arXiv subjects

Sebastian Onder

Publications and source records attributed to Sebastian Onder.

9 recordsLinked to original sources

arXiv2026

Spectacular - A Modular DAQ System for Microdosimetry

Particle therapy using light ions like protons, helium ions or carbon ions enables precise tumor targeting with enhanced biological effectiveness while minimizing damage to healthy tissue. Successful treatment planning depends not only on the absorbed dose but also on quality of the radiation, as quantified by the linear energy transfer (LET). Microdosimetry provides a direct experimental determination of such quantities by measuring the energy deposited per incoming particle in micrometer-sized solid-state detectors representing the relevant biological scales. However, the small signal amplitudes and high particle rates in therapeutic ion beams challenge existing readout systems, which are not sufficiently optimized for reliable operation with respect to pileup and signal-to-noise ratio (SNR). To address this, a modular data acquisition (DAQ) system was developed to accelerate the design of custom readout electronics and sensor characterization in microdosimetry and related spectroscopic applications. Centered around a Xilinx Zynq system-on-chip, it combines real-time processing, high-bandwidth streaming, and high-resolution digitization (16 bit at 100 MSas$^{-1}$) to enable advanced digital signal processing. The platform integrates programmable power supplies, a bias-voltage filter, test-pulse generators, and flexible I/O. Detector and preamplifier front-ends are hosted on interchangeable daugtherboards connected via a standardized interface, allowing different hardware configurations and readout algorithms to be evaluated on the same platform. The Spectacular DAQ system was successfully tested at the MedAustron ion therapy facility with custom charge-sensitive amplifiers and a diamond microdosimeter. First results demonstrate the feasibility and provide a proof-of-concept for potential future applications in clinical practice.

physics.ins-det↗

arXiv2026

A frontend ASIC for Microdosimetry

Recent clinical evidence shows a correlation between linear energy transfer (LET) and tumor control in carbon ion radiotherapy. This prompts the direct inclusion of LET into the treatment planning. Currently, LET is mainly extracted from simulations. Good clinical practice requires adopting measurement routines that correlate with LET, such as microdosimetry. In this work, we describe an application-specific integrated circuit (ASIC) for reading out microdosimeteric sensors. The ASIC is designed for input capacitances up to 3 pF. It contains four readout channels, each with a different saturation charge ranging from 75 fC to 3.2 pC. In the 75 fC range, at 1 pF input capacitance and a shaping time of 1 microseconds, the ASIC has an equivalent noise contribution (ENC) below 15 electrons at ambient temperature. This low noise level is expected to enable new measurement possibilities, including the assessment of microdosimetric proton spectra in the low-LET region of the entrance channel, as well as studying the contribution of delta electrons.

hep-ex↗

arXiv2026

Measurements of the micro-spill structure of medical cyclotron and synchrotron beams and its impact on pulse pileup

Detector characterization and instrumentation testing are often performed at cyclotron and synchrotron facilities, many of which were originally developed for medical applications in cancer therapy. For particle physics experiments requiring a single-particle resolution, pileup can significantly degrade data quality, making precise knowledge of the beam time structure essential for selecting appropriate readout parameters. However, such information is often unavailable from the facilities and challenging to determine experimentally. Here, we report measurements of the spill time structure at two medical accelerator facilities using a silicon carbide (SiC) particle sensor coupled to a high-frequency readout system. Owing to its high carrier saturation velocity and the tolerance to large bias voltages, SiC is well suited for fast readout and measurements requiring precise timing. Using a 6 GHz readout with custom SiC diodes, we characterize the micro-spill structure of both cyclotron and synchrotron beams on a sub-nanosecond timescale. The measured arrival-time distributions exhibit modulation with the accelerator RF frequencies, reflecting features of the extraction process. The resolved micro-spill structure enables quantitative estimation of pileup contributions and provides design constraints for future readout electronics. The presented results emphasize the importance of the characterization of the beam time-structure characterization for the development of precise readout systems.

physics.ins-det↗

arXiv2025

Design and simulation of a 4H-SiC low gain avalanche diode with trench-isolation

We present the design and simulation of a 30 $\mathrm{\mu m}$ thick 4H-SiC Low Gain Avalanche Diode (LGAD) optimized for high-voltage operation. A 2.4 $\mathrm{\mu m}$ thick epitaxially grown gain layer enables controlled internal amplification up to 1 kV reverse bias, while maintaining full depletion below 500 V. Electrical characteristics, including I-V, C-V, and gain behavior, were simulated in Synopsys Sentaurus Technology Computer-Aided Design (TCAD) using a quasi-1D geometry and verified across process-related variations in gain layer parameters. To ensure high-voltage stability and proper edge termination, a guard structure combining deep etched trenches and deep $p^+$ junction termination extension (JTE) implants was designed. TCAD simulations varying the guard structure dimensions yielded an optimized design with a breakdown voltage above 2.4 kV. A corresponding wafer run is currently processed at IMB-CNM, Barcelona.

physics.ins-det↗

arXiv2025

In-situ Radiation Damage Study of Silicon Carbide Detectors Subjected to Clinical Proton Beams

Silicon carbide (SiC) planar PiN diodes from two different manufacturers were irradiated with 252.7 MeV protons from a medical synchrotron. Over the course of two 8h irradiation shifts, the samples were exposed to increasing fluences ranging from 1.4e+11 to 3.5e+13 p+/cm^2. Electrical characterizations, including IV and CV measurements, were performed both before and after irradiation using probe stations, and for selected samples even in-situ between fluence steps directly at the irradiation facility. The results show a gradual compensation of the effective epitaxial doping concentration with each incremental fluence step, observed as a reduction in capacitance before full depletion and confirmed by the extracted effective doping concentration. From these measurements, linear donor removal rates are determined for all sample groups, with values ranging from 4.2/cm to 6.4/cm. These findings provide a quantitative basis for understanding radiation-induced charge carrier removal in 4H-SiC devices and are relevant for predicting the performance and lifetime of future radiation-hard detector technologies, including 4H-SiC LGADs.

physics.ins-det↗

arXiv2025

Extraction of Electron and Hole Drift Velocities in thin 4H-SiC PIN Detectors using High-Frequency Readout Electronics

Silicon carbide (SiC) has been widely adopted in the semiconductor industry, particularly in power electronics, because of its high temperature stability, high breakdown field, and fast switching speeds. Its wide band gap makes it an interesting candidate for radiation-hard particle detectors in high-energy physics and medical applications. Furthermore, the high electron and hole drift velocities in 4H-SiC enable devices suitable for ultra-fast particle detection and timing applications. However, currently, the front-end readout electronics used for 4H-SiC detectors constitute a bottleneck in investigations of the charge carrier drift. To address these limitations, a high-frequency readout board with an intrinsic bandwidth of 10 GHz was developed. With this readout, the transient current signals of a 4H-SiC diode with a diameter of 141 $\mathrm{\mu m}$ and a thickness of 50 $\mathrm{\mu m}$ upon UV-laser, alpha particle, and high-energy proton beam excitation were recorded. In all three cases, the electron and hole drift can clearly be separated, which enables the extraction of the charge carrier drift velocities as a function of the electric field. These velocities, for the first time directly measured, provide a valuable comparison to Monte-Carlo simulated literature values and constitute an essential input for TCAD simulations. Finally, a complete simulation environment combining TCAD, the Allpix$^2$ framework, and SPICE simulations is presented, in good agreement with the measured data.

physics.ins-det↗

arXiv2025

Characterizing the Delivered Spill Structure of Medical Proton and Carbon-Ion Beams at MedAustron using a High Frequency Silicon Carbide Readout

Medical synchrotrons are often used for testing instrumentation in high-energy physics or non-clinical research in medical physics. In many applications of medical synchrotrons, such as microdosimetry and ion imaging, precise knowledge of the spill structure and instantaneous particle rate is crucial. Conventional ionization chambers, while omnipresent in clinical settings, suffer from limitations in charge resolution and integration time, making single-particle detection at high dose rates unfeasible. To address these limitations, we present a beam detection setup based on a silicon carbide (SiC) sensor and a monolithic microwave integrated circuit (MMIC), capable of detecting single particles with a full width at half maximum (FWHM) pulse duration of 500 ps. At the MedAustron ion therapy center, we characterized the spill structure of proton and carbon-ion beams delivered to the irradiation room beyond the timescale of the maximum ion revolution frequency in the synchrotron. The resulting data offer valuable insights into the beam intensity at small time scales and demonstrate the capabilities of SiC-based systems for high-flux beam monitoring.

physics.ins-det↗

arXiv2022

First Results for the pLGAD Sensor for Low-Penetrating Particles

Silicon sensors are the go-to technology for high-precision sensors in particle physics. But only recently low-noise silicon sensors with internal amplification became available. The so-called Low Gain Avalanche Detector (LGAD) sensors have been developed for applications in High Energy Physics, but lack two characteristics needed for the measurement of low-energy protons (<60 keV): a thin entrance window (in the order of tens of nm) and the efficient amplification of signals created near the sensor's surface (in a depth below 1 um). In this paper we present the so-called proton Low Gain Avalanche Detector (pLGAD) sensor concept and some results from characterization of the first prototypes of the sensor. The pLGAD is specifically designed to detect low-energy protons, and other low-penetrating particles. It will have a higher detection efficiency than non-silicon technologies, and promises to be a lot cheaper and easier to operate than competing silicon technologies.

physics.ins-det↗

arXiv2022

Deep Learning for Anomaly Detection in Log Data: A Survey

Automatic log file analysis enables early detection of relevant incidents such as system failures. In particular, self-learning anomaly detection techniques capture patterns in log data and subsequently report unexpected log event occurrences to system operators without the need to provide or manually model anomalous scenarios in advance. Recently, an increasing number of approaches leveraging deep learning neural networks for this purpose have been presented. These approaches have demonstrated superior detection performance in comparison to conventional machine learning techniques and simultaneously resolve issues with unstable data formats. However, there exist many different architectures for deep learning and it is non-trivial to encode raw and unstructured log data to be analyzed by neural networks. We therefore carry out a systematic literature review that provides an overview of deployed models, data pre-processing mechanisms, anomaly detection techniques, and evaluations. The survey does not quantitatively compare existing approaches but instead aims to help readers understand relevant aspects of different model architectures and emphasizes open issues for future work.

cs.LG↗