{"id":"898c8003-8f51-48e1-81ab-62caaa19c8aa","arxiv_id":"2607.19066","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A newly developed 14-bit, 1 GSa/s, 64-channel readout system passes prototype tests for the Jinping neutrino experiment, with 0.1 MeV threshold and 0.3 ns timing drift.","lead":"This paper reports a new 64-channel readout system, PDS1500, built for the Jinping neutrino experiment and tested on its 1-ton prototype. It claims lower noise, stable timing, and a 0.1 MeV energy threshold compared with a commercial system.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Comparative performance claims rely on an uncharacterized PMT signal splitter; without splitter-fidelity data, the noise, waveform, and threshold comparisons may not isolate PDS1500 performance.","rationale":"The reader's weakest assumption is the same one I find most load-bearing. The paper's main evidence for PDS1500 superiority is differential, and the only coupling between the two systems is the splitter. Without splitter characterization, the factor-of-three noise comparison, waveform consistency, and energy-threshold comparison could be biased. I did not find an internal inconsistency in the electronics design itself; the clock architecture is plausible and the timing-drift result is supported by repeated runs. The ENOB/SNR values are asserted without method, which I would list as a secondary concern, but it is not as damaging because it affects a single specification rather than the entire cross-system comparison. The proposed swap test is cheap and decisive: if the performance differences persist after swapping splitter ports, the splitter is exonerated and the comparative claims stand; if they do not, the central 'fully satisfies' claim is overstated. Therefore I agree with the conditional verdict and recommend no change.","tokens_in":13876,"tokens_out":5309,"duration_ms":60117,"concrete_test":"Swap the two splitter output ports: connect the port previously feeding CAEN to PDS1500 and vice versa, then repeat the synchronous data-taking for baseline noise (Fig. 10), waveform chi2 (Eq. 1), and the natural-radioactivity energy-threshold run (Fig. 17). If the one-third noise ratio, chi2~1, and the 0.1 MeV threshold move with the splitter port rather than with PDS1500, the comparative results are artifacts. As a supporting check, characterize the splitter ports directly with a fast oscilloscope and a network analyzer, comparing gain, 10-90% risetime, propagation delay, and S21/S11 over 0-500 MHz.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central validation strategy is simultaneous comparison of PDS1500 and CAEN V1751 reading the same PMT pulses through a two-way splitter (Sec. 3.1, Fig. 7). The paper asserts that the splitter 'ensures that the waveforms of the two output signals are consistent' but provides no measurement of the splitter's bandwidth, port-to-port gain match, return loss, or propagation-delay dispersion. Every quantitative comparison supporting the headline—baseline noise one-third (Fig. 10), waveform chi2/NDF near 1 (Eq. 1 / Fig. 12), energy threshold 0.1 MeV vs 0.5 MeV (Fig. 17), and energy-resolution agreement (Fig. 19)—is a differential measurement made through this uncharacterized component. If the two splitter paths differ in attenuation, frequency response, or impedance mismatch, the differences attributed to PDS1500 could be partly or wholly produced by the splitter. This is an internal gap: the paper's own claims require identical split paths. The ENOB/SNR values in Table 1 are also asserted without a reported measurement method, a separate missing support, but the splitter is the more load-bearing weakness because it underlies the comparative performance claims. The timing-drift (Sec. 3.3.4) and data-loss (Sec. 3.3.1) measurements are less affected because they do not require waveform equality across the two paths.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports the development and testing of PDS1500, a new 64-channel readout system for the Jinping Neutrino Experiment (JNE), offering 1 GSa/s sampling, 14-bit resolution, a 64 Gbps data path, and a deterministic clock distribution architecture. The system was deployed on the upgraded JNE-1ton prototype, and its performance was compared with the commercial CAEN V1751 digitizer using a two-way PMT signal splitter. The authors claim zero data loss in a 1000 ns window, baseline noise one-third of the CAEN level, 0.3 ns timing drift across power cycles, an energy threshold of 0.1 MeV, and consistent energy resolution at ~9%, concluding that the system fully satisfies JNE requirements and is scalable to 3000 channels.","tokens_in":14285,"tokens_out":5560,"duration_ms":57935,"significance":"If the reported figures are substantiated, PDS1500 represents a credible readout solution for a next-generation large liquid-scintillator neutrino detector. The paper contains direct comparative measurements for several key metrics, and the design choices (star-distributed clocks, high-throughput PCIe readout, low-noise front end) are well motivated. However, several load-bearing details are missing: the splitter used for all comparative measurements is not characterized, the ENOB/SNR specifications lack a measurement description, and the simulation supporting the detector-limited energy-resolution conclusion is undocumented. These gaps currently prevent the paper from fully supporting its central claim.","major_comments":[{"comment":"The comparative validation relies entirely on a two-way PMT signal splitter feeding the same pulses to the CAEN V1751 and PDS1500. The paper only states that the splitter 'ensures that the waveforms of the two output signals are consistent' and gives no measurement of its bandwidth, port-to-port gain match, return loss, or propagation-delay dispersion. All differential metrics—baseline noise ratio (Fig. 10), waveform χ²/NDF (Fig. 12), energy threshold (Fig. 17), and energy resolution (Fig. 19)—depend on the two split paths being identical. If the paths differ in attenuation or frequency response, the differences attributed to PDS1500 would be inseparable from splitter effects. Please provide splitter specifications and a standalone characterization, or alternatively use a measurement configuration that does not require an uncharacterized analog component.","section":"Section 3.1, Figure 7"},{"comment":"Table 1 lists ENOB = 10.6 bit and SNR ≈ 65.6 dB with a caption stating these are 'derived from the actual measurement results presented in the following sections,' but no section presents an ENOB or SNR measurement. Without the measurement method (e.g., sine-wave curve fit, FFT-based SNR, input frequency, record length), these headline dynamic-performance numbers are unverifiable. The authors should add a measurement description and results, or clearly label the values as design estimates.","section":"Table 1, Section 3.3"},{"comment":"The normalization factor 0.04858 for PDS1500-to-CAEN amplitude conversion is calibrated from the peak-height ratio distribution (Fig. 9) using the same synchronous events that are subsequently used for the baseline-noise comparison (Fig. 10) and the waveform-fidelity χ² test (Eq. 1, Fig. 12). This couples the amplitude calibration to the performance comparison: any gain mismatch between the two split paths or between the two systems is absorbed into the factor, so the factor-of-three noise improvement and χ²/NDF ≈ 1 are not fully independent validations. An independent calibration source (e.g., a precision pulser) or a validation on a separate dataset should be provided.","section":"Sections 3.3.2 and 3.3.3, Eq. (1)"},{"comment":"The energy-threshold comparison confounds the electronics system with the multichannel trigger threshold setting. The PDS1500 spectrum obtained with Tr_multi=10 is compared against spectra with Tr_multi=45 from both the CAEN system and the PDS1500 itself. The lower threshold attainable with PDS1500 is a genuine system-level advantage (the CAEN DAQ crashed when the threshold was lowered), but the 0.1 MeV versus 0.5 MeV comparison does not isolate the electronics from the trigger-setting change. Please either present a comparison at the same trigger threshold where possible, or clearly frame the result as a combined electronics-plus-DAQ threshold capability rather than a difference in electronics alone.","section":"Section 3.3.5, Figure 17"},{"comment":"The simulation used to argue that energy resolution saturates near 9% beyond 9 bits and is therefore detector-limited is not described. Figure 19(b) shows curves with and without 'Baseline fluct. = 0.5 ADC bin', but the simulation parameters (light yield, PMT response, noise model, event selection, and how the 'intrinsic 9%' was set) are not given. Since the conclusion that 14-bit resolution provides no energy-resolution benefit rests on this simulation, the simulation setup must be documented, or the claim should be limited to the measured data.","section":"Section 3.3.6, Figure 19(b)"}],"minor_comments":[{"comment":"The tabular header contains a typo: 'PDS11500' should be 'PDS1500'.","section":"Table 1"},{"comment":"The sentence 'the PDS1500 system has a time drift of only 0.3 nslower than the experimental time step' is missing a space; should read '0.3 ns lower'.","section":"Section 3.3.4"},{"comment":"The 'zero data loss' claim is based on only two comparison runs. A more thorough statistical statement, including the expected loss rate over longer operation, would strengthen the claim.","section":"Section 3.3.1"},{"comment":"The axis labels for the histogram and the top-row plots are unclear (e.g., '3 10×Entries' and '3 −10×Peak height ratio'). Please reformat to standard scientific notation.","section":"Figure 9"},{"comment":"The single-channel threshold is described as 'approximately one single-photoelectron signal'; the conversion from CAEN ADC units to photoelectrons should be justified or referenced.","section":"Section 3.3.5"}],"recommendation":"major_revision","confidential_remarks":"The paper describes a credible and timely engineering effort, but the comparative validation is currently compromised by the uncharacterized splitter and by several missing measurement/simulation details. These issues are addressable in revision: the authors can supply splitter characterization, an ENOB measurement description, an independent calibration check, and a documented simulation. I do not recommend rejection because the underlying system appears well designed and the headline claims are plausible, but the manuscript in its present form overstates the strength of the evidence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a genuine and useful engineering contribution. The PDS1500 is a 64-channel 1 GSa/s 14-bit digitizer with a deterministic zero-delay clock, built on commercial FPGAs/ADCs, and the authors have validated it on the upgraded JNE-1ton prototype against a commercial CAEN V1751. The zero-data-loss measurement, the 0.3 ns vs 16 ns timing-drift comparison, and the 0.1 MeV threshold demonstration are concrete and, as far as I can tell, honestly reported. The waveform chi2/NDF near 1 across 53 channels is a nice confirmation that the new system reproduces the CAEN data after trivial normalization. I believe the central claims — that this system meets JNE's readout requirements and could scale to 3000 channels — are plausible and mostly supported.\n\nThe soft spots are real but not fatal. The splitter that feeds both systems is described in one line and never characterized: no bandwidth, port-to-port gain mismatch, return loss, or delay dispersion. The paper says it 'ensures' consistency, but doesn't show it. For the baseline noise comparison this matters less, because baseline noise is mostly electronics noise rather than signal-path attenuation. For the waveform chi2 and the energy-threshold comparison, however, an uncharacterized splitter leaves real uncertainty: if the two paths differ in gain slope or phase, some of the differences attributed to the systems could be the splitter. I don't think this invalidates the paper, but it's the first thing a referee should ask for.\n\nTwo other gaps: the ENOB/SNR numbers in Table 1 are presented without a measurement description, despite the caption saying they're derived from actual measurements; and the energy-resolution simulation in Fig. 19(b) is a black box. Also, the energy-threshold comparison uses different multichannel trigger settings (Tr multi=10 for PDS1500 vs 45 for CAEN), which is fine as an honest capability demonstration, but it should be stated more explicitly that the comparison is about trigger configurability rather than raw electronics noise.\n\nThe normalization factor of 0.04858 is fitted from the same peak-height dataset that is used for the subsequent noise and waveform comparisons. That's a minor circularity — the factor absorbs any channel-to-channel gain difference, so the noise comparison after scaling could be slightly optimistic for the new system. I'd like to see the factor derived from a separate calibration run, or at least checked against a known charge injection.\n\nCitation pattern looks fine: the references are appropriate for the field, the prior work on the 1-ton prototype is cited, and the claims about JNE requirements are contextualized. No missing citation that I noticed.\n\nBottom line: this is a paper I would send to a serious referee. It's an instrumentation paper with real hardware, real measurements, and a plausible path to the full detector. The revision needs to add splitter characterization, ENOB/SNR methodology, and a bit more detail on the simulation; after that I'd be comfortable with it.\n\nReading group: maybe — not a field re-organizer, but useful if anyone in the group does fast nuclear/particle DAQ.","headline":"A solid, workmanlike instrumentation paper — the PDS1500 system is real and mostly validated, but the uncharacterized splitter and missing measurement details for ENOB/SNR are genuine gaps that need a revision before I'd rely on the comparative numbers.","tokens_in":14726,"tokens_out":4277,"would_cite":true,"duration_ms":43672,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper claims that PDS1500, a 64-channel 1 GSa/s 14-bit waveform readout, passes every electronics requirement of the Jinping Neutrino Experiment as demonstrated on the JNE-1ton prototype.","keywords":["neutrino detector readout","waveform digitization","clock distribution","zero-delay mode","baseline noise","energy threshold","liquid scintillator detector","PMT readout electronics"],"falsifier":"Characterize the splitter independently—measure each output against the input with a fast pulse generator and a network analyzer—or repeat the comparison with the two systems fed directly from the same calibrated source without a splitter. If the splitter's outputs differ by more than the claimed one-third baseline-noise ratio, or if direct-fed A/B tests do not reproduce the 0.3 ns drift and 0.1 MeV threshold, the central claim fails.","tokens_in":13828,"feed_emoji":"⚡","tokens_out":6071,"duration_ms":60671,"temperature":0.7,"pith_summary":"The paper reports a purpose-built readout system, PDS1500, for the Jinping Neutrino Experiment's 3000-channel slow-liquid-scintillator detector. The central claim is that this single-crate 64-channel waveform digitizer satisfies every stated electronics requirement: 1 GSa/s sampling at 14 bits, a deterministic zero-delay clock that keeps all channels synchronized, no data loss in a 1000 ns acquisition window, baseline noise one-third that of the commercial system previously used, timing drift of 0.3 ns across power cycles, and a trigger threshold of 0.1 MeV. Validation came from bench tests and deployment on the upgraded JNE-1ton prototype, with direct comparison against a commercial digitizer on the same PMT signals. The result matters because reaching a 0.1 MeV threshold is what would open the detector to low-energy solar neutrinos from the pp chain and 7Be, and the modular design is claimed to scale to the full 3000-channel detector.","feed_headline":"New readout hits 0.1 MeV threshold for neutrino experiment","feed_subtitle":"Baseline noise drops to a third and power-cycle drift to 0.3 ns in tests against a commercial system.","key_machinery":"The load-bearing object is the deterministic zero-delay clock distribution. A timing board distributes a 50 MHz reference over the backplane to eight digitizer boards; each board's LMK04832 jitter cleaner (a clock chip) operates in zero-delay mode, feeding its 250 MHz output back into the PLL so the derived 1 GHz ADC sampling clock locks to the same phase as the incoming reference. Because the global trigger is also distributed in equal-length star topology, every channel samples the same phase of the PMT waveform simultaneously, eliminating the ±1 sampling-clock-cycle ambiguity and the need for offline waveform cropping. This mechanism is what produces the 0.3 ns power-cycle drift, zero-dea","core_discovery":"On its own terms, the paper establishes that a newly designed electronics chain—eight 8-channel 14-bit 1 GSa/s digitizer boards, a trigger/timing board, a high-speed backplane, and a PCIe acquisition board in one crate—meets all JNE specifications. The decisive feature is the clock network: a 50 MHz reference is distributed in a star topology, and each digitizer uses a zero-delay-mode jitter cleaner so the 1 GHz ADC sampling clocks and the global trigger stay phase-locked across all channels. Tested on the JNE-1ton prototype, the PDS1500 recorded 1000 ns waveforms with no dead-time losses, showed baseline noise about one-third of the reference system after amplitude normalization, held chann","pith_inferences":["The splitter-based comparison is the paper's unexamined hinge; until the splitter's own bandwidth, impedance, and phase consistency are published, the factor-of-three noise advantage and waveform agreement should be treated as system-level results rather than pure electronics performance.","The zero-delay star-topology clock scheme is not specific to JNE; it could plausibly be adopted by other large liquid-scintillator or water-Cherenkov detectors that need sub-nanosecond multi-crate synchronization.","A testable extension is to verify whether the 0.3 ns PLL alignment step is deterministic per power cycle; if not, a continuous phase monitor or power-on alignment calibration would be needed for the full experiment's long-term stability.","The claim that 14-bit depth does not improve energy resolution is specific to this detector's light yield; in a higher-light-yield configuration, or for single-photon counting and Cherenkov/scintillation separation, the extra bits could matter more than the paper's saturation curve suggests."],"forward_implications":["The full 3000-channel JNE detector can be read out by 64 such crates with a two-stage optical network, at a total estimated power of about 7.5 kW.","A 0.1 MeV energy threshold makes pp-chain and 7Be solar-neutrino measurements accessible, since the previous setup's efficiency fell sharply below about 0.5 MeV.","No post-acquisition inter-board alignment or cropping is needed, simplifying offline processing and removing a source of systematic timing error in vertex reconstruction.","A trigger rate up to 50 kHz and 64 Gbps per-crate throughput are sufficient for the waveform-based, high-precision event reconstruction JNE plans.","Since the electronics contribution to energy resolution saturates by about 9 bits, the 14-bit readout's value is in waveform fidelity and threshold, not in improving the 1-ton prototype's energy resolution."],"fun_headline_variants":["0.1 MeV neutrino threshold met by new 64-channel readout","New readout hits 0.1 MeV, baseline noise cut to a third","JNE readout achieves 0.1 MeV and 0.3 ns timing drift","Zero dead time, 0.1 MeV threshold in new JNE readout","64 Gbps readout hits 0.1 MeV floor for JNE"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The comparative measurements assume the PMT signal splitter sends identical waveforms to both systems; if the two split paths differ in gain, bandwidth, or phase, the baseline-noise and waveform-agreement comparisons could reflect the splitter rather than the new electronics.","fun_headline_variants_meta":{"raw":{"variants":["0.1 MeV neutrino threshold met by new 64-channel readout","New readout hits 0.1 MeV, baseline noise cut to a third","JNE readout achieves 0.1 MeV and 0.3 ns timing drift","Zero dead time, 0.1 MeV threshold in new JNE readout","64 Gbps readout hits 0.1 MeV floor for JNE"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00055,"raw_usage":{"total_tokens":2512,"prompt_tokens":844,"completion_tokens":1668,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":588,"completion_tokens_details":{"reasoning_tokens":1563}},"tokens_in":588,"tokens_out":1668,"duration_ms":13118,"temperature":1.0,"reasoning_tokens":1563,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T13:31:58.674538+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Characterize the splitter independently—measure each output against the input with a fast pulse generator and a network analyzer—or repeat the comparison with the two systems fed directly from the same calibrated source without a splitter. If the splitter's outputs differ by more than the claimed one-third baseline-noise ratio, or if direct-fed A/B tests do not reproduce the 0.3 ns drift and 0.1 MeV threshold, the central claim fails.","supporting_citations":[],"review_version":1}