The DEPFET sensor with signal compression (DSSC) is a 1-megapixel imager developed for the European XFEL. It is designed to detect X-rays with photon energies between 250 eV and 6 keV, and provides a peak frame-rate of 4.5 MHz. The smallest independent unit of the detector, called module, is composed by two sensors, forming a matrix of 512 by 128 active DEPFET pixels, bump-bonded to 16 readout ASICs. The in-pixel electronics comprises a trapezoidal-shaping filter with programmable gain and timing, an ADC with finely tunable gain and offset, and an 800-word long memory. The characterization of the first CMOS-DEPFET based prototype modules showed equivalent noise charge (ENC) better than 10e−rms at megahertz frame-rate and room temperature. In this work, we present the results of a measurement campaign carried out to disentangle the noise contributions of the acquisition chain. We exploited the front-end capabilities and measured the ENC using the knife-edge method, varying the gain of single stages at a time. The measured ENC vs gain were fitted and the gain-dependent and -independent components extracted, and the contributions at key points of the acquisition chain presented.
Noise Analysis of the CMOS-DEPFET Front-End of the DSSC Imager
Porro, M.
2024-01-01
Abstract
The DEPFET sensor with signal compression (DSSC) is a 1-megapixel imager developed for the European XFEL. It is designed to detect X-rays with photon energies between 250 eV and 6 keV, and provides a peak frame-rate of 4.5 MHz. The smallest independent unit of the detector, called module, is composed by two sensors, forming a matrix of 512 by 128 active DEPFET pixels, bump-bonded to 16 readout ASICs. The in-pixel electronics comprises a trapezoidal-shaping filter with programmable gain and timing, an ADC with finely tunable gain and offset, and an 800-word long memory. The characterization of the first CMOS-DEPFET based prototype modules showed equivalent noise charge (ENC) better than 10e−rms at megahertz frame-rate and room temperature. In this work, we present the results of a measurement campaign carried out to disentangle the noise contributions of the acquisition chain. We exploited the front-end capabilities and measured the ENC using the knife-edge method, varying the gain of single stages at a time. The measured ENC vs gain were fitted and the gain-dependent and -independent components extracted, and the contributions at key points of the acquisition chain presented.I documenti in ARCA sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.



