{"id":"9e436091-119a-4de3-8266-7fbdd88ab09b","arxiv_id":"2607.13026","paper_version":1,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":5.5,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"External MEMS microphones on the transducer track acoustic radiation force maxima within ~30 µm and retain ≥98.3% of peak force across resonance modes n=5–8.","lead":"Transducer-mounted MEMS microphones can sense acoustic field state in resonant levitators without sensors inside the cavity. This may enable compact closed-loop control of levitation resonance as conditions change.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified beyond the abstract-only limitation already flagged by the reader; the central empirical claim is concrete and internally consistent as stated.","rationale":"The reader's weakest-assumption diagnosis is exactly the load-bearing concern: generalization from controlled distance sweeps to the broader disturbance set and closed-loop operation is asserted as motivation but only sparsely evidenced in the abstract. No stronger internal flaw (e.g., an unstated boundedness assumption, circular definition of “force maximum,” or inconsistent sampling) is visible from the abstract alone. The quantitative thresholds (30 µm, 98.3%) are concrete and falsifiable once the full data appear; until then the CONDITIONAL / LOW-confidence verdict is the correct holding position. No adjustment is warranted.","tokens_in":2131,"tokens_out":512,"duration_ms":4990,"concrete_test":"When the full text becomes available, extract the raw distance-sweep data for modes n=5–8 and recompute the offset between channel-mean microphone-voltage maxima and balance-measured force maxima; confirm that every mode satisfies ≤30 µm and ≥98.3% force retention, and that the microphone peak width is narrower than the current peak width. If any mode fails either threshold, the headline claim weakens.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's strongest claim is a controlled empirical co-location result: under transducer-reflector distance sweeps for modes n=5–8, channel-mean MEMS microphone voltage maxima lie within two sampled increments (≤30 µm) of measured acoustic radiation force maxima and retain ≥98.3% of peak force, while localizing more sharply than peak-to-peak transducer current. That claim is self-contained, quantitative, and does not rest on an unstated mathematical assumption or circular derivation. The only material soft spot is the one the reader already identified: the abstract reports the force/current comparison solely on distance sweeps, with only limited one-off phase, envelope, and tilt experiments; it does not demonstrate that the same co-location holds under the full set of operating disturbances (wavelength, temperature, object insertion, mechanical alignment) or in closed-loop use. Because the full text, methods, figures, and multi-condition data are unavailable, no deeper internal inconsistency can be verified or refuted. The claim as written therefore stands or falls on the (unseen) experimental evidence rather than on a hidden logical flaw.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The manuscript investigates transducer-mounted external MEMS microphones as non-intrusive sensors of acoustic field state in resonant acoustic levitators, without placing sensors inside the cavity. Using a linear microphone configuration, the authors perform transducer–reflector distance sweeps over resonance modes n = 5–8 and compare channel-mean microphone voltage with acoustic radiation force (precision balance) and peak-to-peak transducer current. They report that microphone-voltage maxima occur within two sampled distance increments (at most 30 µm) of the force maxima, retain at least 98.3% of the corresponding maximum force, and localize the force peak more sharply than transducer current. Limited additional experiments address phase under a frequency shift, envelope modulation during object oscillation, and channel-dependent responses under transducer–reflector tilt. The abstract frames these results as a basis for compact transducer-side feedback and notes possible transfer to transducer–transducer and array architectures.","tokens_in":2319,"tokens_out":1080,"duration_ms":25939,"significance":"If the reported co-location of external MEMS microphone amplitude maxima with independently measured radiation-force maxima holds under the stated conditions, the work offers a practical, non-intrusive relative observable for resonance-related field-state assessment in compact levitators, including architectures without a passive reflector. The central comparison is empirical and non-circular: force is an external benchmark, not derived from the microphone signal. Quantitative claims (≤30 µm co-location, ≥98.3% force retention, sharper localization than current) are concrete and falsifiable. The principle, if validated more broadly, would be useful for closed-loop operation where in-cavity sensors are undesirable.","major_comments":[{"comment":"The problem statement in the abstract motivates sensing by shifts of the optimum distance and resonant condition with wavelength, temperature, object insertion, and mechanical alignment, yet the force/current co-location evidence is reported only for controlled transducer–reflector distance sweeps (modes n = 5–8). The one-off phase, envelope, and tilt experiments described do not establish that microphone-amplitude maxima remain co-located with force maxima under those disturbances, nor do they demonstrate closed-loop use. This gap is load-bearing for the claim that the method supports reliable operation and compact transducer-side feedback under the stated operating conditions; either multi-disturbance force comparisons or a narrowed claim scope is needed.","section":"Abstract (distance-sweep results vs. disturbance framing)"},{"comment":"The abstract states that ring measurements showed channel-dependent responses under transducer–reflector tilt but “did not provide a calibrated or unique tilt estimate.” As written, this undercuts any implication that the same sensing approach currently resolves mechanical alignment. The manuscript should either supply a calibrated tilt estimator with uniqueness/error analysis or explicitly limit the tilt result to a qualitative channel-sensitivity demonstration so that the central claim remains proportionate to the evidence.","section":"Abstract (ring / tilt paragraph)"},{"comment":"The co-location bound “within two sampled distance increments, or at most 30 µm” is only as strong as the sampling grid, error bars, and force-measurement uncertainty. Without reported sampling increments, temperature control, object-insertion statistics, and force/current measurement uncertainties (methods and figures not available in this abstract-only review), it is not possible to judge whether the ≤30 µm and ≥98.3% force-retention figures are limited by sampling or by true physical offset. These quantities are load-bearing for the quantitative claim and must be fully specified and error-bounded in the full manuscript.","section":"Abstract (≤30 µm / ≥98.3% claims)"}],"minor_comments":[{"comment":"Clarify in the abstract whether “channel-mean” is an unweighted average across microphones and how many channels enter the mean for the linear configuration.","section":"Abstract"},{"comment":"State explicitly the sampled distance increment (µm) so that “two sampled distance increments, or at most 30 µm” is self-contained without requiring the full methods.","section":"Abstract"},{"comment":"The transferability remark to transducer–transducer and array architectures is reasonable as outlook but should be clearly labeled as untested so it is not read as an experimental result of this study.","section":"Abstract (closing sentence)"}],"recommendation":"major_revision","confidential_remarks":"This assessment is based solely on the abstract; the full text, methods, figures, and multi-condition data were not available. The central distance-sweep co-location claim appears concrete and non-circular as stated, and I found no internal inconsistency in the abstract. The main risk is over-claim relative to evidence (disturbances and closed-loop). If the full paper supplies rigorous methods, error bars, and either multi-disturbance force comparisons or appropriately narrowed claims, the work could be suitable after revision. Fit for physics.app-ph seems reasonable given the applied sensing focus."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The one thing worth knowing is that they put MEMS microphones on the transducer side, outside the cavity, and show that the channel-mean voltage peaks sit within two sampling steps (≤30 µm) of the independently measured radiation-force peaks for modes n=5–8, retaining ≥98.3% of peak force and localizing more sharply than transducer current. That is a concrete, useful sensor result for resonant levitation, not a physics rewrite.\n\nWhat is new is the specific packaging and the quantitative head-to-head against force and current under controlled distance sweeps. Acoustic sensing and current-based resonance indicators already exist; the contribution is the off-axis, transducer-mounted relative observable that does not put anything inside the cavity and still co-locates with force. The comparison is not circular: force comes from a precision balance, current is electrical, and the mic signal is treated as a relative indicator. The abstract is clear about the limited phase, envelope, and tilt experiments—they are proof-of-principle, not full calibration.\n\nThe soft spot is exactly the one the reader flagged and the stress-test confirmed: the load-bearing numbers are from distance sweeps only. The motivation talks about wavelength, temperature, object insertion, and alignment, plus closed-loop and reflector-free architectures, but the abstract does not show multi-disturbance validation or closed-loop data. That is a real but proportionate gap for a methods paper; it does not sink the co-location claim as stated. No free parameters, no invented entities, no formal proofs or shipped code in what we have.\n\nThis is for people building or controlling acoustic levitators who need compact, non-intrusive feedback. Worth a serious referee once the full methods, error bars, and figures are in hand. I would send it to peer review rather than desk-reject; the empirical claim is sharp enough to deserve that time even if broader validation is still needed.","headline":"Solid applied instrumentation result: external MEMS mics track force maxima within ~30 µm on distance sweeps; generalization to closed-loop and multi-disturbance use is still thin.","tokens_in":2995,"tokens_out":492,"would_cite":false,"duration_ms":5487,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"External MEMS microphones on a levitator transducer track acoustic force peaks within 30 micrometers without sensors inside the cavity.","keywords":["acoustic levitation","MEMS microphones","resonant levitators","acoustic radiation force","non-intrusive sensing","transducer-side feedback","field-state monitoring"],"falsifier":"Repeat the transducer-reflector distance sweep while simultaneously recording microphone voltage, radiation force on a balance, and peak-to-peak current; if the channel-mean microphone maximum systematically sits more than two sampling steps (greater than 30 micrometers) away from the force maximum or retains less than 98 percent of peak force for modes n=5-8, the central claim fails.","tokens_in":3010,"feed_emoji":"🎤","tokens_out":680,"duration_ms":5279,"temperature":0.7,"pith_summary":"Resonant acoustic levitators only hold objects well when the transducer-to-reflector gap and drive conditions sit at the right resonance; those conditions drift with wavelength, temperature, inserted objects, and mechanical alignment. Putting sensors inside the cavity is awkward for compact closed-loop systems and impossible for architectures that lack a passive reflector. This paper shows that microphones mounted on the transducer itself, outside the cavity, can still report the acoustic field state. In distance sweeps over several resonance modes, the mean microphone voltage peaked within two sampling steps (at most 30 micrometers) of the true acoustic radiation force maximum measured by a balance, retaining at least 98.3 percent of peak force, and localized that maximum more sharply than the usual peak-to-peak transducer current. Limited phase and envelope checks further suggest the microphones can hint at correction direction and capture object-induced field changes. The practical claim is that compact, non-intrusive transducer-side feedback becomes feasible for keeping levitators on resonance.","feed_headline":"External mics track levitator force peaks within 30 µm","feed_subtitle":"Transducer-mounted MEMS sensors keep 98% of peak force without any probe inside the cavity","key_machinery":"Linear array of transducer-mounted MEMS microphones used as off-axis external sensors: their channel-mean amplitude (and, secondarily, phase and envelope) serve as relative observables of cavity resonance state without any sensor inside the levitation volume.","core_discovery":"Transducer-mounted external MEMS microphones acquire relative acoustic signals whose channel-mean voltage maxima coincide with acoustic radiation force maxima to within two sampled distance increments (at most 30 micrometers) for resonance modes n=5-8, retaining at least 98.3 percent of the corresponding maximum force and localizing the force peak more sharply than peak-to-peak transducer current.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["MEMS mics hit levitator force peaks within 30 µm","External mics lock force max to ≤30 µm of true peak","Transducer mics retain 98.3% force at 30 µm accuracy","Off-cavity MEMS sensors find force peaks to 30 µm","Microphone voltage peaks align force max within 30 µm"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"That the co-location of microphone amplitude peaks with force peaks seen in controlled distance sweeps will still hold under the real operating disturbances (wavelength, temperature, object insertion, tilt) and in closed-loop use without sensors inside the cavity.","fun_headline_variants_meta":{"raw":{"variants":["MEMS mics hit levitator force peaks within 30 µm","External mics lock force max to ≤30 µm of true peak","Transducer mics retain 98.3% force at 30 µm accuracy","Off-cavity MEMS sensors find force peaks to 30 µm","Microphone voltage peaks align force max within 30 µm"]},"model":"grok-4.5","effort":"low","cost_usd":0.006116,"raw_usage":{"total_tokens":1586,"prompt_tokens":850,"num_sources_used":0,"completion_tokens":94,"cost_in_usd_ticks":61160000,"prompt_tokens_details":{"text_tokens":850,"audio_tokens":0,"image_tokens":0,"cached_tokens":0},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":642,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":850,"tokens_out":94,"duration_ms":4905,"temperature":1.0,"reasoning_tokens":642,"cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-15T01:27:46.461851+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Repeat the transducer-reflector distance sweep while simultaneously recording microphone voltage, radiation force on a balance, and peak-to-peak current; if the channel-mean microphone maximum systematically sits more than two sampling steps (greater than 30 micrometers) away from the force maximum or retains less than 98 percent of peak force for modes n=5-8, the central claim fails.","supporting_citations":[],"review_version":1}