We mainly perform maintenance, operation and upgrade of the 1 GeV electron linac.  High-gradient accelerating structures and RF guns as one of the key technologies for future light sources have also been studied. Our first priority is to enhance stability and reliability of the linac. More than two staff members are assigned to each linac component, in order to fix machine troubles as soon as possible and not to delay the beam injections to the synchrotron and NewSUBARU.

Since the completion of SPring-8, we have been working actively to stabilize the linac. Thus the SPring-8 linac leads the world's electron linacs in beam stability. The constantly improving reliability of the linac is essential to sustain long-term top-up operation with the least interruption

We maintain and improve an injector section of the linac, that is a part between the electron gun and the first accelerating structure.
Our activities are listed as follows:

• The previous complicated RF system for the injector section with two different klystrons was simplified to have one klystron in order to improve stability and reliability of the RF system.

• A bunch length monitor system with a streak camera was designed and installed to optimize phases of the high power RF's that is fed to buncher cavities.

• A beam deflector to kick out the field emission currents from a grid of long-time used cathode was designed and installed to reduce dark currents from the gun.

• A compact electron gun modulator that was newly developed  for high reliability will  replace the present one soon.

We maintain and improve accelerator structures, vacuum systems,  and electric magnet systems. We have also studied high-gradient acceleration and RF couplers for accelerating structures. Our activities are the followings:

• Development of the RF gun test equipment.

• Based on the collaborative research with KEK on high-gradient acceleration, A chemical method for processing RF cavities was developed aiming at reduction of RF discharges.  We finally achieved the world record of a maximum field strength on cathode.

• A new bending magnet that enables a fast pattern excitation was designed and replaced a conventional block type magnet in order to switch an electron beam between the two rings at higher frequency.

• A high power waveguide switch was developed. The waveguide switch will be used to put a backup klystron on line when the first klystron failes.

We maintain and improve klystron amplifiers, an RF reference generator system, a trigger system, a klystron drive system, an RF monitor system and so on. Our activities are;

• Reinforcement of the klystron modulator control system.

• Stabilization of the klystron amplifiers in RF amplitude and phase.

• Reduction of the RF phase variation at the end of the klystron drive line.

• Development of a synchronous RF oscillator (generator) that generates 2856MHz burst waves synchronizing a gun trigger pulse.

• Introduction of an energy compression system (ECS) to stabilize the beam energy.

We develop and maintain devices (beam monitors or signal processing circuits) and programs to diagnose electron beams, including important beam monitors. Application developments are as follows:

• Strip-line type beam position monitor.

• Fast current transformer for short pulse beams.

• Wire grid monitor to observe beam profiles measuring secondary electron charges.

• Optical transition radiation (OTR) monitor to detect fine beam profiles.

• Stabilization of the beam energy and position by feedback control applying beam monitors.

• Automatic beam adjustment such as automatic RF phasing.

The SPring-8 users expect a uniform bunch pattern in the stored beam; that is, the linac is requested to provide a beam in which every pulse has a stable current and energy. The top-up operation since 2004 increasingly emphasizes the importance of stability and reliability of the linac.

At the begining of the SPring-8 operation, the linac had beam energy variations of about 1%. We started the beam stabilization in 1998: Investigation of the fluctuation and several kinds of improvement had been carried out, then the stabilization was almost completed in 2005, resulting in an energy instability of 0.02% rms even in long term. We have basically reinforced the components of the linac, minimizing introduction of feedback control.

The present beam performance is presented in the table.

References

T. Asaka　"Research on stabilization of large linear accelerator", Thesis for Ph. D., Kyoto Univ. (2004).

H. Hanaki, T. Asaka, H. Dewa, T. Kobayashi, A. Mizuno, S. Suzuki, T. Taniuchi, H. Tomizawa, K. Yanagida, "Beam Stabilization in The SPring-8 Linac", APAC2004.

H. Hanaki, T. Asaka, H. Dewa, T. Kobayashi, A. Mizuno, S. Suzuki, T. Taniuchi, H. Tomizawa, K. Yanagida, "Enhancements of Machine Reliability and Beam Quality in SPring-8 Linac for Top-Up Injection into Two Storage Rings", PAC2005.

The long klystron drive line (see Fig.1) that feeds RF powers to eleven klystrons, is a nitrogen-filled rectangular wave guide, in which the RF phase is easily varied by its deformation due to its body temperature variation.  We reduced the phase variation by three measurements; the room temperature stabilization by improving air conditioners, thermal insulation of the waveguide, and fine regulation of the nitrogen pressure. Now the room temperature in the klystron gallery is well regulated with fluctuation of less than 1 degree even in mid winter.

The improved regulation of the klystron cooling water results in the phase variation  of less than 0.5 degrees at the klystron's output.

We have also improved the PFN voltage regulation system of the klystron modulator. The long-period variation of the PFN voltage was reduced to 0.03% rms, which corresponds to the RF power and phase variation  of about 0.08% rms and 0.2 degrees rms, respectively.

Fig. 1: Diagram of linac RF system

References

H. Sakaki, H. Yoshikawa, T. Hori, T. Sogo, N. Adachi, "Statistical Analysis and Control of the Electron Beam Energy Fluctuation at Linear Accelerator", Trans. Society of Instrument and Control Engineers 35-10, 1283  (1999), (in Japanese).

T. Asaka, H. Hanaki, T. Hori, T. Kobayashi, A. Mizuno, H. Sakaki, S. Suzuki, T. Taniuchi, K. Yanagida, H. Yokomizo, H. Yoshikawa, "Stabilization fo the rf system at the SPring-8 linac", Nucl. Instr. and Meth. A488, 26 (2002).

S. Suzuki, T. Asaka, H. Dewa, H. Hanaki, T. Kobayashi, A. Mizuno, T. Taniuchi, H. Tomizawa, K. Yanagida, "Improvement of SPring-8 Linac Toward Top-Up Operation", 30th Linear Accelerator Meeting in Japan (2005).

A trigger pulse for the electron gun is generated by counting the 508.58 MHz master signal for the storage ring. Therefore, a 1 ns beam ejected from the gun has not been synchronized with the linac's 2856MHz RF, which has no harmonic relation with the ring's frequency.

Thus the asynchronous 2856 MHz RF formed two or three bunches along with the gun trigger timing.  This unstable bunching caused random variation of the beam loading of the accelerating RF field, and then the beam current and energy center were consequently different for every shot.

We solved this asynchronousity by introducing an arbitrary waveform generator (AWG) whose oscillation can be started by an external trigger (see Fig. 2 left). A principle of this circuit is as follows:

• Sinusoidal waves of 89.25 MHz with duration of 290 ms, are programmed in the AWG. 

• The 1 Hz beam trigger causes the generator to start oscillating by referring to the external 508.58-MHz clock, and thus a 89.25 MHz burst signal is created by synchronizing with the 508.58 MHz reference.

• This intermediate signal is multiplied by 32 to generate the 2856 MHz reference.

This synchronization stabilized the beam current variation, and then reduced the energy fluctuation to less than 0.01% rms. Figure 2 right presents the current variations of 250 ps single-bunch beam. Comparison of the beam currents  in the left and right halves of the graph explains that the new synchronous oscillator stabilized the beam currents in the buncher.

References

Y. Kawashima, T. Asaka, T. Takashima, "New synchronization method of arbitrary different radio frequencies in accelerators", Phy. Rev. Special Top. Accelerator Beams 4, 082001 (2001).

T. Asaka, Y. Kawashima, T. Takashima, T. Kobayashi, T. Ohshima, H. Hanaki, "Method for stabilizing beam intensity and energy in the SPring-8 linac", Nucl. Instr. and Meth. A516, 249 (2004).

We introduced a conventional ECS which is mainly composed of a magnetic chicane section and an accelerating structure. A magnetic chicane expands a bunch length according to its energy distribution.  An accelerating structure provides the energy modulation of the electron bunches and then compress the energy spread.  The principle of ECS is illustrated in Fig. 3. An ECS also reduces the energy fluctuation according to the same principle.

The installed ECS enabled the high current injection of 350 mA for the 40 ns beam because the energy spread due to the beam loading was compressed and then the beam loss was greatly reduced. Figure 3 right clearly shows that the ECS compensated the beam energy variations caudsed by voltage drops of klystron modulators.

Since the beam energy is sensitive to the phase of the RF fed into the ECS's accelerating structure, the RF phase is stabilized by introduction of PLL technique into the independent drive line (see Fig.1) for the ECS klystron.

Reference

T. Asaka, H. Dewa, H. Hanaki, T. Kobayashi, A. Mizuno, S. Suzuki, T. Taniuchi, H. Tomizawa, K. Yanagida, "Performance of the Energy Compression System at the SPring-8 Linac", EPAC2002.

A quasi nondistractive thin-screen profile monitor using OTR was installed in the center of the chicane section to observe the beam energy and energy spread before being compensated by the ECS.

Beam position monitors were installed in the linac and the beam transport lines to the two rings. Automatic beam position and energy adjustment is executed by feedback control applying the BPMs.

The details of these activities are described in the later section.

The field emission at electron guns or in RF cavities causes dark currents. A part of the dark currents, generated in the injector section of the linac, is accelerated up to 1 GeV and then injected into RF buckets to be vacant. These faint satellite bunches emit unwanted SR lights,  which disturb accurate experiments of users.

The electron gun of the linac accelerates the grid emission currents, emitted from a grid of cathode assembly, and then forms the dark currents. We installed a beam deflector which kicks out the unwanted grid emission currents between the gun and the prebuncher as shown in Fig. 4. Figure 5 demonstrates that the deflector has filtered out the faint charges around the main bunch. Now we are trying to suppress the dark currents generated in the first accelerating structure.

Fig. 4: Beam deflector and its high voltage pulses

Fig. 5: Satellite distributions of stored single bunch beam.

RF powers for injector section were attenuated to be as low as possible to minimize the field emission for this measurement.

References

A fast beam current transformer (CT) was developed and installed for measurements of 1-ns pulsed beams. A signal of the CT is strongly protected form environmental noise, therefore the CT presents a high signal to noise ratio. The CT has good frequency response. Its rise time is shorter than 300 ps with a small ringing.

Reference

K. Yanagida, H. Yoshikawa, S. Suzuki, A. Mizuno, H. Sakaki, T. Taniuchi, T. Hori, T. Kobayashi, T. Asaka and H. Yokomizo, "FAST BEAM CURRENT MONITOR FOR SPring-8 LINAC", AIP Conf. Proc. 392, 765 (1997).

A wire grid monitor was developed for emittance measurement.  Tungsten wires whose diameter is 0.3 mm are used. A single-wire grid monitor was installed in injector section, and five-wires grid monitors were installed in other section. A signal processor is a charge sensitive amplifier that detects secondary emission charge when an electron beam hits on the wire. A commercially available charge sensitive amplifier CS-507 (CLEAR PULSE) is used and takes on a  principal process in the signal processor.

Reference

K. Yanagida, Y.Itoh, T. Hori, S. Suzuki, H. Yoshikawa, A. Mizuno, H. Sakaki, A. Kuba and H. Yokkomizo, "EMITTANCE MONITOR FOR SPring-8 LINAC", 17th Int. Linac Conf., Tsukuba, 920 (1994).

An OTR monitor was installed for an observation of optical transition radiation generated when a beam was passing through a thin foil. The thin foil is composed of a 12.5-µm Kapton foil with 0.4-µm evaporated aluminum. The thin foil enables us to measure a beam profile at any time because increase of emittance is suppressed small when the thin foil is inserted in a beam trajectory.

Reference

T. Asaka, T. Kobayashi, S. Suzuki, H. Hanaki, K. Yanagida and A. Yamashita, "DEVELOPMENT OF THE QUASI-NON-DESTRUCTIVE BEAM SCREEN MONITOR", 27th Linear Accelerator Meeting in Japan (2002).  (in Japanese)

A beam position monitor (BPM) without beam destruction was developed. The BPM for non-dispersive section has 27-mm stripline pickup electrodes and has a diameter of 32 mm. A signal processor is principally consists of 2856-MHz band-pass filters and logarithmic detection circuits that includes AD8313 (ANALOG DEVICES). A resolution of position measurement is 16 µm.

Reference

K. Yanagida, T. Asaka, H. Dewa, T. Fukui, H. Hanaki, T. Kobayashi, T. Masuda, A. Mizuno, S. Suzuki, R. Tanaka, T. Taniuchi, H. Tomizawa and A. Yamashita, "INSTALLATION OF THE SPRING-8 LINAC BPM SYSTEM", LINAC2002.

A program was developed in order to match twiss parameters for the NewSUBARU beam transport line effectively. The program outputs emittances and twiss parameters using beam sizes that are obtained by the wire grid monitors and screen monitors. The program also provides appropriate currents of quadrupole magnets for the twiss parameter matching.

Reference

K. Yanagida, S. Suzuki, T. Asaka, A. Mizuno, T. Taniuchi, T. Hori, T.Kobayashi, H. Akimoto and H. Hanaki, "TWISS PARAMETER MATCHING FOR THE BEAM TRANSPORT LINE OF THE SPring-8 LINAC", 24th Linear Acceleraor Meeting in Japan, Sapporo, 82 (1999) (in Japanese).

Programs have been developed to tune automatically the linac. The main program is a feedback program that stabilizes beam positions at the locations of BPM's. This program suppresses the fluctuations within ±50 µm for long term in the non-dispersive sections as shown in Fig. 8. While in the dispersive section the energy fluctuations are reduced to less than ±0.03%.

Reference

K. Yanagida, T. Asaka, H. Dewa, H. Hanaki, T. Kobayashi, A. Mizuno, S. Suzuki, T.　 Taniuchi, H. Tomizawa, "BEAM INSTRUMENTATION USING BPM SYSTEM OF THE SPring-8 LINAC", LINAC2004.

K. Yanagida, T. Asaka, H. Dewa, H. Hanaki, T. Kobayashi, A. Mizuno, S. Suzuki, T.　 Taniuchi, H. Tomizawa, "BEAM POSITION FEEDBACK AND AUTOMATIC PHASING OF SPring-8 LINAC", 30th Linear Accelerator Meeting in Japan (2005). (in Japanese)

The top-up operation of the SPring-8 and NewSUBARU storage rings has been maintained since 2004. The SPring-8 linac has been improved to realize frequent beam injections into the two synchrotrons at short intervals and to maintain stable top-up injection over one month.

The important activities are summarized below, however, the stabilization of the linac is described in the section 3.1.

References

S. Suzuki, T. Asaka, H. Dewa, H. Hanaki, T. Kobayashi, A. Mizuno, T. Taniuchi, H. Tomizawa, K. Yanagida, "Improvements of SPring-8 Linac towards Top-up Operation", EPAC 2004.

S. Suzuki, T. Asaka, H. Dewa, H. Hanaki, T. Kobayashi, A. Mizuno, T. Taniuchi, H. Tomizawa, K. Yanagida, "Improvement of SPring-8 Linac Toward Top-Up Operation", 30th Linear Accelerator Meeting in Japan (2005). (in Japanese)

The previous block-type bending magnet, that distributed beams to the SPring-8 and NewSUBARU rings, was replaced with the new laminate-type bending magnet which can be momentarily excited at short intervals.

In 2002, we reduced the RF repetition rate from 60 to 10 Hz for electric power saving. As a result, we saw remarkable room temperature drifts in winter. These temperature variations caused the phase drifts in the drive line mentioned in 3.1. We discontinued inadequate operation and function of the air conditioners which had caused excessive reduction of the ventilated air temperature.

Figure 9 presents the recent stabilized room temperatures in winter 2005.

We have optimized the unique beam trajectory in the linac for injections into two rings. This optimization resulted in one united operation parameter set which realizes frequent beam distribution at short intervals.

We improved the trigger system of the linac to prepare standby klystrons which are feeding  RF powers into the accelerating structures but do not accelerate beams. The standby klystron can be immediately set on line instead of a failed one in a minimum down time.

The high gradient acceleration in accelerating structures or RF gun cavities is essential to achieve the following advances of accelerators;  higher beam energy, downsizing and very low emittance beams.  We have been studying the high gradient as follows.

A video spectroscopic camera system applying a spectroscope and an image intensifier, which can observe spectral distributions shot by shot, was developed.  We have observed the RF breakdown induced in S-band accelerating structure for the KEKB linac and the RF gun cavities. The obtained spectra resulted in identification of ions generated in the RF discharges.　These analyses consequently provided information on the copper surface contamination.

References

H. Tomizawa, T. Taniuchi, H. Hanaki, Y. Igarashi, S. Yamaguchi, A. Enomoto, "Spectrographic Approach for Diagnosing RF Breakdown in Accelerating RF Structures", Applied Surface Science 235, 214 (2004).

We treated an RF gun cavity by chemical etching to remove the surface contamination and oxide layer which are suspected to induce RF breakdowns. The RF conditioning of the treated cavity showed remarkable reduction of the RF breakdowns and conditioning time. We finally obtained the world record of 187 MV/m as the field gradient at cathode surface of S-band RF gun.

Reference

H. Tomizawa, T. Taniuchi, H. Dewa, A. Mizuno, T. Moriwaki, Y. Ikemoto, S. Suzuki, H. Hanaki, N. Kumagai, M. Kimura, "Effects of Chemical Etching as a Surface Treatment for Accelerating Structures Made of Copper", 29th Linear Accelerator Meeting in Japan (2004). (in Japanese).

Reference

T. Taniuchi, T. Asaka, T. Kobayashi, S. Suzuki, H. Dewa, H. Tomizawa, H. Hanaki, A. Mizuno, K. Yanagida, "RF Characteristics of Waveguide Coupler for Traveling Wave Structures", 29th Linear Accelerator Meeting in Japan (2004). (in Japanese)

We have been developing photocathode RF guns as electron sources ejecting the high quality electron beams with low emittance or short bunch length. At present (Aug. 2005), a minimum emittance otained is 1.7 π mmmrad at the beam charge of 0.09 nC. Details of this development is described in the section of projects.