A comprehensive CCD detector test setup for noise characterization and precise thermal control

This paper describes the design, development, and end-to-end testing of a novel setup for characterizing the noise performance of silicon-based, photon-counting detectors, using delta-doped detectors developed at JPL’s Microdevices Laboratory, in collaboration with Teledyne-e2v. Combined with Atomic-Layer Deposition (ALD) coatings, delta-doped detectors have been demonstrated to achieve high Quantum Efficiency (QE) in the Ultraviolet, enabling several UV missions (e.g., FIREBall-2, UVEX, SPARCS, SHIELDS and proposals (e.g., Hyperion, UVScope, Nox, Eos, and others). Understanding the noise performance of these detectors and developing strategies and designs to minimize that noise is important for future applications like the Habitable Worlds Observatory (HWO). For instance, delta-doped Electron Multiplying Charge Couple Devices (EMCCDs) have been identified as target technology for HWO. This test setup is designed to investigate the impact of the ambient temperature of the detector on its noise performance, specifically, the dark current plateau observed below -110°C, which limits the Signal-to-Noise Ratio (SNR) that can be achieved in the Photon-Counting mode and has implications on noise performance of based sil25 icon detectors. The test bench incorporates a delta-doped Teledyne-e2v CCD201-20 readout with a Nüvü CCCP v3 controller at 1 MHz and features a dual cooling system: a cryocooler for the detector and a liquid nitrogen-cooled thermal shroud for its ambient environment. We describe the design of the test setup, including independent thermal control for the detector and shroud, as well as its validation, optimization, and characterization process for the testbed. First measurements of dark current were taken across a range of detector temperatures (183 K to 143 K), shroud temperatures (298 K to 180 K), and substrate voltages. Results from these first measurements indicate that the dark current below 173 K without a cold shroud is notably reduced when the shroud is cooled to 230 K. The test bed allows for further characterization of the noise performance of the detector and optimization of readout sequences and operations for low Clock-Induced Charge (CIC).