Beverly F. Payne

Micromachined Convective Accelerometers in Standard Integrated Circuits Technology

This letter describes an implementation of micromachined accelerometers in standard complimentary metal–oxide–semiconductor technology. The devices operate based on heat convection and consist of microheaters and thermocouple or thermistor temperature sensors separated by a gap which measure temperature difference between two sides of the microheater caused by the effect of acceleration on free gas convection. The devices show a small linearity error of <0.5% under tilt conditions (±90°), and <2% under acceleration to 7g(g≡9.81u200am/s2). Sensitivity of the devices is a nearly linear function of heater power. For operating power of ∼ 100 mW, a sensitivity of 115 μV/g was measured for thermopile configuration and 25 μV/g for thermistor configurations. Both types of devices are operable up to frequencies of several hundred Hz.

Comparison of results of calibrating the magnitude of the sensitivity of accelerometers by laser interferometry and reciprocity

The development of a new electrodynamic shaker permits the calibration of accelerometers by two independent and absolute methods using the same shaker. Minimizing the uncertainty in the calibration of accelerometers by the reciprocity method requires minimization of the cross-axis component of the induced motion, and the distortion of its waveform. Minimizing the uncertainty in the calibration of accelerometers by the laser interferometric fringe-counting method requires minimization of the mechanical coupling between the moving element of the shaker and the optical components of the interferometer. The design of the new shaker provides for minimal distortion and cross-axis motion in combination with a very large mechanical impedance between the moving element of the shaker and the optical components of the interferometer. The shaker is equipped with dual coils and two retractable magnets to provide for reciprocity measurements without having to attach a secondary source of vibration located external to the moving element. This paper presents a comparison of calibration results obtained using the two methods and describes the shaker and calibration systems.

Errors in accelerometer calibration using laser interferometry due to harmonic distortion and cross motion in the applied motion

The development of mechanical vibration generators (shakers) used in the precise calibration of accelerometers requires minimizing the amount of distortion and cross (transverse) motion present in the vibration of the moving element of the shaker. It is well known that both distortion and cross motion can introduce significant errors in the calibration of accelerometers. Potential effects on measured sensitivity due to the presence of harmonic distortion were determined experimentally. Varying amounts of harmonic distortion were systematically introduced into the drive signal of a shaker. Changes in the measured sensitivity of an accelerometer were noted for these varied amounts of distortion and compared with theoretical predictions. Measurements of the sensitivity of an accelerometer were also made using moving elements having varying amounts of cross motion in order to experimentally determine potential effects due to the presence of cross motion on measured sensitivity. The results of this study should be helpful in the development of error budgets for accelerometer calibration systems and in the design of shakers used as a part of such systems.

Modeling and Experimental Analysis of Piezoelectric Shakers for High‐Frequency Calibration of Accelerometers

Piezoelectric shakers have been developed and used at the National Institute of Standards and Technology (NIST) for decades for high‐frequency calibration of accelerometers. Recently, NIST researchers built new piezoelectric shakers in the hopes of reducing the uncertainties in the calibrations of accelerometers while extending the calibration frequency range beyond 20 kHz. The ability to build and measure piezoelectric shakers invites modeling of these systems in order to improve their design for increased performance, which includes a sinusoidal motion with lower distortion, lower cross‐axial motion, and an increased frequency range. In this paper, we present a model of piezoelectric shakers and match it to experimental data. The equations of motion for all masses are solved along with the coupled state equations for the piezoelectric actuator. Finally, additional electrical elements like inductors, capacitors, and resistors are added to the piezoelectric actuator for matching of experimental and theore...

Laser interferometer and reciprocity calibration of accelerometers using the NIST Super Shaker

The development of the NIST Super Shaker permits calibration of accelerometers by two independent and absolute methods on the same shaker. Minimizing the uncertainty of reciprocity calibrations imposes stringent requirements for distortion and cross-axis motion. Laser interferometer calibrations require very low coupling between the shaker and the interferometer components. The design of the Super Shaker provides for very low distortion and cross-motion with very low mechanical coupling between the shaker and optical components. The shaker is equipped with dual coils and two retractable magnets to provide for reciprocity measurements without attaching a driving shaker. These features enhance the convenience with which the Super Shaker can be used to perform both reciprocity and laser interferometer calibrations. This paper describes the shaker and calibration system and gives a comparison of calibration results from the two methods.

Piezoelectric shaker developments for calibration of accelerometers at extended frequencies

Piezoelectric shakers are well suited to generate motion suitable for use with laser interferometric methods at frequencies of 3 kHz and above. One advantage of piezoelectric shakers is the higher achievable accelerations and displacement amplitudes as compared to electrodynamic (ED) shakers. Another advantage is the use of a solid ceramic construction rather than a compliant suspension system which results in much lower mechanical noise. Piezoelectric shakers have been developed and used at NIST for many years for high frequency calibration of accelerometers. Previous papers have documented the performance of these shakers for the calibration of accelerometers over the range of 3 kHz to 25 kHz using interferometric calibration methods. Piezoelectric shakers with solid ceramic construction coupled with new high resolution data acquisition systems have made it possible to extend the calibration of accelerometers to smaller amplitudes and a broader range of frequencies. These factors have also allowed calib...

Single point optical calibration of accelerometers at NIST

Typical accelerometer calibrations by laser interferometer are performed by measuring displacement at three places on the shaker table. Each of these measurements, made along the perimeter of the accelerometer, requires repositioning and realigning of the interferometer. This is done to approximate the actual displacement of the accelerometer. Using a dual-coil shaker with a small moving element and two coaxially-located and rigidly-attached mounting tables allows placing the accelerometer on one table and measuring displacement directly on the center axis of the second table. This was found to work effectively at lower frequencies, up to about 5 kHz, with mounting tables of conventional materials such as stainless steel. However, for higher frequencies the use of steel results in unwanted relative motion between the two mounting tables. Mounting tables of beryllium with nickel coating have been used at NIST to overcome this difficulty. This paper shows the calibration results of single point, on-axis measurements, using fringe counting and sine-approximation methods. The results compare favorably with three point measurements made by fringe disappearance using a conventional piezo-electric shaker at frequencies up to 15 kHz.

High‐frequency acceleration comparison among NIST, CENAM, and INMETRO

This paper presents the results obtained in an international laboratory comparison among NIST/USA, CENAM/Mexico, and INMETRO/Brazil. This acceleration comparison was focused on the calibration of the charge sensitivity of a single‐ended standard accelerometer in the frequency range from 3 kHz to 10 kHz. The measurements were carried out by laser interferometry in accordance with the standard ISO 16063‐11:1999 and with a preagreed protocol. This comparison supplements the former comparison SIM.AUV.V‐K1, which covered the frequency range from 50 Hz to 5 kHz.

Charge amplifier calibration comparison among NIST, CENAM, and INMETRO

This paper presents the results obtained in an international laboratory comparison among NIST/USA, CENAM/Mexico and INMETRO/Brazil which was focused on the electric calibration of a charge amplifier B & K model 2626 from 50 Hz to 5 kHz. One of the main concerns of this comparison was to conclude all measurements within the period of 1 month, in order to minimize the influence of long‐term temporal instabilities. Each participating laboratory carried out the measurements in accordance with a calibration protocol, which was established to avoid the effect of additional sources of uncertainty. The results are compared with the ones obtained during the former comparison SIM.AUV.V‐K1 and the conclusions are presented.

Reference values for the sensitivity of standard accelerometers used in intercomparisons

The National Metrology Institutes (NMIs) of five countries in North America and South America participated in an interlaboratory comparison involving the calibration of the magnitude of the sensitivity of three standard accelerometers. This comparison was performed by laboratories within the framework of the Interamerican Metrology System (SIM). One of the key values to be obtained in any interlaboratory comparison is an estimate of the reference values for the artifact being calibrated. Three statistical methods have been used to obtain candidate reference values and associated uncertainties from the SIM intercomparison data: an average of means method; a method based on the ISO Guide to the Expression of Uncertainty in Measurement; and a maximum likelihood method. Reference values and associated uncertainties obtained using the three methods are presented and compared, as well as useful graphical displays with resulting qualitative conclusions.