ANL, Lemont, Illinois, USA
SLAC, Menlo Park, California, USA
We present the design for a rapid proton energy modulator with radio-frequency (RF) accelerator cavities. The energy modulator is designed as a multi-cell one-meter long accelerator working at 2.856 GHz. We envision that each individual accelerator cavity is powered by a 400 kW low-voltage klystron to provide an accelerating / decelerating gradient of 30 MV/m. We have performed beam dynamics simulations showing that the modulator can provide ± 30MeV of beam energy change, with an energy spread of 3 MeV for a 7 mm long (full length) proton bunch. A prototype experiment of a single cell is in preparation at the Next Linear Collider Test Accelerator (NLCTA) at SLAC.
BNL, Upton, New York, USA
The electron LINAC injector of NSLS-II synchrotron light-source runs both Single-Bunch beam and long Multi-Bunch beam of up to 150 bunches. The key component for achieving this dual injector beam mode support capability is a high-speed rf modulator (or RFM) in the LINAC RF electronics front-end, which performs the necessary rf control and the beam loading compensation of different injection beams. The original LINAC rf electronics front-end successfully supported the machine commissioning and meets the basic needs of the machine operation. The upgrade being pursued is focused on improving the RFM control performance through replacing the current analog implementation in the RFM with a much more capable digital implementation, while still maintaining the necessary control bandwidth that is required for long and short Multi-Bunch beams. A variety of modern COTS rf transmission/reception DSP technology will be incorporated in the new design. The improvement in the reliability of network connection between the RFM’s and their host server is another focus in the upgrade, and the solution includes the adoption of the COTS TCP/IP and other communication protocol offload engines.
BNL, Upton, New York, USA
The NSLS-II synchrotron light-source at BNL uses three S-band, 45MW klystrons in its injector LINAC. At the core of the klystron station design is the novel solid-state switching modulators (or SSM). Compared to the conventional PFN klystron modulators, the main advantages of the SSM include the compact size requiring a smaller footprint in the LINAC, and a very flat top in the produced klystron HV pulses. The flatness of the HV pulses is very important for NSLS-II LINAC that runs multi-bunch beams to keep the beam energy dispersion within the tolerance. The principle of the SSM is fairly simple. It uses a large number of relatively low-voltage switched charging capacitor cells (or SU’s) in parallel. A specially designed, high step-up ratio, pulse transformer in the oil-tank with the same number of primary windings (as SU’s) combines the power from all the SU’s, and steps up to the required ~300kV klystron beam voltage. The operation experience at NSLS-II has proven the performance and reliability of the SSM’s. The BNL Model K2 SSM’s are currently being upgraded to Model K300 to run more powerful, and more cost-effective Canon’s E37302A klystrons.
ANL, Lemont, Illinois, USA
Euclid Beamlabs LLC, Bolingbrook, USA
The RF system of Argonne Wakefield Accelerator (AWA) facility has grown over the years from one RF power station into 4 RF power stations. The demand for RF power keeps growing as the capability of AWA continues to grow. Now the 5th RF station is needed to fulfill the RF power needs of AWA facility. Some details regarding the construction and commissioning of the 5th RF station of AWA facility are documented in this paper.
CCR, San Mateo, California, USA
Fermilab, Batavia, Illinois, USA
CPI, Beverley, Massachusetts, USA
Calabazas Creek Research, Inc., Fermilab, and Communications & Power Industries, LLC, developed a 100 kW peak, 10 kW average, 1.3 GHz, magnetron-based, RF system for driving accelerators. Efficiency varied between 81% and 87%. Phase locking uses a novel approach that provides fast amplitude and phase control when coupled into a superconducting accelerator cavity [1]. The system was successfully tested at Fermilab and produced 100 kW in 1.5 ms pulses at a repetition rate of 2 pps. A locking bandwidth of 0.9 MHz was achieved with a drive signal of 269 W injected through a 4 port circulator. The phase locking signal was 25 dB below the magnetron output power. The spectrum of the phase locked magnetron was suitable for driving accelerator cavities. Phase modulation was demonstrated to 50 kHz (the limit of the available driver source). The average power was limited by available conditioning time. Scaling indicates 42 kW of average power should be achievable. Estimated cost is less than $1/Watt of delivered RF power, exclusive of power supplies or modulators. System design and performance measurements will be presented.
[1] B. Chase, R. Pasquinelli, E. Cullerton, P. Varghese, "Precision Vector Control of a Superconducting RF Cavity driven by an Injection Locked Magnetron," Jou. of Instrumentation, Vol. 10 March 2015.
Arizona State University, Tempe, USA
Yale University, Beam Physics Laboratory, New Haven, Connecticut, USA
Omega-P, Inc., New Haven, Connecticut, USA
Abstract: A future European light source CompactLight is being proposed to extend FEL operation further into the x-ray region than other light sources by using a linac operating at X-band (12 GHz) with a short Ka-band (36 GHz) section for linearizing bunch phase space. The Ka-band system requires a high-power RF amplifier, synchronized with the main X-band source. We report here on design of a third-harmonic klystron amplifier for this application. Our design employs a four-cavity system with a multi-cell extended interaction output cavity. Initial simulation results indicate that more than 10 MW of 36-GHz power can be obtained with an efficiency exceeding 20%, and with 12-GHz drive power of 30 W. A preliminary design for a proof-of-principal experimental test of this concept is described
Yale University, Beam Physics Laboratory, New Haven, Connecticut, USA
Omega-P, Inc., New Haven, Connecticut, USA
Main components, which are the electron gun, cavity-chain, magnetic system, and partially- grounded depressed four-stage collector, of a novel klystron were conceptually designed. This klystron is to deliver 50 kW CW at 952.6 MHz and to serve as a microwave power source for ion acceleration at the Electron Ion Collider (EIC) being developed at Thomas Jefferson National Accelerator Facility. The efficiency is 80%, a number to which the power consumption by the solenoid and filament are already factored in. The tube is a combination of proven technologies put together: it uses multiple beams to have its perveance low to boost beam-power to RF-power efficiency. It uses a partially grounded depressed collector to recover energy thereby increasing the overall efficiency. A low operating voltage of 14kV makes the tube more user-friendly avoiding need for costly modulators and oil insulation. A sectioned solenoid is used to insure superb beam-matching to all components downstream of the electron gun, increasing the tube performances. Details of the components designs will be presented.
CCR, San Mateo, California, USA
CPI, Palo Alto, California, USA
Calabazas Creek Research, Inc. and Communications & Power Industries, LLC are developing multiple beam triodes to produce more than 200 kW of RF power at extremely low cost and efficiencies exceeding 85%. RF power is achieved by installing the triode inside coaxial input and output cavities at the desired frequency. The multiple beam triodes developed in this program will provide RF power from 350 MHz to 700 MHz using the appropriate, tuned, resonant cavities. This program is using eight grid-cathode assemblies to achieve 200 kW with a target efficiency exceeding 80%. A 350 MHz RF source would be approximately 36 inches high, 18 inches in diameter and weigh approximately 150 pounds. This is significantly smaller than any other RF source at this frequency and power level. The gain is limited to approximately 14 dB, so a single beam triode-based source will serve as a driver. The combined cost and efficiency will still exceed the performance of other comparable RF sources, including solid state sources. Design issues, include grid cooling, uniformity of RF electric fields on the grids, and efficiency, will be discussed.
SLAC, Menlo Park, California, USA