{"id":"e90a64ee-356e-4507-b783-8676628b85d5","arxiv_id":"2412.12521","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The Tsinghua tabletop Kibble balance has reached component integration with characterized magnet, gravity, weighing, and current-source subsystems, but no end-to-end mass measurement is reported yet.","lead":"This progress report describes new components for a compact tabletop Kibble balance at Tsinghua University, including a split magnet with improved field uniformity, a capacitor-sensor weighing cell, and a two-stage stable current source. It reports measured component performance but does not yet report a full mass calibration result.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Split-gap reproducibility is the load-bearing risk: Eq. (5) requires Bl stable at 5e-8 between weighing and velocity, but THU-2's gap-induced repeatability is unquantified.","rationale":"The paper is a component-level progress report; its central claim is a future 50 ug target rather than a demonstrated end-to-end result. That target corresponds to 5e-8 relative in the mass equation, so the quantity that must be stable and self-consistent at that level is Bl between the weighing and velocity phases. The most concrete threat identified in the manuscript itself is the split-surface air gap: Section III-A reports that a 0.14 mm gap in THU-1 produced considerable ripple in the profile, and attributes THU-2's improvement to tighter assembly and a thinner coating, but gives no quantitative residual or repeatability data. Because the split magnet is meant to be opened routinely, gap width after reassembly directly controls magnetic reluctance and hence Bl at the coil. A gap change that is not identical in the weighing and velocity phases is not canceled by the symmetric current scheme in Eqs. (1)-(5); it enters as a first-order mass bias. The Hall-probe radius mismatch (81.5 vs 80.5 mm) is secondary because the actual velocity measurement uses the coil itself, but it weakens the component-level evidence for flatness at the coil. The proposed open/close repeatability test would settle whether the operating state is reproducible enough to support the target. The reader's weakest assumption captures the same concern; I differ only in de-emphasizing the radius mismatch. The conditional verdict is appropriate: the concern is real but addressable in the planned integration and measurement campaign.","tokens_in":9428,"tokens_out":14336,"duration_ms":143689,"concrete_test":"On magnet THU-2, place a Hall probe at r=80.5 mm and the planned weighing height; perform 10 open/close cycles, recording Bl after each closure and during a simulated ±5 mm velocity scan. If the cycle-to-cycle standard deviation of Bl at the weighing position, or the difference between static and moving profiles, exceeds 5e-8 relative, then the gap-reproducibility assumption behind Eq. (5) is falsified at the 50 ug target level.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section III-A explicitly shows that a 0.14 mm residual air gap in magnet THU-1 produces 'considerable ripple' in the magnetic profile, and attributes THU-2's improvement only to a thinner Ni coating and tighter screws. No residual-gap measurement, no profile uncertainty, and no open/close repeatability data are given. Because the split design is intended for routine opening to insert or remove the coil, gap width after reassembly is effectively a free parameter that controls magnetic reluctance and hence Bl at the coil. In Eq. (5), m is proportional to Bl(I+ - I-)/g; the 50 ug/1 kg target requires Bl to be reproducible between the fixed-magnet weighing phase and the moving-magnet velocity phase at the 5e-8 relative level. A gap-induced Bl change that differs between phases is not canceled by the symmetric current treatment in Eqs. (1)-(4); it enters as a direct mass bias. The r=81.5 mm Hall-probe profile being assumed representative of the r=80.5 mm coil is a secondary weakness, since the in-situ U/v measurement uses the actual coil, but it further weakens the component-level evidence for flatness at the operating radius.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript reports progress over two years on the Tsinghua tabletop Kibble balance, a compact instrument targeting mass calibration from 10 g to 1 kg with an uncertainty below 50 µg. The paper summarizes component-level developments: a split permanent-magnet system using Sm2Co17 and NdFeB with inner-yoke compensation, claimed to increase the vertical field-uniformity range by over 50%; an absolute gravity determination at the site, g0 = (980 139 580.5 ± 1.4) µGal with a combined 5.4 µGal (k = 2) uncertainty; a capacitive-sensor weighing cell with 0.35 nm displacement resolution and 275 N/m stiffness; a two-stage bipolar current source with nA/A-level stability in about 30 minutes; and an interferometer design for velocity measurement. The measurement equations are the standard one-mode, two-phase Kibble balance relations. No end-to-end mass measurement is reported; the paper is explicitly an integration-status update.","tokens_in":9656,"tokens_out":5845,"duration_ms":47258,"significance":"If the component results withstand scrutiny, they represent useful progress toward a compact, open-hardware Kibble balance. The paper's strengths include measured magnetic profiles (despite missing uncertainty bars), a gravity transfer with two independent analysis methods and stated uncertainties, an Allan-deviation characterization of the current source, and a clear description of the split-magnet concept with its attractive/repulsive force scheme. The open-hardware commitment is a practical contribution. The principal limitation is that the load-bearing reproducibility of the split magnet's flux integral across open/close cycles is not quantified, which is essential for the 5e-8-level Bl consistency required by Eq. (5). The paper is therefore best read as an honest progress report rather than a demonstration of the target uncertainty.","major_comments":[{"comment":"The split-magnet design is intended for routine opening and closing to insert or remove the coil, and the paper acknowledges in Section III-A that a 0.14 mm air gap in magnet THU-1 produced 'considerable ripple' above the splitting surface. For THU-2, however, no residual-gap measurement, no open/close repeatability data, and no uncertainty on the magnetic profile are reported. The mass equation (5) requires the flux integral Bl to be identical between the weighing phase (magnet fixed) and the velocity phase (magnet moving) at roughly the 5e-8 relative level for the 50 µg/1 kg target. A gap-induced change in Bl that differs between the two phases is not canceled by the symmetric current treatment in Eqs. (1)-(4) and enters directly as a mass bias. The claim that the split design provides a 'precision link' between the phases needs quantitative support, such as repeated Bl or flux-integral measurements over multiple open/close cycles and a gap-sensitivity uncertainty contribution.","section":"Section III-A and Eq. (5)"},{"comment":"The central quantitative claim that the inner-yoke compensation increases the uniform field range by over 50% is supported only by profiles plotted without error bars or repeatability trials. The difference between the compensated and uncompensated curves could be comparable to measurement scatter. Please provide the measurement uncertainty of the Hall-probe profiles, or repeated-profile data, and report the field range with a stated criterion, for example ΔB/B below 1e-4, rather than by visual inspection.","section":"Section III-A, Fig. 2(b)"},{"comment":"The measured magnetic profile is taken at mean radius ra = 81.5 mm, whereas the coil mean radius is 80.5 mm. Since the paper uses this profile as evidence of field flatness at the coil, the 1 mm radial offset needs to be justified, for example by showing the radial field gradient or by measuring at 80.5 mm. Without this, the component-level evidence for flatness at the operating radius is incomplete, even though the eventual U/v velocity measurement will use the true coil and can in principle bypass this particular limitation.","section":"Section III-A"},{"comment":"The comparison of the capacitor sensor and optical sensor is presented with normalized signals, but no noise bandwidth, measurement time, or repeatability is given, and no error bars appear. The conclusion that the capacitor sensor has 'significantly improved position measurement resolution' is the design rationale for the weighing cell, but the quantitative support is insufficient. The paper should report displacement noise density or Allan deviation for both sensors under the same conditions, and derive the force resolution using Eq. (6).","section":"Section III-C, Fig. 4(b)"}],"minor_comments":[{"comment":"The word 'geometical' should be 'geometrical' in the text near Eq. (1).","section":"Section II"},{"comment":"The term 'Allen deviation' should be 'Allan deviation'.","section":"Section III-D and Fig. 6(b)"},{"comment":"The phrase 'The coil concern cube' appears to be a typo for 'corner cube'; please correct it.","section":"Section III-E"},{"comment":"The FEA curve labeled 'proposed design with air gap' should specify the gap width used, so that it can be compared directly with the THU-1 measurement at 0.14 mm.","section":"Fig. 2(b)"},{"comment":"The statement that higher displacement sensitivity 'can reduce the sensitivity requirement of the flexure hinge' is unclear; the phrase 'sensitivity requirement' should be defined or rephrased.","section":"Section III-C"},{"comment":"The sentence 'The anticipation is the creation of a precision-oriented and robust mass-realization instrument in the near future' is awkwardly phrased and should be rewritten for clarity.","section":"Section IV"}],"recommendation":"major_revision","confidential_remarks":"The paper is a progress report and does not claim an end-to-end result. The split-gap reproducibility issue is the main technical risk and is fixable by a relatively straightforward repeatability experiment, so I would frame the revision request as a need for additional data rather than a conceptual flaw. The heavy reliance on the authors' own accepted companion papers is acceptable, but the present paper should stand alone for the claims it makes. Overall, the manuscript fits the journal's scope and the component-level work is of interest to the Kibble-balance community."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a solid component-level progress report, not a completed metrology result. The magnet work is the real news — first measured demonstration of the inner-yoke field-flattening idea, plus a split magnet that can be opened and closed by hand thanks to the attractive-to-repulsive force transition. The gravity tie is credible, with two independent analysis methods and a stated 5.4 µGal (k=2) budget, and the two-stage current source shows nA/A stability in the Allan deviation data. The capacitive weighing cell prototype is also plausible, with 0.35 nm displacement resolution and 275 N/m stiffness.\n\nWhat is not here yet is the end-to-end 50 µg/1 kg result. The paper is honest that the components are not integrated. That is fine for a progress report. The soft spot worth naming is the split magnet. The paper itself shows that a 0.14 mm air gap in THU-1 produces 'considerable ripple' in the field profile, and attributes THU-2's improvement to a thinner coating and tighter screws, but gives no residual-gap measurement, no profile uncertainty, and no open/close repeatability data. This matters because the magnet is explicitly designed for routine opening to insert or remove the coil. Gap width after reassembly directly changes the reluctance and hence Bl at the coil. Equation (5) needs Bl to be the same in the weighing and velocity phases; a gap-induced Bl shift that differs between the two phases is a direct mass bias, not canceled by the symmetric current treatment. For the 50 µg/1 kg claim, that means Bl reproducibility at the 5e-8 relative level, and there is no evidence yet of that. The r=81.5 mm Hall profile being taken as representative of the r=80.5 mm coil is a secondary concern; the in-situ U/v measurement will use the real coil, but the component-level flatness evidence is weakened.\n\nThe comparisons in Fig. 2(b) and 4(b) also lack error bars, which makes some of the improvement claims hard to judge. These are fixable, but they should be addressed.\n\nOverall: the component work is real, the citations are fair (the inner-yoke idea is self-cited but tested here with new FEA and measurements), and the measurement equations are standard with no fitted parameters. I would bring it to the reading group as a useful status report for people tracking tabletop Kibble balances and compact magnet design. It deserves a serious referee — a good journal should send it out, with the expectation that the gap-repeatability question and missing error bars are handled in revision.","headline":"Solid component progress on a tabletop Kibble balance, but the split-magnet gap repeatability is the unresolved risk that stands between the parts and the 50 µg promise.","tokens_in":10195,"tokens_out":3527,"would_cite":true,"duration_ms":32071,"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 Tsinghua tabletop Kibble balance reports that its magnet, gravity measurement, weighing cell, and current source now meet the component-level targets needed to calibrate 10 g to 1 kg masses with uncertainty below 50 µg.","keywords":["Kibble balance","mass realization","tabletop instrument","Planck constant","permanent magnet system","gravitational acceleration","capacitive weighing cell","current source"],"falsifier":"Perform a 1 kg mass calibration with the fully integrated balance and compare the result with an independent traceable mass value established by a separate primary realization method; a disagreement beyond the claimed 50 µg uncertainty would refute the central target. A more direct component check is to measure the magnetic profile at the coil radius of 80.5 mm while the magnet is closed, compare it with the 81.5 mm profile, then repeat after repeated open/close cycles with various screw torques: if a gap-induced ripple appears, the weighing-velocity equivalence assumed in Eq. (5) fails.","tokens_in":9249,"feed_emoji":"⚖️","tokens_out":11898,"duration_ms":91092,"temperature":0.7,"pith_summary":"This paper reports component-level progress on a compact, tabletop Kibble balance whose stated goal is to calibrate masses from 10 g to 1 kg with measurement uncertainty below 50 µg. It argues that the four hardest subsystems now meet their share of that target: a permanent magnet with $Bl \\approx 400\\,\\mathrm{Tm}$ and a field-uniformity range increased by more than 50% through inner-yoke compensation, a local gravity value with 5.4 µGal expanded uncertainty, a capacitive weighing cell resolving 0.35 nm of displacement at 275 N/m stiffness, and a two-stage current source with nA/A-level stability in about 30 minutes. A sympathetic reader should care because a compact, open-hardware Kibble balance would make primary mass realization practicable beyond large national metrology laboratories, extending the revised SI kilogram definition into routine calibration work.","feed_headline":"Tabletop Kibble balance targets 50-microgram mass calibration","feed_subtitle":"Magnet, gravity, weighing cell, and current source each meet their share of a compact kilogram-realization instrument.","key_machinery":"The load-bearing object is the one-mode, two-phase (OMTP) Kibble balance measurement equation, Eq. (5): $m = ((U_+/v_+ + U_-/v_-)(I_+ - I_-))/(2g) - (\\Delta f_+ - \\Delta f_-)/g$. It links a mass to voltage, velocity, current, and gravity through a single magnetic geometric factor $Bl$ that is assumed identical in the weighing and velocity phases. The paper's key mechanical innovation is inner-yoke compensation: 0.4 mm-thick, 5 mm-high rectangular additions at the two ends of the inner yoke flatten the vertical magnetic profile and enlarge the uniform field region by over 50% at almost no manufacturing cost. The other carrying components are the split permanent magnet whose upper segment can be hand-opened and magnetically levitated, the capacitive-sensor weighing cell that allows a stiffer 275 N/m flexure hinge while retaining nanometre displacement resolution, and the two-stage current source that suppresses output drift to nA/A level.","core_discovery":"The central claim is that the Tsinghua tabletop Kibble balance has advanced from design to working subsystems, with each key ingredient of the Kibble measurement equation now demonstrated. Using the one-mode, two-phase (OMTP) scheme, the mass is determined from $m = ((U_+/v_+ + U_-/v_-)(I_+ - I_-))/(2g) - (\\Delta f_+ - \\Delta f_-)/g$, so the experiment's task is to make the magnetic geometric factor $Bl$, the currents $I_\\pm$, the voltages $U_\\pm$, the velocities $v_\\pm$, the gravity $g$, and the residual forces $\\Delta f_\\pm$ consistent. The paper reports a magnet pair with $Bl \\approx 400\\,\\mathrm{Tm}$ and a measured field profile whose uniform range grows by over 50% when thin rectangles are added to the inner yoke, a gravity value of $g_0 = (980\\,139\\,580.5 \\pm 1.4)\\,\\mu\\mathrm{Gal}$ with a final expanded uncertainty of 5.4 µGal, a weighing cell with 0.35 nm displacement resolution and 275 N/m stiffness, and a two-stage bipolar current source with nA/A-class short-term stability. The paper concludes that these components can be integrated into a compact instrument for mass realization from 10 g to 1 kg with uncertainty below 50 µg.","pith_inferences":["The paper does not yet demonstrate the end-to-end 50 µg figure; that claim will stand or fall on integration, especially on whether the magnetic profile measured at 81.5 mm mean radius matches the field seen by the coil at 80.5 mm during closed operation.","A testable extension is to cycle the split magnet open and closed with different screw torques and coating thicknesses, then compare integrated $Bl$ from velocity runs with the weighing-phase value; the ripple seen with a 0.14 mm gap in the first magnet should reappear if gap control is the limiting factor.","If the inner-yoke compensation is shape-robust, the same 0.4 mm by 5 mm rectangle geometry could be tried on other compact magnet designs without full re-optimization, giving a quick uniformity boost to tabletop Kibble balances.","The gravity transfer method, with its referenceless polynomial tide fit, could be checked by a future absolute gravimeter campaign at the same site; a drift in $g_0$ beyond the quoted 5.4 µGal expanded uncertainty would propagate directly into mass readings."],"forward_implications":["If the component results carry through integration, the assembled balance will calibrate masses from 10 g to 1 kg with uncertainty below 50 µg, making the kilogram definition accessible in a compact instrument.","The inner-yoke compensation, verified experimentally for the first time, can be applied to other permanent-magnet Kibble balances as a low-cost way to widen the usable field region.","The magnet's easy split-open operation reduces the maintenance burden of coil changes, supporting the project's open-hardware goal of letting other laboratories copy the design.","The capacitive weighing cell's combination of 0.35 nm resolution and 275 N/m stiffness points toward force comparators with several-kilogram dead-load capacity and microgram-level repeatability.","The two-stage current source demonstrates that nA/A-level stability can be reached in about 30 minutes with commercial instruments, meeting the velocity-phase stability requirement of the OMTP scheme."],"supporting_citations":[{"why":"Supplies the one-mode, two-phase measurement scheme on which Eqs. (1)-(5) are based.","marker":"[8]"},{"why":"Defines the original Tsinghua tabletop design and the open-hardware, compact-balance objectives being updated.","marker":"[17]"},{"why":"Analyzes the coil-current effect whose suppression is required for the current-symmetry cancellation in Eq. (2).","marker":"[21]"},{"why":"Reviews magnet-system uncertainties for compact Kibble balances and motivates the field-uniformity and gap-control work.","marker":"[22]"},{"why":"Details the design, manufacture, assembly, and testing of the two compact magnet systems, including the measured magnetic profiles and $Bl$ values.","marker":"[23]"},{"why":"Proposes the inner-yoke compensation method that this paper applies and experimentally verifies for the first time.","marker":"[24]"},{"why":"Determines the local absolute gravitational acceleration, vertical gradient, and associated uncertainty used as $g$ in the mass equation.","marker":"[25]"},{"why":"Describes the two-stage bipolar current source whose nA/A stability is reported here.","marker":"[27]"},{"why":"Provides the interferometric displacement measurement scheme used for the $U/v$ measurement.","marker":"[28]"}],"fun_headline_variants":["Tsinghua tabletop Kibble balance: all main subsystems now proven","Compact Kibble balance: magnet, gravity, weighing, current all proven","Tsinghua's tabletop Kibble balance targets sub-50 µg from 10 g to 1 kg","Open-hardware Kibble balance: compact kg realization within reach"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing assumption is that the magnetic field profile measured at the 81.5 mm mean radius is the same field the coil actually experiences at 80.5 mm during both weighing and velocity phases, with the magnet's split seam closed tightly enough to add no field ripple; if the gap changes the field seen in one phase but not the other, the mass equation is biased at the microgram level.","fun_headline_variants_meta":{"raw":{"variants":["Tsinghua tabletop Kibble balance: all main subsystems now proven","Compact Kibble balance: magnet, gravity, weighing, current all proven","Tsinghua's tabletop Kibble balance targets sub-50 µg from 10 g to 1 kg","Open-hardware Kibble balance: compact kg realization within reach"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001428,"raw_usage":{"total_tokens":5790,"prompt_tokens":1005,"completion_tokens":4785,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":621,"completion_tokens_details":{"reasoning_tokens":4700}},"tokens_in":621,"tokens_out":4785,"duration_ms":30539,"temperature":1.0,"reasoning_tokens":4700,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T13:59:41.911754+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Perform a 1 kg mass calibration with the fully integrated balance and compare the result with an independent traceable mass value established by a separate primary realization method; a disagreement beyond the claimed 50 µg uncertainty would refute the central target. A more direct component check is to measure the magnetic profile at the coil radius of 80.5 mm while the magnet is closed, compare it with the 81.5 mm profile, then repeat after repeated open/close cycles with various screw torques: if a gap-induced ripple appears, the weighing-velocity equivalence assumed in Eq. (5) fails.","supporting_citations":[{"cited_title":"The BIPM Kibble balance for realizing the kilogram definition,","cited_arxiv_id":null,"evidence_quote":"Supplies the one-mode, two-phase measurement scheme on which Eqs. (1)-(5) are based."},{"cited_title":"Design of the Tsinghua openhardware tabletop Kibble balance,","cited_arxiv_id":null,"evidence_quote":"Defines the original Tsinghua tabletop design and the open-hardware, compact-balance objectives being updated."},{"cited_title":"Coil-current effect in Kibble balances: analysis, measurement, and optimization,","cited_arxiv_id":null,"evidence_quote":"Analyzes the coil-current effect whose suppression is required for the current-symmetry cancellation in Eq. (2)."},{"cited_title":"The irony of the magnet system for Kibble balances—a review,","cited_arxiv_id":null,"evidence_quote":"Reviews magnet-system uncertainties for compact Kibble balances and motivates the field-uniformity and gap-control work."},{"cited_title":"A compact magnet system for the Tsinghua tabletop Kibble balance,","cited_arxiv_id":null,"evidence_quote":"Details the design, manufacture, assembly, and testing of the two compact magnet systems, including the measured magnetic profiles and $Bl$ values."},{"cited_title":"A simple improvement for permanent magnet systems for Kibble balances: More flat field at almost no cost,","cited_arxiv_id":null,"evidence_quote":"Proposes the inner-yoke compensation method that this paper applies and experimentally verifies for the first time."},{"cited_title":"A determination of the local gravi- tational acceleration for the Tsinghua tabletop Kibble balance,","cited_arxiv_id":null,"evidence_quote":"Determines the local absolute gravitational acceleration, vertical gradient, and associated uncertainty used as $g$ in the mass equation."},{"cited_title":"A bi-polar current source with high short-term stability for Tsinghua tabletop Kibble balance,","cited_arxiv_id":null,"evidence_quote":"Describes the two-stage bipolar current source whose nA/A stability is reported here."},{"cited_title":"A new interferomet- ric system for the BIPM Kibble balance,","cited_arxiv_id":null,"evidence_quote":"Provides the interferometric displacement measurement scheme used for the $U/v$ measurement."}],"review_version":1}