Fast High-Voltage Switching for Beam Kickers in Particle Accelerators
In the intricate ballet of particle accelerators, the ability to rapidly inject, extract, or dump a beam of charged particles is a fundamental operational requirement. This task falls to devices known as kickers or fast beam extractors. These are essentially high-voltage switches that, when triggered, generate a transient magnetic or electric field that deflects the beam onto a new trajectory. The speed at which this field must be established is often breathtaking, with rise times measured in nanoseconds for some applications. In my five decades of involvement with accelerator technology, I have come to regard the fast high-voltage pulser that drives these kickers as one of the most demanding and critical components in the entire accelerator complex. Its performance dictates the cleanliness of the injection process, the efficiency of beam extraction, and, in the case of a beam dump, the very safety of the machine. The design of these pulsers is a relentless pursuit of higher voltage, faster switching, and greater reliability.
The principle of a beam kicker is straightforward. A fast-rising current pulse is sent through a ferrite magnet or applied to a set of electrodes, creating a deflecting field. The beam, traveling at nearly the speed of light, must encounter this field only when it is at its full amplitude. Any rise-time tail or ringing in the pulse will result in some particles being only partially deflected, potentially causing them to be lost in the vacuum chamber or to strike sensitive components. For a perfect, lossless kick, the pulse should be a perfect rectangle, rising from zero to its full voltage in an infinitesimally short time, remaining perfectly flat for the duration of the beam passage, and then falling just as rapidly. The reality, of course, is a constant battle against the parasitic inductance and capacitance of the circuit. The key component at the heart of these pulsers is the high-voltage switch. For decades, the workhorse was the hydrogen thyratron, a gas-filled tube capable of handling tens of kilovolts and kiloamps with a reasonably fast rise time. However, the thyratron has limitations, including a finite lifetime, a requirement for a heated reservoir, and a statistical jitter in its triggering that can be tens of nanoseconds, which is unacceptable for many modern applications.
The advent of solid-state technology has revolutionized this field. Today, the state of the art for fast, high-voltage switching is the use of stacks of semiconductor devices, such as MOSFETs or avalanche transistors, connected in series. The challenge is to turn all of these series-connected devices on simultaneously. If one device turns on a few nanoseconds later than the others, it will momentarily have the full stack voltage across it and will be destroyed. Achieving simultaneous turn-on requires meticulous attention to gate drive circuitry. Each device in the stack must have its own isolated gate drive, and the trigger signal must be delivered to each of these drives with picosecond-level timing accuracy. This is typically accomplished using magnetic cores or optical fibers to distribute the trigger pulse. The devices themselves must be matched for their switching characteristics, and the layout of the printed circuit board must be designed to minimize stray inductance, which limits the rise time. With modern solid-state stacks, rise times of a few nanoseconds at voltages of 10 kV or more are achievable, representing a dramatic improvement over the thyratron.
Beyond the switch, the pulse-forming network that stores the energy and shapes the pulse is equally critical. For the fastest kickers, a simple Blumlein line or a pulse-forming line is often used. This is essentially a charged transmission line that is discharged into the load by the switch. The impedance of the line must be matched to the load impedance to prevent reflections that would distort the pulse. The line is charged to the desired voltage by a high-voltage DC power supply, which must be capable of delivering the charging current quickly between pulses. In a high-repetition-rate accelerator, this charging supply must be a fast, regulated unit in its own right. The entire pulser assembly must be carefully shielded to prevent the intense electromagnetic interference generated by the fast switching from disrupting nearby sensitive electronics. The high-voltage connections to the kicker magnet itself must be designed as coaxial structures to maintain a controlled impedance and to prevent radiation. In my long career, I have seen the design of these pulsers evolve from a black art practiced by a few specialists into a sophisticated branch of power electronics. Yet, the fundamental challenge remains the same: to command a force that can, in the blink of an eye, redirect a beam of particles traveling at near-light speed, and to do so with a precision and reliability that is essential for the advancement of high-energy physics and the operation of modern light sources.
