SBIR-STTR Award

The limited evolution of RF source technology since its inception and the dependence on high voltage components for generating high power, has kept the price and size of these devices very high. The use of magnetrons to overcome these issues introduces non-linearity, poor frequency stability, inability to control phase, and a very low Mean Time Between Failure (MTBF) rate. A radically different approach that optimizes all aspects of the RF sources is therefore required. The goal of building an economic, compact yet high-power RF source, can be achieved through massive integration of highly efficient small klystron sources. These can leverage novel, cost-effective modulators, with fast rise and fall times and built with off-the-shelf electronic components. We propose such a klystron design for achieving the $2/kW goal. Our company, TibaRay Inc. is actively developing a RF component including RF sources for use in our novel medical accelerator application. These are based on recent technology developed at SLAC that involve new approaches to designing highly efficient resonant cavities and for manufacturing them. The proposed RF source will be designed to achieve 70% efficiency while maintaining a compact form factor and light weight. It will operate at X-band frequencies that area a standard for medical and industrial applications as well as for large research accelerators. In Phase I we plan to optimize every design aspect of the klystron (beam voltage, current, cathode diameter, etc). We will also simulate and study the electromagnetic, RF-beam and thermal behavior of the device. Emphasis will be placed on manufacturing ease and reliability. This work will serve as the basis for Phase II where, if approve, we will develop these concepts into commercially available RF sources. We will develop the plan for the production and sale of these devices. Starting from industrial and medical accelerators to large national and international accelerator facilities, there is a significant need for lowering the cost of the RF sources and making them light and compact. Currently, the cost of the high-power RF sources well exceeds the cost of the rest of accelerator facilities. Our proposed solution holds the promise to become a ubiquitous solution for majority of such accelerator systems.

We proposed to build a high-efficiency, high-power, cost-efficient RF source based on the work performed under Phase I of this program. High peak power RF sources are the costliest component of accelerator systems and the requirement is for the cost to go down to $2/W average power (or $2/kW peak power for duty factor of 0.1%) Our approach involves producing high output power for the end system by performing end-to-end optimization of each sub-system. This involves combining a novel self-consistent global optimization algorithm with a choice of hardware components that are mass produced and work at low voltages. We took a multi-pronged approach to minimizing the cost of the individual sub-systems. In Phase I we have optimized the electrostatic focusing to get higher efficiencies. We started with Eizel lenses alternating with focusing cavities, but we realized that we can also use asymmetric lens to boost the energy of the bunched beam with improved beam confinement system efficiency. By combining magnetic focusing at the initial stages with electrostatic focusing and post acceleration, we expect to reach the targeted 1.1 MW power. We designed a flat cathode that reduces the cost of the electron gun by eliminating the curved shape that is expensive to produce. We have simulated the electron trajectories and their interaction with the first cavity to optimize the location and shape of this cavity and the gun parameters. The RF window design was adapted to accept an off-the-shelf mass-produced ceramic window for RF power output. Commercially available low-cost solid-state RF IC was selected for the low power input RF. An efficient and compact klystron body with optimized cavities was designed. An efficient beam dump with optimized shielding will further add to the efficiency in cost and form-factor. We have developed the necessary electrical and mechanical designing and manufacturing capabilities in-house. By designing the parts for manufacturing, we will reduce the cost and logistical challenges of external production. The design will include the split-cell cavity concept that we are currently using for manufacturing of our other RF components. This allows for significantly reduced parts count. Separately, we have developed a high-power modulator that enables not only the driving of the klystron gun but also provides the high voltage necessary for the focusing lenses. The design of this modulator is based on a well-known Marx-bank design but with the addition of pulse forming network on each stage of the bank to reduce the amount of storage capacitors and consequently reduce the system cost. This will be used to provide the electrostatic focusing and post acceleration. During Phase II of this project, we will build a prototype klystron. We will generate the mechanical design followed by machining, sourcing of off-the-shelf components, brazing and bonding of the parts. Finally, we will test the system with to provide at least 1.1 MW output power. By aggressively redesigning every aspect of the RF source with the goal of high performance and low cost, we will be able to meet the goals of this program.