High count rate effects in event processing for XRISM/Resolve x-ray microcalorimeter

Misaki Mizumoto,Masahiro Tsujimoto,Renata S. Cumbee,Megan E. Eckart,Yoshitaka Ishisaki,Caroline A. Kilbourne,Edmund Hodges-Kluck,Maurice A. Leutenegger,Frederick S. Porter,Makoto Sawada,Yoh Takei,Yuusuke Uchida,Shin'ya Yamada,XRISM Resolve team
DOI: https://doi.org/10.1117/12.2628784
2023-12-25
Abstract:The spectroscopic performance of x-ray instruments can be affected at high count rates. The effects and mitigation in the optical chain, such as x-ray attenuation filters or de-focusing mirrors, are widely discussed, but those in the signal chain are not. Using the Resolve x-ray microcalorimeter onboard the XRISM satellite, we discuss the effects observed during high count rate measurements and how these can be modeled. We focus on three instrumental effects that impact performance at high count rate: CPU limit, pile up, and electrical cross talk. High count rate data were obtained during ground testing using the flight model instrument and a calibration x-ray source. A simulated observation of GX 13+1 is presented to illustrate how to estimate these effects based on these models for observation planning. The impact of these effects on high count rate observations is discussed.
Instrumentation and Methods for Astrophysics
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the specific situation where the performance of the XRISM/Resolve X - ray micro - calorimeter is affected under high count - rate conditions and its solutions. Specifically, the article explores three main instrumental effects observed during high - count - rate measurements: CPU limitation, pulse pile - up, and electrical cross - talk, and attempts to model and mitigate these effects through experimental and simulation methods. ### 1. CPU Limitation At high count - rates, the CPU processing events may reach the limit of its processing capacity, resulting in data loss or processing delay. The article experimentally verifies whether Resolve can meet the requirement of processing more than 200 events per second and discusses how to correct the count - rate exceeding the CPU processing capacity through lost - event telemetry data. ### 2. Pulse Pile - Up When X - ray pulses are too dense, adjacent pulses may overlap and cannot be correctly distinguished, resulting in spectral shape distortion and count - rate deviation. The article analyzes the relationship between RISE TIME and energy and uses the SLOPE DIFFER screening method to identify and remove pile - up events, thereby improving the spectral shape quality. ### 3. Electrical Cross - Talk At high count - rates, the pulse energy in one pixel may be partially deposited into adjacent pixels, causing interference to other originally uncontaminated pulses. The article experimentally evaluates the change in energy resolution at different count - rates and proposes a method to restore energy resolution by discarding events affected by cross - talk. ### Experiment and Simulation The article combines experimental data and simulation methods to describe in detail the effects of these effects and provides specific models for observation planning. For example, by simulating the observation of GX 13 + 1, it shows how to estimate the effects under high count - rate conditions based on these models for more accurate observation planning. ### Summary The main purpose of this paper is to study and model the impact of high count - rates on the performance of the XRISM/Resolve X - ray micro - calorimeter through experimental and simulation methods, especially the three effects of CPU limitation, pulse pile - up, and electrical cross - talk, and propose corresponding mitigation measures to ensure accurate data can be obtained in actual astronomical observations. ### Formula Summary - **FWHM Recovery Formula**: \[ \text{FWHM recovery}=\sqrt{\text{FWHM}_{\text{NoXTalkCut}}^2-\text{FWHM}_{\text{XTalkCut}}^2} \] - **Proportion of Effective Exposure Time Reduction (Caused by Electrical Cross - Talk)**: \[ \alpha_{\text{XTalk}}=\exp\left(-\frac{f_X}{613}\right)\times\left(1 - 2.358\times 10^{-4}f_X+6.919\times 10^{-7}f_X^2\right) \] The formulas are slightly different for the first - level and second - level electrical cross - talk cuts. Through these studies, the author hopes to provide theoretical basis and technical support for future high - count - rate X - ray observations.