plane and tuning of the current in the vertical beam-alignment coil were used. It was found that compensation of distortions of the instantaneous orbit in the vertical plane at the injection radius over the combined two revolutions was insufficient. As the beam moved from the outer edge of the working region of the magnetic field to the inner edge, the amplitude of the vertical oscillations reached \(1.5\ \mathrm{cm}\). This fact indicated a nonuniformity of the distortions of the median plane of the magnetic field along the radius. For a detailed study of the topography of the median magnetic plane, the following method was applied. The moment of injection was chosen so that the instantaneous orbit coincided with the equilibrium orbit. In this case the beam, undergoing radial oscillations, in the first three revolutions actually probed the entire working region. Analysis of the spatial trajectories of the particles made it possible to conclude that the main cause of the vertical oscillations of the beam was distortions of the median magnetic plane in the straight sections of the accelerator, with the principal contribution to the oscillation amplitude being made by the straight section containing the resonator. The vertical oscillations of the beam were eliminated after changing the design of the coils that compensate the radial component of the magnetic field in the straight sections. Concentrated coils were replaced by distributed ones. The law of radial distribution of the coil turns was selected on a model. As a result of the measures taken, the amplitude of the vertical oscillations of the beam was reduced to \(3\ \mathrm{mm}\).

Fig. 4. Signal from the crystal in the presence of the quasi-betatron regime.

Careful correction of the magnetic field was a prerequisite for obtaining the quasi-betatron regime. The calculated duration of the quasi-betatron regime in the synchrotron is $4.4\ \mu\text{s}$. During this time the electrons make about 140 revolutions in the chamber. A NaI(Tl) crystal with a photomultiplier and an internal $\beta$-target [1] serve to indicate the quasi-betatron regime. Initially, the injection instant was chosen so that the beam struck the $\beta$-target on the first revolution. Then, by adjusting the injection instant, the first revolution of the beam was moved away from the $\beta$-target. In this case the signal from the $\beta$-target lagged in time behind the injection instant, which indicated the presence of particle circulation. The calculated duration of the quasi-betatron regime was achieved by fine selection of the injection instant and the angle of entry. The signal from the crystal in the presence of the quasi-betatron regime is shown in Fig. 4. The first peak corresponds to the injection instant, the second to the end of the quasi-betatron regime.

The efficiency of particle capture into the quasi-betatron regime was estimated by comparing oscillograms from the first current flag and from the $\beta$-target. At an injection current of $40\ \text{mA}$, the capture coefficient was 10%.

Obtaining the quasi-betatron regime is an important stage preceding the switching on of the high-frequency system.

REFERENCES

1. V. P. Anokhin, L. G. Kositsyn. Beam and radiation indication system in a 300 MeV electron synchrotron. In the present collection.

2. Electron Accelerator at 680 MeV. Collection of articles, Gosatomizdat, 1962.