{"id":"c8c076be-9c5d-4f2a-adb9-0727ef81f343","arxiv_id":"1908.02418","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A feedback-sustained MEMS mode pumps four other modes through simultaneous internal resonances, forming a five-mode frequency-locked comb.","lead":"A MEMS beam with five mechanical modes was driven into self-oscillation in its second flexural mode, pulling the other modes into 1:2 and 1:3 frequency ratios to generate a five-mode frequency-locked comb. The work shows that internal resonances can turn a single feedback-driven oscillator into a multimode frequency reference on a chip.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper asserts a five-mode frequency-locked comb, but its own phase data (Fig. 3d) show modes 1 and 3 are not phase-coherent, and no frequency-ratio error analysis is provided; the frequency-locked part of the central claim is therefore unsupported.","rationale":"The reader's CONDITIONAL verdict is appropriate, but the most load-bearing gap is slightly different from the reader's stated weakest assumption. The filter-isolation concern is real, but it concerns the excitation mechanism; the more central gap is the frequency-locked nature of the five-mode comb. The paper's own evidence shows phase incoherence in the odd modes and uses the word 'prospect' for a constant ratio with unlocked phase, which is a self-acknowledged limitation. The strongest claim in the abstract ('five modes frequency-locked comb is generated') is stronger than what the data establish. A single numerical analysis of the recorded phases and ratios would settle it. I give credit for the direct experimental observations (Duffing plateaus, Mathieu threshold fit, Lissajous figures) which support internal resonances and simultaneous mode activation; the problem is the 'frequency-locked comb' wording, not the observations themselves. If the proposed test shows stationary ratios, the abstract is fine; if not, the contribution should be described as multimodal internal resonances without claiming a locked comb.","tokens_in":7191,"tokens_out":13971,"duration_ms":161151,"concrete_test":"From the raw time series behind Figs. 3(c)-(d), demodulate each mode with a phase-locked reference at f2 (or use Hilbert transforms) and compute the instantaneous phase differences φ4 − 3φ2, φtor − 2φ2, φ1 − 0.5φ2, and φ3 − 1.5φ2 over the whole recorded interval, together with frequency-ratio estimates fk/f2 from zero-crossings or short-time FFTs. Check whether all wrapped phase differences are stationary within noise while f2 is swept through the locking range, and whether each fk/f2 stays within the mutual locking bandwidth of the rational value. If any phase difference wraps monotonically or any ratio drifts outside the bandwidth, the 'frequency-locked comb' claim is not supported and the paper should be revised to a weaker claim of concurrently excited near-commensurate modes.","verdict_should_be":"UNCHANGED","load_bearing_attack":"To support a frequency-locked comb, one must show the five modal frequencies maintain the claimed commensurate ratios (f4 ≈ 3f2, ftor ≈ 2f2, f1 ≈ f2/2, f3 ≈ 3f1 ≈ 3f2/2) over the operating interval. The paper never plots these ratios or their deviations. The body instead reports that after the 2:1 parametric transition, modes 1 and 3 'do not show a very constant phase relation' and their phase differences have error bars spanning [−π, π] (Fig. 3d). The authors call this a 'time-dependent phase difference' and, in the final paragraph, offer only the 'additional prospect of having the even and odd modes in a constant frequency ratio, but with an unlocked phase.' For an exactly commensurate ratio n:m, the wrapped combination nφ2 − mφk is constant when noiseless and bounded under phase locking; full-range phase wandering is evidence that the frequencies are not actually locked in the ratio, or that frequent cycle slips occur. Hence the abstract's central claim that 'a five modes frequency-locked comb is generated' rests on a frequency-locking assertion that the data do not verify and that the text itself hedges to a 'prospect.' This is load-bearing: if the odd-mode frequencies merely sit near, but not in, the rational ratios, the headline object is a set of concurrently excited modes rather than a frequency-locked comb.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports a MEMS clamped-clamped beam oscillator in which a feedback loop is closed around the second flexural mode, driving it into self-oscillation. The authors observe that, as the oscillator frequency is pulled by tuning the band-pass filter, the second mode engages in a 1:3 internal resonance with the fourth flexural mode, a 2:1 parametric interaction with the first flexural mode, and a nominally 2:1 interaction with a torsional mode, while the first and third modes also interact. Plateaus in the frequency-amplitude curves, a Mathieu-type parametric threshold fit, and Lissajous figures are presented as evidence of internal resonances. The central claim, stated in the abstract and conclusion, is that this constitutes a 'five modes frequency-locked comb' generated from a single self-sustained mode. The paper also proposes a reduced two-mode analytical model for the second and fourth modes.","tokens_in":7576,"tokens_out":4691,"duration_ms":54631,"significance":"If fully substantiated, the demonstration of simultaneous multiple internal resonances in a self-sustained MEMS oscillator would be a useful contribution to the growing literature on mechanical frequency combs and multimodal MEMS oscillators. The paper has clear strengths: the experimental observation of amplitude plateaus indicating internal resonances, the extraction of a three-wave coupling coefficient from the parametric threshold, and the phase-anchoring behavior of the even and torsional modes. These provide credible evidence for internal-resonance-mediated energy transfer. However, the paper's headline claim of a 'frequency-locked comb' is not established by the presented data, and the analytical model is explicitly compared only after a free scaling parameter. The work is therefore of interest but requires substantially strengthened evidence before the central claim can be accepted.","major_comments":[{"comment":"The central claim that a 'five modes frequency-locked comb' is generated is not supported by the data and is internally inconsistent with the text's own phase analysis. In Fig. 3(d), the phase differences involving modes 1 and 3 have error bars spanning the full [−π, π] range and are described as a 'time-dependent phase difference'; the final paragraph explicitly offers only the 'additional prospect of having the even and odd modes in a constant frequency ratio, but with an unlocked phase.' Furthermore, the quoted frequencies do not by themselves establish exact commensurability: with f2 = 519 kHz and ftor = 1083 kHz, ftor/f2 ≈ 2.09, not 2.00, and f1/f2 ≈ 0.61, not 0.50. To support the abstract's claim, the authors should report measured frequency-ratio deviations over the operating interval (for example |2f1 − f2|/f2, |3f1 − f3|/f3, |2f2 − ftor|/ftor, and |3f2 − f4|/f4) together with a phase-coherence statistic. As written, the data demonstrate concurrent excitation of five modes, but not a frequency-locked comb.","section":"Abstract and final paragraph; Fig. 3(d)"},{"comment":"The interpretation that modes 1, 3, 4, and the torsional mode are excited only through internal resonances and parametric pumping depends on the band-pass filter in the feedback loop rejecting direct excitation of those modes at all operating frequencies. The paper states that the filter is inserted 'to insure that only the second mode goes into self-oscillation' but does not report the filter's measured attenuation at 319, 953, 1564, and 1083 kHz, nor a control experiment with the loop open or with the filter center frequency far from the second mode. Without such data, electrical feedthrough or filter leakage could in principle contribute to the observed multimode spectra, which would weaken the internal-resonance attribution. This point should be addressed quantitatively.","section":"Feedback-loop filter assumption, Section II and Fig. 2"},{"comment":"The analytical model is presented as supporting the frequency-pulling behavior, but the comparison in Fig. 2(c) is made after stating that 'the quantitative response is left as a free parameter.' This means the model curve is not a parameter-free prediction and cannot quantitatively validate the mechanism. In addition, the model produces a negative-detuning branch that is not observed experimentally, and the fourth-mode data depart strongly from the model near its natural frequency. The authors should state explicitly which parameter is being used as the free scaling factor, how the model curve is normalized, and what quantitative or falsifiable statement the model is intended to support. If the model is meant only as a qualitative scaling argument, the text should say so and the analytical support for the comb claim should be downgraded accordingly.","section":"Eq. (2) and Fig. 2(c)"}],"minor_comments":[{"comment":"There is a typographical error: 'reasonnable' should be 'reasonable.'","section":"Final paragraph"},{"comment":"The phrase 'To use the second fluxural mode' contains a typographical error: 'fluxural' should be 'flexural.'","section":"Section II, text near Fig. 2"},{"comment":"The notation in Eq. (1) is ambiguous: the Hamiltonian Hjk is defined only through the sum of a three-wave and a four-wave term, but the equation then uses ∂Hjk/∂xk without specifying the sign convention or showing how the derivative acts on both terms. Please write out the resulting coupling forces explicitly.","section":"Eq. (1)"},{"comment":"The insets in Fig. 2(b) are described in the caption as showing the change in the Brownian response, but the main text says the quality factor is plotted as a function of loop gain. Please clarify what quantity is shown in the insets and on which axis.","section":"Fig. 2(b)"},{"comment":"The torsional mode is quoted at 1083 kHz while the second flexural mode is quoted at 519 kHz; since 1083/519 ≈ 2.09, the text's statement that this corresponds to a 1:2 ratio with the second flexural mode requires an explicit detuning value or a correction, because the nominal frequencies do not satisfy the ratio.","section":"Fig. 1 and Section II"}],"recommendation":"major_revision","confidential_remarks":"The paper builds heavily on the authors' prior work [12] and the novelty relative to that work should be stated more crisply. The main concern, however, is substantive: the abstract's 'frequency-locked comb' claim goes beyond what the phase data and frequency-ratio information support. I would recommend that the editors require either a quantitative frequency-ratio and phase-coherence analysis or a careful rewriting of the claim to describe concurrent multimode excitation with unlocked odd-mode phases."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick read: the experiment is real and the multimode internal-resonance configuration is new, but the 'five-mode frequency-locked comb' in the abstract is bigger than what the data show. The paper's own phase data and final paragraph undercut that phrase.\n\nHere's what is actually new: simultaneous 1:2, 1:3, and 2:1 internal resonances in a self-sustained MEMS oscillator, with even-odd mode coupling achieved by breaking the longitudinal symmetry. The plateaus in Fig. 1, the parametric threshold fit in Fig. 3(b), and the phase anchoring of modes 4 and the torsional mode after the 2:1 transition are convincing experimental fingerprints. The analytical reduction is honest: the two-mode model captures qualitative scaling, the Mathieu equation fits the parametric onset, and the authors state plainly that the model breaks down when all five modes interact.\n\nThe load-bearing soft spot is the 'frequency-locked' assertion. No frequency-ratio error analysis is provided, and the phase differences for modes 1 and 3 wander over the full [-pi, pi] range after parametric onset. In an exact commensurate lock, the wrapped phase combination should stay constant or bounded; full-range wandering means the odd modes are not locked to the even modes in the stated ratios, at least not in the sense the abstract implies. To their credit, the authors admit the odd modes show no constant phase relation and end with a 'prospect' of a constant ratio with unlocked phase. That is a reasonable interpretation, but it is not what the abstract says. Also, the band-pass filter leakage is an assumption; direct electrical feedthrough to other modes is not fully ruled out, though the threshold behavior suggests internal coupling dominates.\n\nMinor concerns: the quantitative model agreement uses a free scaling parameter, which is disclosed, and the self-citation to [12] is appropriate since this builds directly on that work.\n\nBottom line: this is a solid experimental contribution that deserves a serious referee, but the claims need revision. Either measure and report the frequency-ratio deviations across the operating band, or reword the central claim to 'near-commensurate multimode oscillation' or 'multimode frequency comb with unlocked odd-mode phases.' I would send it out, with a request for those changes.","headline":"Real experimental novelty in multimode internal resonance, but the 'five-mode frequency-locked comb' claim outruns the phase data and needs either new measurements or softer wording.","tokens_in":8093,"tokens_out":2419,"would_cite":true,"duration_ms":25105,"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 single self-oscillating MEMS mode can lock four other vibration modes—three flexural and one torsional—into a five-mode frequency-locked comb through internal resonances.","keywords":["internal resonance","MEMS oscillator","multimode frequency comb","self-sustained oscillator","parametric resonance","mode coupling","Duffing oscillator","symmetry breaking"],"falsifier":"With the loop gain held just below the self-oscillation threshold and the filter centered on the second mode, measure the four other modes: if any of them appears at measurable amplitude, direct electrical driving is contaminating the supposed internal-resonance comb.","tokens_in":6990,"feed_emoji":"⚙️","tokens_out":10927,"duration_ms":109936,"temperature":0.7,"pith_summary":"The paper reports a microelectromechanical beam oscillator in which one gain-feedback loop drives the second flexural mode into self-oscillation, and nonlinear internal resonances then lock four other modes to it at rational frequency ratios. The couplings include 1:2 and 1:3 internal resonances plus a 2:1 parametric pumping of the first flexural mode, so a single pump produces a five-mode frequency-locked pattern. Because the beam's longitudinal symmetry is deliberately broken, even and odd mode families can interact, which is what makes the simultaneous multimodal locking possible. The paper's conclusion is that one feedback-driven mode can generate a five-mode frequency-locked comb purely from internal resonances.","feed_headline":"One MEMS mode locks four others into a five-mode comb","feed_subtitle":"Internal resonances at 1:2, 1:3 and 2:1 let one feedback mode pump four others into lock at once.","key_machinery":"The load-bearing device is the modal-interaction Hamiltonian of Eq. (1), written $H_{jk} = \\kappa_{jk} x_k^2 x_j + g_{jk} x_k^3 x_j$, together with a coupled Duffing equation per mode: the $\\kappa_{jk}$ terms are three-wave mixing couplings and the $g_{jk}$ terms are four-wave mixing couplings. These coupling terms are what transfer energy between modes whose natural frequencies sit near rational ratios. On the experimental side, the feedback loop consists of a laser Doppler vibrometer readout passed through a filter-amplifier and back to the device, with the band-pass filter intended to let only the second flexural mode self-oscillate and with loop gain as the control parameter; the single-sided DC bias is what activates even-odd coupling. For the 2:1 parametric interaction, the reduced equation $\\ddot{x}_1 + \\gamma_1 \\dot{x}_1 + (\\omega_{1,0}^2 + \\kappa_{12} A_2 \\cos(\\omega t + \\varphi_2))x_1 = 0$ is a Mathieu equation, and fitting its threshold gives the value of $\\kappa_{12}$ used to map the parametric-oscillation region.","core_discovery":"On a 150-micrometer piezoelectric clamped-clamped beam, the paper places the second flexural mode in self-oscillation through a gain-feedback loop, with a steep band-pass filter selected so that only that mode is driven directly. Under a DC bias that breaks the beam's longitudinal symmetry, the first four flexural modes (near 319, 519, 953, and 1564 kHz at low drive) and the first torsional mode (near 1083 kHz) are coupled by 1:3, 1:2, and 2:1 internal resonances. Sweeping the filter's center frequency at constant loop gain pulls the self-oscillation frequency over roughly 30% and pumps the non-driven modes through modal interactions; near 620 kHz the first mode enters parametric oscillation through the 2:1 resonance, while the fourth and torsional modes develop a tristable phase relation and the first and third modes show a time-dependent phase. The paper concludes that one feedback-driven mode can generate a five-mode frequency-locked comb purely from internal resonances.","pith_inferences":["If the band-pass filter truly isolates the second mode, the same symmetry-breaking recipe should transfer to other multimode resonators with near-commensurate spectra, producing single-pump locked tone sets in devices beyond this particular beam.","A decisive control experiment would be to repeat the measurement with symmetric, two-sided actuation and no DC strain bias: the even-odd couplings should vanish and the first and third modes should drop out of the locked pattern.","The absence of sidebands next to the time-dependent phase of the odd modes suggests the phase instability is a slow drift rather than a conventional Hopf or SNIPER bifurcation; long time traces or phase-locked detection could distinguish these cases.","The measured phase tristability of the fourth and torsional modes is a sensitive fingerprint of the coupling parameters $\\kappa_{jk}$ and $g_{jk}$, and fitting it could replace the currently free quantitative amplitude parameter."],"forward_implications":["A mechanical frequency comb can be generated with one self-oscillating mode as the only pump, requiring no multi-tone external drive.","The self-oscillation frequency can be tuned over about 30% at constant loop gain by sweeping the filter, which opens a wide-range, single-knob tuning method for internally resonant oscillators.","Breaking longitudinal symmetry doubles the accessible interaction space by coupling even and odd mode families, allowing 1:2, 1:3, and 2:1 resonances to be active simultaneously.","Within the locked comb, the fourth and torsional modes show a tristable phase that is anchored, while the first and third modes remain frequency-locked but phase-unlocked, so spectral locking and phase coherence are independent properties.","The simplified two-mode model captures the qualitative scaling of the self-oscillating response until the 2:1 parametric onset, beyond which a full multimode treatment is needed to account for the observed dynamics."],"supporting_citations":[{"why":"It is the preceding open-loop two-mode internal-resonance study whose fitted parameters and interaction framework this work extends to a closed self-oscillating loop.","marker":"[12]"},{"why":"It supplies the nonlinear interaction terms inserted into the Duffing equations that form the paper's modal interaction Hamiltonian.","marker":"[11]"},{"why":"It provides the canonical Mathieu-equation treatment used to model and fit the 2:1 parametric resonance threshold for the first mode.","marker":"[23]"},{"why":"It demonstrates a mechanical frequency comb in MEMS, the phenomenon this paper aims to produce from one feedback-driven mode and internal resonances.","marker":"[14]"},{"why":"It reports coherent energy transfer between modes in an internally resonant self-oscillator, supporting the claim that one pumped mode can drive several others.","marker":"[17]"},{"why":"It shows that internal resonance in a self-sustained MEMS oscillator can reduce frequency noise, supplying the practical motivation for studying the feedback-loop configuration.","marker":"[3]"},{"why":"It gives the fitting procedure used to extract the linear and Duffing modal parameters that feed the model and its simulations.","marker":"[20]"},{"why":"It provides the expected quadratic jump-down frequency versus amplitude relation against which the observed plateaus are read as signatures of internal resonance.","marker":"[21]"}],"fun_headline_variants":["One MEMS mode pumps four into a five-mode comb","Five modes locked by one: MEMS internal resonances","Single feedback mode drives five-mode comb in MEMS","1:2,1:3,2:1 resonances lock five MEMS modes at once"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The band-pass filter in the feedback loop lets only the second flexural mode self-oscillate, with no electrical feedthrough or filter leakage driving the other four modes, so their oscillations must come solely from internal nonlinear coupling.","fun_headline_variants_meta":{"raw":{"variants":["One MEMS mode pumps four into a five-mode comb","Five modes locked by one: MEMS internal resonances","Single feedback mode drives five-mode comb in MEMS","1:2,1:3,2:1 resonances lock five MEMS modes at once"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000162,"raw_usage":{"total_tokens":1215,"prompt_tokens":900,"completion_tokens":315,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":516,"completion_tokens_details":{"reasoning_tokens":240}},"tokens_in":516,"tokens_out":315,"duration_ms":3916,"temperature":1.0,"reasoning_tokens":240,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:44:28.043593+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"With the loop gain held just below the self-oscillation threshold and the filter centered on the second mode, measure the four other modes: if any of them appears at measurable amplitude, direct electrical driving is contaminating the supposed internal-resonance comb.","supporting_citations":[{"cited_title":"Houri, D","cited_arxiv_id":null,"evidence_quote":"It is the preceding open-loop two-mode internal-resonance study whose fitted parameters and interaction framework this work extends to a closed self-oscillating loop."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the nonlinear interaction terms inserted into the Duffing equations that form the paper's modal interaction Hamiltonian."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It provides the canonical Mathieu-equation treatment used to model and fit the 2:1 parametric resonance threshold for the first mode."},{"cited_title":"Ganesan, C","cited_arxiv_id":null,"evidence_quote":"It demonstrates a mechanical frequency comb in MEMS, the phenomenon this paper aims to produce from one feedback-driven mode and internal resonances."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It reports coherent energy transfer between modes in an internally resonant self-oscillator, supporting the claim that one pumped mode can drive several others."},{"cited_title":"Antonio, D","cited_arxiv_id":null,"evidence_quote":"It shows that internal resonance in a self-sustained MEMS oscillator can reduce frequency noise, supplying the practical motivation for studying the feedback-loop configuration."},{"cited_title":"Davidovikj, F","cited_arxiv_id":null,"evidence_quote":"It gives the fitting procedure used to extract the linear and Duffing modal parameters that feed the model and its simulations."},{"cited_title":"Anderson, S","cited_arxiv_id":null,"evidence_quote":"It provides the expected quadratic jump-down frequency versus amplitude relation against which the observed plateaus are read as signatures of internal resonance."}],"review_version":1}