The Data Acquisition System for LZ

TL;DR

The paper describes the design and capabilities of the LZ experiment's data acquisition system, which handles signals from nearly 800 PMTs with high speed and efficiency for dark matter detection.

Contribution

It introduces a custom FPGA-based DAQ system capable of processing high-rate data from 1276 channels in a large-scale dark matter detector.

Findings

Maximum data rate of ~1500 MB/s demonstrated

System utilizes only 33% resources during calibration

Digital filters operate with throughput over 595,000 MB/s

Abstract

The Data Acquisition System of the LZ experiment, the 10-tonne dark matter detector to be installed at the Sanford Underground Research Facility (SURF), will collect signals from 788 photomultiplier tubes (PMTs). Because the signals from the time projection chamber PMTs will be passed through dual-gain amplifiers, the DAQ system will collect waveforms from a total of 1276 channels, using custom built, 32-channel, FPGA-based digital signal processors. The appropriately conditioned signals will be digitized at 100 MHz with 14-bit resolution. Based on actual measurements with a small-scale prototype system, the LZ DAQ is expected to be able to handle a maximum sparsified data rate of ~1500 MB/s. During calibrations, it is estimated that only 33% of the system resources are utilized. The digital filters that are used for data selection operate with an aggregate throughput in excess of 595,000 MB/s. Data selection decisions are based on, for example, the amount of scintillation (S1) and photoluminescence S2 light, S1 and S2 hit-patterns, and total energy deposition.