Integrated circuits (ICs) are nowadays a core element in many research areas, and it’s not infrequent to observe breakthroughs in medical science, physics or biology as direct consequence of an innovation leap in the electronic instrumentation. In this respect, a key research target in the extended field of Information Technology is to anticipate, and not just to follow, the needs of the scientific community and to enable the development of future applications by taking the most out of the collected raw data. The increasing demand for performances in radiation detectors, driven by cutting-edge research in nuclear physics, astrophysics and medical imaging systems, is causing not only a proliferation in the variety of the radiation sensors themselves, but also a growing necessity of tailored solutions for front-end readout circuits and for subsequent signal elaboration, storage and digital transmission. In high-resolution x-ray Semiconductor Radiation Detectors (SRD), the complexity and importance of such functionalities require Application Specific Integrated Circuits (ASICs), that challenge the designer with some crucial topics in both architecture definition and transistor level implementation.

The aim of the presented project is to study novel solutions for such readout ASICs, defining new possible paths for improving the state-of-the-art in terms of two key parameters: energy resolution and power consumption, that compose the most important trade off in the design specification of the front-end electronics.  Noise and its sources have been widely studied in literature, and there are many ways to model its behaviour in electronic systems, depending on the application of interest. Nevertheless, the realization of an ultra-low noise measurement chain implies an extensive effort in the research and elimination unexpected noise sources, and usually requires an underestimated experimental knowledge which spreads from high-density PCB design, parasitic components estimation and RF shielding techniques. For radiation detectors, being the input signal typically expressed in terms of charge, the most common parameter for the evaluation of the goodness of a processing chain, is the Equivalent input Noise Charge (ENC). Relevant blocks in the noise performance definition, such as charge sensitive amplifiers (CSA), are thus to be designed – and carefully arranged during the physical layout and PCB assembly phase – in order to minimise the noise contributions arising from both active and passive components. In the first part of this work, a newly designed CSA will be presented, optimizing the noise performance - and thus the spectroscopic resolution - down to few equivalent noise electrons r.m.s., with specific focus on sub-microsecond signal filtering, addressing the growing interest in high-luminosity experiments.  In the second part of this work, the topic of low-power design for radiation instrumentation is presented; indeed, despite being a subject mainly addressed by portable electronic devices, and in general by battery supplied applications, reducing the power budget is also demanded by the strongly increasing number of channels in high-resolution pixel arrays, where a lower thermal dissipation can be crucial for the reduction of thermal noise. This is especially important in radiation instrumentation for astrophysics experiments, where a dedicated bulky cooling system can significantly impact the overall Size-Weight-and-Power (SWaP) budget of the payload in scientific satellites. An ASIC designed for the THESEUS (Transient High Energy Sources and Early Universe Surveyor) experiment will be presented, providing a comprehensive case study of a complete low-power radiation detection processing chain. Eventually, the experimental results collected with the two ICs obtained using a finely-controlled laboratory setup will be presented, and the final integration on the top-level radiation detection instrument will be deeply described and commented.