Method for monitoring state of micromechanical sensor, micromechanical sensor and device

Abstract

The invention relates to a method for monitoring the state of a micromechanical sensor, to a micromechanical sensor and to a device, comprising a sprung-supported seismic mass (SM) which is designed to convert a physical input variable into an electrical sensor output signal of the micromechanical sensor, the method comprises the following steps: evaluating (S1) the sensor output signal of the micromechanical sensor (SEN) in order to detect harmful vibrations that interfere with the recalibration of the micromechanical sensor (SEN); and, upon detection of harmful vibrations, interrupting (S2) the recalibration of the micromechanical sensor (SEN).

Description

Technical Field

[0001] The present invention relates to a method and a device for condition monitoring of a micromechanical sensor and, in particular, to a vibration recognition method for verifying the rest position in a micromechanical sensor for its correct calibration. Background Art

[0002] MEMS (Micro-Electro-Mechanical Systems) are small sensors, typically made of silicon, such as accelerometers or gyroscopes. They are used in a variety of applications in automotive and consumer electronics. Several influences can contribute to a deterioration in sensor signal errors after soldering and during the service life of micromechanical sensors. Several conventional methods are known to compensate for these influences.

[0003] US Pat. No. 6,498,996 B1 and US Pat. No. 11,333,499 B2 discuss vibration as a source of error in sensors or gyroscopes. Both documents describe how oscillations can lead to bias or distortion. Oscillations can be characterized during factory calibration in terms of amplitude and frequency. Derived calibration coefficients are then used to compensate for the deviations caused by vibration.

[0004] Document US Pat. No. 11,333,499 B2 describes a compensating test mass for compensating for distortions caused by vibrations. Summary of the Invention

[0005] According to a first aspect, the present invention provides a method for condition monitoring of a micromechanical sensor.

[0006] Accordingly, the present invention provides a method for condition monitoring of a micromechanical sensor having a resiliently supported seismic mass (SM), which is configured to convert a physical input variable into an electrical sensor output signal of the micromechanical sensor, the method comprising the following steps:

[0007] Analyzing and processing the sensor output signals of micromechanical sensors for interference detection Unwanted vibrations that recalibrate micromechanical sensors; and

[0008] As soon as an unwanted vibration is detected, the recalibration of the micromechanical sensor is interrupted.

[0009] The method according to the invention is not intended to compensate for vibrations, but rather to identify unwanted vibrations in order to execute or interrupt a compensation mechanism or a sensor recalibration.

[0010] The method according to the invention increases the reliability and measuring accuracy of the micromechanical sensor and makes it significantly more robust to external environmental influences, in particular to vibrations.

[0011] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, a sensor output signal of the micromechanical sensor is evaluated with regard to whether at least one predefined abort criterion for interrupting a recalibration of the micromechanical sensor is met.

[0012] By using and defining different stop criteria, the method according to the invention can be used flexibly for a large number of different types of applications and micromechanical sensors of different designs.

[0013] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, an abort criterion for interrupting the recalibration of the micromechanical sensor comprises an exceeding of a defined threshold value by a sensor output signal of the micromechanical sensor.

[0014] The appropriate threshold value can be obtained through trial operation.

[0015] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, a sensor output signal of the micromechanical sensor is buffered for evaluation of the sensor output signal.

[0016] This simplifies data processing and allows the parallel or simultaneous application of different abort criteria, which can also be logically linked to each other.

[0017] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, the micromechanical sensor has at least one micromechanical gyroscope, which provides an angular velocity signal as a sensor output signal.

[0018] Gyroscopes are widely used and are particularly sensitive to vibrations. Therefore, the condition monitoring according to the invention significantly increases the reliability of the monitored micromechanical gyroscopes.

[0019] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, the micromechanical sensor has at least one micromechanical acceleration sensor, which provides an acceleration signal of a linear acceleration as a sensor output signal.

[0020] Micromechanical acceleration sensors are also widely used and are sensitive to vibrations. Therefore, the condition monitoring according to the present invention significantly improves the reliability and accuracy of the monitored micromechanical acceleration sensors.

[0021] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, a sensor output signal of the micromechanical sensor is subjected to a drift analysis and / or statistical tests of the mean, standard deviation, minimum and maximum values ​​to determine whether at least one predefined abort criterion for interrupting recalibration of the micromechanical sensor is met.

[0022] The plurality of available signal evaluation methods allows a flexible adaptation of the method according to the invention to micromechanical sensors of different designs for different fields of application.

[0023] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, a sensor output signal of the micromechanical sensor is evaluated by means of a fast Fourier transform FFT or a wavelet transform in order to determine distortions in an undesired frequency range.

[0024] The additional signal evaluation in the frequency domain increases the reliability of the method according to the invention with regard to detecting the influence of harmful environmental influences, in particular vibrations, on the recalibration of the micromechanical sensor.

[0025] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, recalibration of the micromechanical sensor is initiated by a start signal actively triggered by a user and / or by at least one defined sensor event and / or automatically at defined time intervals.

[0026] This allows for flexible use of the method according to the invention in a variety of application scenarios.

[0027] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, a recalibration of the micromechanical sensor is initiated after the sensor has been installed in the device.

[0028] This allows for robust recalibration of the micromechanical sensor at the end of an assembly line or manufacturing process for a device.

[0029] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, a recalibration of the micromechanical sensor is initiated during operation of the sensor.

[0030] This enables efficient and permanent recalibration of the micromechanical sensor on site during use at the customer's location.

[0031] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, a sensor output signal of the micromechanical sensor is evaluated both before and after a sensor recalibration of the micromechanical sensor to detect harmful vibrations.

[0032] This is suitable if the design of the micromechanical sensor does not allow condition monitoring of the micromechanical sensor during its recalibration.

[0033] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, a recalibration of the micromechanical sensor is performed if the sensor output signal of the micromechanical sensor does not meet a defined abort criterion neither before nor after the sensor recalibration of the micromechanical sensor.

[0034] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, a sensor output signal of the micromechanical sensor is evaluated during sensor recalibration in order to detect unwanted vibrations.

[0035] This has the advantage that recalibration can be carried out uninterruptedly at any time without interrupting the vibration recognition running continuously in the background.

[0036] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, a recalibration of the micromechanical sensor is performed if a sensor output signal of the micromechanical sensor does not meet a defined abort criterion during a recalibration of the sensor.

[0037] According to a second aspect, the present invention also provides a corresponding micromechanical sensor.

[0038] Accordingly, the present invention provides a micromechanical sensor having a resiliently supported seismic mass and a state monitoring unit, wherein the seismic mass is configured to convert a physical input variable into an electrical sensor output signal of the micromechanical sensor, and wherein the state monitoring unit is designed to analyze and process the sensor output signal of the micromechanical sensor for detecting harmful vibrations that interfere with the recalibration of the micromechanical sensor and automatically interrupt the recalibration of the micromechanical sensor once harmful vibrations are detected.

[0039] In one possible embodiment of the micromechanical sensor, the micromechanical sensor comprises at least one micromechanical gyroscope, which generates an angular velocity signal as the sensor output signal.

[0040] In one possible embodiment of the micromechanical sensor, the micromechanical sensor has at least one acceleration sensor which generates an acceleration measurement signal as a sensor output signal.

[0041] Other micromechanical sensors can also be monitored, in particular those comprising a sensing mass SM, such as pressure sensors.

[0042] According to another aspect, the present invention also provides a corresponding device.

[0043] Accordingly, the present invention provides a device having at least one micromechanical sensor, which has a resiliently supported seismic mass, which is configured to convert a physical input variable into an electrical sensor output signal of the micromechanical sensor, and the micromechanical sensor has a state monitoring unit, which is designed to analyze the sensor output signal of the micromechanical sensor for detecting harmful vibrations that interfere with the recalibration of the micromechanical sensor and automatically interrupt the recalibration of the micromechanical sensor once harmful vibrations are detected.

[0044] The device can be any electronic device having one or more micromechanical sensors. The method according to the present invention can be controlled by a controller of the device. The controller can include, for example, a processor or an application-specific integrated circuit (ASIC). The condition monitoring unit can also be integrated into the controller of the device. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In the following, possible embodiments of the method according to the invention and of the micromechanical sensor according to the invention are described in detail with reference to the accompanying drawings.

[0046] The accompanying drawings show:

[0047] Figure 1A A simple flow chart for presenting a possible embodiment of the method according to the invention;

[0048] Figure 1B A schematic diagram of one possible exemplary embodiment of a sensor, Figure 1A The method according to the invention presented in the embodiment is applied in the sensor;

[0049] Figure 1C A schematic diagram for illustrating a possible implementation variant of the method according to the invention;

[0050] Figure 2 Signal diagram for explaining a possible embodiment of the method according to the invention;

[0051] Figures 3A-3C A diagram for illustrating a possible embodiment of the method according to the invention;

[0052] Figure 4-5 Signal diagram for explaining a possible embodiment of the method according to the invention;

[0053] Figure 6-10 Signal diagram for explaining further possible specific embodiments of the method according to the invention. DETAILED DESCRIPTION

[0054] As in accordance with Figure 1AAs can be seen from the flow chart of FIG. 1 , the method according to the invention essentially has two main steps S.

[0055] Figure 1A The method shown in is used for condition monitoring of a micromechanical sensor SEN having a resiliently mounted seismic mass (SM) which is provided for converting a physical input variable into an electrical sensor output signal of the micromechanical sensor.

[0056] exist Figure 1A In a first step S1 of the method shown in , a sensor output signal of a monitored micromechanical sensor SEN is evaluated in order to detect unwanted vibrations which interfere with the recalibration of the micromechanical sensor.

[0057] In a preferred embodiment of the method, the sensor output signal of the micromechanical sensor SEN is first buffered in a data memory for analysis and processing. To this end, in a possible implementation, the sensor output signal is sampled by an analog-to-digital converter, and the sampled sensor output signal values ​​are buffered as samples in the data memory for subsequent analysis and processing in step S1.

[0058] In step S1, a buffered sensor output signal of the micromechanical sensor SEN is evaluated to determine whether at least one predefined abort criterion AK for aborting a recalibration of the micromechanical sensor is met. In one possible embodiment, the abort criterion AK used for aborting a recalibration of the micromechanical sensor includes the sensor output signal of the micromechanical sensor SEN exceeding a defined threshold value SW.

[0059] In one possible embodiment of the method for condition monitoring of a micromechanical sensor SEN, in step S1, a drift analysis and / or statistical tests of the mean, standard deviation, minimum and maximum values ​​are performed on the sensor output signal of the micromechanical sensor SEN to determine whether at least one predefined abort criterion AK for interrupting the recalibration of the micromechanical sensor SEN is met.

[0060] In one possible embodiment of the method for condition monitoring of a micromechanical sensor SEN, in step S1 a sensor output signal of the micromechanical sensor is evaluated by means of a fast Fourier transform FFT or a wavelet transform to determine distortions in an undesired frequency range.

[0061] exist Figure 1A In the second step S2 of the method for condition monitoring of a micromechanical sensor SEN shown in , a recalibration of the micromechanical sensor SEN is automatically interrupted or not performed as soon as an unwanted vibration is detected in step S1 .

[0062] In one possible embodiment or variant of the method for condition monitoring of a micromechanical sensor SEN, in step S1, a sensor output signal of the micromechanical sensor SEN is evaluated both before and after a sensor recalibration of the micromechanical sensor SEN to detect unwanted or interfering vibrations. In this first variant, a recalibration of the micromechanical sensor SEN is only performed if the sensor output signal of the micromechanical sensor SEN meets a defined abort criterion AK neither before nor after the sensor recalibration of the micromechanical sensor SEN.

[0063] In an alternative embodiment or variant of the method for condition monitoring of a micromechanical sensor, in step S1, a sensor output signal of the micromechanical sensor SEN is evaluated during ongoing sensor recalibration to detect unwanted vibrations. In this second variant, a recalibration of the micromechanical sensor SEN is performed if the sensor output signal of the micromechanical sensor SEN does not meet a defined abort criterion AK during the sensor recalibration.

[0064] In one possible embodiment of the method for condition monitoring of a micromechanical sensor, the monitored micromechanical sensor SEN has at least one micromechanical gyroscope, which provides an angular velocity signal WGS as a sensor output signal.

[0065] In one possible embodiment of the method for condition monitoring of a micromechanical sensor SEN, the monitored micromechanical sensor SEN has at least one micromechanical acceleration sensor which provides an acceleration signal of a linear acceleration as a sensor output signal.

[0066] In one possible embodiment of the method for condition monitoring of a micromechanical sensor SEN, recalibration of the micromechanical sensor SEN is initiated by a start signal actively triggered by a user and / or by at least one defined sensor event and / or automatically at defined time intervals.

[0067] In another possible embodiment of the method for condition monitoring of a micromechanical sensor, a recalibration of the micromechanical sensor SEN is initiated after the sensor has been installed in the device.

[0068] In a further possible embodiment of the method for condition monitoring of a micromechanical sensor SEN, a recalibration of the micromechanical sensor SEN is initiated during ongoing sensor operation.

[0069] According to another aspect, the present invention also provides a micromechanical sensor SEN, which has a resiliently supported seismic mass SM and an integrated or connected condition monitoring unit, wherein the seismic mass is configured to convert a physical input variable into an electrical sensor output signal of the micromechanical sensor SEN, and the condition monitoring unit is designed to analyze and process the sensor output signal of the micromechanical sensor SEN for detecting harmful vibrations that interfere with the recalibration of the micromechanical sensor SEN, and automatically interrupt the recalibration of the micromechanical sensor SEN once harmful vibrations are detected.

[0070] In one possible embodiment of the micromechanical sensor, the micromechanical sensor SEN has at least one micromechanical gyroscope, which generates an angular velocity signal WGS as a sensor output signal.

[0071] In another possible embodiment of the micromechanical sensor, the micromechanical sensor SEN has an acceleration sensor which generates an acceleration measurement signal as a sensor output signal.

[0072] Figure 1B A gyroscope or rotation rate sensor is schematically shown, which can be monitored by means of the method according to the invention.

[0073] Using a gyroscope as a sensor, the rotation rate or angular velocity ω is usually measured, rather than the orientation or rotation angle. If the gyroscope rotates around itself once per second, an angular velocity WG of 360 degrees per second is measured.

[0074] The rotation rate sensor is an inertial sensor that can measure the angular velocity WG without an external reference. The vibration gyroscope uses the Coriolis effect, which is generated due to forced oscillation (vibration) in the MEMS and the presence of the angular velocity WG.

[0075] The method according to the invention relates to measures for recalibrating a micromechanical sensor SEN, whose sensor structure comprises at least one deflectable sensor component SM for detecting a physical input variable and converting it into an electrical sensor signal.

[0076] exist Figure 1BIn the case of the gyroscope shown in Figure 1 as a rotation rate sensor SEN, recalibration or sensitivity compensation can be performed. The relationship between the actual angular velocity (WG) and the gyroscope output signal (measured angular velocity) is called the sensitivity error (English: Sensitivity Error). The sensitivity of the gyroscope depends on the reaction of the test mass or seismic mass SM to the Coriolis force FC, which can vary in each manufactured MEMS gyroscope due to small differences in the production process. The sensitivity of the MEMS gyroscope is fine-tuned to a value of 1 during production. After adjustment, the sensitivity is not exactly equal to 1 because the micromechanical sensor SEN is exposed to different electromagnetic and mechanical environments after adjustment. Additional sensitivity deviations can occur due to welding and over the service life of the micromechanical sensor SEN. This leads to a worsening of the sensitivity error of the micromechanical sensor SEN and can be compensated by calibrating or recalibrating the sensor, or using a sensor recalibration method (English: Component Re-Trim).

[0077] The monitored micromechanical sensor SEN can be, for example, an inertial sensor for detecting acceleration or rotational rate. The sensor structure of an inertial sensor generally comprises at least one resiliently supported seismic mass SM, the deflection of which is detected.

[0078] In the case of an acceleration sensor, these deflections are directly attributable to the acceleration to be detected. In the case of a micromechanical rotation rate sensor SEN or a gyroscope, as in Figure 1B As schematically shown in FIG, for detection purposes, seismic mass SM of a micromechanical rotation rate sensor SEN is actively excited to oscillate in an excitation plane. A rotational movement of micromechanical sensor SEN about an axis parallel to the excitation plane and oriented perpendicularly to the excitation direction can then be detected as a deflection of seismic mass SM perpendicular to the excitation plane, because this rotational movement induces a Coriolis force FC acting on seismic mass SM perpendicularly to the excitation plane.

[0079] In particular, when such an inertial sensor is to detect accelerations or rotational rates in more than one spatial direction, the sensor structure of the micromechanical rotational rate sensor SEN or gyroscope usually includes not just one but multiple seismic masses SM, whose suspension is then particularly flexible in different spatial directions.

[0080] The sensor structure of the micromechanical rotation rate sensor SEN comprises at least one seismic mass SM as a detection element, which is connected to the rest of the sensor structure via a spring assembly. Figure 1BThe yaw rate sensor is equipped with drives, by means of which the resiliently mounted detector element or seismic mass SM can be set into oscillation at a defined excitation frequency ωA.

[0081] The detection element SM oscillates along the x direction and rotates along the y direction (in Figure 1B When an axis (oriented perpendicular to the image plane) undergoes rotational motion, a Coriolis force FC is generated. This Coriolis force FC acts in the z-direction and is proportional to the rotational velocity Ω of the rotational motion. Due to this Coriolis force FC, the sensor element SM, which oscillates in the x-direction, is additionally excited to oscillate in the z-direction. The frequency of this oscillation in the z-direction corresponds to the excitation frequency ωA, but is phase-shifted by 90° from the excitation oscillation of the sensor element SM in the x-direction. The oscillatory motion of the sensor element SM in the z-direction can be detected as a sensor output signal and fed to a signal processing unit SVE associated with or integrated in the micromechanical sensor SEN to determine the rotational velocity Ω of the rotational motion.

[0082] exist Figure 1B In the exemplary embodiment shown in FIG, the sensor signal is detected capacitively by means of a detection electrode DE, which is arranged opposite the detection element or the seismic mass SM in the z direction, specifically at a distance A, as shown in FIG. Figure 1B As schematically shown in FIG. , the detection electrode DE and the detection element SM together form a detection capacitor. The distance A changes periodically with the oscillating motion of the detection element SM in the z direction caused by the Coriolis force FC. The capacitance change of the detection capacitor associated with this is Figure 1B The sensor signal of the rotation rate sensor or gyroscope schematically shown in FIG is output.

[0083] The sensor signal of this micromechanical rotation rate sensor SEN (e.g. Figure 1B The sensor signal of the micromechanical rotation rate sensor SEN can be influenced by a number of other, often undesirable, parameters (as schematically shown in FIG). Fluctuations in the layer thickness of the sensor structure arising from the MEMS manufacturing process play a role, for example. The packaging can also influence the sensor signal of the micromechanical rotation rate sensor SEN, for example, if mechanical stresses are introduced into the sensor structure.

[0084] The calibration process serves to recalibrate a micromechanical sensor SEN having a micromechanical sensor structure with at least one deflectable sensor element or a deflectable seismic mass SM, which is provided for detecting a physical input variable and converting it into an electrical sensor signal.

[0085] Figure 1BThe detection electrode DE of the micromechanical sensor SEN shown in FIG. 1 is arranged to detect a test deflection of the seismic mass SM as an electrical sensor response signal S el The sensor SEN is equipped with a sensor for correcting the electrical sensor signal S el The signal processing device SVE can be embedded in the sensor or integrated therein. In one possible implementation, the electrical sensor outputs a signal S el The correction of is performed based on at least one predetermined initial fine-tuning value Cal(0), which is selected such that, under a defined sensor excitation, manufacturing-related deviations of the sensor signal from the target sensor signal are compensated. The signal processing device SVE provides a corrected electrical sensor signal S el ‘ .

[0086] In one possible implementation, Figure 1B The recalibration of the micromechanical sensor SEN shown in FIG comprises a plurality of calibration steps. First, a defined electrical test excitation signal, in particular a defined DC voltage U , is applied to the sensor structure by means of at least one test electrode TE. DC , which results in a test deflection of the seismic mass SM of the micromechanical sensor SEN. The test deflection is detected by means of the detector electrode DE of the micromechanical sensor SEN as a capacitance change, which provides a corresponding electrical sensor response signal S el (i). Based on the sensor response signal S el A fine adjustment correction value ΔCal(i) is determined for an initial fine adjustment value Cal(0) based on a predetermined relationship between the fine adjustment correction value and the fine adjustment correction value. Finally, at least one current fine adjustment value Cal(i) for correcting the sensor signal is determined based on the at least one initial fine adjustment value Cal(0) and the determined fine adjustment correction value ΔCal(i).

[0087] Figure 1B The micromechanical rotation rate sensor SEN schematically shown in FIG. 1 is equipped with a test electrode TE, which is preferably arranged in the same plane as the detection electrode DE, but laterally thereto. By applying a defined voltage U between the seismic mass SM serving as the detection element and the test electrode TE, a defined electrical test excitation can be applied to the detection element or seismic mass SM in the z-direction. The corresponding sensor response signal, or sensor output signal Sel, is detected as a change in capacitance between the detection electrode DE and the seismic mass SM. The sensor response signal Sel(i) can also be referred to as the electrical sensitivity of the sensor SEN at a measuring time i.

[0088] In one possible embodiment, a DC voltage U is applied between the detection element SM and the test electrode TE, so that the corresponding electrostatic force acts in the z direction (i.e. in the detection direction, that is, in the direction of the Coriolis force FC). Due to the arrangement and geometry of the test electrode TE, this force is modulated by the driven oscillating motion of the seismic mass SM. Therefore, the electrostatic force has the same frequency f and the same phase as the driven oscillating motion of the seismic mass SM. However, it is phase-shifted by 90° from the Coriolis force FC caused by the rotational motion. In other words, by applying the DC voltage U, an orthogonal force is electrically induced. The resulting deflection of the seismic mass SM can be detected separately by demodulating the sensor output signal. The amplitude of this part of the sensor signal is then output as a static electrical sensor response signal Sel.

[0089] By recalibrating, Figure 1B The sensor sensitivity of the micromechanical sensor SEN shown in FIG is repeatedly adjusted back to the initial sensor sensitivity, ie the physical sensitivity can be repeatedly corrected as closely as possible to its new component value by recalibration.

[0090] Figure 1B The recalibration of the micromechanical sensor SEN shown in FIG. 1 can easily compensate for the influence of mechanical stresses caused by assembly on the measurement signal or sensor output signal and is therefore preferably performed before the micromechanical sensor SEN is recalibrated. Figure 1B The micromechanical rotation rate sensor SEN shown in FIG is used after being assembled in the application environment. However, recalibration of the micromechanical sensor SEN can also be performed in situ during ongoing sensor operation of the micromechanical sensor SEN to compensate for the influence of changing environmental conditions on the measurement signal or sensor output signal. To this end, electrical excitation of the sensing mass SM of the micromechanical sensor SEN can be performed, for example, at regular intervals or automatically triggered by a defined sensor event, such as after the micromechanical sensor SEN receives a shutdown signal and shortly before shutting down the rotation rate sensor. Finally, recalibration of the micromechanical sensor SEN can also be initiated by the user of the device in which the micromechanical rotation rate sensor SEN is installed.

[0091] by Figure 1B Taking the sensor recalibration of the micromechanical sensor SEN shown in FIG as an example, the electrodes for measuring the sensitivity deviation of the angular velocity WG are particularly sensitive to the angular motion induced by the vibration.

[0092] There are two different methods or variants for carrying out a recalibration of such a micromechanical sensor SEN.

[0093] Variant 1: The electrodes can be sensitive either to the test signal or the experimental signal or to the angular velocity WG, since they are phase-shifted by 90°.

[0094] During sensor recalibration, the test signal TS must be detected to calculate the gain compensation coefficient. Therefore, the angular velocity WG cannot be measured during sensor recalibration. However, the correct criterion for the motion threshold SW is based on the processed angular velocity signal WGS. In a first variant, in step S1 of the method according to the invention, this angular velocity signal is measured before and after performing the sensor recalibration, as also described in Figure 1C As shown schematically in Figure 1 (Sensor amplitude in dps (degrees / second) versus time t). Sensor recalibration is only performed if stability (+, +) is present both before and after. If there is no stability before or after, sensor recalibration is aborted.

[0095] If a predefined motion threshold SW of the processed angular velocity signal WGS is not reached both before and after the sensor recalibration or a corresponding abort criterion AK is not met, the sensor recalibration is performed, otherwise the sensor recalibration is aborted in step S2 of the method according to the invention, as also in step S3. Figure 1C As schematically shown in .

[0096] Variant 2: The electrodes are sensitive not only to the test signal TS but also to the angular velocity WG. In this second variant, the motion threshold SW can be evaluated in step S1 of the method according to the present invention during sensor recalibration. If the motion threshold SW of the processed angular velocity signal WGS is not reached or the corresponding abort criterion AK is not met, sensor recalibration is performed in step S2; otherwise, the sensor recalibration is aborted.

[0097] Conventionally used sensor recalibration abort thresholds SW are often based on linear acceleration measurements. These thresholds SW are often measured to be tight and can lead to aborting sensor calibration even when such abort is unnecessary. Consequently, applying such simple abort criteria AK, for example in SMT assembly lines (e.g., by smartphone or wearable device manufacturers), can lead to unnecessary errors and production with high failure rates, increased test times, or both. Using linear acceleration thresholds as abort criteria AK makes it impossible to identify critical angular movements that worsen the sensitivity error of the micromechanical sensor SEN.

[0098] In addition to using the angular velocity signal WGS as a criterion for the motion threshold value SW, it is also possible to use this signal during a recalibration of the sensor as an alternative thereto.

[0099] In a second method or variant (Variant 2), the measured signal or sensor output signal of the micromechanical sensor SEN contains not only information about the test signal TS but also information about the angular velocity WG caused by the external force F. By analyzing the measurement signal or sensor output signal of the micromechanical sensor SEN, it is possible to distinguish which portion of the sensor output signal is attributable to or caused by the test signal TS and which portion of the sensor output signal is attributable to or caused by the external force or vibration. Therefore, a termination criterion AK can be a detected high contribution of the external force F or vibration in the sensor output signal of the micromechanical sensor SEN during sensor recalibration. If the influence of the external force F on the measurement signal or sensor output signal is too great, the sensor recalibration is automatically terminated in step S2; otherwise, the sensor recalibration continues.

[0100] According to one aspect, the present invention provides an apparatus having at least one micromechanical sensor SEN with a resiliently supported seismic mass (SM), which is configured to convert a physical input variable into an electrical sensor output signal of the micromechanical sensor SEN, and an embedded or connected condition monitoring unit, which is configured to evaluate the sensor output signal of the micromechanical sensor SEN, detect unwanted vibrations that interfere with recalibration of the micromechanical sensor SEN, and automatically interrupt recalibration of the micromechanical sensor SEN upon detection of unwanted vibrations. The apparatus is, for example, a portable user terminal, such as a mobile phone.

[0101] Depending on the application, different compensation methods or recalibration methods can be implemented for acceleration sensors, gyroscopes or other IMUs (Inertial Measurement Units).

[0102] One possible compensation method involves fast offset compensation, or fast offset calibration (FOC), of the acceleration sensor. A MEMS acceleration sensor SEN consists of a spring-mass-damper system, wherein a test mass SM deflects when subjected to an external acceleration force F. The deflection of the test mass SM can be measured using capacitive detection electrodes and appropriate signal processing. The offset of a MEMS acceleration sensor SEN is often susceptible to variations in mechanical housing stresses, such as those caused by soldering the sensor to a printed circuit board, requiring offset compensation at the end of the production line or in the field. In offset compensation methods, such as fast offset calibration (FOC), the actual offset is measured by recording acceleration signals. Therefore, the acceleration sensor must be in a stationary position and must not be subject to vibrations during the detection of the signals required to determine the offset. In one possible embodiment of the method according to the present invention, the correct stationary position is verified by measuring a suitable signal and comparing it to a threshold value. Offset compensation, or fast offset calibration (FOC), is performed in the acceleration sensor. This offset compensation can be performed, for example, in the customer's production line to compensate for offset drift after the micromechanical sensor has been soldered on a printed circuit board.

[0103] Another possible compensation method involves bias compensation in the gyroscope. The gyroscope measures the angular velocity WG, which can then be used, for example, for position and orientation measurement or for image stabilization. MEMS gyroscopes, as micromechanical sensors SEN, operate based on the sinusoidal Coriolis force FC, which is generated by the oscillation of a test or seismic mass SM in combination with an orthogonal angular velocity input. The dynamic overall system is typically a mass-spring-damper system with two degrees of freedom (2DOF), where energy is transferred to the sensor mode via the rotation-induced Coriolis force FC, which is proportional to the angular velocity. If the gyroscope is in its rest position, the bias corresponds to the measured angular velocity WG.

[0104] Offset compensation can be performed in the micromechanical rotation rate sensor SEN or the gyroscope. This compensation is performed, for example, in the customer's production line to compensate for offset deviations after the micromechanical sensor SEN has been soldered onto a printed circuit board.

[0105] Another possible compensation method involves recalibration of the sensitivity and bias in the accelerometer. In end-of-line fine-tuning and testing, the accelerometer signal is fine-tuned in a familiar environment. In addition to the signal in normal operating mode, other signals are recorded, such as the built-inexcitation reaction on electrostatic force (BITE), a determined test mode for measuring the basic capacitance C0, the original bias and original sensitivity in different FE modes, and tests for accurately determining the mechanical parameters of the MEMS component (such as natural frequency and damping ratio). All results can be stored in the device's data memory as a reference for calibration on site. On site, these tests or a subset of them are performed and the device is recalibrated with the help of correlation techniques or machine learning ML algorithms, and the sensitivity error and bias error are minimized.

[0106] In one possible implementation, Figure 1A The sensor recalibration in step S2 of the method according to the invention shown in can use physical test signals (eg test quadrature signals) to measure and correct sensitivity deviations and thus minimize sensitivity errors of the micromechanical sensor SEN.

[0107] Movement of the sensor, in particular vibrations, can affect the movement of the electrodes and thus the signal. This can lead to a deterioration in the measurement of the sensor parameters, which must be compensated for. This is the case, for example, in compensation methods that use sensor recalibration.

[0108] To ensure compliance with defined environmental conditions, the method according to the invention includes built-in condition monitoring. In order to identify vibrations that interfere with the sensor signal, a detection mechanism with a predefined threshold value SW for harmful vibrations is used in one possible embodiment.

[0109] The unwanted vibrations to be detected can occur in six dimensions, namely three directions of linear motion and three directions of angular motion.

[0110] The three linear motion directions can be covered by corresponding micromechanical acceleration sensors, and the angular motion direction can be covered by three corresponding micromechanical gyroscopes.

[0111] If vibrations or mechanical oscillations below a defined threshold SW are detected by the detection mechanism, compensation or recalibration is performed, otherwise Figure 1A The compensation or recalibration is terminated in step S2 of the method shown in FIG.

[0112] One challenge is to provide suitable termination criteria AK, in particular a suitable motion threshold SW.

[0113] Existing acceleration sensors can be used to monitor mechanical influences on the micromechanical sensor SEN. If multiple axes are available during the method, the axes not used during recalibration can be used to monitor environmental vibration or motion influences on the micromechanical sensor SEN. In some sensor configurations, multiple axes are available for each direction. Examples include different sensor cores for each axis, optimized for different acceleration ranges or signal bandwidths. Therefore, each time a signal is recalibrated, all other signals are also monitored and analyzed or evaluated.

[0114] exist Figure 1A In the signal analysis process in step S1 of the method shown in FIG, various signal monitoring methods can be used. In one possible implementation, these signal monitoring methods include a drift analysis over time for detecting motion, statistical methods for checking the mean, checking the standard deviation, and monitoring minimum and maximum values ​​in order to detect undesirable influences on the micromechanical sensor SEN.

[0115] The recorded sensor signal of the micromechanical sensor SEN can be evaluated in step S1 of the method by means of a fast Fourier transform FFT or a wavelet transform in order to determine distortions in an undesired frequency range.

[0116] In general, all monitored signals can be used in a classification method to distinguish whether the environmental conditions are suitable for applying the recalibration method or whether the environmental conditions are used as abort criteria.

[0117] For each of the three angular movement directions, preferably a gyroscope signal of the micromechanical rotation rate sensor SEN is used. The processed angular movement signal can thus be considered a reliable measure.

[0118] By using a suitable predefined threshold value SW based on the processed linear acceleration signal or angular velocity signal, it can be ensured that the compensation method or recalibration is correctly applied.

[0119] The method according to the invention can be used to identify the correct rest pose for sensitivity and bias recalibration of a micromechanical acceleration sensor SEN.

[0120] In this case, the sensor output signal of the micromechanical acceleration sensor SEN is recorded and a numerical signal characteristic curve or signal signature is calculated. The signal characteristic curve is not limited to Figure 2The difference between the minimum and maximum signal values ​​shown in can be extended to any other signal characteristic curve, for example, determined by a moving average or a fast Fourier transform (FFT). In any case, the derived signal characteristic curve must be compared with a suitable threshold value to determine whether the recalibration meets the specifications.

[0121] For the three angular motion directions, the gyroscope output signal is used as a recognition mechanism. The gyroscope output signal is processed to derive the threshold SW.

[0122] For sensor recalibration applications, different signals from angular vibration measurements with different amplitudes and different frequencies f for all three axes can be studied and analyzed in detail. It follows that, in one possible implementation, the following processed angular velocity signal WGS predicts a reliable threshold value SW.

[0123] During a defined time period (this time period is usually equal to the duration of the sensor recalibration operation, for example 100 ms to 1000 ms), the angular velocity signal WGS is recorded and the maximum angular velocity value WGS is determined. max With the minimum angular velocity value WGS min The WGS difference between them.

[0124] An example of this is Figure 2 The signal diagram (channel values ​​recorded before the CRT period at 2 Hz and 50 dps stimulation) is shown in:

[0125] WGS-Diff = WGSmax–WGSmin

[0126] according to Figure 2 The processed angular velocity signal WGS is obtained as a difference WGS-Diff of 62.7 dps-(-59 dps)=121 dps.

[0127] After a detailed analysis, a threshold value SW of approximately 50 dps can be derived for this example for the processed angular velocity signal WGS, wherein the sensor calibration duration is 500 ms.

[0128] As it is Figure 1B As shown in Figure 1, the electrodes of a micromechanical gyroscope react sensitively to angular movement because their function is to detect the deflection of the probe mass SM caused by the angular velocity. Therefore, angular movement caused by undesired vibrations during calibration can influence the test signal and thus lead to a worsening of sensitivity errors.

[0129] In one possible implementation, correct performance of the sensor calibration is guaranteed when gain drift (GD) = < 2LSB.

[0130] The gain drift GD is derived from the signal sensed by the test electrode TE and determines the factor used for sensitivity compensation. A good correlation between the processed angular velocity signal WGS and the gain drift GD is observed at a critical frequency f of 2 Hz. From this, a suitable threshold SW of approximately 50 dps can be derived, such as for different directions X, Y, and Z. Figure 3A 、 3B , as can be seen in the signal diagram of 3C (gyroscope Delta versus gain drift GD, where the sensor is excited at 2 Hz in the Y direction).

[0131] A disadvantage of the first method or variant (Variant 1) of the method according to the present invention is that the angular velocity signal WGS cannot be determined during sensor recalibration. Therefore, there is still a minimal risk that the threshold value SW is not sufficiently adhered to before and after sensor recalibration (that is, the threshold value SW is exceeded during sensor recalibration).

[0132] However, since sensor recalibration typically lasts less than a second, it is relatively unlikely that the mechanical motion / vibration signal is visible only during sensor recalibration, and not before or after.

[0133] There is another (second) method (Variant 2) to reliably predict the correct application of sensor recalibration.

[0134] The signals recorded by the electrodes make it possible to identify the presence of mechanical vibrations which lead to a deterioration of the sensitivity error compensation by recalibration of the sensor.

[0135] According to Figure 4 (with a test quadrature signal of 2 Hz and 125 dps mechanical vibration) and according to Figure 5 Comparison of signal plots (test quadrature signals with 20 Hz and 125 dps mechanical vibration) shows that the average value (AVG average) of the signal recorded by the electrodes (here, quadrature) depends significantly on its phase. A fluctuating average value AVG of the electrode-recorded signal is an indicator of an insufficient resting position and renders sensitivity error compensation through sensor recalibration unreliable.

[0136] An algorithm that can distinguish between these two conditions can be applied during sensor recalibration and provide appropriate indicators.

[0137] This offers two advantages over the first approach:

[0138] On the one hand, the recording of the angular velocity signal WGS before and after is eliminated, which reduces the time required for sensor recalibration and reduces the time required to ensure sensor recalibration.

[0139] On the other hand, using the in-situ signal directly during sensor recalibration avoids the risk of using only the signal from shortly before or after sensor recalibration and potentially overlooking unwanted motion during the sensor recalibration period.

[0140] There is the possibility of analyzing the measured signal during the recalibration of the sensor.As mentioned above, the measured signal is influenced not only by the test signal but also by the external force F.

[0141] It is shown below how the measured signal behaves for different angular velocities WG during sensor recalibration.

[0142] Next, a possible way of processing the measurement signal or sensor output signal of the micromechanical sensor SEN is shown, and a series of features which can be used to interrupt the standard AK are proposed.

[0143] exist Figure 6 Measured measurement signal values ​​are shown for different sensor recalibration runs when no external force F acts on the micromechanical sensor SEN.

[0144] The sample index S of the measured signal is plotted on the x-axis. index (English: Sample Index), and plot the measured value in dps on the y-axis. Figure 6 In the exemplary signal value variation shown in FIG, the dps values ​​are between 386.2 dps and 396 dps for different runs.

[0145] According to Figure 7 Compared with the signal change process of Figure 6 In the signal value curve shown in , no drift of the measurement signal or of the sensor output signal can be detected even during the sensor calibration run.

[0146] exist Figure 7 The measured signals for a single sensor calibration run are recorded in , in which an oscillating force F was applied to the micromechanical sensor SEN. In this case, the micromechanical sensor SEN was rotated about one of its axes, and the direction of rotation was changed with a period of 2 Hz. Figure 7 It can be clearly seen in FIG5 that the external force F has an influence on the measured signal, which causes the measured signal to drift with the basic rotation frequency of the external force F.

[0147] In accordance with Figure 7In the case of varying signal values, the simple mean value formation required for sensor calibration is unreliable.

[0148] In the following description, a possibility is presented: to evaluate the drift of the measurement signal in order to decide whether a sensor recalibration of the micromechanical sensor SEN should be aborted.

[0149] As mentioned above, the measured value signal profile can be averaged for sensor recalibration.

[0150] The arithmetic mean M is defined in equation (1) as:

[0151]

[0152] Where N is the number of samples, and a n It is the value of the nth sample (English: Sample) or the sampling value.

[0153] A disadvantage of using equation (1) in the context of sensor recalibration is that no time information is available. This means that only a single value can be used for each sensor recalibration run.

[0154] In order to determine whether the arithmetic mean value M has drifted, a plurality of sensor recalibration runs must therefore be repeated in this method, which prolongs the overall calibration time of the micromechanical sensor SEN.

[0155] Due to multiple sensor recalibration runs The repetition of leads to a higher time expenditure for analysis and execution, so in a preferred embodiment of the method according to the invention, an efficient method for determining the analysis within a loop (Durchgang), namely the iterative calculation of the arithmetic mean M, is implemented.

[0156] The iterative calculation of the arithmetic mean M can be derived from equation (1).

[0157] The iterative calculation of the arithmetic mean M (which is also called the cumulative mean) is given by:

[0158]

[0159] M1=a1 (3)

[0160] Using equation (2) allows for a better understanding of the evolution of the arithmetic mean value M over the entire sensor recalibration, as shown in Figure 8 、 9 As shown in the process of change.

[0161] exist Figure 8 In the figure, we draw Figure 6 The cumulative average value M of the measured test quadrature signal n Cumulative mean M n The sample index S index is plotted on the x-axis, while the cumulative mean M given in dps is n is plotted on the y-axis. Figure 8 As can be seen in the figure, the cumulative mean M n In the example shown, it converges within the range (Korridor) of 390.9 dps to 391.1 dps.

[0162] exist Figure 9 In the figure, we draw Figure 7 The cumulative average value M of the measured test quadrature signal. Figure 8 Compared with the change process in Figure 9 The cumulative mean value M can be clearly identified during the change of n With sample index S index Drift. If the cumulative average value M n (like Figure 9 If drift occurs, then in one possible embodiment of the method according to the present invention, the sensor recalibration is aborted in step S2; otherwise, the sensor recalibration may be continued.

[0163] Next, we introduce a feature that can be used to determine the cumulative mean M n However, other features can also be used. For the evaluation of the arithmetic mean, it is not interesting how strongly the measured signal fluctuates due to the external force F, but whether the influence of the external force F is averaged out during the sensor recalibration operation.

[0164] One feature that provides this information is the cumulative mean M n The number of how often the minimum and maximum values ​​of the cumulative average are updated. For n=1, the minimum and maximum values ​​of the cumulative average are set to the same value, i.e., to the value a1. When calculating M2 in equation (2), it is checked whether the minimum and maximum values ​​need to be updated. If the minimum value needs to be updated, the first counter Z1 for updating the minimum value is increased or incremented by one. If the maximum value needs to be updated, the second counter Z2 for updating the maximum value is increased or incremented by one. If neither the minimum value of the cumulative average nor the maximum value of the cumulative average needs to be updated, the values ​​of the two counters Z1 and Z2 remain unchanged. The maximum values ​​of the two updated counters Z1 and Z2 represent a feature that can be used to determine the cumulative average M n How strongly does the sample index S index Drift. When this feature is applied to Figure 8The curve of the cumulative average, minimum and maximum values ​​up to the sample index S index is updated to five. Therefore, the eigenvalue is five. In contrast, Figure 9 The cumulative mean M in n The minimum and maximum values ​​are updated more frequently.

[0165] exist Figure 10 In FIG, the feature values ​​FV (English: feature value) of different sensor recalibration runs are plotted, which are divided into two groups, namely “Case 1” and “Case 2”.

[0166] Case 1 consists of all sensor recalibration runs without applying external forces to the micromechanical sensors SEN, i.e. according to Figure 8 All runs.

[0167] Case 2 includes an all-sensor recalibration run in which an oscillating rotational force F is applied to the micromechanical sensor SEN.

[0168] Figure 10 Each circle in represents a sensor recalibration run for Case 1.

[0169] Figure 10 Each asterisk in represents a sensor recalibration run for Case 2.

[0170] As in Figure 10 As can be seen in , the circles (case 1) and the asterisks (case 2) form two groups that are well separated from each other.

[0171] The two groups only slightly overlap when the eigenvalue (FV) is around 50. For this reason, a simple threshold based on the eigenvalue FV can be used to distinguish the two groups.

[0172] If more statistics are available, it is also possible to decide one of the two groups based on their probability density functions. Furthermore, the above features can also be combined with other features, which leads to decisions based on vectors rather than scalars.

[0173] The decision itself—whether to abort or continue with the sensor recalibration in step S2 of the method according to the invention—can be a hard decision or a soft decision.

[0174] Soft decisions provide additional information about the certainty of the decisions made. This additional information can be used to gain insights into the quality of sensor recalibration, for example, in SMT assembly lines. By monitoring soft decisions across different sensors, it is also possible to monitor changes in the environment in which sensor recalibration is performed. A decrease in decision reliability indicates an increase in interfering environmental factors, particularly vibrations, and allows countermeasures to be taken before sensor recalibration is aborted.

[0175] Fast bias calibration FOC can be used for various customer gyroscopes and accelerometers.

[0176] The method according to the present invention uses a detection mechanism to identify vibrations or tremors that lead to a deterioration of the original offset.

[0177] Signal monitoring for recalibration of micromechanical accelerometers can be used in a variety of applications, depending on their configuration and the possibilities for internal state monitoring and data analysis. In particular, high-performance IMUs may have powerful processors capable of performing statistical analysis, FFT, wavelet transform, and classification algorithms, which can be supported by input from machine learning (ML) algorithms.

[0178] Sensor recalibration of micromechanical sensors SEN can be used in various applications:

[0179] For example, sensor recalibration can be used in an SMT assembly line to compensate for sensitivity errors after soldering. In this case, the algorithm for deciding whether to abort the sensor recalibration in step S2 of the method according to the present invention can run on an external computer. In this example application, there are virtually no storage or performance constraints.

[0180] Sensor recalibration can also be used continuously at the end customer or in the field (e.g., in a smartphone or smartwatch in use) to compensate for sensitivity errors that occur during the operating life due to changes in humidity, temperature, and / or mechanical loads. Sensor recalibration can also be used periodically in the field. The algorithm for deciding whether to abort sensor recalibration in step S2 of the method according to the present invention can run on the sensor's MCU or on a host computer.

[0181] Other embodiments are possible. For example, in one possible implementation, the frequency with which sensor recalibration is aborted or interrupted due to the fulfillment of an abort criterion AK is monitored. In one possible implementation, this frequency is counted. If the abort occurs too frequently and / or the micromechanical sensor SEN cannot be recalibrated within a certain time interval, a corresponding warning message can be generated and transmitted, for example, to a controller of the device.

[0182] In one possible embodiment, the device has a user interface. This user interface can be used by the user to edit and adjust the termination criteria AK and logically link them together for the relevant application. The user interface can also display a warning message regarding possible unsuccessful recalibration of the micromechanical sensor SEN installed in the device.

Claims

1. A method for condition monitoring of a micromechanical sensor (SEN) having a spring-mounted seismic mass (SM), the seismic mass being configured to convert a physical input variable into an electrical sensor output signal of the micromechanical sensor (SEN), the method comprising the following steps: analyzing (S1) the sensor output signal of the micromechanical sensor (SEN) for detecting unwanted vibrations that interfere with recalibration of the micromechanical sensor (SEN); and As soon as an unwanted vibration is detected, the recalibration of the micromechanical sensor (SEN) is interrupted ( S2 ).

2. The method for condition monitoring of a micromechanical sensor according to claim 1, wherein: The sensor output signal of the micromechanical sensor (SEN) is evaluated with respect to whether at least one predefined abort criterion for interrupting a recalibration of the micromechanical sensor (SEN) is met.

3. The method for condition monitoring of a micromechanical sensor according to claim 2, wherein: The abort criterion for interrupting the recalibration of the micromechanical sensor (SEN) comprises that the sensor output signal of the micromechanical sensor (SEN) exceeds a defined threshold value.

4. The method for condition monitoring of a micromechanical sensor according to claim 1 , wherein: The sensor output signal of the micromechanical sensor (SEN) is buffered for evaluation of the sensor output signal.

5. The method for condition monitoring of a micromechanical sensor according to any one of the preceding claims 1 to 4, wherein: The micromechanical sensor (SEN) has at least one micromechanical gyroscope, which provides an angular velocity signal as a sensor output signal.

6. The method for condition monitoring of a micromechanical sensor according to any one of the preceding claims 1 to 5, wherein: The micromechanical sensor (SEN) has at least one micromechanical acceleration sensor, which provides an acceleration signal of a linear acceleration as a sensor output signal.

7. The method for condition monitoring of a micromechanical sensor according to any one of the preceding claims 1 to 6, wherein: The sensor output signal of the micromechanical sensor (SEN) is subjected to a drift analysis and / or statistical tests of mean, standard deviation, minimum and maximum values ​​to determine whether at least one predefined abort criterion for interrupting recalibration of the micromechanical sensor (SEN) is met.

8. The method for condition monitoring of a micromechanical sensor according to any one of the preceding claims 1 to 7, wherein: The sensor output signal of the micromechanical sensor (SEN) is evaluated by means of a fast Fourier transform (FFT) or a wavelet transform in order to determine distortions in an undesired frequency range.

9. The method for condition monitoring of a micromechanical sensor according to any one of the preceding claims 1 to 8, wherein: The recalibration of the micromechanical sensor (SEN) is initiated by an actuation signal actively triggered by a user and / or by at least one defined sensor event and / or automatically at defined time intervals.

10. The method for condition monitoring of a micromechanical sensor according to any one of the preceding claims 1 to 9, wherein: A recalibration of the micromechanical sensor (SEN) is initiated after the sensor has been assembled in a device.

11. The method for condition monitoring of a micromechanical sensor according to any one of the preceding claims 1 to 10, wherein: A recalibration of the micromechanical sensor (SEN) is initiated during ongoing sensor operation.

12. The method for condition monitoring of a micromechanical sensor according to any one of the preceding claims 1 to 11, wherein: The sensor output signal of the micromechanical sensor (SEN) is evaluated both before and after a sensor recalibration of the micromechanical sensor in order to detect unwanted vibrations.

13. The method for condition monitoring of a micromechanical sensor according to claim 12, wherein: A recalibration of the micromechanical sensor (SEN) is performed if the sensor output signal of the micromechanical sensor does not meet a defined abort criterion either before or after the sensor recalibration of the micromechanical sensor.

14. The method for condition monitoring of a micromechanical sensor according to any one of the preceding claims 1 to 13, wherein: During sensor recalibration, the sensor output signal of the micromechanical sensor (SEN) is evaluated in order to detect unwanted vibrations.

15. The method for condition monitoring of a micromechanical sensor according to claim 14, wherein: If the sensor output signal of the micromechanical sensor does not meet a defined abort criterion during the sensor recalibration, a recalibration of the micromechanical sensor (SEN) is performed.

16. A micromechanical sensor (SEN) having a spring-supported seismic mass (SM) and a condition monitoring unit, wherein the seismic mass is configured to convert a physical input variable into an electrical sensor output signal of the micromechanical sensor, the condition monitoring unit being designed to analyze the sensor output signal of the micromechanical sensor (SEN) for detecting harmful vibrations that interfere with the recalibration of the micromechanical sensor (SEN) and to automatically interrupt the recalibration of the micromechanical sensor (SEN) upon detection of harmful vibrations.

17. The micromechanical sensor according to claim 16, wherein: The micromechanical sensor (SEN) has at least one micromechanical gyroscope, which generates an angular velocity signal as a sensor output signal.

18. The micromechanical sensor according to claim 16, wherein: The micromechanical sensor (SEN) has at least one acceleration sensor which generates an acceleration measurement signal as a sensor output signal.

19. A device comprising a micromechanical sensor (SEN) according to any one of claims 16 to 18.

Images