Search

We have begun a two-year study on the problem of designing high-power pulse modulators for the induction cavities of a recirculating heavy-ion linac driver for inertial fusion. Recent studies have shown the potential of a recirculator for major cost reduction. However, important issues exist including beam stability and control, and severe tolerances. We are addressing the need for efficient, cost-effective modulators. Our example conceptual design uses a 3 ms, linearly rising dipole field and constant energy/turn for 96 turns. The modulator pulses must vary in duration and spacing by a factor of 4. For each 264 kV module we propose to use 50–100 Metglas cores, each driven by a number recently developed, solid-state, on-off switches. Several circuit methods are being studied, including delayed triggering of groups of switches and selective charging of pulse networks. Here we report on initial circuit studies and on preliminary experiments to characterize promising switches and related components. Plans are underway to assemble a scaled accelerating module and simulate an ion beam with a separte high-voltage modulator.

The authors report on initial experimental results of a high-efficiency L-band proof-of-principle microwave amplifier experiment. The prototype under experimental investigation employs a 120 kV, 20A, 5mm diameter, 4 {mu}sec pencil beam. The pencil beam is made to spiral through the interaction with a rotating mode microwave cavity. Efficient energy extraction from the spiral beam is obtained in an output cavity also supporting a rotating mode. The system is immersed in a uniform magnetic field with its amplitude properly adjusted to permit synchronization between the temporal oscillations of the spiral beam and the fields of the rotating modes. This device should be capable of providing coherent, efficient, multimegawatt microwave radiation for a variety of applications appropriate to its frequency range.

This paper presents several new approaches for producing bunched electron beams. The need for higher power injectors and microwave sources for various applications has in turn created the need for higher emission current density electron guns with a means for bunching at the gum. Four different cathode/gun approaches are presented. The first three utilize external methods of bunching. They are based on: field emission, photoemission and ferroelectric emission. The fourth is based on secondary emission and utilizes self-bunching. The latter is emphasized in this presentation. This self-bunching electron gun produces naturally bunched electron beams with the use of microwave fields to generate secondary emission and multipacting of electrons. This type of electron gun is called the Micro-Pulse Gun (MPG). One wall of an RF cavity is made opaque to an input RF electric field but is partially transparent to electrons in order to extract the beam. FMT has studied using simulation codes the resonant bunching process which gives rise to high current densities (0.01-5 kA/cm/sup 2/), high charge bunches (up to 500 nC/bunch), and short pulses (1-100 ps) for frequencies from 1 to 12 GHz. The beam pulse width is nominally /spl sim/5% of the RF period. Experimental verification of self-bunching, cold electron emission, long life, and tolerance to contamination has been achieved. Measurements of current density, bunch length and lifetime are presented for the MPG.

A novel concept for a high-current-density micro-pulse electron gun has been studied. The concept employs the resonant amplification of an electron current by secondary electron emission in an RF cavity. We have studied this "multipacting" process in detail, including space charge, and have found the parameters necessary to make use of the phenomenon to produce a micro-pulsed electron beam. One wall of the cavity is made partially transparent to electrons but opaque to the input RF electric field. It is shown from self-consistent analytic theory and two-dimensional, fully relativistic, electromagnetic particle-in-cell (PIC) simulations that the current density scales with frequency cubed. The natural "bunching", or resonant phase selection of the particles gives rise to high current densities (20-5600 A/cm/sup 2/), high charge bunches (up to 100 nC/bunch for a solid beam and up to 1000 nC for a hollow beam), and short pulses (36-3.5 ps) for frequencies from 1.3 to 8 GHz. The beam pulse width is found to be typically 4.6% of the RF period. Beam extraction through the transparent wall was studied using a 2-1/2 dimensional PIC code. Good beam transmission (52-85%) with low normalized emittance (9-18 mm-mrad) could be achieved. The best normalized emittance per charge is 3 mm-mrad/nC. Tuning to the resonant parameters has been shown to be very tolerant. Electrical breakdown is not a problem, and materials are available to satisfy the secondary emission yield requirements for this device. Applications are accelerator injectors and RF generation at a high harmonic of the fundamental frequency. >

A proof-of-principle experiment has been assembled and is being tested to validate a novel microwave amplifier concept. The method uses transverse deflection modulation of a collimated electron beam which is converted to axial bunching by a magnet designed to bend the beam 105 degrees . Since the mechanism avoids the well-known problem of longitudinal bunching at relativistic velocities, it is suitable for high-voltage beams. Design considerations and preliminary initial results are presented. The results give confidence that the basic mechanism is valid. It is believed that the present results are limited by a combination of finite beam size effects and space charge, both of which tend to dilute effective bunching. >

We report here on a study of circuits for the accelerating modules in a recirculator. Future requirements are challenging: flat-top (or slightly rising) pulses with varying duration; beam currents as high as a kiloampere; a train of 50-100 pulses with instantaneous repetition rate as high as 50 kHz; and high efficiency. An accelerating module providing 200-400 kV to the beam might house 30-100 cores. Our study uses cores of moderate size - we chose 20.3 cm diameter. The pulsers use tapered pulse-forming networks and provide core reset, 20-/spl mu/s recovery and variable duration. A novel addition is a separate, 150-kV, 200-A modulator used to simulate a constant-current ion beam, allowing a direct measurement of the cell dynamic impedance and the modulator-to-beam conversion efficiency. The basic goal of the project is to explore the use of solid-state switches and low-voltage circuits with efficiency and low cost in mind. In this report we describe our circuits and salient results, then offer comments and conclusions.