TL;DR
The JUNO experiment is deploying over 20,000 20-inch photomultiplier tubes (PMTs) to detect neutrinos with unprecedented precision. This paper describes a high-throughput "last-mile" test system at the JUNO site designed to validate these sensitive sensors after they have been waterproofed and transported, ensuring no faulty units are permanently installed in the detector’s compact, inaccessible lattice.
What Problem Does This Paper Address?
The Jiangmen Underground Neutrino Observatory (JUNO) aims to determine neutrino mass ordering, a feat that requires an exceptional energy resolution of 3% at 1 MeV. To achieve this, the detector relies on a massive array of 20,012 20-inch PMTs.
While these PMTs underwent rigorous acceptance testing years ago, they have since undergone a "potting" process (waterproof sealing) and been transported to the site. Because the PMTs are installed with a packing fraction of 78%—leaving only millimeters between them—it is virtually impossible to replace a single unit once the detector is assembled. The authors addressed the need for a final, high-speed functionality check to catch any "infant mortality" or transport damage before the sensors are submerged in the central detector.
Methods
The researchers established a specialized test facility in the JUNO Surface Assembly Building (SAB) consisting of two main components: a large dark room and a dedicated readout room.
- High-Throughput Design: The system can test 60 PMTs simultaneously. By running two 10-hour cycles per day, the team can validate 120 PMTs daily, matching the pace of the overall detector installation.
- Magnetic Shielding: PMT performance is highly sensitive to the Earth's geomagnetic field. The team experimented with various shielding materials, eventually settling on a cost-effective solution: lining the PMT cardboard boxes with a single layer of nanocrystalline ribbons. This achieved ~90% attenuation of the local magnetic field.
- "Test Like You Fly": To ensure the results reflect actual operating conditions, the system uses the final JUNO readout electronics, including "UWBoxes" (underwater electronics boxes) and the same Data Acquisition (DAQ) protocols that will be used in the experimental hall.
- Automated Validation: A LabVIEW-based program automates the testing of Dark Count Rate (DCR) and charge spectra under LED illumination.
Datasets
The "dataset" in this context refers to the population of PMTs being validated:
- Total Population: 20,012 20-inch PMTs (15,012 MCP-PMTs from NNVT and 5,000 dynode PMTs from Hamamatsu).
- Initial Sample: This paper reports on the first 2,000 PMTs tested through the new system.
- Reference Data: Results were compared against the original "Pan-Asia" acceptance test data collected between 2017 and 2021.
Key Insights and Ideas
- The "Cooling Down" Necessity: PMTs exposed to ambient light require significant time in total darkness to reach a stable baseline. The authors found that an 8–10 hour "cooling" period is required before DCR measurements become reliable.
- Magnetic Field Sensitivity: The study quantified a linear relationship between the residual magnetic field and the DCR. Interestingly, the impact of the magnetic field was slightly more pronounced on the MCP (Multi-Channel Plate) PMTs compared to the traditional dynode versions.
- Robustness of the Potting Process: A major "aha" moment was the high level of consistency between the original factory tests and these new on-site tests. This confirms that the waterproofing and shipping processes did not significantly degrade the sensors' performance.
Key Results
- Shielding Effectiveness: The 1-layer nanocrystal shielding reduced the average residual magnetic field from 528 mGs to approximately 62 mGs.
- DCR Stability: The average DCR for NNVT PMTs was 29.3 kHz, and for Hamamatsu PMTs, it was 17.7 kHz. These values align closely with previous measurements (29.3 kHz vs 32.4 kHz for NNVT; 17.7 kHz vs 16.6 kHz for Hamamatsu).
- Failure Rate: Out of the initial 2,000 samples, only two PMTs (~0.1%) were found to have abnormally high or unstable DCR, justifying the need for the test system while proving the overall quality of the batch.
- Charge Response: The system demonstrated a consistent charge response across the entire 12m x 6m dark room, with measured charges scaling predictably based on the distance from the LED light sources.
Limitations and Future Work
The paper presents results for only the first 10% of the total PMT array. While the initial failure rate is very low, the authors note that testing is ongoing. The current system is a "validation" check rather than a full characterization; it focuses on DCR and charge spectra rather than re-measuring every parameter (like transit time spread) already covered in previous years. Future reports will likely provide a comprehensive analysis of the entire 20,000-unit population once installation is complete.
Summary updated