W. Roybal

Modifications to harmonic current bunching of electron beams from RF cavities due to radial boundary conditions

Harmonic current bunching of an electron beam after it passes through a standing-wave RF cavity is well known, with analytic solutions in both the transversely infinite space-charge regime and the ballistic regime. The Wallander formula is often used between these regimes for finite-sized beams, with pretty good accuracy. These formulas describe simple space-charge modes that are launched by the RF cavity, and assume a single space-charge reduction factor. However, upon close examination these formulas do not provide the complete story. Additional space-charge modes are introduced by the radial boundary conditions, and the actual total harmonic current is modified somewhat from the simple formulas, and includes beating effects between the different modes, even in the small-signal, or linear, regime. This beating leads to mode mixing and eventual damping, which also affects the long-term propagation of longitudinal space-charge waves in accelerator structures.

Results and lessons learned from conditioning 1 MW CW 350 MHz coaxial vacuum windows

The reliability of the radio frequency (RF) windows on the Low Energy Demonstration Accelerator (LEDA) is critical to the success of the Accelerator Production of Tritium Program (APT). On the APT accelerator there will be over 1,000 windows, each passing on the order of 250 kW of CW RF power. This power level is well above power levels historically used in RF windows. Based on the high-power RF test results of the RF window prototypes from vendors, the coaxial windows made by EEV Ltd. of Chelmsford, England, were selected for LEDA. This paper describes the high-power RF testing of the 16 EEV coaxial windows. The RF window diagnostic equipment, data acquisition system and test stand are described. The results of the high power RF testing of the windows are presented. The successes and failures in the conditioning, manufacturing and testing techniques of the windows are presented. The conditioning timeline, power profile and the conditioning waveform are also discussed.

Dynamics of Retrograde Electrons Returning From the Output Cavity in Klystrons

Loss of efficiency and tube oscillations have been attributed to electrons returning from the output cavity in klystrons due to excessive output cavity voltages. It is generally believed that the retrograde electrons lead to a relatively large harmonic current component in the input cavity, which overwhelms the input drive. Here, for the first time, detailed simulations describing the dynamics of retrograde particles for a nominal klystron design and cases with the shunt impedance of the output cavity increased, which show persistent harmonic bunching induced by the penultimate cavity, are presented. The maximum retrograde harmonic current is comparable to the current induced in the input cavity without the retrograde particles and can significantly influence the overall klystron behavior

The SNS linac high power RF system design, status, and results

The Spallation Neutron Source being built at the Oak Ridge National Lab in Tennessee requires a 1 GeV proton linac. Los Alamos has responsibility for the RF systems for the entire linac. The linac requires 3 distinct types of RF systems: 2.5-MW peak, 402.5 MHz, RF systems for the RFQ and DTL (7 systems total); 5-MW peak, 805 MHz systems for the CCL and the two energy corrector cavities (6 systems total); and 550-kW peak, 805 MHz systems for the superconducting sections (81 systems total). The design of the SNS Linac RF system was presented at the 2001 Particle Accelerator Conference in Chicago. Vendors have been selected for the klystrons (3 different vendors), circulators (1 vendor), transmitter (1 vendor), and high power RF loads (3 different vendors). This paper presents the results and status of vendor procurements, test results of the major components of the Linac RF system and our installation progress.

An overview of the Low Energy Demonstration Accelerator (LEDA) project RF systems

Successful operation of the Accelerator Production of Tritium (APT) plant will require that accelerator downtime be kept to an absolute minimum. Over 230 separate 1 MW RF systems are expected to be used in the APT plant, making the efficiency and reliability of these systems two of the most critical factors in plant operation. The Low Energy Demonstration Accelerator (LEDA) being constructed at Los Alamos National Laboratory will serve as the prototype for APT. The design of the RF systems used in LEDA has been driven by the need for high efficiency and extremely high system reliability. We present details of the high voltage power supply and transmitter systems as well as detailed descriptions of the waveguide layout between the klystrons and the accelerating cavities. The first stage of LEDA operations will use four 1.2 MW klystrons to test the RFQ and supply power to one test stand. The RFQ will serve as a power combiner for multiple RF systems. We present some of the unique challenges expected in the use of this concept.

Green's Function Simulation of Space–Charge Effects in Electron Beams

In this paper, a new numerical model for space-charge forces in electron beams based on a Greens function approach is described. In this model, the beam is simulated by a series of rings with nonzero thickness and length. The space-charge force on a ring is found by summing over all the interactions with all the other rings, where each interaction force is integrated over the entire volume of each source rings. With proper beam initialization of the simulation parameters, the rings can perfectly form the electron beam, leading to a very smooth and accurate calculation of the space-charge fields. The space-charge fields calculated this way can be more accurate than those found with particle-in-cell (PIC) calculations. The fields can also be distributed onto a mesh, as in a PIC calculation, leading to equivalent accuracy with greatly reduced simulation times. The accuracy of this type of model is demonstrated by comparing the harmonic-current evolution from an RF gap for a transversely cold confined beam to analytic theory and we show its utility with a large-signal annular beam klystron simulation using this technique

Experimental evaluation of 350 MHz RF accelerator windows for the Low Energy Demonstration Accelerator

Radio frequency (RF) windows are historically a point where failure occurs in input power couplers for accelerators. To obtain a reliable, high-power, 350 MHz RF window for the Low Energy Demonstration Accelerator (LEDA) project of the Accelerator Production of Tritium program, RF window prototypes from different vendors were tested. Experiments were performed to evaluate the RF windows by the vendors to select a window for the LEDA project. The Communications and Power Industries, Inc. (CPI) windows were conditioned to 445 kW in roughly 15 hours. At 445 kW a window failed, and the cause of the failure will be presented. The EEV windows were conditioned to 944 kW in 26 hours and then tested at 944 kW for 4 hours with no indication of problems.

Characterization of a klystrode as a RF source for high-average-power accelerators

The klystrode is a relatively new type of RF source that has demonstrated DC-to-RF conversion efficiencies in excess of 70% and a control characteristic uniquely different from those for klystron amplifiers. The different control characteristic allows the klystrode to achieve this high conversion efficiency while still providing a control margin for regulation of the accelerator cavity fields. We present test data from a 267-MHz, 250-kW, continuous-wave (CW) klystrode amplifier and contrast this data with conventional klystron performance, emphasizing the strengths and weaknesses of the klystrode technology for accelerator applications. We present test results describing that limitation for the 250-kW, CW klystrode and extrapolate the data to other frequencies. A summary of the operating regime explains the clear advantages of the klystrode technology over the klystron technology.

Resonant ring for testing of accelerator RF windows

A klystron-driven resonant ring has been designed and assembled at the Los Alamos National Laboratory for use in the Accelerator Production of Tritium Project (APT). The ring was built to test RF windows for the 700 MHz section of the APT accelerator. The ring has been designed to apply an effective power of approximately 1 MW on test windows. Details of ring design, operation and performance are presented.

Quantitative evaluation of magnet hysteresis effects at LANSCE with respect to magnet power supply specifications

The proton beam in the LANSCE accelerator is guided and focused almost exclusively by electromagnets. Magnet hysteresis has had significant impacts on the tuning of the LANSCE accelerator. Magnet hysteresis can also have an impact on magnet power supply (MPS) control, regulation and repeatability requirements. To date, MPS performance requirements have been driven by the requirements on the magnetic fields as determined by the accelerator physicists. Taking hysteresis effects into account can significantly change MPS requirements, as some requirements become more stringent and some are found to be over specified. Overspecification of MPS requirements can result in significant increases in MPS cost. Conversely, the use of appropriate MPS requirements can result in significant cost savings. The LANSCE accelerators more than three decades of operation provide a wide variety of MPS technologies and operational experience. We will survey the LANSCE MPS history and determine how performance specifications can be refined to both reduce costs and improve the operators abilities to control the magnetic fields.