{"id":"f3f5c9c2-9685-44dc-8f95-9cedf5b1b77a","arxiv_id":"2505.23050","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A beam test of a silicon microstrip detector with three floating strips and interleaved bias resistors measured 99.8% efficiency and 7.6 um spatial resolution for MIPs.","lead":"A new silicon microstrip detector puts three 'floating' strips between readout strips and tucks large bias resistors between aluminum readout strips, so it can cover large areas with few electronic channels and low power. In a CERN test beam, a single sensor achieved 99.8% detection efficiency and about 7.6 micrometer hit resolution for fast particles, and the design is now part of the AMS-02 tracker upgrade.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The quoted 7.6 um resolution rests on an imported telescope-subtraction factor (sigma_dut = 0.89 sigma_mea) with no closure test; if the equal-resolution assumption is wrong, the headline intrinsic resolution shifts.","rationale":"The beam test itself is credible: real SPS data, more than 10 million tracks, a clear efficiency definition, and the efficiency claim is robust because it uses a direct telescope prediction and a 109 um search window. The floating-strip charge-sharing effect is qualitatively supported by cluster-size and eta distributions. However, the quantitative resolution claim is the load-bearing result for the 'state of the art' statement and for the AMS-02 Layer-0 suitability, and it rests entirely on the imported 0.89 subtraction in Eq. (8). The paper supplies no derivation and no validation of that factor for this geometry, and the equal-resolution assumption is violated in the very Region-C comparison used to claim the improvement from the floating strips. This does not prove the headline is wrong, but it means the central number is not yet established to the precision implied. A computationally cheap closure test using the existing data would settle it. I therefore keep the reader's CONDITIONAL verdict.","tokens_in":9857,"tokens_out":12151,"duration_ms":134112,"concrete_test":"Re-analyze the beam-test data with a closure test: for each of the four non-DUT X-layers, compute the layer's unbiased residual against a track fitted from the remaining X-layers, infer that layer's resolution, and compare with the 0.89 sigma_mea value from the same layer. Independently, extract the telescope prediction covariance at the DUT plane from the GBL covariance matrix (without assuming equal resolutions) and recompute sigma_dut = sqrt(sigma_mea^2 - sigma_pred^2) for Regions A, B, and C. If any recomputed sigma_dut moves by more than 0.3 um, the area-weighted 7.6 um value should be revised and quoted with a systematic error.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section V C states 'assuming each layer having the same intrinsic spatial resolution, the telescope's contribution can be readily subtracted [33], giving sigma_dut = 0.89 sigma_mea.' That factor is not derived in this paper and is only valid for a particular geometry (e.g., DUT central among five equal-resolution X-layers with sigma_pred = sigma/2), for which sigma_mea^2 = 1.25 sigma_dut^2. The paper does not establish that the 5X/5Y layout satisfies this, nor that the telescope layers have the same resolution as the DUT in the region being measured. The assumption is demonstrably false for Region-C, where the DUT has no floating strips and its intrinsic resolution is roughly twice that of Regions A/B; applying the same 0.89 factor changes the inferred sigma_dut by more than 1.5 um for that region. No systematic uncertainty is quoted for any resolution, and no closure test (e.g., treating a telescope layer as a pseudo-DUT) is reported. Since the abstract's headline 7.6 um is an area-weighted combination of these sigma_dut values, an unreliable subtraction directly propagates into the central claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents the design and beam-test characterization of a p+-in-n silicon microstrip detector (SSD) with three floating strips between readout strips and bias resistors placed between the aluminum readout strips. The sensor, developed for the AMS-02 Layer-0 upgrade, has a 27.25 um strip pitch and 113x80 mm2 area. Beam tests at the CERN SPS with 5 X-layers and 5 Y-layers are used to evaluate charge sharing, detection efficiency, and spatial resolution. The authors report a total detection efficiency of 99.8% and an area-weighted spatial resolution of 7.6 um for MIPs, with the three floating strips improving the intrinsic resolution from 15.2 um to 7.2 um. A double-eta position-finding algorithm is introduced to reduce strip-swapping effects in the high-charge-sharing configuration.","tokens_in":10076,"tokens_out":5965,"duration_ms":60303,"significance":"If the reported performance is robust, this is a valuable engineering result for large-area, power-limited silicon trackers: the floating-strip design improves resolution without adding readout channels, and the bias-resistor layout shortens bond wires in long daisy-chained ladders. The paper is based on direct beam-test measurements with more than ten million reconstructed tracks, which is a strength. The double-eta algorithm and the comparative study of Regions A, B, and C are useful contributions. However, the central numerical claims, especially the headline 7.6 um resolution, depend on an imported telescope-resolution subtraction factor that is not justified in the manuscript, and no uncertainties are quoted for the efficiency or resolution values. These issues affect the conclusions as stated and need to be addressed.","major_comments":[{"comment":"The intrinsic resolution values quoted in the abstract and in Sec. V C depend on the telescope-subtraction step sigma_dut = 0.89 sigma_mea. The manuscript states that this follows from assuming each layer has the same intrinsic resolution and cites Ref. [33], but it does not establish that the 5X/5Y telescope geometry satisfies the conditions under which the 0.89 coefficient was derived, nor that the telescope layers have the same resolution as the DUT. In Region C the DUT has no floating strips, and its measured sigma_mea is 17.23 um, roughly twice the value in Regions A/B, so the equal-resolution assumption is implausible there. If the actual telescope contribution differs from the one implied by the 0.89 factor, the inferred sigma_dut for Region C shifts by more than 1.5 um, and this propagates directly into the area-weighted 7.6 um headline value. Please derive the coefficient for the actual geometry, include a closure test (for example, treating one telescope layer as a pseudo-DUT), or quote a systematic uncertainty that covers the range of plausible telescope resolutions.","section":"Sec. V C, Eq. (8)"},{"comment":"The efficiency and resolution results are quoted without statistical or systematic uncertainties: total efficiency 99.8%, sigma_mea values of 8.1/8.1/17.2 um, sigma_dut values of 7.2/7.2/15.2 um, and the final area-weighted 7.6 um have no error bars. The cluster thresholds th_seed = 3.5 and th_side = 2.0 in Sec. IV A are chosen without a threshold scan or robustness study. Because these thresholds determine cluster membership and therefore affect both efficiency and the residual distributions, the paper should provide statistical errors and a threshold-variation systematic for the efficiency and resolution claims. Such uncertainties are also needed to support the statement that placing bias resistors between the aluminum strips does not affect efficiency or resolution in Region B.","section":"Secs. V B and V C, Figs. 10-13"},{"comment":"The double-eta algorithm selects x* for each layer by comparing the residuals of x12 and x13 to an initial trajectory reconstructed from the x12 values of all layers, including the DUT. The final track is then refit using x*, and the reported resolution is the residual between the DUT and the telescope prediction. Because the same track information is used both to choose the candidate hit position and to evaluate the residual, there is a risk of selection bias that artificially narrows the measured distribution. The paper should demonstrate that the procedure is unbiased, for example by excluding the DUT from the initial trajectory used for the x* selection, or by splitting the data into selection and evaluation samples, and it should quantify any effect on the quoted resolutions.","section":"Sec. IV B and Sec. V C 1"}],"minor_comments":[{"comment":"The text says the Z-shaped structure before the double-eta correction is shown in Fig. 12(c), but the caption identifies (c) as the double-eta residual distribution; the Z-shape should be in Fig. 12(b).","section":"Sec. V C 1, Fig. 12"},{"comment":"The definition of 'hit position inner strip' as the position mapped onto [-54.5, 54.5] um is introduced in Sec. V A but would be clearer if stated before the first reference to Fig. 6.","section":"Sec. V A"},{"comment":"The notation 'Fig.13(a)(c)' in the text should read 'Fig. 13(a)-(c)' for clarity, and the same formatting issue appears in the caption of Fig. 12.","section":"Sec. V C 2"},{"comment":"The variables eta_12 and eta_13 are used in the double-eta algorithm but it is not explicitly stated that they follow the same definition as Eq. (5); adding one sentence would improve the readability of the algorithm.","section":"Sec. IV B"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within scope for physics.ins-det and reports an interesting engineering result for the AMS-02 Layer-0 upgrade. My recommendation is driven by the unresolved telescope-subtraction assumption, the absence of uncertainties on the headline numbers, and the potential bias in the double-eta selection procedure. These issues are load-bearing for the central claims but appear fixable within the scope of the paper, so I am not recommending rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this one if you work on silicon strip detectors for space. The new pieces are real: three floating strips per readout gap, with the large bias resistors physically placed between the aluminum readout strips so that long daisy-chained ladders don't need risky bond wires; and a double-eta position algorithm that fixes the swap problem in the standard eta method. Both are absent from the cited literature. The beam test is well designed — comparing Regions A, B, and C cleanly isolates the effects of the bias resistors and the floating strips, and the double-eta algorithm visibly removes the Z-shaped residual structure, improving sigma_mea from 10.6 to 8.1 um in Region A. The 99.8% efficiency is a direct measurement with a clear track-based definition. That part is solid.\n\nThe soft spot is the telescope subtraction. The paper quotes sigma_dut = 0.89 sigma_mea from Ref. [33], assuming all layers have the same intrinsic resolution. For Regions A and B, where the DUT and telescope layers are the same sensor design, that is plausible, though they never run a closure test (e.g., treating a telescope layer as a pseudo-DUT). For Region C, the DUT has no floating strips and its intrinsic resolution is roughly twice as poor, so the equal-resolution assumption is demonstrably wrong there. Applying the same 0.89 factor to the 17.2 um measured residual likely biases the quoted 15.2 um intrinsic resolution by more than a micron, and since the 7.6 um headline is area-weighted, that systematic uncertainty propagates into the central claim. Also, no uncertainty is quoted for the efficiency (99.8% with no error bar), and the cluster thresholds are tuned on the same data without a stability check. The 'state of the art' wording in the abstract is not backed by any quantitative comparison.\n\nNone of this is fatal. The design and algorithm stand on their own, and the efficiency claim is largely unaffected. But the resolution number should be treated as provisional until the telescope subtraction is validated or recalibrated.\n\nThis paper is for detector engineers building silicon trackers for space, especially the AMS-02 Layer-0 upgrade. It deserves a serious referee: send it to review with a request for systematic uncertainties, a justification or re-derivation of the telescope-subtraction factor for each region, and a toned-down abstract.","headline":"Credible engineering result with a genuinely new strip layout and a new position algorithm, but the headline resolution rests on an unvalidated telescope-subtraction factor and no systematic uncertainties.","tokens_in":10752,"tokens_out":2658,"would_cite":true,"duration_ms":28464,"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":"A silicon microstrip detector with three floating strips reaches 99.8% efficiency and 7.6 µm resolution for MIPs.","keywords":["Silicon microstrip detector","Floating strips","Bias resistors","Test beam","Detection efficiency","Spatial resolution","Double-eta algorithm","AMS-02 Layer-0"],"falsifier":"Re-analyze the same beam-test data with an alignment and track fit that treats each telescope layer's resolution as a free parameter, or repeat the measurement with a higher-resolution reference telescope; if the fitted intrinsic DUT resolution then moves by more than about half a micron, the quoted 7.6 µm area-weighted resolution is not robust.","tokens_in":9642,"feed_emoji":"🛰️","tokens_out":5165,"duration_ms":47475,"temperature":0.7,"pith_summary":"This paper describes a silicon microstrip detector designed for large-area trackers where power is scarce, such as the planned Layer-0 upgrade of the AMS-02 cosmic-ray experiment. The design's key idea is to place three electrically floating strips between every two readout strips, so that charge from a passing particle is shared among more channels without adding readout electronics. A dedicated iterative \"double-eta\" algorithm then converts the shared charge into a hit position. Beam test results on a single sensor show a total detection efficiency of 99.8% for minimum-ionizing particles and an area-weighted spatial resolution of 7.6 µm, with the three floating strips improving the intrinsic resolution from 15.2 µm to 7.2 µm. If these results hold in the full ladder configuration, the design would give a long, low-power silicon tracker sharp resolution with a small channel count.","feed_headline":"Three floating strips give 99.8% efficiency and 7.6 µm resolution","feed_subtitle":"A low-power silicon tracker layout for the AMS-02 Layer-0 upgrade keeps precision without extra readout channels.","key_machinery":"The central mechanism is the floating-strip geometry: one of every four p+ strips is AC-coupled to an aluminum readout strip, and the three intervening strips are left electrically floating, so a particle's charge is capacitively shared to neighboring readout strips and the cluster size grows (mean 2.03 in Region A versus 1.38 in Region C). The double-eta algorithm is the companion reconstruction: it first finds the highest-signal strip and compares the two candidate neighboring strips, using both eta values to resolve which neighbor truly shared the charge, which removes the \"swapping\" distortion near readout strips. The bias-resistor layout places the large polysilicon resistors for the floating strips between aluminum readout strips, reducing bonding-wire length on daisy-chained ladders; the comparison of Regions A and B shows this layout introduces only about a 1% increase in cluster size and no resolution loss.","core_discovery":"The paper claims that a silicon microstrip detector can simultaneously satisfy three constraints that usually conflict: large sensitive area, low power (few readout channels), and high spatial resolution. The claimed discovery is that three floating p+ strips between adjacent readout strips more than double the charge-sharing effect, and that this improved sharing translates directly into spatial resolution: the measured residual distribution narrows from 17.2 µm to 8.1 µm (intrinsic 15.2 µm to 7.2 µm) when the floating strips are present. The paper further claims that placing the bias resistors of the floating strips between two aluminum readout strips—a layout needed to keep bonding wires short on long ladders—does not degrade either detection efficiency (99.8% in all three regions) or spatial resolution (7.2 µm in Regions A and B). Area-weighting Regions A, B, and C gives an overall spatial resolution of 7.6 µm for MIPs.","pith_inferences":["The paper tests a single sensor in a lab beam; the real test not covered here is whether the 7.6 µm resolution and 99.8% efficiency survive when 8–12 sensors are daisy-chained into a ladder, where per-channel capacitance and noise change.","The same floating-strip layout could be tried at different strip pitches; a short simulation of charge-sharing fractions could predict whether more or fewer floating strips per readout gap would improve resolution further.","If the 0.89 telescope-subtraction factor is not exactly valid for this 5X/5Y geometry, the intrinsic-resolution numbers shift by a fraction of a micron, while the efficiency claim is essentially independent of that factor and depends mainly on the 3.5-sigma seed threshold."],"forward_implications":["A large-area tracker can keep readout channel count low while achieving resolution near that of finer-pitch detectors, because floating strips do the interpolation in hardware.","The AMS-02 Layer-0 ladders of 8, 10, or 12 daisy-chained SSDs can use this design with short bonding wires and the required roughly 100-times larger bias resistance.","The double-eta algorithm removes a specific failure mode (strip swapping) that otherwise broadens residuals to 10.6 µm, and it can likely be applied to other high-charge-sharing strip detectors.","The three floating strips cut the collected-charge most probable value from about 65 LSB to 46 LSB, yet efficiency stays at 99.8%, so the design tolerates lower charge collection when noise is low."],"supporting_citations":[{"why":"Supplies the eta-algorithm formalism and the theoretical link between charge sharing and spatial resolution that underlies the floating-strip design.","marker":"[18]"},{"why":"Provides the 0.89 coefficient used to subtract the telescope resolution from the measured residual to obtain the intrinsic DUT resolution.","marker":"[33]"},{"why":"Defines the AMS-02 Layer-0 upgrade context and the ladder architecture whose power and length constraints motivate the bias-resistor layout.","marker":"[16]"},{"why":"Explains charge-collection losses and the role of bias-resistor and coupling-capacitance changes in strip sensors, used to interpret the MPV drop in floating-strip regions.","marker":"[19]"},{"why":"The General Broken Lines track fitter used to reconstruct trajectories and measure residuals in the beam test.","marker":"[29]"},{"why":"Specifies the IDE1140 front-end readout chips used on the frame board, which set the gain and peaking time of the measurements.","marker":"[23]"}],"fun_headline_variants":["Floating strips give 7.6 µm resolution with 99.8% efficiency","Low-power silicon strips: 7.6 µm resolution, 99.8% efficiency","Triple floating strips double charge share, reach 7.6 µm precision","Power-limited SSD: 7.6 µm resolution, 99.8% efficiency","Floating-strip tracker hits 7.6 µm at low power, high efficiency"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quoted intrinsic resolutions rest on the assumption that all ten telescope layers have the same intrinsic resolution, so their contribution can be removed by one fixed factor (0.89) taken from an earlier beam-test study; if that factor or the identical-layer assumption is off, the stated 7.2 µm and 7.6 µm numbers change, even though the 99.8% efficiency claim does not.","fun_headline_variants_meta":{"raw":{"variants":["Floating strips give 7.6 µm resolution with 99.8% efficiency","Low-power silicon strips: 7.6 µm resolution, 99.8% efficiency","Triple floating strips double charge share, reach 7.6 µm precision","Power-limited SSD: 7.6 µm resolution, 99.8% efficiency","Floating-strip tracker hits 7.6 µm at low power, high efficiency"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001559,"raw_usage":{"total_tokens":6174,"prompt_tokens":839,"completion_tokens":5335,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":455,"completion_tokens_details":{"reasoning_tokens":5223}},"tokens_in":455,"tokens_out":5335,"duration_ms":31606,"temperature":1.0,"reasoning_tokens":5223,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T12:54:50.535472+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-analyze the same beam-test data with an alignment and track fit that treats each telescope layer's resolution as a free parameter, or repeat the measurement with a higher-resolution reference telescope; if the fitted intrinsic DUT resolution then moves by more than about half a micron, the quoted 7.6 µm area-weighted resolution is not robust.","supporting_citations":[{"cited_title":"Lubelsmeyer et al","cited_arxiv_id":null,"evidence_quote":"Supplies the eta-algorithm formalism and the theoretical link between charge sharing and spatial resolution that underlies the floating-strip design."},{"cited_title":"Fast align- ment of a complex tracking detector using advanced track mod- els.Comput","cited_arxiv_id":null,"evidence_quote":"Provides the 0.89 coefficient used to subtract the telescope resolution from the measured residual to obtain the intrinsic DUT resolution."},{"cited_title":"DAMPE silicon tracker on-board data compression algorithm.Chin","cited_arxiv_id":null,"evidence_quote":"Defines the AMS-02 Layer-0 upgrade context and the ladder architecture whose power and length constraints motivate the bias-resistor layout."},{"cited_title":"The AMS-02 Silicon Tracker after 500 days in space.PoS, Vertex2012:052, 2013","cited_arxiv_id":null,"evidence_quote":"Explains charge-collection losses and the role of bias-resistor and coupling-capacitance changes in strip sensors, used to interpret the MPV drop in floating-strip regions."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The General Broken Lines track fitter used to reconstruct trajectories and measure residuals in the beam test."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Specifies the IDE1140 front-end readout chips used on the frame board, which set the gain and peaking time of the measurements."}],"review_version":1}