SAMEER, Mumbai
Sokendai, Ibaraki
A successful development of a 6 MeV electron radiotherapy machine at SAMEER, India was reported earlier*. Now a 15 MeV electron linac prototype is designed, developed and tested at our site. We have measured a beam current of 80 mA at the X-ray target attached to the linac. Energy gained by electrons in a cavity chain of about 1.2 m length is measured to be more than 15 MeV using a 6 MW klystron power source. An RF window capable of handling 12kW average power is attached to the linac tube and it is cooled by water. The final linac parameters measured were at par with the designed values. A high voltage modulator and control console for the linac are designed and developed in house. This paper will describe key aspects of the design and development process of the complete system. Also future applications are planned like-dual energy dual mode linac for radiotherapy, cargo scanning system and compact compton X-ray source using this technology is briefed in this paper.
* R.Krishnan et. al. "S band linac tube developmental work in SAMEER", FR5REP083, PAC09, Vancouver, Canada.
CERN, Geneva
The construction of Linac4, a 160 MeV H- Linac, is the first step in upgrading the LHC injector chain. Unlike CERN's present injector linac, Linac4 will inject into the subsequent synchrotron via charge exchange injection. In a first stage, it will inject into the existing CERN PS Booster. At a later stage, Linac4 has the option to be extended by a superconducting linac (SPL) which could then inject into a new synchrotron (PS2). Construction of Linac4 has started in 2008, and beam operation is presently planned for 2014. An overview of the Linac4 main parameters and design choices is given, and the status of the construction reported.
IKP, Mainz
During the last three years at the 1.5 GeV Harmonic Double Sided Microtron (HDSM)* of MAMI a lot of improvements concerning the longitudinal operation of the accelerator were tested and installed. To monitor the rf power dissipated in the accelerating sections, their cooling water flow and its temperature rise are now continuously logged. Phase calibration measurements of the linacs and the rf-monitors revealed nonlinearities of the high precision step-motor driven waveguide phase shifters. They were recalibrated to deliver precise absolute values. Thereby it is now possible to measure not only the first turn's phase very exactly, but also determine the linac's rf-amplitude within an error of less than 5%, using the well known longitudinal dispersion of the bending system. These results are compared to the thermal load measurements. For parity violating experiments the beam energy has to be stabilised to some ppm. A dedicated system measuring the time-of-flight through a bending magnet is now used in routine operation and controls the output energy via the proper linac phases.
* K.-H. Kaiser et al., NIM A 593 (2008) 159 - 170, doi:10.1016/j.nima.2008.05.018
RIKEN/SPring-8, Hyogo
The pulse width of an X-ray laser at XFEL/SPring-8 is several tens femto-seconds, which requires reference rf signals to have the same time-stability. The reference signals with a low phase-noise oscillator are sent to instruments in 19" racks developed along an accelerator by an optical fiber system. The temperature drift of the fiber makes phase shifts of the reference signals. Therefore, the fiber is put in a thermal-insulated duct. By feeding temperature-controlled water (26.1 ± 0.1 deg. C) in a pipe attached to the duct, the fiber temperature was kept to be 26.2 ± 0.08 deg. C at the ambient temperature change of 29.1 ± 1.7 deg. C. From this temperature controllability, the phase shifts of the signals through a 400 m fiber of a thermal coefficient of 5 ps/km/K are 160 fs. Further reduction of the shifts is required and will be achieved by a fiber-length feedback control in a future plan. Vibration of the fiber also degrades the quality of the signals. The fiber is embedded on a vibration buffer material. A test to evaluate the effect of the vibration to the transmitted signal phase was carried out. The test result will be also shown in this paper.
DESY, Hamburg
The Free Electron Laser in Hamburg (FLASH) is a user facility providing high brilliant laser light for experiments. It is also a unique facility for testing the superconducting accelerator technologies. FLASH cavities are operating at pulsed mode. There is a filling stage to build the RF voltage in the cavities and then follow a flattop for beam operation. By the limitation of the klystron pulse length the filling time of the cavities is limited to several hundred microseconds. In order to fill the cavities to the dedicated voltage usually large RF power is required for the filling stage. For European XFEL during RF operation the klystrons will be working quite near the saturation point for better efficiency. So lowering the unnecessary klystron peak power under closed loop operation is very important for close-limitation operation. The paper will present the method which allows decreasing the required klystron peak power as well as the reflected power by filling the cavity in resonance. Simulation results will be presented as well as experimental demonstrations at FLASH.
IPM, Tehran
CERN, Geneva
The pulse compressor is located after the klystron to increase the power peak by storing the energy at the beginning and releasing it near the end of klystron output pulse. In the CTF3 [1] pulse compressors a doubling of the peak power is achieved according to our needs and the machine parameters. The magnitude of peak power, pulse length and flatness can be controlled by using a phase modulator for the input signal of klystrons. A C++ code is written to simulate the pulse compressor behaviour according to the klystron output pulse power. By manually changing the related parameters in the code for the best match, the quality factor and the filling time of pulse compressor cavities can be determined. This code also calculates and sends the suitable phase to the phase modulator according to the klystron output pulse power and the desired pulse length and peak power.
KEK, Ibaraki
S1 global will be operated at STF in KEK, where total 8 cavities will be installed. The digital llrf system to control the vector sum of the field gradients to be flat has been developed. All the digital llrf system including rf monitoring, piezo-control system will be shown. The new llrf system suitable for the DRFS scheme, which is also studied during S1 grobal, is also under development.
TUL-DMCS, Łódź
DESY, Hamburg
DESY FLASH accelerator is composed of 6 accelerating modules. The single accelerating module contains 8 superconducting resonant cavities. Since FLASH operation is dedicated for various energy physics experiments such as high current beam acceleration or SASE tuning, the sc cavities are Lorentz force detuned when operated with high gradient accelerating fields*. The ACC 3, 5 and 6 cryomodules are equipped with piezo tuners allow compensating of dynamic detuning during the RF pulse. In order to assure the simultaneous control of all available piezo tuners a distributed, multichannel digital and analogue piezo control system was applied. The paper describes the main parts of the system as well as its efficiency measurements obtained during high current beam acceleration (9 mA tests) performed in DESY. The piezo tuners were operable for 23 cavities for several hours. Moreover, the first piezo sensor measurements using double stack piezos installed in ACC 6 cryomodule are briefly demonstrated.
*M. Grecki, A. Andryszczak, T. Poźniak, K. Przygoda, S. Sękalski,
"Compensation of Lorentz Force Detuning For SC Linacs (With Piezo
Tuners)", Proceedings of EPAC 2008, pp. 862-864.
I-Tech, Solkan
In this article we are presenting tests and development of digital low level RF control system Libera LLRF. Libera LLRF is a digital system small in size but powerful in terms of performance as tests revealed. Size of unit matches industrial standards and is in 19" 2U sustainable metal box that fits into racks. Development of the Libera LLRF reflects needs of accelerator's and their operators. With its capabilities it is a system that is able to control RF at 4th generation light sources. Concept of the Libera LLRF system also enables implementation of operator's own solutions in controlling RF. During preparations for testing Libera LLRF's features proved to be useful since little time was needed to install and operate the system. In some cases its features and capability enabled operators to identify and quickly resolve problems that were accelerator's components related.
SLAC, Menlo Park, California
The Linac Coherent Light Source (LCLS) at SLAC is in full user operation and has met the stability goals for stable lasing. The 250pC bunch can be compressed to below 100fS before passing through an undulator. In a new mode of operation a 20pC bunch is compressed to what is believed to be about 10fS. Experimenters are regularly using this shorter X-ray pulse and getting pristine data. The 10fS bunch has timing jitter on the order of 100fS. Physicists are requesting that the RF system achieve better stability to reduce timing jitter. Drifts in the RF system require longitudinal feedbacks to work over large ranges and errors result in reduced performance of the LCLS. This paper describes the new RF system being designed to help diagnose and reduce jitter and drift in the SLAC linac.
SLAC, Menlo Park, California
CERN, Geneva
The LHC Low Level RF system (LLRF) is a complex multi-loop system used to regulate the superconductive cavity gap voltage as well as to reduce the impedance presented by RF stations to the beam. The RF system can have a profound impact on the stability of the beam; a mis-configured RF system has the potential of causing longitudinal instabilities, beam diffusion and beam loss. To configure the RF station for operation, a set of parameters in the LLRF multi-loop system have to be defined. Initial system commissioning as well as ongoing operation requires a consistent method of computer based remote measurement and model-based design of each RF station feedback system. This paper describes the suite of Matlab tools used for configuring the LHC RF system during the start up in Nov2009-Feb2010. We present a brief overview of the tool, examples of commissioning results, and basics of the model-based design algorithms. This work complements our previous presentation [1], where the algorithms and methodology followed in the tools were described.
[1] D. Van Winkle et. al. 'Feedback Configuration Tools for LHC Low Level RF System,' PAC'09, Vancouver, Canada, May 2009, THZCH03, p. 249 (2009); http://www. JACoW.org.
RIKEN/SPring-8, Hyogo
JASRI/SPring-8, Hyogo-ken
This paper describes the present status of a precise temperature-regulation system for the C-band accelerator at XFEL (X-ray Free Electron Laser)/SPring-8. It is essential to maintain a constant temperature of an rf cavity for stable lasing. We therefore installed a heater-assembly unit into a cooling water circuit of each rf cavity. By controlling the heater power, the temperature of the cavity can be stabilized. We constructed a prototype of this system at the SCSS (SPring-8 Compact SASE Source) test accelerator to check its feasibility for the XFEL. The prototype significantly contributes to a stable supply of SASE to users. For the XFEL, we simplified this system in consideration of cost and controllability. For example, to make one regulation system simultaneously controlling two C-band accelerating structures was tried. Keeping a temperature variation as tight as ±0.02 K at any operational mode could be achieved by this system. The preliminary test results of the system are also reported in this paper.
CERN, Geneva
The LHC ACS RF system is composed of 16 superconducting cavities, eight per ring, housed in a total of four cryomodules each containing four cavities. Each cavity is powered by a 300 kW klystron. The ACS RF power control system is based on industrial Programmable Logic Controllers (PLCs), but with additional fast RF interlock protection systems. Operational performance and reliability are described. A full set of user interfaces, both for experts and operators has been developed, with user feedback and maintenance issues as key points. Operational experience with the full RF chain, including the low level system, the beam control, the synchronisation system and optical fibres distribution is presented. Last but not least overall performance and reliability based on experience with beam are reviewed and perspectives for future improvement outlined.
JASRI/SPring-8, Hyogo-ken
RIKEN/SPring-8, Hyogo
The injector of the 8 GeV linac generates an electron beam of 1 nC, accelerates it up to 30 MeV, and compresses its bunch length down to 20 ps. Even slight RF instability in its multi-stage bunching section fluctuates the bunch width and the peak current of an electron beam and it accordingly results in unstable laser oscillation in the undulator section. The acceptable instabilities of the RF fields in the cavities, which permit 10% rms variation of the peak beam current, are only about 0.01% rms in amplitude and 120 fs rms in phase according to beam simulation. The long-term RF variations can be compensated by feedback control of the RF amplitude and phase, the short-term or pulse-to-pulse variations, however, have to be reduced as much as possible by improving RF equipment such as amplifiers. Thus we have carefully designed and manufactured the RF cavities, amplifiers and control systems, giving the highest priority to the stabilization of the short-term variations. Components of the injector will be completed by the end of the April 2010, and the injector will be perfected in the summer 2010. We will present the performance of the completed devices in the conference.
SLAC, Menlo Park, California
BNL, Upton, Long Island, New York
High level applications often suffer from poor performance and reliability due to lengthy initialization, heavy computation and rapid graphical update. Service oriented architecture (SOA) is trying to separate the initialization and computation from applications to distributed service providers. Heavy computation such as beam tracking will be done periodically on a dedicated server and data will be available to client applications at all time. Industrial standard service architecture can help to improve the reliability and maintainability of the service providers. Robustness will also be improved by reducing the complexity of individual client applications.
ELETTRA, Basovizza
The single pass FEL facility FERMI@ELETTRA, in construction at the ELETTRA Synchrotron Radiation Laboratory in Trieste, requires very short electron bunches with a very high beam quality at the entrance of the undulator chain. To linearize the longitudinal phase space before the bunch compression, mitigating the effects of Coherent Synchrotron Radiation (CSR), a 4th harmonic accelerating section (12 GHz) will be installed before the first magnetic chicane. Here an overall description of the X-band system under development is reported.
RIKEN/SPring-8, Hyogo
University of electro-communications, Tokyo
The intensity of SASE generated by undulators is sensitive to the peak intensity fluctuation of an electron bunch. The bunch is formed by velocity bunching in an injector and magnetic bunching in bunch compressors (BC). The peak intensity is sensitive to rf phase and amplitude of off-crest acceleration at injector cavities and 5712 MHz cavities before the BCs. Thus, demanded stabilities of the rf phase and amplitude for stable SASE generation are very tight. These are 0.6 degree (p-p) and 0.06 % (p-p) at the 5712 MHz cavities, respectively. We are constructing a low-level rf (LLRF) system comprising a master oscillator, an optical rf signal transmission system, and a digital rf control system using IQ modulator/demodulator to drive klystrons. To realize the demands, much attention was paid to temperature stabilization for the system. A water-cooled 19-inch rack and a water-cooled cable ducts are employed for almost all part of the system. Temperature stability of the rack was 0.4 K (p-p) even though outside was 4 K (p-p). The phase and amplitude stabilities of the LLRF modules were measured to be 0.30 degree (p-p) and 0.56 % (p-p). These stabilities are sufficient for our demands.
SLAC, Menlo Park, California
The Linac Coherent Light Source (LCLS) at SLAC is an x-ray Free Electron Laser with wavelengths of 0.15 nm to 1.5 nm. The electron beam stability is important for good lasing. While the transverse jitter of the beam is about 10-20% of the rms beam sizes, the jitter in the longitudinal phase space is a multiple of the energy spread and bunch length. At the lower energy of 4.3 GeV (corresponding to the longest wavelength of 1.5 nm) the relative energy jitter can be 0.125%, while the rms energy spread is with 0.025% five times smaller. An even bigger ratio exists for the arrival time jitter of 50 fs and the bunch duration of about 5 fs (rms) in the low charge (20 pC) operating mode. Although the impact to the experiments is reduced by providing pulse-by-pulse data of the measured energy and arrival time, it would be nice to understand and mitigate the root causes of this jitter. The thyratron of the high power supply of the RF klystrons is one of the main contributors. Another suspect is the multi-pacting in the RF loads. Phase measurements down to 0.01 degree (equals 10 fs) along the RF pulse were achieved, giving hints to the impact of the different sources.
CELLS-ALBA Synchrotron, Cerdanyola del Vallès
European industry has participated in the LHC Project for technology development, component design and system construction. A good relationship in academia-industry collaboration has led to successful results for the project. Industry plays an important role for component design, manufacture and system construction in the XFEL project. The long history of academia-industry collaboration in the accelerator field in Europe is presented.
SLAC, Menlo Park, California
The recent progress at the SPEAR3 were the increase in stored current from 100 mA to 200 mA maximum and the top-off injection to allow beamlines to stay open during injection. Presently the booster injects 3.0 GeV beam to SPEAR3 three times a day. The stored beam decays to about 150 mA between the injections. The growing user demands are to increase stored current to the design value of 500 mA, and to maintain it at a constant value within a percent or so. To achieve this goal the booster must inject once every few minutes. For improved injection efficiency, all RF systems at the linac, booster and SPEAR3 need to be phase-locked. These requirements entail a booster RF system upgrade to a scaled down version of the SPEAR3 RF system running at 476.3 MHz with a 1.2 MW cw output power capability. The present booster RF system is basically a copy of the SPEAR2 RF system operating at 358.5 MHz with 80 kW peak power to a 5-cell RF cavity for 1.2 MV gap voltage. We will analyze each subsystem option for their merits within budgetary and geometric space constraints. A substantial portion of the system will come from the decommissioned PEP-II RF stations.
PAL, Pohang, Kyungbuk
The RF system for PLS-II upgrade, of which beam current and emittance are 400 mA and 5.6 nmrad at 3 GeV, becomes much more important compared to PLS. To reduce the HOM intensity in RF cavities for stable beam, a superconducting RF cavity is selected for the PLS-II. The RF system has to compensate beam loss power of 663 kW from 24 bending magnets, 20 insertion devices and other losses by RF HOM and broadband losses along vacuum chambers. For sufficient energy acceptance and lifetime the design RF voltage is 4.5 MV. Three 500 MHz superconducting cavities will be operated from October 2012, following successful commissioning with PLS NC cavities from July 2011. For the 3 SRF cryomodules, a 700 W class He cryogenic system will be prepared in 2011. The design of PLS-II SRF system including cryogenic system will be reported in the paper.
CLASSE, Ithaca, New York
Cornell University has developed and fabricated a SCRF injector cryomodule for the acceleration of a high current, low emittance beam in the Cornell ERL injector prototype. This cryomodule is based on superconducting rf technology with five 2-cell rf cavities operated in the cw mode, supporting beam currents of up to 100 mA. After a rework of this cryomodule in 2009 to implement several improvements, it is now in beam operation again. In this paper we report on latest results and discuss future test plans.
SLAC, Menlo Park, California
Superconducting RF is increasingly important for particle accelerators. A lot of effort has been made in the SRF material research recently, aiming to find the superconducting materials with better performance. We developed a testing system using a resonant cavity with high quality factor and an interchangeable wall for the testing of different materials. The system is capable for high power RF cryogenic test to find the critical magnetic field at different temperature. The facility can be also used on testing the low temperature properties of the normal conducting material. Different Cu, Nb and MgB2 samples have been tested. In this paper, we will present the most recent development of the system, along with a discussion on the recent testing results.
ISIR, Osaka
Hiroshima University, Faculty of Science, Higashi-Hirosima
The 40 MeV L-band electron linac at the Institute of Scientific and Industrial Research, Osaka University is extensively used for various applications on advanced beam sciences including radiation chemistry by means of pulse radiolysis and development of the free electron laser in the THz region. It was constructed in 1975-78 and has been remodeled sometimes for improving its performance. The most recent one was made in 2002-2004 for higher operational stability and reproducibility, resulting in significant advances in the studies. We will report the present status of the linac and results of its performance evaluation.
SLAC, Menlo Park, California
Linac Energy Management (LEM) is a control system program which calculates, and optionally implements, magnet setpoint settings (BDESs) following a change in Energy (such as a change in the number, phase, and amplitude of active klystrons). The change is made relative to those magnets' existing BDES setpoints by a factor encoding the change in energy. LEM is necessary because changes in the number, phase, and amplitude of the active klystrons (the so-called "Klystron complement") change the beam's rigidity, and therefore, to maintain constant optics, one has to change focusing gradients and bend fields. This paper describes the basic process and some of the implementation lessons learned for LEM at the LCLS.
JASRI/SPring-8, Hyogo-ken
RIKEN/SPring-8, Hyogo
Low electro-magnetic noise interference (EMI) is required to the klystron modulator power supply for XFEL/SPring-8 project in order to realize the highly stable beam operation with aid of various feedback loops using high-performance beam monitors. The dominant noise source is the thyratron switching noise, associated with its rapid voltage swing of 50 kV maximum. To suppress the noise leakage, special care was taken to the enclosure design of klystron modulator, i.e., using thick steel plates a monocoque enclosure was fabricated, in which all of the high power circuits was installed. The rapid image current flows on the inner surface, thus EMI was minimized. A special co-axial feed-though was developed for filtering the conducted noise on power line for thyratron and klystron heaters. In this presentation, we will report the details of the devices and the results of the noise suppression.
University of Cassino, Cassino
INFN-Pisa, Pisa
The possible issues related to the use of last-generations Insulated Gate Bipolar Transistors (IGBTs) switches into a Marx-topology klystron modulator are discussed. Experimental results obtained from two cells Marx prototypes using two different solutions, including single device and series connected devices both hard-switched, are presented. The use of single high voltage device per cell allowed us to obtain lower on-state voltage drop but much slower switching times. On the other side the series connection of lower voltage IGBTs results in much faster commutations and lower devices costs accompanied by a larger on state voltage drop.
RIKEN/SPring-8, Hyogo
XFEL/SPring-8 will use 72 line type modulator pulse-power supply for 66 C-band klystrons, 4 S-band, one L-band and pulsed 500 kV electron gun. In order to make the size smaller to fit the space available in the high gradient C-band accelerator, we have developed all in one box design of modulator. Using metal monocok design, filled with oil, it becomes possible to fit all circuitry: PFN, thyratron, pulse transformer, klyston socket, and protection circuit into a metal box of W 1m x L 1.7m x H 1m, which provides strong support for massive klystron and solenoide with lead shield and functions as superior EM shiled. We developed high precission HV charger for PFN, which has stability better than 100 ppm.pp. Modulator and PFN chargers are under mass prodution.
KEK, Ibaraki
This paper describes a long-pulse 1.3 GHz klystron modulator that was recently developed for the Superconducting RF Test Facility (STF) at High Energy Accelerator Research Organization (KEK). The modulators is a direct-switched-type design with a 1:15 step-up transformer and a bouncer circuit to compensate for the output pulse droop within ±0.5%; it can drive a klystron with up to 10 MW peak power, 1.5 ms rf pulse width, and up to 5 pps repetition rate. The main features of this modulator are the use of four 50 kW switching power supplies in parallel to charge the storage capacitors to 10 kV, self-healing-type capacitor to realize a compact storage capacitor bank, and a highly reliable IGBT switch which enables elimination of a crowbar circuit. Design considerations and its performance are presented. An IEGT (Injection Enhanced Gate Transistor) switch, composed of six series devices with a rating of 4.5 kV and 2100 A-DC, has been also developed and tested for R&D to realize a compact modulator.
Diversified Technologies, Inc., Bedford, Massachusetts
Diversified Technologies, Inc.(DTI) has delivered a fully solid state Klystron Power Supply for the ISIS Front End Test Stand to Rutherford Appleton Laboratory in the UK. The new pulsed power supply drives a Toshiba E3740A klystron in preparation for construction of a system to demonstrate high quality intense chopped beams. DTI's system represents a significant advance in solid-state high power accelerator technology based upon a hard switch developed for the US Department of Energy (DOE) to meet similar requirements for the International Linear Collider(ILC). The system includes two 220 kW switching power supplies, a 110 kV solid state hard switch pulse modulator, mod anode and filament power supplies, klystron fault protection, and interfaces to the ISIS controls. This paper will address the design and construction of the KPS system, as well as test results from the installation at RAL in May 2009.
SLAC, Menlo Park, California
SLAC National Accelerator Laboratory has built and is currently operating a first generation prototype Marx klystron modulator to meet ILC specifications. Under development is a second generation prototype, aimed at improving overall performance, serviceability, and manufacturability as compared to its predecessor. It is designed around 32 cells, each operating at 3.75 kV and correcting for its own capacitor droop. Due to the uniqueness of this application, high voltage gate drivers needed to be developed for the main 6.5 kV and droop correction 1.7 kV IGBTs. The gate driver provides vital functions such as protection of the IGBT from over-voltage and over-current, detection of gate-emitter open and short circuit conditions, and monitoring of IGBT degradation (based on collector-emitter saturation voltage). Gate drive control, diagnostic processing capabilities, and communication are digitally implemented using an FPGA. This paper details the design of the gate driver circuitry, component selection, and construction layout. In addition, experimental results are included to illustrate the effectiveness of the protection circuit.
CERN, Geneva
University of Oslo, Oslo
The physics experiments at CLIC will require that the machine scans lower than nominal centre-of-mass energy. We present different options to achieve this and discuss the implications for luminosity and the machine design.
CERN, Geneva
INFN/LNF, Frascati (Roma)
CEA, Gif-sur-Yvette
Uppsala University, Uppsala
The objective of the CLIC Test Facility CTF3 is to demonstrate the key feasibility issues of the CLIC two-beam technology: the efficient generation of a very high current drive beam and its stable deceleration in 12 GHz resonant structures, to produce high-power RF pulses and accelerate the main beam with an accelerating gradient of 100 MV/m. The construction and commissioning of CTF3 has taken place in stages from 2003. Many milestones had already been reached, including the first demonstration at the end of 2009 of a factor 2 x 4 re-combination of the initial drive beam pulse, thus reaching a beam current of 25 A. In this paper we summarise the commissioning highlights and the issues already validated at the earlier stages. We then show and discuss the latest results obtained, in view of the completion of the CLIC feasibility demonstration due for the end of 2010.
SLAC, Menlo Park, California
A Marx-topology klystron modulator is under development for the International Linear Collider (ILC) project*. It is envisioned as a lower cost, smaller footprint, and higher reliability alternative to the present, bouncer-topology, baseline design. The application requires 120 kV (±0.5%), 140 A, 1.6 ms pulses at a rate of 5 Hz. The Marx constructs the high voltage pulse by combining, in series, a number of lower voltage cells. The Marx employs solid state elements; IGBTs and diodes, to control the charge, discharge and isolation of the cells. Active compensation of the output is used to achieve the voltage regulation while minimizing the stored energy. The developmental testing of a first generation prototype, P1, has been completed. This modulator has been integrated into a test stand with a 10 MW L-band klystron, where each is undergoing life testing. Development of a second generation prototype, P2, is underway. The P2 is based on the P1 topology but incorporates an alternative cell configuration to increase redundancy and improve availability. Status updates for both prototypes are presented.
* ILC Reference Design Report, http://www.linearcollider.org/cms/?pid=1000437
INFN/LNF, Frascati (Roma)
Istituto Nazionale di Fisica Nucleare, Milano
Rome University La Sapienza, Roma
ENEA C.R. Frascati, Frascati (Roma)
In the frame of the SPARC-X project, the energy of the Photo-Injector SPARC, in operation at INFN-LNF, will be upgraded from 180 to 250 MeV by replacing a low gradient S-band traveling wave accelerating section with two C-band units, designed and developed at LNF. The new system will consist of a 50 MW klystron, supplied by a pulsed modulator, to feed the high gradient C-band structures through a RF pulse compressor. This paper deals with the design of the full system, the C-band R&D activity and study of the related beam dynamics.
RIKEN/SPring-8, Hyogo
JASRI/SPring-8, Hyogo-ken
C-band (5712 MHz) accelerator is used as the main accelerator of the XFEL in SPring-8. Since the C-band generates a high accelerator gradient, as high as 35 MV/m, the total length of the 8-GeV accelerator fits within 400 m, including the injector and three bunch compressors. We use 64 C-band rf units, which consists of 128 accelerating structures, 64 rf pulse compressors, 64 klystrons, waveguide components, etc. Mass-production of these high power rf components has been almost completed. Production quality is confirmed by the high power rf test. Installation of the C-band components started in August 2009. So far, about half of the components have been installed on schedule. The accelerating structures are aligned with about 0.1 mm accuracy. By the date of the IPAC'10 conference, we will almost complete the installation. In this presentation, we will report the construction status.
RIKEN/SPring-8, Hyogo
JASRI/SPring-8, Hyogo-ken
We report the high power rf test results of C-band accelerator system for X-ray free electron laser (XFEL) in SPring-8 site. In XFEL main accelerator, 64 C-band systems will be used in total, whose components are under mass production at several industries in Japan. We performed high power RF test with three sets of the mass-produced components in XFEL test bunker. We operate the C-band components with the accelerating gradient, as high as 40 MV/m. We measured the high voltage breakdown rate and the dark current emission.
KEK, Ibaraki
In order to improve the capture efficiency of the positron produced at the target in present KEKB Injector linac, a new project has just started to utilize L-band (1298MHz) RF. The present S-band (2856MHz) capture cavities and successive three RF units are to be replaced by those of L-band. The specifications of the L-Band system should fulfill the demands of a positron damping ring downstream which is also to be under study for super KEKB project. Besides the whole design work of the system, our present ongoing work is rather concentrated on establishing L-Band RF source and accelerating structures.
POSTECH, Pohang, Kyungbuk
KAPRA, Cheorwon
NFRI, Daejon
An intense L-band electron linac is now being commissioned at ACEP (Advanced Center for Electron-beam Processing in Cheorwon, Korea) for irradiation applications in collaboration with POSTECH (Pohang University of Science and Technology) and KAPRA (Korea Accelerator and Plasma Research Association). It is capable of producing 10-MeV electron beams with average 30-kW. For a high-power capability, we adopted the L-band traveling-wave structure operated with a 2π/3 mode. The RF power is supplied by the pulsed 25-MW and average 60-kW klystron with the matched pulse modulator and the inverter power supplies. The accelerating gradient is 4.2 MV/m with the beam current of 1.45 A which is fully beam-loaded condition. The solenoidal magnetic field is 700 Gauss to focus the electron beam and suppress the BBU instability. In this paper, we present commissioning status with details of the accelerator system.
CERN, Geneva
In the CTF3 accelerator, the RF produced by each of ten 3 GHz klystrons goes through waveguides, RF pulse compressors and splitters. The RF phase and power transformation of these devices depend on their temperature. The quantitative effect of the room temperature variation on the RF was measured. It is the major source of undesired changes during the CTF3 operation. An RF phase-loop and a compressor temperature stabilization are developed to suppress the phase fluctuation and the power profile change due to the temperature variation. The implementation is transparent for operators it does not limit anyhow the flexibility of RF manipulations. Expected and measured suppression characteristics will be given. As well RF measurement dependence on the temperature will be mentioned.
LLNL, Livermore, California
SLAC, Menlo Park, California
A Mono-Energetic Gamma-Ray Compton scattering light source is being developed at LLNL. The electron beam for the interaction will be generated by a X-band RF gun and LINAC at the frequency of 11.424 GHz. High power RF in excess of 500 MW is needed to accelerate the electrons to energy of 250 MeV or greater. Two high power klystrons, each capable of generating 50 MW, 1.5 msec pulses, will be the main RF sources for the system. These klystrons will be powered by state of the art solid-state high voltage modulators. A RF pulse compressor, similar to the SLED II pulse compressor, will compress the klystron output pulse with a power gain factor of five. For compactness consideration, we are looking at a folded RF line. The goal is to obtain 500 MW at output of the compressor. The compressed pulse will then be distributed to the RF gun and to six traveling wave accelerator sections. Phase shifter and amplitude control are located at the RF gun input and additional control points along the LINAC to allow for parameter control during operation. This high power RF system is being designed and constructed. In this paper, we will present the design, layout, and status of this RF system.
SLAC, Menlo Park, California
KEK, Ibaraki
INFN/LNF, Frascati (Roma)
We report results of ongoing high power tests of single cell standing wave structures. These tests are part of an experimental and theoretical study of rf breakdown in normal conducting structures at 11.4 GHz. The goal of this study is to determine the accelerating gradient capability of normal-conducting rf powered particle accelerators. The test setup consists of reusable mode-launchers and short test structures powered by SLAC's XL-4 klystron. We have tested structures of different geometries, cell joining techniques, and materials, including hard copper alloys and molybdenum. We found that the behavior of the breakdown rate is reproducible for different structures of the same geometry and material. The breakdown rate dependence on peak magnetic fields is stronger than on peak surface electric fields for structures of different geometries.
SLAC, Menlo Park, California
The phenomenon of rf breakdown presents a technological limitation in the application of high-gradient particle acceleration in normal conducting rf structures. Attempts to understand the onset of this phenomenon and to study its limits with different materials, cell shapes, and pulse widths has been driven in recent years by linear collider development. One question of interest is the role magnetic field plays relative to electric field. A design is presented for a single, non-accelerating, rf cavity resonant in two modes, which, driven independently, allow the rf magnetic field to be increased on the region of highest electric field without affecting the latter. The design allows for the reuse of the cavity with different samples in the high-field region. Available high-power data will also be presented.
LNLS, Campinas
In November 2010, LNLS plans to replace the two 50 kW UHF klystron valves which currently provides power to the RF cavities installed in the storage ring. Thanks to a close collaboration with the Syncrotron Soleil started in 1999, LNLS adapted the characteristics of the French project to 476 MHz. The choice of the transistor, the design of the combiners and details on power supplies will be reported, as well as the power tests performed with the two amplifiers using a resistive load.
DESY, Hamburg
BINP SB RAS, Novosibirsk
For the European XFEL project horizontal multi-beam klystrons will be installed in the XFEL tunnel and will be connected to the double tank pulse transformers. Both, the klystron and pulse transformer need for the normal operation to be filled with oil. To avoid the possible oil leakage during connection of the klystron and transformer tank inside tunnel, the connection module (CM) was proposed. The CM will be mounted on the support platform of the klystron and through the tube socket connected to the guns electrodes outside of the tunnel and will transported to the tunnel together with klystron. The connection to the pulse transformer tank will be done only with HV cable, because the CM has inside it the filament transformer. To reduce the weight and volume of the oil the design of filament transformer was done as high frequency coaxial one with coupling factor of 0.58 and working frequency about 1 kHz. The CM has the built-in current and voltage monitors. In this paper we give an overview about design and test result of the CM together with klystron.
KEK, Ibaraki
ERL (Energy Recovery Linac) of 5GeV energy is a future plan in KEK and in order to study the technical feasibility, construction of a compact ERL machine (cERL) is considered. Beam energy and current of cERL are 245MeV and 100mA, respectively. As 1.3 GHz frequency and super conducting cavity are chosen for the RF system, similar technology with KEK STF is employed. From 2008, KEK started the preparation of cERL and one RF unit of injector linac is introduced in this fiscal 2009. A new cw klystron of 300kW out put power, 150kW Y-type circulator and high power water load were developed in FY2009. DC power supply was under manufacturing. Preliminary test of HLRF and the high power couplers are scheduled in the Photon Factory site by making use of the old DC power supply. For main accelerator, we also introduced a 30kW IOT and a 35kW klystron and a DC power supply. At the same time, cERL is determined to be constructed in the East Counter Hall in KEK and the design layout is preceded. In this paper, the recent RF source development of cERL is described. Layout of the east counter hall, where cERL is constructed, is progressed and shown in this report.
CERN, Geneva
SLAC, Menlo Park, California
CEA, Gif-sur-Yvette
IAP/RAS, Nizhny Novgorod
The CLIC RF frequency has been changed in 2008 from the initial 30 GHz to the European X-band 11.9942 GHz permitting beam independent power production using klystrons for CLIC accelerating structure testing. A design and fabrication contract for five klystrons at that frequency has been signed by different parties with SLAC. France (CEA Saclay) is contributing a solid state modulator purchased in industry to the CLIC study. RF pulses over 120 MW peak at 230 ns length will be obtained by using a novel SLED I type pulse compression scheme designed and fabricated in Nizhny Novgorod, Russia. The X-band power test stand has been installed in the CLIC Test Facility CTF3 for independent structure and component testing in a bunker, but allowing, in a later stage, for powering RF components in the CTF3 beam lines. The design of the facility, results from commissioning of the RF power source and the performance of the Test Facility are reported.
Lancaster University, Lancaster
Cockcroft Institute, Lancaster University, Lancaster
2D Photonic crystals (PC) with defects can act as standing-wave resonators, which offer benefit of high mode selectivity for building novel RF sources. We introduce our work on designing two-cavity single-beam and multi-beam klystrons using triangular lattice metallic PCs. We present the cold test results of the stub-coupled single-beam structure, which show that at resonance a very low reflection can be obtained, and the waves are well confined. We also present bead-pull measurement results of field strengths in the defect, using modified perturbation equation for small unit dielectric cylinder, which are in very good agreement to numerical results. A 6-beam klystron cavity is designed as a 6-coupled-defect structure with a central stub, which only couples to the in-phase mode at the lowest frequency. Finally, we present a feasibility discussion of using this multi-defect PC structure to construct an integrated klystron-accelerator cavity, along with numerical results showing a peak acceleration field of 22MV/m can be achieved.
SLAC, Menlo Park, California
The Linac Group at SLAC is actively pursuing a broad range of R&D to improve the reliability and reduce the cost of the L-band (1.3 GHz) rf system proposed for the ILC linacs. Current activities include the long-term evaluation of a 120 kV Marx Modulator driving a 10 MW Multi-Beam Klystron, design of a second-generation Marx Modulator, testing of a sheet-beam gun and beam transport system for a klystron, construction of an rf distribution system with remotely-adjustable power tap-offs, and development of a system to combine the power from many klystrons in low-loss circular waveguide where it would be tapped-off periodically to power groups of cavities. This paper surveys progress during the past few years.
SLAC, Menlo Park, California
A 12 GHz 50MW X-band klystron is under development at the SLAC National Accelerator Laboratory Klystron Department. The klystron will be fabricated to support programs currently underway at three European Labs; CERN, PSI, and INFN Trieste. The choice of frequency selection was due to the CLIC RF frequency changing from 30 GHz to the European X-band frequency of 11.9942 GHz in 2008. Since the Klystron Department currently builds 50MW klystrons at 11.424 GHz known collectively as the XL4 klystrons, it was deemed cost-effective to utilize many XL4 components by leaving the gun, electron beam transport, solenoid magnet and collector unchanged. To realize the rf parameters required, the rf cavities and rf output hardware were necessarily altered. Some improvements to the rf design have been made to reduce operating gradients and increase reliability. Changes in the multi-cell output structure, waveguide components, and the window will be discussed along with testing of the devices. Five klystrons known as XL5 klystrons are scheduled for production over the next two years.
SLAC, Menlo Park, California
The SLAC National Accelerator Laboratory Klystron Department is developing a 10 MW, 5 Hz, 1.6 ms, 1.3 GHz plug-compatible Sheet-Beam Klystron as a less expensive and more compact alternative to the ILC baseline Multiple-Beam Klystron. Earlier this year a beam tester was constructed and began test. Device fabrication issues have complicated the analysis of the data collected from an intercepting cup for making beam quality measurements of the 130 A, 40-to-1 aspect ratio beam. Since the goal of the beam tester is to confirm 3d beam simulations it was necessary to rebuild the device in order to mitigate unwanted effects due to imperfect focusing construction. Measurements are underway to verify the results of this latest incarnation. Measurement will then be made of the beam after transporting through a drift tube and magnetic focusing system. In the klystron design, a TE oscillation was discovered during long simulation runs of the entire device which has since prompted two design changes to eliminate the beam disruption. The general theory of operation, the design choices made, and results of testing of these various devices will be discussed.
RISE, Tokyo
KEK, Ibaraki
ISIR, Osaka
AIST, Tsukuba, Ibaraki
At Waseda University, we have been studying a high quality electron beam generation and its application experiments with Cs-Te photocathode RF-Gun. We have already succeeded in generating a stable high-charged single-bunch electron beam. To generate more intense electron beam, we designed a multi-bunch electron linac and developed the multi-pulse UV laser which irradiates to the cathode. The target values of the number of electron bunch and bunch charges are 100 bunches/train and 800 pC/bunch, respectively. In addition, we adopted the method of the amplitude modulation of the incident RF pulse to the S-band klystron in order to compensate the energy difference in each bunch because of the slow rise time of acceleration voltage in cavity and beam loading effect in the accelerating structure. In this conference, we will report design properties of our multi-bunch electron linac, the results of the multi-bunch electron beam diagnosis and the energy difference compensation using the RF amplitude modulation method.
Siemens AG, Healthcare Technology and Concepts, Erlangen
The concept of a compact particle accelerator capable of delivering accelerating fields upto 100MV/m using a direct drive RF LINAC is explored. Such a machine consists of a succession of RF cavities with the RF power being supplied from a ring of solid state RF transistors placed around the cavity circumference. To achieve the required accelerating fields 3 core technologies are presented. (i) The solid-state transistors are used to drive the wall currents in the cavities so achieving a direct drive of the cavity. This allows unprecedented powers to be reached (>GW class) as well as enabling independent phase control of the individual cavities. Central to the implementation is the design of the RF drive consisting of distributed SiC vJFET modules delivering 750kA at 800V per cavity. (ii) A High Gradient Insulator structure is required to hold an electric field of >100MV/m. In contrast to a conventional HGI, the concept utilizes a vacuum insulated grading layer structure. (iii) A chopper and injection system allow the formation of proton bunches with a spatial emissivity <3ns and an injection field of up to 100MV/m.