02 dez 2024
10:00 Master's Defense Room 53 of IC2
Theme
Energy-Efficient Readout System for Radiation Detector ASICs
Student
Guilherme Paulino
Advisor / Teacher
Lucas Francisco Wanner
Brief summary
In many scientific research experiments, X-ray imaging techniques are used in synchrotron accelerators to enable advanced studies of various physical aspects in the fields of biology, chemistry, and materials science at the micro and nano scale. And in the last two decades, hybrid photon-counting pixel detector chips have been widely used for the purpose of enabling spectral and multidimensional imaging in these contexts. However, with the evolution of fourth-generation synchrotron light sources, which allow a considerable increase in brightness and a dramatic increase in the coherent X-ray radiation flux, it becomes necessary to develop large-area cameras, with different sensor array geometries, which can have vacuum-compatible water-cooled cooling systems and must meet the specifications according to the requirements of each application. This work focuses on optimizing the energy efficiency of Medipix3RX readout chips used in the context of large-area X-ray photon-counting detectors. We introduced several power optimization techniques, including dynamic voltage and frequency scaling (DVFS), and the incorporation of sleep modes into the sensor operation. These methods significantly reduced power consumption, thus allowing us to simplify cooling systems and increase durability without compromising system robustness and reliability. Experimental results demonstrate that applying dynamic voltage and frequency scaling can reduce power consumption by 5%, while the clock gating technique can reduce it by up to 23%. Furthermore, implementing a sleep mode during periods of inactivity reduces heat dissipation, improving sensor temperature stability, while reducing overall power requirements by up to 40% compared to continuous operation. A detailed characterization of the internal parameters of the Medipix3RX analog front-end allowed us to identify optimal operating points that balance low power consumption with high performance, which is crucial for maintaining low noise levels and fast count rates in challenging experimental environments.
Examination Board
Headlines:
| Lucas Francisco Wanner | IC / UNICAMP |
| Rodolfo Jardim de Azevedo | IC / UNICAMP |
| Alessandra Tomal | IFGW/UNICAMP |
Substitutes:
| Sandro Rigo | IC / UNICAMP |
| José Augusto Miranda Nacif | IEF/UFV |