Method for recalibrating a micromechanical sensor, and recalibrateable sensor

Abstract

Recalibrating a micromechanical sensor. The sensor is assigned a signal processing device for correcting the sensor signal on the basis of at least one previously determined initial trim value that is selected such that, given a defined sensor excitation, a production-related deviation of the sensor signal from a target sensor signal is compensated. The method for recalibrating the sensor includes: applying a defined electrical test excitation signal to the sensor structure, acquiring the corresponding sensor response signal, ascertaining a trim correction value for the at least one initial trim value on the basis of a previously determined relation between the sensor response signal and the trim correction value, and determining at least one current trim value for correcting the sensor signal, the determination of the at least one current trim value taking place on the basis of the at least one initial trim value and the ascertained trim correction value.

Claims

1. A method for recalibrating a sensor, the sensor having a micromechanical sensor structure having at least one deflectable sensor structure element for detecting and converting a physical input variable into an electrical sensor signal, a first circuit element for applying a defined electrical test excitation signal to the micromechanical sensor structure that causes a test deflection of the deflectable sensor structure element, and a second circuit element for acquiring the test deflection of the deflectable sensor structure element as an electrical sensor response signal, the sensor being assigned a signal processing device for correcting the sensor signal based on at least one previously determined initial trim value that is selected such that, given a defined sensor excitation, a production-related deviation of the sensor signal from a target sensor signal is compensated, the method the method for recalibrating the sensor comprising the following steps: a. applying a defined electrical test excitation signal to the micromechanical sensor structure; b. acquiring the corresponding sensor response signal; c. ascertaining a trim correction value for the at least one initial trim value based on a previously determined relation between the sensor response signal and the trim correction value; and d. determining at least one current trim value for correcting the sensor signal, the determination of the at least one current trim value taking place based on the at least one initial trim value and the ascertained trim correction value.

2. The method as recited in claim 1, wherein the ascertaining of the trim correction value is based on a change ΔS.sub.el(i) of the sensor response signal S.sub.el(i) in relation to a sensor response signal S.sub.el(0) determined previously for the test excitation signal, using the following formula:
ΔS.sub.el(i)=[S.sub.el(i)−S.sub.el(0)]/S.sub.el(0).

3. The method as recited in claim 1, wherein the relation between the sensor response signal or a change of the sensor response signal, relating to the sensor response signal determined previously for the test excitation signal and the trim correction value, on which the ascertaining of the trim correction value is based, is previously determined empirically based on measurements carried out on a multiplicity of sensors of the same type as the sensor.

4. The method as recited in claim 2, wherein the ascertaining of the trim correction value ΔCal(i) is based on a linear relation having the form
ΔCal(i)a1*S.sub.el(i)+b1 or
ΔCal(i)a2*S.sub.el(i)+b2, where a1, b1, or, respectively, a2, b2, are previously determined characterizing constants.

5. The method as recited in claim 2, wherein in the ascertaining of the trim correction value ΔCal(i) at least one additional influencing variable Par.sub.j is taken into account, so that
ΔCal(i)=f(S.sub.el(i),Par.sub.j) or
ΔCal(i)=f(ΔS.sub.el(i),Par.sub.j). where f(S.sub.el(i), Par.sub.j) and f(ΔS.sub.el(i), Par.sub.j) are the correlation functions.

6. The method as recited in claim 5, wherein the at least one influencing variable Par.sub.j is a previously determined part-specific parameter, Par.sub.j being the sensor response signal S.sub.el(0) previously determined for the test excitation signal and/or the previously determined initial trim value Cal(0).

7. The method as recited in claim 5, wherein the at least one influencing variable Par.sub.j is a parameter capable of being acquired during sensor operation.

8. The method as recited in claim 1, wherein the recalibration of the sensor is initiated by a start signal that is actively triggered by a user, and/or is initiated by at least one defined sensor event, and/or is automatically initiated at defined time intervals.

9. The method as recited in claim 1, wherein the recalibration of the sensor is initiated after the sensor is installed in a device.

10. The method as recited in claim 1, wherein the recalibration of the sensor is initiated during a sensor operation.

11. The method as recited in claim 2, wherein the ascertaining of the trim correction value ΔCal(i) takes place in situation-dependent fashion in that the ascertaining of the trim correction value ΔCal(i) is based on different previously determined relations
ΔCal(i)=f.sub.inst(S.sub.el(i)) or ΔCal(i)=f.sub.run(S.sub.el(i))
or
ΔCal(i)=f.sub.inst(ΔS.sub.el(i)) or ΔCal(i)=f.sub.run(ΔS.sub.el(i)) depending on an installation or sensor operation. where f.sub.inst (S.sub.el(i)), f.sub.run (S.sub.el(i)), f.sub.inst (ΔS.sub.el(i)), f.sub.run (ΔS.sub.el(i)) are the correlation functions.

12. A sensor, comprising: a. a micromechanical sensor structure having at least one deflectable sensor structure element for detecting and converting a physical input variable into an electrical sensor signal; b. a first circuit element configured to apply a defined electrical test excitation signal to the micromechanical sensor structure that causes a test deflection of the deflectable sensor structure element; and c. a second circuit element configured to acquire the test deflection of the deflectable sensor structure element as an electrical sensor response signal; wherein the sensor is assigned a signal processing device configured to correct the sensor signal based on at least one previously determined initial trim value that is selected such that, given a defined sensor excitation, a production-related deviation of the sensor signal from a target sensor signal is compensated; wherein the signal processing device is configured to: access a previously determined, stored relation between the sensor response signal and a trim correction value for the at least one initial trim value; and determine a current trim value for the correction of the sensor signal based on the at least one initial trim value and the ascertained trim correction value.

13. The sensor as recited in claim 12, wherein at least one storage device is provided to store:r a. the at least one previously determined initial trim value Cal(0) and/or b. at least one sensor-specific parameter determined before the sensor is operated, that is to be taken into account in the ascertaining of the trim correction value; and/or c. the sensor response signal previously determined for the test excitation signal; and/or d. the at least one currently ascertained trim correction value, and/or e. the determined current trim value.

14. The sensor as recited in claim 12, wherein at least one storage device is provided to store the previously determined relation between the sensor response signal and a trim correction value for the at least one initial trim value.

15. The sensor as recited in claim 12, wherein at least one interface is provided for a trigger signal for initiating a recalibration of the sensor including the determination of the current trim value for the correction of the sensor signal.

16. The sensor as recited in claim 12, wherein the sensor is a micromechanical rotational rate sensor, or an acceleration sensor, or a magnetic field sensor, or a pressure sensor.

Description

DRAWINGS

(1) As mentioned above, the measures according to the present invention for recalibrating are not limited to micromechanical sensors of a particular detection type, as long as the sensor structure includes at least one deflectable sensor structure element for detecting a physical input variable and converting it into an electrical signal. Thus, the measures according to the present invention are equally suitable for the recalibration of micromechanical rotational rate sensors, acceleration sensors, magnetic field sensors, or pressure sensors. Advantageous specific embodiments and developments of the present invention are explained in more detail below in relation to the example of a micromechanical rotational rate sensor, on the basis of the Figures.

BRIEF DESCRIPTION OF THE DRAWINGS

(2) FIG. 1 shows a schematic representation of the main components of a rotational rate sensor that are used for the realization of the present invention.

(3) FIG. 2 shows a diagram with the results of measurements of characterizing parts in order to determine the correlation between the relative change in the electrical sensor sensitivity ΔS.sub.el and the relative change in the physical sensor sensitivity ΔS.sub.ph(i).

(4) FIG. 3 shows a diagram illustrating the example method according to the present invention.

DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS

(5) The sensor structure of a micromechanical rotational rate sensor includes at least one seismic mass as detection element, connected to the rest of the sensor structure via a spring system. The detection element is designated 1 in the schematic representation of FIG. 1. The rotational rate sensor is equipped with a drive with the aid of which the spring-mounted detection element 1 can be set into oscillation with a particular excitation frequency ω.sub.A. The type and functioning of the drive is not important for the realization of the present invention, for which reason they are not explained and shown in more detail in FIG. 1. FIG. 1 indicates only the direction of this driven oscillation, by arrows 2. This direction is referred to in the following as the x direction 2.

(6) When there is a rotational movement of detection element 1, oscillating in x direction 2, about an axis oriented in the y direction (here perpendicular to the plane of the image), a Coriolis force F.sub.c occurs. This Coriolis force F.sub.c acts in the z direction (indicated in FIG. 1 by arrows 3), and is proportional to the rate of rotation Ω of the rotational movement. Through this Coriolis force F.sub.c, detection element 1 oscillating in x direction 2 is additionally excited to an oscillation in z direction 3. The frequency of this oscillation in z direction 3 corresponds to the excitation frequency ω.sub.A, but the oscillation is phase-shifted by 90° relative to the excitation oscillation of detection element 1 in x direction 2. The oscillating movement of detection element 1 in z direction 3 is acquired as sensor signal, and is supplied to a signal processing device 7 assigned to the sensor in order to determine therefrom the rate of rotation Ω of the rotational movement.

(7) In the exemplary embodiment explained here, the acquisition of the sensor signal is done capacitively, using a detection electrode 4 that is situated opposite detection element 1 in z direction 3, at a distance 6. Together with detection element 1, it forms a detection capacitor. Distance 6 varies periodically with the oscillating movement, caused by Coriolis force F.sub.c, of detection element 1 in z direction 3. The associated change in capacitance of the detection capacitor is acquired as sensor signal.

(8) The sensor signal of such a micromechanical rotational rate sensor is in addition influenced by a number of further, mostly undesirable, parameters. For example, fluctuations in the layer thicknesses of the sensor structure, which occur due to the MEMS production process, play a role. The packaging can also have an effect on the sensor signal if for example it causes mechanical tensions to be carried into the sensor structure.

(9) In order to counteract falsifications of the sensor signal resulting from production factors, micromechanical rotational rate sensors are, in practice, compensated or calibrated at the manufacturer. There, an initial trim value Cal(0) is determined with the aid of which, for a defined sensor excitation, a production-related deviation of the sensor signal from a target sensor signal can be compensated. This sensor trimming works particularly well if a part-specific initial trim value Cal(0) is ascertained for each sensor on the basis of defined mechanical stimuli. For this purpose, each sensor has to be subjected to a defined mechanical rotational movement in order to then determine an individual initial trim value Cal(0) from the corresponding sensor signals.

(10) This procedure has only very limited suitability for compensating disturbing influences that occur after installation and during sensor operation, such as installation-related mechanical tensions in the sensor structure, and/or environmental influences. According to the present invention, it is therefore provided to recalibrate the sensor after its installation and/or in the field, so that the sensor signal can also be purified of disturbing influences that occur later and that change over the operating life of the sensor.

(11) For this purpose, the rotational rate sensor shown schematically in FIG. 1 is equipped with test electrodes 5 that here are configured in a plane with detection electrode 4, but are situated laterally thereto. By applying a defined voltage between detection element 1 and test electrodes 5, a defined electrical test excitation in z direction 3 can be exerted on detection element 1. Here, the corresponding sensor response signal S.sub.el is easily acquired as a change of capacitance between detection electrode 4 and detection element 1. Sensor response signal S.sub.el(i) is also referred to as the electrical sensitivity of the sensor at measurement time i.

(12) In the exemplary embodiment described here, a DC voltage U is applied between detection element 1 and test electrodes 5, so that the corresponding electrostatic force acts in z direction 3, i.e., in the detection direction, that is in the direction of Coriolis force F.sub.c. As a result of the configuration and geometry of test electrodes 5, this force is modulated with the driven oscillation movement of detection element 1. As a result, this electrostatic force has the same frequency and also the same phase as the driven oscillating movement. However, it is phase-shifted by 90° to the Coriolis force F.sub.c caused by a rotational movement. That is, through the application of the DC voltage U, a quadrature force is electrically induced. The resulting deflection of detection element 1 can be separately acquired through demodulation of the sensor signal. The amplitude of this portion of the sensor signal is then outputted as static sensor response signal S.sub.el.

(13) Advantageously, the electrical sensitivity S.sub.el(i) ascertained in this way is set into relation to an electrical initial sensitivity S.sub.el(0) of the sensor that has been determined in part-specific fashion, for example during final measurements at the manufacturer.

(14) Therefore, in the exemplary embodiment described here the relative change of the electrical sensitivity ΔS.sub.el(i) is always considered, where

(15) $\Delta S_{\mathrm{el}}\left(i\right):=\frac{S_{\mathrm{el}}\left(i\right)-S_{\mathrm{el}}\left(0\right)}{S_{\mathrm{el}}\left(0\right)}$

(16) According to an example embodiment of the present invention, it is a correction value ΔCal(i) for the initial trim value Cal(0) is determined using the thus ascertained electrical sensor sensitivity, or the relative change in the electrical sensor sensitivity ΔS.sub.el(i). This is done on the basis of a previously determined correlation function between the electrical sensor sensitivity S.sub.el(i), or ΔS.sub.el(i), and the physical sensor sensitivity S.sub.ph(i), or ΔS.sub.ph(i); in the case of a rotational rate sensor, this is thus the sensor sensitivity to rotational movements.

(17) Here it is to be noted that in the recalibration method according to the present invention, in principle any test excitation signal may be used, as long as for the recalibration of a particular sensor the same, defined test excitation signal is always used that was also used as the basis for determining the correlation between the electrical sensor sensitivity and the physical sensor sensitivity. For the exemplary embodiment described above, this means that, for example, a defined alternating voltage signal could also be applied, as test excitation signal, to the capacitor system of test electrode 5 and detection element 1 in order to modulate the deflection of detection element 1. In each case, the movement or deflection of detection element 1 and the corresponding sensor response signal S.sub.el permit inferences to be made concerning the performance properties of the sensor, and in particular concerning changes in the sensor sensitivity with regard to the initial sensor sensitivity determined at the manufacturer.

(18) In the exemplary embodiment described here, the determination of the trim correction value ΔCal(i) is based on the correlation, shown in FIG. 2, between the relative change in the electrical sensor sensitivity ΔS.sub.el and the relative change in the physical sensor sensitivity ΔS.sub.ph(i), the physical initial sensitivity S.sub.ph(0) corresponding to the target value of the sensor sensitivity after the initial trimming of the sensor, and being defined as 1. In the example embodiment, care is to be taken that, in this case, what have to be correlated with one another are not the absolute sensitivities, but rather only their changes due to soldering influences or lifespan influences.

(19) For the determination of this correlation, in preparation for mass production measurements are carried out of the relative change of the electrical sensor sensitivity ΔS.sub.el and the corresponding relative change in the physical sensor sensitivity ΔS.sub.ph(i) for an adequately large number of characterizing parts. These parts will have undergone a soldering process in order to ascertain the change in the physical sensor sensitivity due to soldering stress. In the measurement diagram of FIG. 2, the measurement results are plotted for the relative change in the electrical sensor sensitivity ΔS.sub.el and the corresponding relative change in the physical sensor sensitivity ΔS.sub.ph. On the basis of these measurement points, a linear correlation function was determined:
ΔS.sub.ph=f(ΔS.sub.el)=const.Math.ΔS.sub.el

(20) FIG. 2 illustrates that, already on the basis of such a linear scaling function—without taking into account further input parameters—a significant reduction in the sensitivity drift can be achieved. The drift in the physical sensitivity ΔS.sub.ph without recalibration is shown here as double arrow 10, while the sensitivity drift after the recalibration according to the present invention is shown by double arrow 20.

(21) From the previously described measurements of characterizing parts, it may also result that the correlation between the electrical sensor sensitivity ΔS.sub.el and the physical sensor sensitivity ΔS.sub.ph is better approximated by a higher-order function, or that the sensor sensitivity is a function of one, or also a plurality of, further influencing variables Par.sub.j. If these are part-specific influencing variables Par.sub.j, which can be ascertained for example in the end-of-line compensation, it is recommended to store these in a storage device of the sensor, or at least in a storage device to which the signal processing device of the sensor has access. In the case of a rotational rate sensor, for example the raw sensitivity or quadrature may be such part-specific influencing variables Par.sub.j.

(22) In some cases, however, it may also be appropriate to take into account so-called real-time parameters in the description of the correlation between the electrical sensor sensitivity ΔS.sub.el and the physical sensor sensitivity ΔS.sub.ph. These are influencing variables Par.sub.j measured in the field, such as the sensor temperature.

(23) FIG. 3 illustrates the procedure for the determination according to an example embodiment of the present invention of a correction value ΔCal(i) for the initial trim value ΔCal(0) taking as a basis a previously determined correlation function f( ) between the electrical sensor sensitivity ΔS.sub.el and the physical sensor sensitivity ΔS.sub.ph.

(24) First, at the manufacturer, in a “sensor final test”-phase, an initial trim value Cal(0) for calibrating the sensor is determined, so that the sought physical sensor sensitivity is achieved. The physical initial sensor sensitivity is then determined as S.sub.ph(0)=1. By applying the defined electrical test excitation signal U to the thus calibrated sensor, in addition the electrical initial sensor sensitivity S.sub.el(0) is acquired.

(25) The initial trim value Cal(0), the electrical initial sensor sensitivity S.sub.el(0), and the correlation function f( ) are stored in a storage device NVM (non-volatile memory) of the sensor or of the associated ASIC.

(26) The installation of the sensor on a circuit board, which requires a soldering process, very frequently results in a change in the physical initial sensor sensitivity, so that S.sub.ph(1)≠1. Here, S.sub.ph(1) stands for the physical sensor sensitivity in a first measurement after the soldering process. For the compensation of the soldering-related change in the physical sensor sensitivity, the defined electrical test excitation signal U is applied to the rotational rate sensor, in order to first acquire the electrical sensor sensitivity S.sub.el(1) after the soldering process. Using S.sub.el(1), the relative change in the physical sensor sensitivity ΔS.sub.el(1) is then calculated, in order also to determine the relative change in the physical sensor sensitivity ΔS.sub.ph(1), using the correlation function f( ). In this way, finally a corresponding trim correction value ΔCal(1) can also be ascertained and a further influencing variable Par(1) is also taken into account here. The compensation of the deviation of the sensor sensitivity from a sought sensor sensitivity then takes place using a new trim value Cal(1), where
Cal(1)=Cal(0)+ΔCal(1).

(27) In the exemplary embodiment shown here, the recalibration occurring to the present invention is also designated the CRT (Component Re-Trim) correction.

(28) The method according to the present invention also permits a recalibration of the sensor in the field. Here, environmental influences and aging may result in a change in the physical initial sensor sensitivity, so that S.sub.ph(2)≠1. Here, S.sub.ph(2) stands for the physical sensor sensitivity in a second measurement carried out during the use of the sensor in the field. For the recalibration, the defined electrical test excitation signal U is again applied to the rotational rate sensor in order to first acquire the current electrical sensor sensitivity S.sub.el(2). Using this, the relative change in the physical sensor sensitivity ΔS.sub.el(2) with regard to the electrical initial sensor sensitivity S.sub.el(0) is then calculated in order to as well determine the relative change in the physical sensor sensitivity ΔS.sub.ph(2) using the correlation function f( ). In the ascertaining of the corresponding trim correction value ΔCal(2), here a further influencing variable Par(2) is also taken into account. The compensation of the deviation of the sensor sensitivity from the sought sensor sensitivity then takes place using a new trim value Cal(2), where:
Cal(2)=Cal(0)+ΔCal(2).

(29) Using the recalibration according to the present invention, the sensor sensitivity is thus always brought back to the initial sensor sensitivity, i.e., the physical sensitivity is always corrected, as far as possible, to its initial factory value by the recalibration according to the present invention. The method according to the present invention enables, in a simple manner, the compensation of the influence of installation-related mechanical stresses on the measurement signal, and is therefore preferably applied after the rotational rate sensor is installed in an environment of use. However, during sensor operation it is also still possible to perform recalibration in order to compensate the influence of changing environmental conditions on the measurement signal. For this purpose, the electrical excitation of the detection mass can take place for example at regular time intervals, or can be automatically triggered by defined sensor events, such as shortly before the switching off of the rotational rate sensor after the sensor has received the switch-off signal. Finally, the recalibration according to the present invention can also be initiated by the user of the device in which the rotational rate sensor is installed.