Pith. sign in

REVIEW 5 major objections 4 minor 2 cited by

Towards Navigation-Grade and Deployable Optomechanical Accelerometry

T0 review · 5 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper claims that an optomechanical accelerometer built around a differential strain-sensing Mach-Zehnder interferometer on a bulk-micromachined proof mass reaches navigation-grade bias stability, with a projected path to 330 ng/√Hz.

desk verdict Strong new architecture, solid NEA/bandwidth results, but the 6.3 µg bias instability is not established as an inertial bias—navigation-grade claim is premature. read the letter →

arxiv 2505.11751 v2 pith:3NVPMPX5 submitted 2025-05-16 physics.optics physics.app-phphysics.ins-det

Chang Ge , Daniel Dominguez , Allison Rubenok , Michael Miller , Matt Eichenfield This is my paper · ORCID
classification physics.opticsphysics.app-phphysics.ins-det
keywords optomechanicalaccelerometerMach-Zehnderinterferometerbulkmicromachinedproofmassbiasinstabilitynoise-equivalentaccelerationinertialnavigationsiliconnitridewaveguidedynamicrange
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Navigation-grade accelerometers need noise near $1\ \mu\mathrm{g}/\sqrt{\mathrm{Hz}}$ and bias drift below $10\ \mu\mathrm{g}$. The paper argues that the usual optomechanical approach—tiny proof masses inside high-Q optical cavities—makes sensors too sensitive to wavelength, temperature, and package stress to hold calibration in the field. The proposed replacement is a large, rigid bulk-micromachined proof mass (93.4 kHz resonance, 22.5 pm/g displacement) read out by an integrated differential strain-sensing Mach-Zehnder interferometer (DSMZI). The authors measure $4.2\ \mu\mathrm{g}/\sqrt{\mathrm{Hz}}$ resolution over a 66 kHz bandwidth and $6.3\ \mu\mathrm{g}$ bias instability at 243 s, with a projected floor of $330\ \mathrm{ng}/\sqrt{\mathrm{Hz}}$ after better fiber coupling and balanced detection. If correct, this would make optomechanical accelerometers deployable in real environments rather than laboratory demonstrations.

What carries the argument

The load-bearing object is the DSMZI-on-BMPM: a differential strain-sensing Mach-Zehnder interferometer (DSMZI) integrated on the surface of a bulk-micromachined proof mass (BMPM), with waveguides routed along the sensing tethers so that proof-mass displacement becomes strain and a differential optical phase. The proof mass is intentionally ultra-rigid, at a 93.4 kHz fundamental resonance (22.5 pm/g), and the key identities are the sub-resonant susceptibility $\mathrm{d}x/\mathrm{d}a \approx 1/\Omega_s^2$ and the ratio $\frac{\mathrm{d}x/\mathrm{d}a}{\mathrm{d}x/\mathrm{d}F_B} = m$, which show that a heavier proof mass suppresses body-force and package-stress drift. The differential phase is $\Delta\phi = 2[k\chi a + \Delta(kL)_{PE} + \Delta(kL)_{MB}]$, where the photoelastic and moving-boundary terms capture strain-induced changes in the propagation constant. A finite-element model evaluates those terms from the exact strain profile and predicts the acceleration-to-power responsivity and noise floors. The differential geometry is what cancels common laser, thermal, and packaging drift while keeping the readout broadband.

What would settle it

Re-run the bias-instability test with balanced detection and a power-stabilized laser while continuously monitoring the laser output; if the floor near 243 s moves with laser power or disappears, the claimed $6.3\ \mu\mathrm{g}$ belongs to the optical readout, not to the accelerometer.

Watch

Extended reading notes

Core claim

The central discovery is that a navigation-grade optomechanical accelerometer does not need a high-Q optical resonator; sensitivity can come from an interferometric strain readout on a deliberately stiff proof mass. The device is a 16.7 mg silicon proof mass suspended by tethers, with a 93.4 kHz fundamental resonance; acceleration displaces it by 22.5 pm/g, and the tether strain is measured by a silicon-nitride DSMZI whose arms wrap the tethers up to four times. Because the readout is differential, common-mode effects such as laser frequency drift, temperature changes, and package stress partially cancel, and because the proof mass is large, displacement per unit body force is small, which suppresses bias drift. The measured off-resonance resolution is $4.2\ \mu\mathrm{g}/\sqrt{\mathrm{Hz}}$ over 66 kHz, the optical bandwidth is 17 nm, the temperature range exceeds 20 °C, the projected dynamic range is 165.4 dB, and the measured bias instability is $6.3\ \mu\mathrm{g}$ at 243.4 s for the two-pass device ($20.3\ \mu\mathrm{g}$ at 579.3 s for the four-pass device). The model predicts a shot-noise-limited floor of $1.15\ \mu\mathrm{g}/\sqrt{\mathrm{Hz}}$ at 10 mW per detector, so the reported performance is claimed to be limited by insertion loss and amplifier noise rather than by the sensing principle.

Load-bearing premise

The load-bearing premise is that the measured $6.3\ \mu\mathrm{g}$ Allan floor is a true acceleration bias of the sensor rather than drift of the single-ended photodiode readout, since the run had no balanced detection, no laser intensity stabilization, and ambient air.

Editorial extensions

If this is right

  • Navigation-grade thresholds—about $1\ \mu\mathrm{g}/\sqrt{\mathrm{Hz}}$ resolution and below $10\ \mu\mathrm{g}$ bias instability—become reachable for chip-scale optomechanical accelerometers without vacuum, cryogenic cooling, or laser locking.
  • Geometric scaling of the same design covers resonances from below 50 kHz to 1 MHz, giving projected resolutions from $0.259\ \mu\mathrm{g}/\sqrt{\mathrm{Hz}}$ to $31.3\ \mu\mathrm{g}/\sqrt{\mathrm{Hz}}$ for vibrometry, impact testing, and condition monitoring.
  • Because the optical bandwidth is 17 nm, the sensor can run on a simple fixed-wavelength laser, eliminating the tunable, frequency-stabilized source and control electronics that block deployment of cavity-based designs.
  • Improved fiber-chip coupling and balanced detection should lower the noise floor to about $330\ \mathrm{ng}/\sqrt{\mathrm{Hz}}$, approaching the $130\ \mathrm{ng}/\sqrt{\mathrm{Hz}}$ thermomechanical limit of the 105 kHz design.
  • The architecture points toward vibratory Coriolis gyroscopes: adding piezoelectric actuators to drive motion orthogonal to the sense mode could extend the same low-drift readout to rotation-rate sensing for GPS-free navigation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: if the $6.3\ \mu\mathrm{g}$ floor persists in a balanced-detection run, this sensor enters the bias-instability class of commercial navigation-grade MEMS accelerometers.
  • Editorial inference: the fitted velocity random walk of $89.2\ \mu\mathrm{g}/\sqrt{\mathrm{Hz}}$ versus the measured $4.2\ \mu\mathrm{g}/\sqrt{\mathrm{Hz}}$ NEA implies a slow optical noise source dominated the Allan run; a reference photodiode would settle that.
  • Beyond the paper, a decisive environmental test is to sweep temperature across the full $>20$ °C range while keeping wavelength in the 17 nm band and logging Allan deviation; sub-$10\ \mu\mathrm{g}$ bias throughout would prove deployability.
  • Beyond the paper, the 'heavy stiff proof mass plus differential strain readout' rule likely transfers to optomechanical force, pressure, and magnetometry sensors, though no such demonstration is included here.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 4 minor

Summary. The paper presents an optomechanical accelerometer architecture based on a differential strain-sensing Mach–Zehnder interferometer (DSMZI) integrated on a bulk-micromachined proof mass (BMPM) with a 93.4 kHz resonant frequency. The authors report an off-resonance noise-equivalent acceleration (NEA) of 4.2 µg/√Hz over a 66 kHz bandwidth, a 6.3 µg bias instability at 243 s for a two-pass device, a 17 nm optical bandwidth, a >20 °C temperature operating range, and a projected dynamic range of 165.4 dB. They argue that the combination of high rigidity and differential optical readout makes the sensor insensitive to common causes of bias instability and thereby constitutes a path toward navigation-grade, deployable optomechanical accelerometers.

Significance. If the bias instability result were established as a true acceleration bias, the reported 6.3 µg value would be the lowest for an optomechanical accelerometer and would represent a meaningful step toward navigation-grade performance. The work also demonstrates a useful design principle—using a high-stiffness proof mass with an integrated differential interferometer—that broadens the optical and temperature operating bandwidths relative to cavity-based devices. The experimental characterization is substantial, including wavelength and temperature scans, Allan deviation measurements, and a comparison with prior work. The central limitation is that the bias instability claim is not yet convincingly separated from optical readout drift, and the projected performance figures (330 ng/√Hz, 165.4 dB dynamic range) are not measured but extrapolated.

major comments (5)
  1. [Appendix F.3, Section II] The Allan deviation analysis reports a fitted velocity random walk (VRW) of 89.2 µg/√Hz for the four-pass device, while the same device's NEA is 4.2 µg/√Hz. If the white noise in the Allan run were the accelerometer's acceleration noise, these two quantities should be equal. The 21-fold excess indicates that the Allan measurement was dominated by an unidentified noise source—likely laser intensity noise or polarization drift in the single-ended photodiode readout—rather than by acceleration noise. Consequently, the bias instability minimum at 579.3 s (20.3 µg) inherits this contamination and cannot be attributed to an inertial bias without additional discriminating measurements.
  2. [Section II, Appendix E] The bias instability measurement was performed with a single-ended photodiode on one interferometer output, without balanced detection, laser intensity stabilization, or polarization control. The DSMZI's differential readout suppresses common-mode strain but does not reject laser RIN or power drift in a single output. Appendix E concedes that intensity stabilization and polarization-maintaining fibers would be needed to improve bias instability, which is consistent with the interpretation that the measured Allan deviation minimum is readout-limited rather than acceleration-limited. The manuscript should directly address this and provide a test that distinguishes optical readout drift from true acceleration bias.
  3. [Table III, Section II] The 6.3 µg bias instability is obtained from the two-pass device, which has an NEA of 29 µg/√Hz and a path-length mismatch of 27.9 µm, whereas the four-pass device has an NEA of 4.2 µg/√Hz but a bias instability of 20.3 µg. The fact that the less sensitive device shows the lower Allan minimum suggests the result may reflect wavelength sensitivity or scale-factor drift rather than an intrinsic inertial property. The paper should explain why the two-pass device's Allan minimum is more than three times lower despite its larger NEA.
  4. [Appendix G, Section III] The claimed dynamic range of 165.4 dB is based on the projected 330 ng/√Hz noise floor, not on the measured NEA of 4.2 µg/√Hz. Using the measured NEA, the dynamic range would be approximately 143 dB at 1 Hz bandwidth. Moreover, the experimentally demonstrated linear dynamic range is 81.6 dB, limited by the actuator. The abstract and discussion should not present 165.4 dB as an experimentally supported value without clearly labeling it as a projection.
  5. [Section II, Appendix C] The measured acceleration responsivity is reported to be within a factor of 5.6 of the FEM prediction, yet this factor is not discussed. Since the projected path to 330 ng/√Hz and the dynamic-range calculation both rely on the model's phase sensitivity (5.76×10⁻⁴ rad/g per pass), the authors should quantify the uncertainty this discrepancy introduces into the projected performance and explain whether it originates from the optical model, the mechanical model, or the calibration.
minor comments (4)
  1. [Section I] There is a typo in the sentence 'This optimization leads to optimizng the ratio'—'optimizng' should be 'optimizing'.
  2. [Appendix A] Equation (A2) appears to be missing factors of the mass in the damping and spring terms; as written it is dimensionally inconsistent with the Langevin equation in (A1).
  3. [Figure 3 caption] The caption states that the Allan deviation shows a minimum at 579.3 seconds, but Section II reports both 579.3 s (four-pass) and 243.4 s (two-pass) values. The caption should specify that it refers to the four-pass device to avoid ambiguity.
  4. [Table I] The table lists 'Our devices' with two values in parentheses for several columns, but the footnote only explains asterisks; a brief legend would clarify that the parenthetical values correspond to the two-pass device.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: headline metrics are direct measurements, and the FEM phase-sensitivity model is independently compared rather than fitted.

full rationale

The paper's central claims are experimental measurements: NEA of 4.2 µg/√Hz, bias instability of 20.3/6.3 µg, 66 kHz operational bandwidth, >20 °C temperature range, and optical bandwidth of 5/17 nm. These are obtained directly from photodiode, lock-in, PSD, and Allan-deviation measurements; none are outputs of a fitted model. The FEM model predicts a phase sensitivity of 5.76e-4 rad/g per pass from geometry, photoelastic constants, and moving-boundary effects, and the measured responsivity is compared with this prediction and found to disagree by a factor of 5.6, so the model is not being tuned to reproduce the measurement. The theoretical NEA values are computed from shot, detector, and thermal noise formulas with stated assumptions (e.g., 10 mW per detector, balanced detection) and are not used to produce the measured 4.2 µg/√Hz value. The 165.4 dB dynamic range is an explicitly 'expected' projection based on the modeled scale factor and a projected 330 ng/√Hz noise floor, while the measured dynamic range of 81.6 dB is separately reported. The self-citations [8,9] support the differential-readout insensitivity argument, but that argument is also backed in this paper by direct wavelength and temperature scans and by Eq. (2), so the citations are not load-bearing. The VRW/NEA mismatch noted in Appendix F.3 is a validity concern about optical-readout noise contamination, not a circular derivation. Therefore no load-bearing step reduces by construction to its own input.

Assumptions & free parameters 3 free parameters · 7 assumptions · 0 invented entities

The central results are experimental; the theoretical model is used for projections and is not fitted to the data. The main quantities fitted to data are the mechanical Q-factor and the effective path-length mismatch. Axioms are standard oscillator theory, fluctuation-dissipation, photoelastic/moving-boundary perturbation theory, and the measurement assumptions described in the appendices. No new physical entities are introduced.

free parameters (3)
  • Mechanical Q-factor = 54.3
    Extracted from the resonance lineshape fit (Fig. 3b inset) and used in the thermal noise limit (0.33 µg/√Hz) and in the model comparison. This is a measured device property, not a model parameter chosen to force agreement.
  • Effective path-length mismatch (4-pass device) = 121.4 µm
    Inferred from the FSR of the wavelength-dependent transmission fringes; used to explain the 5 nm optical bandwidth and the temperature sensitivity of the 4-pass device.
  • Effective path-length mismatch (2-pass device) = 27.9 µm
    Inferred from the FSR of the wavelength-dependent transmission fringes; used to explain the 17 nm optical bandwidth and the temperature sensitivity of the 2-pass device.
assumptions (7)
  • standard math The accelerometer's mechanical response is a 1-DoF damped harmonic oscillator (Eq. A1-A3).
    Used throughout the theory and data analysis (e.g., Eq. 1, Appendix A) to define susceptibility, resonance frequency, Q, and to convert measured response to acceleration sensitivity.
  • domain assumption The sense mode at 93 kHz is the lowest-frequency mechanical mode driven by acceleration, so the sub-resonant response is undistorted (Section I, Eq. 1).
    The flat frequency response region and the 66 kHz bandwidth claim rely on this. The FEM model includes stabilizing tethers to suppress other modes, but the measured response is not shown across all modes.
  • standard math Fluctuation-dissipation theorem relates thermal force noise to mechanical damping (Eq. B2).
    Used to compute the thermomechanical noise floor of 0.33 µg/√Hz. Cited to [24].
  • standard math Photoelastic and moving-boundary perturbation theory describes strain-induced phase shifts in the waveguides (Eq. C3-C4).
    The FEM model evaluates these integrals to predict dφ/da = 5.76e-4 rad/g per pass. Material photoelastic tensors and refractive indices are taken from literature.
  • domain assumption The MZI output follows P = P0 sin²((Δφ+π/2)/2) and operates near quadrature (Eq. D1-D3).
    The scale factor and shot-noise calculations assume the device is biased at the maximum slope of the transfer function. In the experiment, the wavelength is tuned to a peak scale factor, supporting this.
  • domain assumption The free-space Michelson interferometer accurately calibrates the acceleration applied to the device (Appendix H).
    All scale factor, NEA, and bias instability values are derived from the piezo acceleration calibration. The calibration is only described briefly and has known motion-suppression dips.
  • ad hoc to paper The bias instability measured on a single-ended photodiode output represents the accelerometer's acceleration bias (Appendix F.3).
    The Allan deviation is measured with no balanced detection, no laser intensity stabilization, and no active temperature control. The paper does not demonstrate that optical power drift is negligible.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Navigation-Grade and Deployable Optomechanical Accelerometry." pith.science (2026). https://pith.science/paper/3NVPMPX5

@misc{pith2026250511751,
  author       = {Pith},
  title        = {Pith review of: Towards Navigation-Grade and Deployable Optomechanical Accelerometry},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3NVPMPX5}},
  note         = {Machine review of arXiv:2505.11751}
}
abstract

We design and experimentally demonstrate an architecture for achieving navigation-grade, fiber-packaged optomechanical accelerometers that can operate with a large dynamic range, over a wide temperature range, and without sophisticated laser sources. Our accelerometer architecture is based on a novel set of design principles that take advantage of the strengths of optomechanical accelerometry while eliminating many of its historical weaknesses. Displacement readout is provided by an integrated, differential strain-sensing Mach-Zehnder interferometer (DSMZI) attached to an ultra-rigid, bulk-micromachined proof mass having a 93.4 kHz fundamental resonance frequency (22.5 pm/g displacement). Despite the extreme rigidity, the high displacement sensitivity provides an insertion loss limited 4.2 $\mu g/\sqrt{\mathrm{Hz}}$ acceleration resolution, with a straight-forward path to achieving 330 $n g/\sqrt{\mathrm{Hz}}$ by improving the fiber-to-chip coupling. Further, we show that the combination of high rigidity and intrinsic differential optical readout makes the device insensitive to the common causes of bias instability, and we measure a bias instability of 6.3 $\mu g$ at 243 seconds. The DSMZI provides a 17 nm optical bandwidth and a temperature operating range of greater than 20 $^\circ\mathrm{C}$, both orders of magnitude larger than previous demonstrations of optomechanical accelerometers. The high rigidity and large optical bandwidth yield an expected dynamic range of 165.4 dB. The combination of high acceleration resolution, high dynamic range, low bias instability, and intrinsic insensitivity to wavelength, temperature, and package stresses makes our device well suited for deployment in realistic environments demanded by real-world applications and demonstrates a path for optomechanical accelerometers to ultimately exceed the performance of all other chip-based accelerometers.

Figures

Figures reproduced from arXiv: 2505.11751 by the authors.

Figure 1
Figure 1. (a) Illustrated schematic of the optomechanical accelerometer. (b) Microscope image of the fabricated fiber-packaged [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. (a) FEM model of the displacement field of the device at [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. (a) Experimental Setup: A. 633 nm laser for free space reference interferometer. B. Photodiode for free space reference [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Wavelength and temperature sensitivity plots, with the dashed blue line indicating a 3 dB drop in scale factor. (a, [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: The comparison of Power Spectral Density (PSD) measurements with various input sources. The red curve represents [PITH_FULL_IMAGE:figures/full_fig_p011_5.png]
Figure 6
Figure 6. Figure 6: Allan deviation σ(τ ) as a function of integration time τ , with fitted noise terms: (1) velocity random walk (VRW), (2) bias instability, (3) acceleration random walk (ARW), and (4) rate ramp (RR). Appendix G: Dynamic range analysis From the derivations in section D, …
Figure 7
Figure 7. Figure 7: Frequency dependence of motion suppression and sensitivity. (a) Applied acceleration as a function of frequency, [PITH_FULL_IMAGE:figures/full_fig_p016_7.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Precision optomechanical accelerometer via hybrid test mass integration

    physics.app-ph 2025-08 unverdicted novelty 6.0 of 10

    A pick-and-place bonded 95 mg platinum sphere on a silicon nitride trampoline yields a chip-integrated accelerometer with 5.5 ng/√Hz peak sensitivity at 117 Hz in air.

  2. Optomechanical Accelerometer Search for Ultralight Dark Matter

    hep-ex 2025-09 conditional novelty 5.0 of 10

    A 39 kHz optomechanical membrane accelerometer was operated at 4 K to search for ultralight vector dark matter, yielding no signal and upper limits weaker than existing equivalence-principle bounds.

Reference graph

Works this paper leans on

33 extracted references · 33 canonical work pages · cited by 2 Pith papers

  1. [1]

    A. G. Krause, M. Winger, T. D. Blasius, Q. Lin, and O. Painter, Nature Photonics 6, 768 (2012)

    work page 2012

  2. [2]

    What is an inertial navigation system?

    VectorNav Technologies, “What is an inertial navigation system?” (2023), accessed: 2025-05-13

    work page 2023

  3. [3]

    Higher-Order Composite Resonators for Extra Resilience (HORCREX): Broad agency announce- ment (hr001124s0039),

    Defense Advanced Research Projects Agency (DARPA), “Higher-Order Composite Resonators for Extra Resilience (HORCREX): Broad agency announce- ment (hr001124s0039),” https://sam.gov/opp/ fdc0f3e41f2948f4b8cf0d574ec39ecb/view (2024), accessed: May 13, 2025

    work page 2024

  4. [4]

    Y. L. Li and P. Barker, Journal of Lightwave Technology 36, 3919 (2018)

    work page 2018

  5. [5]

    L. Jin, C. Wang, L. Jin, W. Chen, H. Xu, M. Cui, and M. Li, Applied Optics 61, 5386 (2022)

    work page 2022

  6. [6]

    Huang, J

    Y. Huang, J. G. Flor Flores, Y. Li, W. Wang, D. Wang, N. Goldberg, J. Zheng, M. Yu, M. Lu, M. Kutzer, et al. , Laser & Photonics Reviews 14, 1800329 (2020)

    work page 2020

  7. [7]

    Gerberding, F

    O. Gerberding, F. G. Cervantes, J. Melcher, J. R. Pratt, and J. M. Taylor, Metrologia 52, 654 (2015)

    work page 2015

  8. [8]

    Dominguez, L

    D. Dominguez, L. Hackett, M. Miller, J. Restrepo, K. Casper, and M. Eichenfield, in 2021 IEEE Interna- tional Symposium on Inertial Sensors and Systems (IN- ERTIAL) (IEEE, 2021) pp. 1–4

    work page 2021

Show all 33 references
  1. [9]

    C. Ge, D. Dominguez, M. Miller, and M. Eichenfield, in Frontiers in Optics (Optica Publishing Group, 2024) pp. FW1D–3

    2024

  2. [10]

    Murphy, Analog Devices, Inc (2020)

    C. Murphy, Analog Devices, Inc (2020)

    2020

  3. [11]

    Tandon and K

    N. Tandon and K. Kumar, Noise & Vibration Worldwide 34, 9 (2003)

    2003

  4. [12]

    H. G. Espinosa, J. Lee, and D. A. James, Journal of Fitness Research 4, 13 (2015)

    2015

  5. [13]

    L. B. Soldano and E. C. Pennings, Journal of lightwave technology 13, 615 (1995)

    1995

  6. [14]

    Sheng, Z

    Z. Sheng, Z. Wang, C. Qiu, L. Li, A. Pang, A. Wu, X. Wang, S. Zou, and F. Gan, IEEE Photonics Jour- nal 4, 2272 (2012)

    2012

  7. [15]

    S. G. Johnson, M. Ibanescu, M. Skorobogatiy, O. Weis- berg, J. Joannopoulos, and Y. Fink, Physical review E 65, 066611 (2002)

    2002

  8. [16]

    Appendix d: Elasticity, photoelasticity, and electrooptic effects,

    C. Chen, “Appendix d: Elasticity, photoelasticity, and electrooptic effects,” (2006)

    2006

  9. [17]

    H. Tian, J. Liu, A. Attanasio, A. Siddharth, T. Bl´ esin, R. Ning Wang, A. Voloshin, G. Lihachev, J. Riemens- berger, S. E. Kenning, et al. , Advances in Optics and Photonics 16, 749 (2024)

    2024

  10. [18]

    Murphy, Analog dialogue 51, 1 (2017)

    C. Murphy, Analog dialogue 51, 1 (2017)

    2017

  11. [19]

    Mehmet, S

    M. Mehmet, S. Ast, T. Eberle, S. Steinlechner, H. Vahlbruch, and R. Schnabel, Optics express 19, 25763 (2011)

    2011

  12. [20]

    Y. Xia, A. R. Agrawal, C. M. Pluchar, A. J. Brady, Z. Liu, Q. Zhuang, D. J. Wilson, and Z. Zhang, Nature Photonics 17, 470 (2023). 8

    2023

  13. [21]

    F. Zhou, Y. Bao, R. Madugani, D. A. Long, J. J. Gorman, and T. W. LeBrun, Optica 8, 350 (2021)

    2021

  14. [22]

    Guzm´ an Cervantes, L

    F. Guzm´ an Cervantes, L. Kumanchik, J. Pratt, and J. M. Taylor, Applied Physics Letters 104 (2014)

    2014

  15. [23]

    D. Long, J. Stroud, B. Reschovsky, Y. Bao, F. Zhou, S. Bresler, T. LeBrun, D. Plusquellic, and J. Gorman, APL Photonics 8 (2023)

    2023

  16. [24]

    P. R. Saulson, Physical Review D 42, 2437 (1990)

    1990

  17. [25]

    Johnson, N

    K. Johnson, N. Alshamrani, D. Almutairi, A. Grieco, C. Horvath, J. N. Westwood-Bachman, A. McKinlay, and Y. Fainman, Optics Express 30, 46134 (2022)

    2022

  18. [26]

    Silicon (si) - properties, applications,

    AZoM, “Silicon (si) - properties, applications,” https:// www.azom.com/properties.aspx?ArticleID=53 (2001), accessed: 2025-05-02

    2001

  19. [27]

    AVAR (Adaptive Vector Autoregression) [MATLAB GUI],

    E. Ogier, “AVAR (Adaptive Vector Autoregression) [MATLAB GUI],” https://www.mathworks.com/ matlabcentral/fileexchange/ (2025). 9 Table I. Summary of the specifications of optomechanical accelerometers reported in the literature Paper Mass (mg) fs (kHz) Q-factor Operational Ban...

    2025

  20. [28]

    The Fth is a random force with a white spectrum density added to Fext shown in (A1), it can be expressed as Fth 2 = 4kBTf

    Noise from thermal Brownian motion Thermal noise is one of the fundamental limits to the precision of mechanical measurements. The Fth is a random force with a white spectrum density added to Fext shown in (A1), it can be expressed as Fth 2 = 4kBTf. (B1) Where kB is Boltzmann’...

  21. [29]

    (B3) Where dP da is the scale factor of the device, ω0 is the optical frequency of light, and Ptot is the total power on the detector

    Shot and detector noises Shot noise is generated by the dc power hitting the photodiode, it can be expressed by ashot = (dP da )−1 s 2ℏω0Ptot ηqe . (B3) Where dP da is the scale factor of the device, ω0 is the optical frequency of light, and Ptot is the total power on the dete...

  22. [30]

    The Power Spectral Density (PSD) in dBm/Hz can be expressed as PSD = 10 log10 V 2 rms Z × 1000 − 10 log10(BW)

    Optical power spectral density measurements We use a lock-in amplifier to measure the photodiode output voltage in Vrms with a time constant of 100 ms and an impedance Z of 50 Ω. The Power Spectral Density (PSD) in dBm/Hz can be expressed as PSD = 10 log10 V 2 rms Z × 1000 − 1...

  23. [31]

    In the case of Fabry–P´ erot microcavity accelerometers, the FWHM can be directly calculated as the free spectral range (FSR) divided by the cavity finesse

    Optical bandwidth analysis For cavity-based optomechanical accelerometers, the 3 dB optical bandwidth is equivalent to the optical full width at half maximum (FWHM). In the case of Fabry–P´ erot microcavity accelerometers, the FWHM can be directly calculated as the free spectr...

  24. [32]

    The laser cavity detuning induced by a change in temperature is ∆th =gth∆To, (F1) where gth is the thermo-optical tuning coefficient

    T emperature bound For an example of the temperature range that can be expected from a resonator-based optomechanical accelerometer, we follow the temperature analysis given in [1]. The laser cavity detuning induced by a change in temperature is ∆th =gth∆To, (F1) where gth is ...

  25. [33]

    The overlapped Allan deviation is computed using the MATLAB AVAR GUI [27], which is also used to fit and extract various noise coefficients

    Bias instability and Allan deviation To measure bias instability, the accelerometer is placed in an acoustically shielded enclosure under ambient condi- tions (in air at room temperature) without external acceleration applied. The overlapped Allan deviation is computed using t...

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.