Inductive Adder for the FCC Injection Kicker System

• Main Author: Woog, David Gerd
• Format: Others
• Language: en
• Published: 2020
• Online Access: https://tuprints.ulb.tu-darmstadt.de/11726/13/2020-04-27_Woog_David.pdf; Woog, David Gerd <http://tuprints.ulb.tu-darmstadt.de/view/person/Woog=3ADavid_Gerd=3A=3A.html> (2020): Inductive Adder for the FCC Injection Kicker System.Darmstadt, Technische Universität, DOI: 10.25534/tuprints-00011726 <https://doi.org/10.25534/tuprints-00011726>, [Ph.D. Thesis]

This thesis presents the design, construction, assembly and measurements of an inductive adder (IA) type pulse generator. The IA was built for the future circular collider (FCC) study to investigate the possibility of a fast, high current, high voltage and reliable pulse generator for use in the injection kicker systems. In addition, the possibility of utilizing this technology as part of the injection system of an existing particle accelerator, with a generator voltage of up to 40 kV, is born in mind during the studies for the FCC injection IAs. Kicker magnets are used in particle accelerators to deflect the beam for example during the injection process, to place the injected beam onto the central orbit of a circular accelerator. In order to achieve a high reliability for the FCC, it is necessary to replace the thyratrons presently used in the pulse generators of the kickers systems by pulse generators based on solid-state technology. Solid-state switches are a promising alternative to thyratrons for pulsed power applications. Recent developments have increased the current and voltage ratings of power semiconductors and make it possible to use them for high current and voltage solid state pulse generators. The IA is a promising technology for generating high voltage and high current pulses: it consists of many ground-referenced layers, to achieve the high voltage, and many parallel connected solid state switches to achieve the high current capability. An IA with demanding specifications as required for the FCC injection kicker system has never been built so far. The main challenge for such an IA is the combination of 2.4 kA and 15 kV output waveform, 2.3 μs flattop duration, 6.25 Ω system impedance and a short current rise time of approximately 75 ns (0.5 %-99.5 %). Based on these demanding requirements, an inductive adder was designed and simulated. The specifications for the main components of the IA were defined and sample components were ordered and tested. In addition to off-the-shelf components, such as SiC MOSFETs and gate drivers, some components required a custom made design e.g. pulse capacitors and magnetic cores. After the components were selected, based on detailed analyses of tests and measurements, the hardware structure was designed and manufactured and a prototype IA was assembled. To obtain fast rise times, the height of the mechanical structure was reduced, by applying biasing to the magnetic cores, which made it possible to decrease the required volume of magnetic material in half. An oil insulation was selected to insulate the 15 kV output voltage and realise the low characteristic impedance of 6.25 Ω while keeping the diameter of the magnetic cores within an acceptable range. The principle of passive analogue modulation, together with biasing to reduce the flattop droop was proven. The required output voltage of 15 kV was achieved for a load impedance of 50 Ω with a pulse length of 2.32 μs. The specified layer output current of 2.4 kA was demonstrated with the nominal layer output voltage of 1 kV. Furthermore, a theoretical approach to use the IA in a short-circuit terminated system is discussed. Simulations show that the increase of the current due to pulse reflection from a short-circuit can be prevented by adding a second switch in each branch of the IA to change the impedance of the stack. The demanding requirements of the FCC injection system can be achieved by using the IA with presented technologies such as SiC MOSFETs, oil insulation, magnetic core biasing and passive analogue modulation. Some issues which require future work are discussed and possible improvements are proposed.