研究目的
To analyze the noise performance and components of spectroscopic prototype DEPFET detectors, including shot noise, white noise, 1/f noise, and signal noise, and to parametrize these for optimizing operation modes.
研究成果
The analysis successfully parametrized noise components, showing that prototype DEPFET detectors achieve low noise (e.g., 1.51 e?ENC for 64x64 pixels and 0.95 e?ENC in window mode). The performance is enabled by reduced transistor channel size and extended signal integration. Future work will focus on further reducing 1/f noise through design modifications, such as smaller gate widths in upcoming prototypes.
研究不足
The noise reduction is limited by signal noise from charge carrier generation, which cannot be fully eliminated. Operation in window mode reduces sensor size and dynamic range. The study is based on prototypes, and further optimizations (e.g., gate size reduction) are needed for sub-electron noise.
1：Experimental Design and Method Selection:
The study involved designing experiments to measure noise components using test algorithms and a trapezoidal filter function for correlated double sampling. Theoretical models for noise contributions (e.g., shot noise proportional to dark current) were employed.
2：Sample Selection and Data Sources:
Prototype DEPFET detectors with 64x64 pixels and linear gate design were used, fabricated by the semiconductor laboratory of the Max-Planck-Society. Data came from dark frames (without illumination) and calibrated measurements using X-ray sources like Mn-Kα and Al Kα lines.
3：List of Experimental Equipment and Materials:
DEPFET sensors (130x130 μm2 pixel size, 450 μm thick silicon bulk), VERITAS readout ASIC, cooling system to -80°C, X-ray sources for calibration.
4：Experimental Procedures and Operational Workflow:
Detectors were operated in source follower mode. Noise was measured by varying signal integration times, using different gain settings, and performing operations like window mode to reduce sensor size. Steps included recording dark frames, applying offset and common mode correction, and Gaussian fitting for noise determination.
5：Data Analysis Methods:
Noise components were analyzed using parametrization equations (e.g., ENC^2 = a1/τ A1 + a2 A2 + a3τ A3 for readout noise). Statistical techniques included Gaussian fitting and quadratic subtraction of noise components. Software tools were not specified.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
