METHOD FOR COMPENSATING FOR AN INFLUENCE OF A MAGNETIC INTERFERENCE SOURCE ON A MEASUREMENT OF A MAGNETIC FIELD SENSOR IN A DEVICE AND A DEVICE

Abstract

A method and device for compensating for an influence of a magnetic interference source on a measurement of a magnetic field sensor in a device. In the method, a magnetic flux density M.sub.1 measured with the magnetic field sensor at a measured ambient temperature T.sub.k is compensated for with a compensation factor M.sub.interference of the magnetic interference source according to

M=M.sub.1−M.sub.interference,

where

M.sub.interference=M.sub.0+aM.sub.0(T′.sub.k−T.sub.0)

and M.sub.0 is a magnetic reference flux density relative to a reference temperature T.sub.0, a corresponding to a material parameter, which is defined for a used magnet material of the magnetic interference source, and the measured ambient temperature T.sub.k being corrected using a non-linear delay parameter to a temperature of the magnetic interference source T′.sub.k. The method is used for the axis-based compensation of a temperature drift, the material parameter a being determined individually for each Cartesian axis.

Claims

1. A method for compensating for an influence of a magnetic interference source on a measurement of a magnetic field sensor in a device including a temperature sensor configured to detect an ambient temperature, the method comprising the following steps: compensating, for a magnetic flux density M.sub.1 measured with the magnetic field sensor at a measured temperature T.sub.k, with a compensation factor M.sub.interference of the magnetic interference source according to
M=M.sub.1−M.sub.interference
where
M.sub.interference=M.sub.0+aM.sub.0(T′.sub.k−T.sub.0) and M.sub.0 is a magnetic reference flux density relative to a reference temperature T.sub.0, a corresponds to a material parameter which is defined for a used magnet material of the magnetic interference source, and the measured ambient temperature T.sub.k being corrected to a temperature of the magnetic interference source T′.sub.k using a non-linear delay parameter.

2. The method as recited in claim 1, wherein a filter is implemented with the following equation for ascertaining the temperature of the magnetic interference source T′.sub.k
T′.sub.k=(1−b)T′.sub.k-1+bT.sub.k b being the delay parameter, T.sub.k being the ambient temperature and T′.sub.k-1 being the temperature of the magnetic interference source at a preceding point in time, and a correlation between a change of the ambient temperature ΔT and a change of the temperature of the magnetic interference source ΔT′ being
ΔT′=g(ΔT,b), and the delay parameter b being ascertained by minimizing an error from $b=\underset{b}{\min }\left(\Delta M-a{M}_0\left(\Delta T^{\prime }\right)\right)$ and ΔM indicating a change of the magnetic flux density.

3. The method as recited in claim 1, wherein the material parameter a is determined individually for each Cartesian axis to compensate for a position dependency of the magnetic interference source and of the magnetic field sensor in the device in the measurement of the magnetic flux density M.sub.1.

4. The method as recited in claim 3, wherein the reference flux density M.sub.0 and the measured magnetic flux density M.sub.1 are temporally averaged for determining the material parameter a for each Cartesian axis.

5. The method as recited in claim 3, wherein the material parameter a for three Cartesian axes (a.sub.x a.sub.y a.sub.z).sup.T is ascertained based on a change of the magnetic flux density in three Cartesian axes, an absolute value of change of the magnetic flux density ΔM being obtained from
ΔM=M.sub.1−M.sub.0=√{square root over (m.sub.1,x.sup.2+m.sub.1,y.sup.2+m.sub.1,z.sup.2)}−√{square root over (m.sub.0,x.sup.2+m.sub.0,y.sup.2+m.sub.0,z.sup.2)}, and indicating the measured magnetic flux density M.sub.1 and the reference flux density M.sub.0, the change of the magnetic flux density ΔM in three Cartesian axes being obtained from $\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right]=\left[\begin{array}{c}\Delta m_{1,x}\\ \Delta m_{1,y}\\ \Delta m_{1,z}\end{array}\right]-\left[\begin{array}{c}\Delta m_{0,x}\\ \Delta m_{0,y}\\ \Delta m_{0,z}\end{array}\right],$ a projection matrix RotProjection ${\mathrm{Rot}}_{\mathrm{projection}}=I_{3\times 3}+{\left[\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right]\times \left[\begin{array}{c}\Delta M\\ 0\\ 0\end{array}\right]\right]}_x+{\left[\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right]\times \left[\begin{array}{c}\Delta M\\ 0\\ 0\end{array}\right]\right]}_x^2.Math.\frac{1-\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right].Math.\left[\begin{array}{c}\Delta M\\ 0\\ 0\end{array}\right]}{{.Math.\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right]\times \left[\begin{array}{c}\Delta M\\ 0\\ 0\end{array}\right].Math.}^2}$ being ascertained to map the following equation $\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right]={\mathrm{Rot}}_{\mathrm{projection}}\left[\begin{array}{c}\Delta M\\ 0\\ 0\end{array}\right]$ and, based on the projection matrix RotProjection, ascertaining the material parameter a for three Cartesian axes (a.sub.x a.sub.y a.sub.z).sup.T from $\left[\begin{array}{c}{a}_x\\ a_y\\ a_z\end{array}\right]={\mathrm{Rot}}_{\mathrm{projection}}\left[\begin{array}{c}a\\ 0\\ 0\end{array}\right].$

6. The method as recited in claim 3, wherein the material parameter a is ascertained for three Cartesian axes (a.sub.x a.sub.ya.sub.z).sup.T based on the magnetic flux density in three Cartesian axes, the magnetic flux density in three Cartesian axes being determined from the reference flux density M.sub.0 in three Cartesian axes (m.sub.0,x, m.sub.0,y, m.sub.0,z), the material parameter a for three Cartesian axes (a.sub.x a.sub.y a.sub.z).sup.T being obtained from $\left[\begin{array}{c}{a}_x\\ a_y\\ a_z\end{array}\right]=\left[\begin{array}{c}{m}_{0,x}/M_0\\ m_{0,y}/M_0\\ m_{0,z}/M_0\end{array}\right]a.$

7. The method as recited in claim 6, wherein an influence of a geomagnetic field on the reference flux density M.sub.0 in three Cartesian axes (m.sub.0,x, m.sub.0,y, m.sub.0,z) is taken into consideration.

8. The method as recited in claim 3, wherein the material parameter a is determined for each individual Cartesian axis once during an initialization of the device, and the compensation of the temperature drift of the measured magnetic flux density M.sub.1 being continuously carried out during an operation of the device.

9. A device, comprising: a magnetic interference source; a magnetic field sensor configured to detect a magnetic flux density; a temperature sensor configured to detect an ambient temperature; and a processing unit, which is connected to the magnetic field sensor and to the temperature sensor, the processing unit being configured to compensate for an influence of the magnetic interference source on a measurement of the magnetic field sensor in the device, the processing unit being configured to: compensate, for the magnetic flux density M.sub.1 measured with the magnetic field sensor at a measured temperature T.sub.k, with a compensation factor M.sub.interference of the magnetic interference source according to
M=M.sub.1−M.sub.interference
where
M.sub.interference=M.sub.0+aM.sub.0(T′.sub.k−T.sub.0) and M.sub.0 is a magnetic reference flux density relative to a reference temperature T.sub.0, a corresponds to a material parameter which is defined for a used magnet material of the magnetic interference source, and the measured ambient temperature T.sub.k being corrected to a temperature of the magnetic interference source T′.sub.k using a non-linear delay parameter.

10. The device as recited in claim 9, wherein the device is a hearing device.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The above-described properties, features and advantages of the present invention and the manner in which these are achieved become more clearly and explicitly understandable in conjunction with the following description of exemplary embodiments, which are explained in greater detail in conjunction with the schematic figures.

[0028] FIG. 1 schematically shows a representation of one specific embodiment of a device, in accordance with the present invention.

[0029] FIG. 2A schematically shows a representation of a first specific embodiment of a provided method for a device according to FIG. 1, in accordance with the present invention.

[0030] FIG. 2B schematically shows a second specific embodiment of a provided method for a device according to FIG. 1, in accordance with the present invention.

[0031] FIG. 2C schematically shows a representation of a third specific embodiment of a provided method for a device according to FIG. 1, in accordance with the present invention.

[0032] FIG. 3 schematically shows a fourth specific embodiment of a provided method for a device according to FIG. 1, in accordance with the present invention.

DETAILED DESCRIPTION OF EXAMPLE EMBODIMENTS

[0033] It is noted that the figures are merely schematic in nature and are not true to scale. In this sense, components and elements shown in the figures may be represented as excessively large or reduced for better understanding. It is further noted that the reference numerals in the figures have been selected unchanged if identically designed elements and/or components are involved.

[0034] FIG. 1 schematically shows a representation of one specific embodiment of a device 100. Device 100 includes a magnetic interference source 120, a magnetic field sensor 125, a temperature sensor 135 and a processing unit 130. Magnetic interference source 120 produces an external magnetic field or a magnetic flux density, which influences the measurement of magnetic field sensor 125. Temperature sensor 135, as depicted, is not attached at or in magnetic interference source 120, thus, temperature sensor 135 is designed to detect the ambient temperature prevailing in the surroundings of device 100. Temperature sensor 135 is connected via processing unit 130 to magnetic field sensor 125. Processing unit 130 is designed to process measured data of temperature sensor 135 and of magnetic field sensor 125. In addition, processing unit 130 is designed to carry out the provided method for compensating for the influence of magnetic interference source 120 in device 100 and to determine from the measured data the parameters necessary for the compensation and to provide a value of the compensated magnetic flux density for device 100, for example, to calculate based thereon, directions for an audio-guided navigation of device 100.

[0035] Device 100 in FIG. 1 is designed, for example, as a hearing device 140, hearing device 140 represented being depicted by way of example as a headset. Hearing device 140 may, however, also have an alternative design. Hearing device 140 includes a first headphone 105 and a second headphone 110, first headphone 105 and second headphone 110 each including magnetic interference source 120, which are designed, for example, as loudspeaker permanent magnets of hearing device 140. First headphone 105 and second headphone 110 are connected to one another via a headband 115. Thus, with device 100 depicted, the ambient temperature may be determined with the aid of temperature sensor 135 and not directly a magnet temperature of the permanent magnet of magnetic interference source 120.

[0036] Temperature fluctuations or temperature changes are generally expressed in such devices 100 via the magnetic flux density or the change of the magnetic flux density or the magnetic field or the change of the magnetic field (i.e., the magnetic field strength or the change of the magnetic field strength). FIG. 2A shows a method 200 for compensating for an influence of magnetic interference source 120 on a measurement of magnetic field sensor 125 in device 100 in FIG. 1 according to one first specific embodiment. In this case, method 200 in FIG. 2A maps the steps required for determining a material parameter a, which is defined for a used magnet material of magnetic interference source 120 and describes a correlation between magnetic flux density or magnetic field and temperature. Method 200 as well as the following described methods are carried out, for example, by processing unit 130 in FIG. 1. A relationship between a change of the magnetic flux density (change of the magnetic field) and a change of the temperature may be approximated as a linear model, provided via the following mathematical correlation (1)

M.sub.k−M.sub.0=aM.sub.0(T.sub.k−T.sub.0)

[0037] in this case T.sub.0 corresponding to a reference temperature, which is selected, for example, as room temperature at 20° C. or 25° C. and M.sub.0 corresponding to a reference flux density for reference temperature T.sub.0, T.sub.k and M.sub.k being the magnet temperature and the magnetic flux density at point in time k, and a indicating the material parameter for the used magnet material.

[0038] In this case, material parameter a is independent of the mounting and placement of the permanent magnet of interference source 120, not, however, reference flux density M.sub.0. Even when using the same permanent magnet with the same magnet material, the results are only minimal mounting and placement deviations of the permanent magnet in a magnetic reference flux density M.sub.0 deviating under constant temperature. For this reason, the linear increase is generally divided into a constant value for material parameter a and a variable value for reference flux density M.sub.0.

[0039] Accordingly, a calibration of the measurement of magnetic reference flux density M.sub.0 by magnetic field sensor 125 in FIG. 1 to the aforementioned reference temperature T.sub.0 initially takes place in a first method step 205 in FIG. 2A.

[0040] In a second method step 210, a multiple measured data set is provided. The multiple measured data set includes, for example, a large number of measurements of the ambient temperature with the aid of temperature sensor 135 and a large number of measurements of the magnetic flux density with the aid of magnetic field sensor 125 in FIG. 1. This means, the variables T.sub.k and M.sub.k in equation 1 now map an entire time series of various points in time k. In a third method step 215, material parameter a is ascertained based on first and second method step 205, 210 according to the following equation (2)

[00007]$a=\frac{M-M_0}{M_0\left(T-T_0\right)}$

[0041] In this case, M and T correspond to the multiple measured data set provided in second method step 210.

[0042] Since, however, a correlation between the magnetic flux density or a change of the magnetic flux density (the magnetic field or the magnetic field strength or the change of the magnetic field or the change of the magnetic field strength) and the temperature or the change of the temperature in equations 1 and 2 in each case references the magnet temperature, it is necessary to initially convert the measured ambient temperature (or the time series of measured ambient temperature values) into the magnet temperature of the magnetic interference source 120, in order to be able to further correctly apply the linear correlation of the variables in the aforementioned equations. This approach is schematically represented in FIG. 2B, by adding a non-linear delay according to Newton's law of cooling. FIG. 2B shows a method 300 according to one second specific embodiment, in which, similar to second method step 210 of method 200 in FIG. 2A, a multiple measured data set is provided in a first method step 310. In this case, the multiple measured data set may be designed similarly as explained above. In a second method step 320 in FIG. 2B, a temperature drift of a measured magnetic flux density across a temperature-dependent delay parameter is taken into consideration. This delay parameter indicates, in particular, a non-linear correlation between the measured ambient temperature and magnet temperature T′.sub.k and is ascertained based on material parameter a, which has been obtained from third method step 215 in FIG. 2A.

[0043] To ascertain the temperature of magnetic interference source T′.sub.k, a filter is implemented with the following equation (3)

T′.sub.k=(1−b)T′.sub.k-1+bT.sub.k

[0044] b corresponding in this case to the delay parameter, T.sub.k corresponding to the ambient temperature and T′.sub.k corresponding to the magnet temperature at point in time k, and T′.sub.k-1 corresponding to the magnet temperature at a preceding point in time. The filter is designed, for example, as an exponential filter or as an IIR filter (IIR: Infinite Impulse Response). A correlation between a change of ambient temperature ΔT and a change of the temperature of magnetic interference source ΔT′ may be formulated with the aid of equation (4)

ΔT′=g(ΔT,b).

[0045] Delay parameter b in FIG. 2B may be ascertained by minimizing an error from equation (5)

[00008]$b=\underset{b}{\min }\left(\Delta M-a{M}_0\left(\Delta T^{\prime }\right)\right)$

[0046] ΔM indicating a change of the magnetic flux density and M.sub.0 referring to the aforementioned reference flux density.

[0047] Methods 200, 300 FIGS. 2A and 2B are carried out only once, for example, during the development or during the manufacture of device 100 in FIG. 1. Material parameter a and delay parameter b, in particular, each have constant values for a particular magnetic material of magnetic interference source 120 and of a fixed positioning of temperature sensor 135 in FIG. 1 with respect to permanent magnets of magnetic interference source 120. These constant values are utilized as input variables for subsequent calculations.

[0048] FIG. 2C shows a method 400 according to one third specific embodiment, in which material parameter a is determined individually for each Cartesian axis of three-axis magnetic field sensor 125 in order to compensate for a position dependency of magnetic interference source 120 and of magnetic field sensor 125 in device 100 in FIG. 1 during the measurement of a magnetic flux density M.sub.1. This position dependency may result in the measured flux density (or field strength) varying depending on the placement of interference source 120 or of sensor 125. The material parameter ascertained with the aid of method 200 in FIG. 2A is a function only of the magnet material and once designed device 100 is finished, the material parameter determined with method 200 is suitable for all devices 100 of the same model. However, the temperature drift for each axis may only be considered by ascertaining material parameter a for each Cartesian axis and this determination must take place individually for each device 100, since there is no standard axis-dependent material parameter a for all devices of a model. To ascertain the axis-based material parameter, two different methods may be utilized, which are explained in FIG. 2C, one method being based on a change of the magnetic flux density in three axes and the other method being based on the magnetic flux density in three axes.

[0049] A first method step 403 in FIG. 2C forms the provision of the measured magnetic flux density M.sub.1. In a second method step 407, processing unit 130 in FIG. 1 checks whether an interference of the surroundings is present. This may manifest itself, for example, in an additional influencing of a degree of the change of the magnetic flux density (change of the degree of the magnetic field), so that this interference is no longer able to reflect a temperature change alone. It is therefore checked whether or not the magnetic flux density (or whether the magnetic field strength) changes. In interference-free surroundings, in particular, the magnetic flux density (the magnetic field strength) remains constant. If no interference is present, which is indicated by an n, then the material parameter is ascertained for three Cartesian axes (a.sub.xa.sub.ya.sub.z).sup.T based on the change of the magnetic flux density in three Cartesian axes. For this purpose, two measurements of the magnetic flux density are selected in a third method step 409 at various ambient temperatures, for example, reference flux density M.sub.0 at temperature T.sub.0 [m.sub.0,x,m.sub.0,y,m.sub.0,z].sup.T for one, and magnetic flux density (M.sub.1 at temperature T.sub.1 [m.sub.1,x,m.sub.1,y,m.sub.1,z].sup.T, T.sub.1 being different from T.sub.0.

[0050] In a fourth method step 411, the measured magnetic flux densities are temporally averaged in order, for example, to reduce the noise influence. For example, the averaging time may be 1 second. In a fifth method step 413, the aforementioned flux densities are selected, where the orientation of device 100 therefor should not be changed, so that the geomagnetic field along the Cartesian axes does not change and is correctable by subtraction. In a sixth method step 417, a degree of the change of magnetic flux density ΔM is obtained from equation (6)

ΔM=M.sub.1−M.sub.0=√{square root over (m.sub.1,x.sup.2+m.sub.1,y.sup.2+m.sub.1,z.sup.2)}−√{square root over (m.sub.0,x.sup.2+m.sub.0,y.sup.2+m.sub.0,z.sup.2)}

[0051] M.sub.1 corresponding in this case to the measured flux density and M.sub.0 corresponding to the reference flux density.

[0052] The change of magnetic flux density ΔM in three Cartesian axes is obtained from equation (7)

[00009]$\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right]=\left[\begin{array}{c}\Delta m_{1,x}\\ \Delta m_{1,y}\\ \Delta m_{1,z}\end{array}\right]-\left[\begin{array}{c}\Delta m_{0,x}\\ \Delta m_{0,y}\\ \Delta m_{0,z}\end{array}\right]$

[0053] In a seventh method step 419, a projection matrix Rot.sub.Projection is ascertained according to equation (8)

[00010]${\mathrm{Rot}}_{\mathrm{projection}}=I_{3\times 3}+{\left[\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right]\times \left[\begin{array}{c}\Delta M\\ 0\\ 0\end{array}\right]\right]}_x+{\left[\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right]\times \left[\begin{array}{c}\Delta M\\ 0\\ 0\end{array}\right]\right]}_x^2.Math.\frac{1-\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right].Math.\left[\begin{array}{c}\Delta M\\ 0\\ 0\end{array}\right]}{{.Math.\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right]\times \left[\begin{array}{c}\Delta M\\ 0\\ 0\end{array}\right].Math.}^2},$

[0054] I.sub.3×3 corresponding to the unit matrix, [ ].sub.x indicating the notation for the corresponding cross product matrix, in order to map the following equation (9)

[00011]$\left[\begin{array}{c}\Delta m_x\\ \Delta m_y\\ \Delta m_z\end{array}\right]={\mathrm{Rot}}_{\mathrm{projection}}\left[\begin{array}{c}\Delta M\\ 0\\ 0\end{array}\right]$

[0055] and to ascertain in an eighth method step 430 material parameter a for three Cartesian axes (a.sub.xa.sub.ya.sub.z) from equation (10)

[00012]$\left[\begin{array}{c}{a}_x\\ a_y\\ a_z\end{array}\right]={\mathrm{Rot}}_{\mathrm{projection}}\left[\begin{array}{c}a\\ 0\\ 0\end{array}\right]$

[0056] based on projection matrix Rot.sub.Projection. In this case, material parameter a, which has been ascertained with the aid of method 200 in FIG. 2A, is used for the calculation.

[0057] A ninth method step 421 in FIG. 2C indicates the calculation method in interfered surroundings (indicated by y). In this case, the degree of change of magnetic flux density ΔM or the change of the magnetic flux density (or the degree of change of the magnetic field strength or the change of the magnetic field strength) is influenced, for example, by an additional magnetic interference. In ninth method step 421, a measurement of magnetic reference flux densities (m.sub.0,x, m.sub.0,y, m.sub.0,z).sup.T is therefore selected and these are averaged in a tenth method step 423, for example, also for an averaging duration of 1 second. This averaging is selected in an eleventh method step 427 in order to ascertain axis-based material parameter a for three Cartesian axes (a.sub.x a.sub.y a.sub.z).sup.T in eighth method step 430 from equation (11)

[00013]$\left[\begin{array}{c}{a}_x\\ a_y\\ a_z\end{array}\right]=\left[\begin{array}{c}{m}_{0,x}/M_0\\ m_{0,y}/M_0\\ m_{0,z}/M_0\end{array}\right]a$

[0058] The selected magnetic reference flux density (m.sub.0,x, m.sub.0,y, m.sub.0,z).sup.T may be influenced by the geomagnetic field, which may result in a strengthening of the temperature effect. Assuming the magnetic flux density in the z direction, m.sub.0,z without the geomagnetic field has a magnetic reference flux density of half the geomagnetic field, so that m.sub.0,z may be written (equation 12) as follows

[00014]$m_{0,z}=\frac{.Math.m_{\mathrm{earth}}.Math.}{2}$

[0059] Device 100 in FIG. 1 is then moved into a particular orientation, so that the entire geomagnetic field is added with an opposite sign to the above equation, so that m.sub.0,z

[00015]$m_{0,z}=-\frac{.Math.m_{\mathrm{earth}}.Math.}{2}$

[0060] includes half the amount of the geomagnetic field or of the magnetic flux density of the geomagnetic field, but with a minus sign. If this result is inserted into the projection matrix, then a strengthening rather than a compensation of the temperature effect in the z direction takes place due to the changed sign. However, since the geomagnetic field or the magnetic flux density of the geomagnetic field is considered to be small (compared to the magnetic interference source, for example, small by a factor of 30), then the strengthening as a result of the minus sign continues to be negligibly small and may accordingly be disregarded.

[0061] One method according to a fourth specific embodiment 500 in FIG. 3 shows the combination of previous methods 200, 300, 400 and expands these by additional steps, in order to ascertain the magnetic flux density M.sub.Komp compensated by the temperature drift and to be able to provide device 100 as the output variable in FIG. 1 for further processing. Methods 200 and 300 in FIGS. 2A and 2B in this case form, for example, a first calculation step 537 in FIG. 3 and method 400 in FIG. 2C forms a second calculation step 539 in FIG. 3. In a first method step 501, ambient temperature T.sub.k is converted into magnet temperature T′.sub.k, based on delay parameter b ascertained in first calculation step 537 in method 300.

[0062] A second method step 530 is designed similarly to eighth method step 430 in FIG. 2C and includes the ascertainment of axis-based material parameter a for three Cartesian axes (a.sub.x a.sub.y a.sub.z).sup.T according to the above calculation methods. Based thereon, and on ascertained magnet temperature T′.sub.k, processing unit 130 in FIG. 1 carries out the axis-based temperature compensation in a third method step 535 in FIG. 3. Third method step 535 thus forms a third calculation step 541 in FIG. 3 on the basis of reference flux density M.sub.0 and reference temperature T.sub.0.

[0063] The idea of compensation involves “resetting” or “moving” the drifted magnetic flux density or the drifted magnetic field to a reference level, which is defined by given reference temperature T.sub.0. This means that the drifted magnetic flux density (or the drifted magnetic field) is reset after the compensation to the corresponding value of the magnetic flux density (value of the magnetic field or value of the magnetic field strength), in which the temperature corresponds to reference temperature T.sub.0, i.e., for example, 20° C. or 25° C. With the aid of reference temperature T.sub.0 and reference flux density M.sub.0, it is possible to determine magnetic flux density M.sub.Komp compensated by the temperature drift as follows for the Cartesian axes

m.sub.Komp,x=m.sub.1,x−a.sub.xM.sub.0(T′.sub.k−T.sub.0),

m.sub.Komp,y=m.sub.1,y−a.sub.yM.sub.0(T′.sub.k−T.sub.0),

m.sub.Komp,z=m.sub.1,z−a.sub.zM.sub.0(T′.sub.k−T.sub.0).

[0064] The present invention has been described in detail with the aid of preferred exemplary embodiments. Instead of the exemplary embodiments described, further exemplary embodiments are possible, which may include additional modifications or combinations of described features. For this reason, the present invention is not restricted by the examples described, since other variations may be derived therefrom by those skilled in the art without departing from the scope of protection of the present invention in the process.