Test study of detector modules based on monolithic active pixel sensor

Abstract

Silicon pixel detectors are widely used as vertex detectors or inner trackers in high-energy collider experiments because of their excellent spatial resolution, extremely high counting capability and low power consumption. The study of the silicon pixel detector based on the monolithic active pixel sensor (MAPS) is one of the R&D schemes for upgrading the inner chamber of the Beijing spectrometer Ⅲ (BESⅢ) drift chamber (MDC). It is planned to build a 1/10 scale prototype of the inner chamber. The detector module is the basic component of the prototype, consisting of ten MIMOSA28 chips thinned to 50 μm, a flexible cable and a carbon fiber support. In order to study the performance of the module, a test system is set up. The system is composed mainly of five-layer modules, readout electronics and data acquisition system. This article focuses on the realization of the continuous data readout method with trigger marking function, the measurement of the noise level of the detector modules, the module response test by a radioactive source, and the study of the hit reconstruction algorithm. The test of trigger readout logic verifies the correctness of the continuous readout method with trigger marking. It can be concluded that analyzing consecutive three frames data (the previous frame, the current frame and the next frame) of each valid trigger will not lead to effective hit data loss. The noise level of the detector module is tested. The results show that the fixed pattern noise (FPN) is 0.253 mV, and the temporal noise (TN) is 1.65 mV. The fake hit rate is less than <inline-formula><tex-math id="M1">\begin{document}$ {10}^{-5} $\end{document}</tex-math><alternatives><graphic xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="17-20210464_M1.jpg"/><graphic xmlns:xlink="http://www.w3.org/1999/xlink" xlink:href="17-20210464_M1.png"/></alternatives></inline-formula> that can be ignored when the chip threshold is set to be above 4σ of noise. Two algorithms for hit reconstruction (i.e. adjacent method and comparison method) are studied and compared. When the average number of fired pixels caused by each hit is 2.562 and more than four hits in each frame of data, the adjacent method can find all of the fired pixels to be faster. During the test, the detector module and the electronics are proved to work well. These studies lay a foundation for further testing the performance of the detector prototype.