Magnetic sensor

Abstract

Magnetic sensor (1) comprising a magnetic field detection unit (3) and a calculation unit (30), wherein the magnetic sensor (1) is configured to generate a value of a detected angle (θs) which has a correspondence relationship to an angle (θ) formed by the direction (DM) of a magnetic field (MF) to be detected at a reference position (PR) with respect to a reference direction (DR), characterized in that: the magnetic field detection unit (3) comprises a plurality of magnetoresistive elements (50) for detecting the magnetic field (MF) and outputs a first signal (S1) and a second signal (S2), wherein the first signal (S1) has a correspondence to the angle formed by the direction (DM) of the magnetic field (MF) with respect to a first direction (D1), and the second signal (S2) has a correspondence to the angle formed by the direction (DM) of the magnetic field (MF) with respect to a second direction (D2); Each of the plurality of magnetoresistive elements (50) comprises: a fixed magnetization layer (53) whose magnetization direction is fixed; a free layer (51) whose magnetization direction changes depending on the direction (DM) of the magnetic field (MF); and a non-magnetic layer (52) arranged between the fixed magnetization layer (53) and the free layer (51); the calculation unit (30) calculates the value of the detected angle (θs) based on the first signal (S1) and the second signal (S2); the value of the detected angle (θs), when the direction (DM) of the magnetic field (MF) rotates with a predetermined period, contains an angular error component that changes with a period of 1 / 4 of the predetermined period; and the distortion ratio dr and the offset angle ϕ are chosen such that a maximum amount of the angular error component changes by 0.1° or less when the strength of the magnetic field (MF) at the magnetic field detection unit (3) changes within a sub-range which is part of a range of 20 to 150 mT and has an upper limit and a lower limit which differ from each other by 30 mT or more, wherein the magnetic field detection unit (3) comprises a first detection circuit (10) for outputting the first signal (S1) and a second detection circuit (20) for outputting the second signal (S2); both the first detection circuit (10) and the second detection circuit (20) comprise a chain of magnetoresistive elements consisting of two or more of the plurality of magnetoresistive elements (50) connected in series; the free layer (51) of each of the plurality of magnetoresistive elements (50) has a first surface in contact with the non-magnetic layer (52) and a second surface opposite the first surface, wherein the second surface has a rotationally symmetric shape of order four and not of order five or higher; the number of two or more magnetoresistive elements (50) that form the chain of magnetoresistive elements is even; the two or more magnetoresistive elements (50) that form the chain of magnetoresistive elements contain one or more pairs of magnetoresistive elements (50); the magnetization directions of the layers with fixed magnetization (53) of two magnetoresistive elements (50) paired together form a predetermined relative angle that is not 0° or 180°; in the first detection circuit (10) the first direction (D1) is either an intermediate direction between the magnetization directions of the layers with fixed magnetization (53) of two paired magnetoresistive elements (50) or a direction opposite to the intermediate direction; and in the second detection circuit (20) the second direction (D2) is either an intermediate direction between the magnetization directions of the layers with fixed magnetization (53) of two paired magnetoresistive elements (50) or a direction opposite to the intermediate direction, • dr is the ratio between the offset distance d and the radius of the reference circle Rc, • Rc is the mean of the minimum and maximum values ​​of R(θa), • R(θa) is the distance from the center of the second surface to a given point on the outer boundary (51a) of the second surface and • The offset angle ϕ is the angle between the magnetization direction (D111) of the layer with fixed magnetization and the first direction D1, Distortion ratio and offset angle (dr, ϕ) selected view from (1%, 25°), (0.1%, 10°), (1%, 30°).

Description

Background of the invention 1. Field of the invention

[0001] The present invention relates to a magnetic sensor for detecting an angle formed by the direction of a magnetic field to be detected with respect to a reference direction. 2. Description of the state of the art

[0002] In recent years, magnetic sensors have been widely used to detect the rotational position of an object in various applications, such as the rotation of a steering wheel or a power steering motor in an automobile. Magnetic sensors are used not only to detect the rotational position of an object but also to detect its linear displacement. Systems using magnetic sensors typically include a magnetic field generation unit to create a target magnetic field whose direction changes in response to the rotation or linear movement of the object. Hereinafter, the target magnetic field is referred to as the target magnetic field. The magnetic field generation unit can be a magnet. The magnetic sensors use magnetic sensing elements to detect the angle formed by the direction of the target magnetic field at a reference position relative to a reference direction.The rotational position or linear displacement of the object is thus detected.

[0003] Among the known magnetic sensors is one that incorporates two bridge circuits (Wheatstone bridge circuits), as described in EP 2455720 A1 and EP 1232400 B1. In such a magnetic sensor, each of the two bridge circuits contains four magnetic sensing elements and outputs a signal that responds to the direction of the target magnetic field. Each magnetic sensing element contains a magnetoresistive (MR) element. The output signals of the two bridge circuits differ in phase by 1 / 4 of the period of the output signals of the bridge circuits. The angle formed by the direction of the target magnetic field with respect to the reference direction is calculated based on the output signals of the two bridge circuits.

[0004] In a magnetic sensor using an MR element, the output signal of the MR element, corresponding to the resistance of the MR element, should ideally have a waveform describing a sinusoidal curve (including sine and cosine waveforms) as the direction of the target magnetic field rotates. However, as described in EP 2455720 A1, it is known that the output signal waveform of the MR element can sometimes be distorted relative to a sinusoidal curve. Distortion of the output signal waveform of the MR element relative to a sinusoidal curve means that the output signal of the MR element contains a harmonic component that is not a fundamental sinusoidal wave. If the output signal of the MR element contains this harmonic component, an error can occur in the angle detected by the magnetic sensor.

[0005] The following describes examples of situations where the output signal waveform of an MR element is distorted. It is assumed here that the MR element is a spin-valve MR element containing a fixed magnetization layer with a fixed magnetization direction, a free layer whose magnetization direction changes depending on the direction of the target magnetic field, and a non-magnetic layer positioned between the fixed magnetization layer and the free layer. Examples of spin-valve MR elements include giant magnetoresistive elements (GMRs) and tunneling magnetoresistive elements (TMRs).Examples of situations in which the output signal waveform of an MR element is distorted include the case where the magnetization direction of the fixed magnetization layer varies due to influence from the target magnetic field or other factors, and the case where the magnetization direction of the free layer deviates from the direction of the target magnetic field due to effects such as free layer shape anisotropy.

[0006] EP 2455720 A1 discloses a method for reducing the error of the detected angle, which is described below. According to the method, the spin-valve MR elements forming the conventional bridge circuit are replaced by chains of MR elements, each chain consisting of a plurality of spin-valve MR elements connected in series. The chains of MR elements each contain one or more pairs of MR elements. The fixed magnetization layers of two paired MR elements have magnetization directions that are rotated in opposite directions by the same angle with respect to the magnetization direction of the fixed magnetization layers of the MR elements forming the conventional bridge circuit.

[0007] EP 1232400 B1 discloses a method for correcting the detected angle by establishing an electrical connection between a main sensor element and two correction sensor elements, wherein the main sensor element has a main reference magnetization axis and each of the two correction sensor elements has a reference magnetization axis which is inclined with respect to the main reference magnetization axis.

[0008] The research conducted by the inventors of the present application revealed that the error of the angle detected by magnetic sensors varies depending on the strength of the target magnetic field. Hereinafter, the error of the detected angle is referred to as the angular error, and the change in the angular error as a function of the target magnetic field strength is referred to as the dependence of the angular error on the magnetic field strength. This dependence of the angular error on the magnetic field strength was even observed in a magnetic sensor equipped with a means for reducing the angular error.

[0009] On the other hand, in a system using a magnetic sensor, the strength of the target magnetic field applied to the sensor is not necessarily constant. For example, a variety of magnetic field generation units can be used with a single magnetic sensor, producing target magnetic fields of varying strengths. Furthermore, when using a magnet as the magnetic field generation unit, temperature fluctuations or deterioration of the magnet can lead to a change in the strength of the target magnetic field, even if the magnet itself remains unchanged. Additionally, using a low-cost magnet or similar material can result in the actual strength of the target magnetic field deviating from the value assumed during system design.

[0010] With conventional magnetic sensors, the angular error, due to its dependence on the magnetic field strength mentioned above, sometimes exhibits significant fluctuations within a predefined range of the target magnetic field strength, which is the expected range during sensor use. As a remedy, a correction process can be applied to the detected angle to reduce the angular error when the target magnetic field strength falls within one of the specified values ​​within this range. However, the angular error is only sufficiently reduced if the target magnetic field strength falls within one of these specified values.

[0011] Another state of the art is known from DE 10 2013 102 165 A1 and from US 2015 / 0 066 426 A1. Object and summary of the invention

[0012] It is an object of the present invention to provide a magnetic sensor for detecting an angle formed by the direction of a target magnetic field with respect to a reference direction, wherein the magnetic sensor achieves a reduced fluctuation of the angular error over a large strength range of the target magnetic field.

[0013] The problem is solved by the magnetic sensor specified in claim 1. Advantageous embodiments of the invention are specified in the dependent claims.

[0014] A magnetic sensor according to the present invention is configured to generate a detected angle value that corresponds to an angle formed by the direction of the target magnetic field at a reference position with respect to a reference direction. The magnetic sensor comprises a magnetic field detection unit and a computation unit. The magnetic field detection unit includes a plurality of magnetoresistive elements for detecting the target magnetic field and outputs a first signal and a second signal. The first signal corresponds to the angle formed by the direction of the target magnetic field with respect to a first direction. The second signal corresponds to the angle formed by the direction of the target magnetic field with respect to a second direction.Each of the multiple magnetoresistive elements contains: a layer with fixed magnetization, whose magnetization direction is fixed; a free layer, whose magnetization direction changes depending on the direction of the target magnetic field; and a non-magnetic layer located between the layer with fixed magnetization and the free layer. The processing unit calculates the value of the detected angle based on the first and second signals.

[0015] According to the magnetic sensor of the present invention, the value of the detected angle includes an angular error component which, when the direction of the target magnetic field rotates with a predetermined period, varies with a period of 1 / 4 of the predetermined period. The magnetic sensor of the present invention is configured such that a maximum amount of the angular error component changes by 0.1° or less when the strength of the target magnetic field at the magnetic field detection unit changes within a sub-range that is part of a range of 20 to 150 mT and has an upper and a lower limit that differ from each other by 30 mT or more.

[0016] According to the magnetic sensor of the present invention, the first signal can contain a first ideal component, a first third harmonic component, and a first fifth harmonic component, and the second signal can contain a second ideal component, a second third harmonic component, and a second fifth harmonic component when the direction of the target magnetic field rotates with a predetermined period. The first ideal component is a component that changes periodically such that it describes an ideal sine wave. The first third harmonic component is an error component equivalent to a third harmonic of the first ideal component. The first fifth harmonic component is an error component equivalent to a fifth harmonic of the first ideal component. The second ideal component is a component that changes periodically such that it describes an ideal sine wave.The second third harmonic component is an error component equivalent to a third harmonic of the second ideal component. The second fifth harmonic component is an error component equivalent to a fifth harmonic of the second ideal component.

[0017] The magnitude of a difference between a third harmonic component ratio and a fifth harmonic component ratio can change by 0.18% or less when the strength of the target magnetic field at the magnetic field detection unit changes within the sub-area, wherein the third harmonic component ratio is the mean of a first ratio and a third ratio, and the fifth harmonic component ratio is the mean of a second ratio and a fourth ratio, and wherein the first ratio is the ratio of the first third harmonic component to the first ideal component when the first ideal component is at its maximum value, the second ratio is the ratio of the first fifth harmonic component to the first ideal component when the first ideal component is at its maximum value, and the third ratio is the ratio of the second third harmonic component to the second ideal component.when the second ideal part takes on a maximum value, and the fourth ratio is the ratio of the second fifth harmonic part to the second ideal part when the second ideal part takes on a maximum value.

[0018] In the magnetic sensor of the present invention, the processing unit can comprise an angle detection value calculation section for calculating the value of the detected angle, and a correction processing section that performs correction processing on the value of the detected angle, thereby generating a corrected value of the detected angle. The maximum magnitude of an error in the corrected value of the detected angle is less than the maximum magnitude of an error in the value of the detected angle. In this case, the correction processing section can perform the correction processing such that half the difference between a maximum and a minimum value of the error in the corrected value of the detected angle is 0.1° or less, regardless of the value of the target magnetic field strength at the magnetic field detection unit within the partial range.

[0019] The magnetic sensor of the present invention can be configured such that the maximum amount of the angular error component changes by 0.05° or less when the strength of the target magnetic field at the magnetic field detection unit changes within the sub-range. In this case, the amount of the difference between the ratio of the third harmonic component and the ratio of the fifth harmonic component can change by 0.09% or less when the strength of the target magnetic field at the magnetic field detection unit changes within the sub-range. The processing unit can include an angle detection value calculation section for calculating the value of the detected angle and a correction processing section that performs a correction processing on the value of the detected angle, thereby generating a corrected value of the detected angle.The maximum magnitude of an error in the corrected value of the detected angle is less than the maximum magnitude of an error in the value of the detected angle itself. In this case, the correction processing section can perform the correction processing such that half the difference between the maximum and minimum values ​​of the error in the corrected value of the detected angle is 0.05° or less, regardless of the value of the target magnetic field strength at the magnetic field detection unit within the sub-area.

[0020] In the magnetic sensor of the present invention, the second direction can be orthogonal to the first direction.

[0021] In the magnetic sensor of the present invention, the magnetic field detection unit can include a first detection circuit for outputting a first signal and a second detection circuit for outputting a second signal. Both the first detection circuit and the second detection circuit can include a chain of magnetoresistive elements consisting of two or more of the plurality of magnetoresistive elements connected in series. The free layer of each of the plurality of magnetoresistive elements has a first surface in contact with the non-magnetic layer and a second surface opposite the first surface. The second surface has a rotationally symmetric shape of order four, not order five or higher. The number of two or more magnetoresistive elements forming the chain of magnetoresistive elements is even.The two or more magnetoresistive elements forming the magnetoresistive element chain contain one or more pairs of magnetoresistive elements. The magnetization directions of the fixed magnetization layers of two paired magnetoresistive elements form a predetermined relative angle that is not 0° or 180°. In the first detection circuit, the first direction is either an intermediate direction between the magnetization directions of the fixed magnetization layers of two paired magnetoresistive elements or a direction opposite to the intermediate direction. In the second detection circuit, the second direction is either an intermediate direction between the magnetization directions of the fixed magnetization layers of two paired magnetoresistive elements or a direction opposite to the intermediate direction.

[0022] According to the magnetic sensor of the present invention, the change in the maximum angular error component is 0.1° or less when the strength of the target magnetic field at the magnetic field detection unit changes within the sub-area. The present invention thus achieves a reduced change in the angular error over a large range of target magnetic field strengths.

[0023] Other and further tasks, features and advantages of the present invention will become more fully apparent from the following description. Brief description of the drawing Fig. Figure 1 is a perspective view showing the general structure of a magnetic sensor system containing a magnetic sensor according to an embodiment of the invention. Fig. Figure 2 is an explanatory diagram illustrating the definitions of directions and angles used in the embodiment of the invention. Fig. Figure 3 is a circuit diagram illustrating the construction of the magnetic sensor according to the embodiment of the invention. Fig. 4 is a function block diagram that shows the structure of a calculation unit of the in Fig. 3 shows the magnetic sensor. Fig. Figure 5 is an explanatory diagram showing the relationship between the magnetization directions of the fixed magnetization layers of two paired MR elements in the Fig. 3 shows the magnetic sensor. Fig. 6 is a top view of a Fig. 3 Wheatstone bridge circuits shown. Fig. Figure 7 is a perspective view of part of a building. Fig. 6 magnetic detection elements shown. Fig. 8A to Fig. 8C are waveform diagrams that show a first example of the relationship between a fourth-order angular error component and a multitude of error components. Fig. 9A to Fig. 9C are waveform diagrams that show a second example of the relationship between the fourth-order angular error component and the multitude of error components. Fig. 10A to Fig. 10C are waveform diagrams that show a third example of the relationship between the fourth-order angular error component and the multitude of error components. Fig. 11A to Fig. 11C are waveform diagrams showing a fourth example of the relationship between the fourth-order angular error component and the multitude of error components. Fig. Figure 12 is a characteristic curve diagram showing the relationship between the magnitude of the difference between a ratio of the third harmonic component and a ratio of the fifth harmonic component and the maximum magnitude of the fourth-order angular error component. Fig. Figure 13 is an explanatory diagram illustrating the shape of the second surface of the free layer of the MR element in the embodiment of the invention. Fig. Figure 14 is an explanatory diagram illustrating the shape of the second surface of the free layer of the MR element in the embodiment of the invention. Fig. Figure 15 is an explanatory diagram illustrating the shape of the second surface of the free layer of the MR element in the embodiment of the invention. Fig. Figure 16 is a characteristic curve diagram showing how the dependencies of the ratio of the third harmonic component and the ratio of the fifth harmonic component on the magnetic field strength change with the change in the shape of the second surface of the free layer of the MR element. Fig. Figure 17 is a characteristic curve diagram showing how the dependence of the maximum magnitude of the fourth-order angular error component on the magnetic field strength changes with the change in the shape of the second surface of the free layer of the MR element. Fig. Figure 18 is a characteristic curve diagram showing how the dependencies of the ratio of the third harmonic component and the ratio of the fifth harmonic component on the magnetic field strength change with the change in the offset angle. Fig. Figure 19 is a characteristic curve diagram showing how the dependence of the maximum magnitude of the fourth-order angular error component on the magnetic field strength changes with the change in the offset angle. Fig. Figure 20 is a characteristic curve diagram showing the dependencies of the ratio of the third harmonic component and the ratio of the fifth harmonic component on the magnetic field strength in a first example. Fig. Figure 21 is a characteristic curve diagram showing the dependence of the maximum magnitude of the fourth-order angular error component on the magnetic field strength in the first example. Fig. Figure 22 is a characteristic curve diagram showing an angular error in the first example. Fig. Figure 23 is a characteristic curve diagram showing an estimated angular error in the first example. Fig. Figure 24 is a characteristic curve diagram showing a corrected angular error in the first example. Fig. Figure 25 is a characteristic curve diagram showing the relationship between an applied field strength and the amplitude of the corrected angular error in the first example. Fig. Figure 26 is a characteristic curve diagram showing the dependencies of the ratio of the third harmonic component and the ratio of the fifth harmonic component on the magnetic field strength in a second example. Fig. Figure 27 is a characteristic curve diagram showing the dependence of the maximum magnitude of the fourth-order angular error component on the magnetic field strength in the second example. Fig. Figure 28 is a characteristic curve diagram showing the dependencies of the ratio of the third harmonic component and the ratio of the fifth harmonic component on the magnetic field strength in a third example. Fig. Figure 29 is a characteristic curve diagram showing the dependence of the maximum magnitude of the fourth-order angular error component on the magnetic field strength in the third example. Detailed description of the preferred embodiment

[0024] A preferred embodiment of the present invention will now be described in detail with reference to the drawing. First, it is noted that Fig. 1 and Fig. 2 Reference is made to describe the general structure of a magnetic sensor system with a magnetic sensor according to the embodiment of the invention. Fig. Figure 1 is a perspective view showing the general structure of the magnetic sensor system according to this embodiment. Fig. Figure 2 is an explanatory diagram illustrating the definitions of directions and angles used in the embodiment.

[0025] As in Fig. As shown in Figure 1, the magnetic sensor system of this embodiment comprises a magnetic sensor 1 according to this embodiment and a magnetic field generation unit 2. The magnetic field generation unit 2 generates a target magnetic field MF. In the Fig. In the example shown in Figure 1, the magnetic field generating unit 2 is a cylindrical magnet. The magnetic field generating unit 2 has an N pole and an S pole, which are arranged symmetrically with respect to an imaginary plane that includes the central axis of the cylinder. The magnetic field generating unit 2 rotates about the central axis of the cylinder. Therefore, the direction of the target magnetic field MF generated by the magnetic field generating unit 2 rotates about a rotation axis C that includes the central axis of the cylinder.

[0026] The magnetic sensor 1 is designed to detect the angle formed by the direction of the target magnetic field MF at a reference position with respect to a reference direction. More precisely, the magnetic sensor 1 generates a value for a detected angle that is proportional to the angle formed by the direction of the target magnetic field MF at the reference position with respect to the reference direction.

[0027] The reference position is located in an imaginary plane parallel to an end face of the magnetic field generating unit 2. This imaginary plane is referred to below as the reference plane. In the reference plane, the direction of the target magnetic field MF generated by the magnetic field generating unit 2 rotates around the reference position. The reference direction lies in the reference plane and passes through the reference position. In the following description, the direction of the target magnetic field MF at the reference position is a direction lying in the reference plane. The direction of the target magnetic field MF at the reference position can simply be referred to as the direction of the target magnetic field MF. The magnetic sensor 1 is arranged so that it points towards the aforementioned end face of the magnetic field generating unit 2.

[0028] The magnetic sensor 1 and the magnetic field generation unit 2 can also be used in ways other than as shown. Fig. The magnetic sensor 1 and the magnetic field generation unit 2 simply need to be configured such that their relative position changes in such a way that the direction of the target magnetic field MF rotates from the perspective of the magnetic sensor 1. For example, the magnetic sensor 1 and the magnetic field generation unit 2, as shown in Figure 1, can be configured as follows: Fig. 1 are arranged as follows: The magnetic sensor 1 rotates while the magnetic field generation unit 2 is stationary; the magnetic sensor 1 and the magnetic field generation unit 2 rotate in opposite directions; or the magnetic sensor 1 and the magnetic field generation unit 2 rotate in the same direction but with different angular velocities.

[0029] Alternatively, instead of the in Fig. 1. A magnet 2 is used as a magnetic field generating unit in the magnet shown in Figure 1, comprising one or more pairs of N and S poles arranged alternately in a ring, wherein the magnetic sensor 1 can be arranged near the outer circumference of the magnet. In this case, at least one of the magnet and the magnetic sensor 1 rotates.

[0030] Alternatively, instead of the in Fig. 1. A magnetic field generating unit 2 is used with the magnet shown in Figure 1. A magnetic tape is used which comprises a plurality of pairs of N and S poles arranged alternately in a linear configuration, wherein the magnetic sensor 1 can be positioned near the edge region of the magnetic tape. In this case, at least one of the magnetic tape and magnetic sensor 1 moves linearly in the direction in which the N and S poles of the magnetic tape are arranged.

[0031] In the various configurations of the magnetic sensor 1 and the magnetic field generation unit 2 described above, there is also a reference plane which has a predefined positional relationship to the magnetic sensor 1, and within the reference plane the direction of the target magnetic field MF rotates around the reference position from the perspective of the magnetic sensor 1.

[0032] The magnetic sensor 1 includes a magnetic field detection unit 3. The magnetic field detection unit 3 contains a plurality of magnetic detection elements for detecting the target magnetic field MF and outputs a first signal S1 and a second signal S2. The first signal S1 is related to the angle formed by the direction of the target magnetic field MF with respect to a first direction. The second signal S2 is related to the angle formed by the direction of the target magnetic field MF with respect to a second direction. Each of the plurality of magnetic detection elements contains at least one magnetoresistive (MR) element for detecting the target magnetic field MF. The magnetic field detection unit 3 thus contains a plurality of MR elements.

[0033] The magnetic field detection unit 3 contains a first detection circuit 10 and a second detection circuit 20. For ease of understanding, the following is presented. Fig. 1. The first and second detection circuits 10 and 20 are represented as separate components. However, the first and second detection circuits 10 and 20 can be integrated into a single component. Furthermore, while in Fig. 1. The first and second detection circuits 10 and 20 are stacked in a direction parallel to the axis of rotation C, the stacking order being opposite to that shown in Fig. The opposite is true for example 1.

[0034] The definitions of the directions and angles used in the embodiment are now given with reference to Fig. 1 and Fig. 2 explained. First, the Z-direction corresponds to the direction parallel to the one in Fig. 1 shown axis of rotation C and from bottom to top in Fig. 1. In Fig. 2 is the Z-direction, defined as the direction out of the plane of the Fig. 2 shown out. Next, the X and Y directions are the two directions that are perpendicular to the Z direction and orthogonal to each other. In Fig. 2. The X direction is shown as the direction to the right, and the Y direction is shown as the direction upwards. Furthermore, the -X direction is the direction opposite to the X direction, and the -Y direction is the direction opposite to the Y direction.

[0035] The reference position PR is the position where the magnetic sensor 1 detects the target magnetic field MF. The reference direction DR is the X-direction. The angle formed by the direction DM of the target magnetic field MF at the reference position PR with respect to the reference direction DR is denoted by the symbol θ. The direction DM of the target magnetic field MF is to point in Fig. 2. Rotate counterclockwise. The angle θ is expressed in positive values ​​when rotating counterclockwise, as seen from the reference direction DR, and in negative values ​​when rotating clockwise, as seen from the reference direction DR.

[0036] The construction of the magnetic sensor 1 will now be described in detail with reference to Fig. 3 described. Fig. Figure 3 is a circuit diagram illustrating the structure of the magnetic sensor 1. The magnetic sensor 1 contains a processing unit 30 and the magnetic field detection unit 3 described above. As described above, the magnetic field detection unit 3 contains a first detection circuit 10 and a second detection circuit 20.

[0037] The first detection circuit 10 generates the first signal S1, which has a correspondence to the angle formed by the direction DM of the target magnetic field MF with respect to the first direction. The second detection circuit 20 generates the second signal S2, which has a correspondence to the angle formed by the direction DM of the target magnetic field MF with respect to the second direction. In this embodiment, the first direction is the X-direction and the second direction is the Y-direction. The second direction is therefore orthogonal to the first direction. Fig. In section 2, the arrow labeled D1 indicates the first direction, and the arrow labeled D2 indicates the second direction. For example, the first signal S1 is a signal that responds to the strength of a component of the target magnetic field MF at the reference position PR, where the component is the first direction D1 (the X-direction). Similarly, the second signal S2 is a signal that responds to the strength of a component of the target magnetic field MF at the reference position PR, where the component is the second direction D2 (the Y-direction).

[0038] If the direction DM of the target magnetic field MF rotates with a predetermined period, the first and second signals S1 and S2 change periodically with a signal period T corresponding to the predetermined period. The second signal S2 differs from the first signal S1 in phase. In this embodiment, the phase difference of the second signal S2 preferably differs from the first signal S1 by an odd multiple of 1 / 4 of the signal period T. However, due to the manufacturing accuracy of the MR elements and other factors, the phase difference between the first signal S1 and the second signal S2 may deviate slightly from an odd multiple of 1 / 4 of the signal period T. In the following description, it is assumed that the phases of the first signal S1 and the second signal S2 satisfy the aforementioned preferred relationship.

[0039] The first acquisition circuit 10 has an output for the first signal S1. The second acquisition circuit 20 has an output for the second signal S2. The processing unit 30 has two inputs and one output. The two inputs of the processing unit 30 are connected to the respective outputs of the first and second acquisition circuits 10 and 20.

[0040] Fig. Figure 4 is a functional block diagram illustrating the structure of the computation unit 30. The computation unit 30 contains an angle detection value calculation section 31, a correction processing section 32, and a correction information storage section 33. Based on the first and second signals S1 and S2, output by the first and second detection circuits 10 and 20, the angle detection value calculation section 31 calculates the value of the detected angle θs, which has a correspondence relationship to the angle θ. The correction processing section 32 performs the correction processing on the value of the detected angle θs, thereby generating a corrected value of the detected angle θt. The maximum magnitude of an error in the corrected value of the detected angle θt is less than the maximum magnitude of an error in the value of the detected angle θs.The correction information storage section 33 contains the correction information to be used by the correction processing section 32 when performing the correction processing. The corrected value of the detected angle θt is the value of the angle θ detected by the magnetic sensor 1. The processing unit 30 can be implemented, for example, as an application-specific integrated circuit (ASIC) or as a microcomputer. A detailed description of the calculation of the value of the detected angle θs by the angle detection value calculation section 31 and the correction processing by the correction processing section 32 is given below.

[0041] As will be described later, the calculation unit 30 can contain neither the correction processing section 32 nor the correction information storage section 33 and can output the value of the detected angle θs as the value of the angle θ detected by the magnetic sensor 1.

[0042] The first detection circuit 10 comprises a Wheatstone bridge circuit 14 and a differential detector 15. The Wheatstone bridge circuit 14 includes a power supply terminal V1, a ground terminal G1, two output terminals E11 and E12, magnetic sensing elements R11 and R12 connected in series, and magnetic sensing elements R13 and R14 connected in series. One end of each of the magnetic sensing elements R11 and R13 is connected to the power supply terminal V1. The other end of magnetic sensing element R11 is connected to one end of magnetic sensing element R12 and the output terminal E11. The other end of magnetic sensing element R13 is connected to one end of magnetic sensing element R14 and the output terminal E12. The other end of each of the magnetic sensing elements R12 and R14 is connected to the ground terminal G1.A supply voltage of a predetermined amount is applied to the power supply terminal V1. The ground terminal G1 is connected to earth. The differential detector 15 outputs a signal value S1 to the processing unit 30, corresponding to the potential difference between the output terminals E11 and E12.

[0043] The second detection circuit 20 has a similar circuit design to the first detection circuit 10. More precisely, the second detection circuit 20 includes a Wheatstone bridge circuit 24 and a differential detector 25. The Wheatstone bridge circuit 24 includes a power supply terminal V2, a ground terminal G2, two output terminals E21 and E22, magnetic detection elements R21 and R22 connected in series, and magnetic detection elements R23 and R24 connected in series. One end of each of the magnetic detection elements R21 and R23 is connected to the power supply terminal V2. The other end of magnetic detection element R21 is connected to one end of magnetic detection element R22 and the output terminal E21. The other end of magnetic detection element R23 is connected to one end of magnetic detection element R24 and the output terminal E22.The other end of each of the magnetic detection elements R22 and R24 is connected to the ground terminal G2. A supply voltage of a predetermined amount is applied to the power supply terminal V2. The ground terminal G2 is connected to earth. The differential detector 25 outputs a signal value, S2, to the processing unit 30, corresponding to the potential difference between the output terminals E21 and E22.

[0044] As described above, each of the plurality of magnetic sensing elements included in the Wheatstone bridge circuits (hereinafter referred to as bridge circuits) 14 and 24 contains at least one MR element. In this embodiment, the MR element is a spin-valve MR element, in particular a TMR element. The TMR element can be replaced by a GMR element. The spin-valve MR element comprises a fixed-magnetization layer with a fixed magnetization direction, a free layer with a magnetization direction that changes depending on the direction DM of the target magnetic field MF, and a non-magnetic layer located between the fixed-magnetization layer and the free layer. In the TMR element, the non-magnetic layer is a tunnel barrier layer. In the GMR element, the non-magnetic layer is a non-magnetic conductive layer.The spin valve MR element changes its resistance value depending on the angle formed by the magnetization direction of the free layer relative to the magnetization direction of the layer with fixed magnetization, and has a minimum resistance value when the said angle is 0°, and a maximum resistance value when the said angle is 180°.

[0045] In this embodiment, each of the plurality of magnetic sensing elements contained in the bridge circuits 14 and 24 consists, in particular, of a chain of MR elements. The chain of MR elements consists of two or more MR elements connected in series. The number of the two or more MR elements forming the chain is even. The two or more MR elements forming the chain contain one or more pairs of MR elements. The magnetization directions of the fixed magnetization layers of two paired MR elements form a predetermined relative angle that is not 0° or 180°.

[0046] In the first detection circuit 10, the first direction D1 is either an intermediate direction between the magnetization directions of the fixed magnetization layers of two paired MR elements, or a direction opposite to the intermediate direction. In the second detection circuit 20, the second direction D2 is either an intermediate direction between the magnetization directions of the fixed magnetization layers of two paired MR elements, or a direction opposite to the intermediate direction.

[0047] Fig. Figure 3 shows an example where the two or more MR elements forming the chain contain only one pair of MR elements, as explained below. Magnetic sensing element R11 consists of a pair of MR elements R111 and R112. Magnetic sensing element R12 consists of a pair of MR elements R121 and R122. Magnetic sensing element R13 consists of a pair of MR elements R131 and R132. Magnetic sensing element R14 consists of a pair of MR elements R141 and R142. Magnetic sensing element R21 consists of a pair of MR elements R211 and R212. Magnetic sensing element R22 consists of a pair of MR elements R221 and R222. Magnetic sensing element R23 consists of a pair of MR elements R231 and R232. The magnetic detection element R24 consists of a pair of MR elements R241 and R242. In Fig. 3. The filled arrows indicate the magnetization directions of the layers with fixed magnetization of the MR elements, and the hollow arrows indicate the magnetization directions of the free layers of the MR elements.

[0048] It will now be on Fig. 5. Reference is made to describe the relationship between the magnetization directions of the fixed magnetization layers of two paired MR elements. Fig. Figure 5 shows that the arrows labeled D111 and D112 indicate the magnetization directions of the layers with fixed magnetization of the MR elements R111 and R112, respectively. The magnetization directions D111 and D112 of the layers with fixed magnetization of the MR elements R111 and R112 are defined such that the intermediate direction between directions D111 and D112 coincides with the first direction D1 (the X-direction). The magnetization direction D111 of the layer with fixed magnetization of the MR element R111 is the direction rotated clockwise by an angle ϕ from the first direction D1. The magnetization direction D112 of the layer with fixed magnetization of the MR element R112 is the direction rotated counterclockwise by an angle ϕ from the first direction D1. In the following, the angle ϕ will be referred to as the offset angle ϕ.

[0049] The magnetization directions of the fixed magnetization layers of MR elements R121 and R122 are defined such that the intermediate direction between these magnetization directions is the -X direction, i.e., the direction opposite to the first direction D1. The magnetization directions of the fixed magnetization layers of MR elements R131 and R132 are also defined such that the intermediate direction between these magnetization directions is opposite to the first direction D1. The magnetization direction of the fixed magnetization layer of MR element R131 is the same as that of the fixed magnetization layer of MR element R121. The magnetization direction of the fixed magnetization layer of MR element R132 is the same as that of the fixed magnetization layer of MR element R122.The magnetization direction of the layers with fixed magnetization of the MR elements R121 and R131 is the magnetization direction D111 of the layer with fixed magnetization of the in . Fig. The magnetization direction of the layers with fixed magnetization of the MR elements R122 and R132 is opposite to the magnetization direction D112 of the layer with fixed magnetization of the in Fig. 5 MR elements shown are opposite to R112.

[0050] The magnetization directions of the fixed magnetization layers of MR elements R141 and R142 are defined such that the intermediate direction between these magnetization directions coincides with the first direction D1 (the X-direction). The magnetization direction of the fixed magnetization layer of MR element R141 coincides with the magnetization direction D111 of the fixed magnetization layer of the Fig. The magnetization direction of the fixed magnetization layer of the MR element R142 matches the magnetization direction D112 of the fixed magnetization layer shown in the diagram. Fig. 5 MR elements shown are R112.

[0051] In Fig. Figure 5 shows that the arrows labeled D211 and D212 indicate the magnetization directions of the layers with fixed magnetization of the MR elements R211 and R212, respectively. The magnetization directions D211 and D212 of the layers with fixed magnetization of the MR elements R211 and R212 are defined such that the intermediate direction between directions D211 and D212 coincides with the second direction D2 (the Y-direction). The magnetization direction D211 of the layer with fixed magnetization of the MR element R211 is the direction rotated clockwise from the second direction D2 by the offset angle ϕ. The magnetization direction D212 of the layer with fixed magnetization of the MR element R212 is the direction rotated counterclockwise from the second direction D2 by the offset angle ϕ.

[0052] The magnetization directions of the fixed magnetization layers of MR elements R221 and R222 are defined such that the intermediate direction between these magnetization directions is the -Y- direction, i.e., the direction opposite to the second direction D2. The magnetization directions of the fixed magnetization layers of MR elements R231 and R232 are also defined such that the intermediate direction between these magnetization directions is opposite to the second direction D2. The magnetization direction of the fixed magnetization layer of MR element R231 is the same as that of the fixed magnetization layer of MR element R221. The magnetization direction of the fixed magnetization layer of MR element R232 is the same as that of the fixed magnetization layer of MR element R222.The magnetization direction of the layers with fixed magnetization of the MR elements R221 and R231 is the magnetization direction D211 of the layer with fixed magnetization of the in . Fig. The magnetization direction of the layers with fixed magnetization of the MR elements R211 shown in Figure 5 is opposite to that of the MR elements R222 and R232. Fig. 5 MR elements shown are opposite to R212.

[0053] The magnetization directions of the fixed magnetization layers of MR elements R241 and R242 are defined such that the intermediate direction between these magnetization directions coincides with the second direction D2 (the Y direction). The magnetization direction of the fixed magnetization layer of MR element R241 coincides with the magnetization direction D211 of the fixed magnetization layer of the Fig. The magnetization direction of the fixed magnetization layer of the MR element R211 shown in diagram 5 is identical. The magnetization direction of the fixed magnetization layer of the MR element R242 matches the magnetization direction D212 of the fixed magnetization layer of the MR element shown in diagram 5. Fig. 5 MR elements shown are R212.

[0054] Considering the manufacturing accuracy of the MR elements and other factors, the magnetization directions of the fixed magnetization layers of the multitude of MR elements in the detection circuits 10 and 20 may differ slightly from those described above.

[0055] In the first detection circuit 10, the potential difference between the output terminals E11 and E12 changes depending on the angle between the direction DM of the target magnetic field MF and the first direction D1 (the X-direction). The first detection circuit 10 thus generates the first signal S1, which has a relationship to the angle formed by the direction DM of the target magnetic field MF with respect to the first direction D1 (the X-direction).

[0056] In the second detection circuit 20, the potential difference between the output terminals E21 and E22 changes depending on the angle between the direction DM of the target magnetic field MF and the second direction D2 (the Y-direction). The second detection circuit 20 thus generates the second signal S2, which has a relationship to the angle formed by the direction DM of the target magnetic field MF with respect to the second direction D2 (the Y-direction).

[0057] An example of the construction of the bridge circuits and the magnetic detection elements will now be given with reference to Fig. 6 and Fig. 7 described. Fig. 6 is a top view of the Fig. 3 bridge circuit shown 14. Fig. Figure 7 is a perspective view of part of a building. Fig. 6 magnetic detection element shown. In the Fig. In the example shown in Figure 6, each of the magnetic sensing elements R11, R12, R13, and R14 of the bridge circuit 14 consists of a chain of MR elements containing eight MR elements 50 connected in series. The plurality of MR elements 50 comprising the chain of MR elements includes four pairs of MR elements 50. The chain of MR elements contains a plurality of lower electrodes 41 and a plurality of upper electrodes 42 by which the eight MR elements 50 are connected in series. Each of the MR elements 50 has an upper end face and a lower end face. The bridge circuit 24 has the same structure as the one shown in Figure 6. Fig. 6 Bridge circuit shown 14.

[0058] The multitude of lower electrodes 41 are arranged on a substrate (not shown). Each of the lower electrodes 41 has a long, slender shape. There is a gap between each pair of adjacent lower electrodes 41. As shown in Fig. As shown in Figure 7, two adjacent MR elements 50 are arranged longitudinally on the upper surface of each lower electrode 41 at positions near opposite ends. Each of the upper electrodes 42 has a long, slender shape and establishes an electrical connection between two MR elements 50 that are arranged on two adjacent lower electrodes 41 and are adjacent to each other.

[0059] As in Fig. As shown in Figure 7, each of the MR elements 50 contains an antiferromagnetic layer 54, a fixed magnetization layer 53, a non-magnetic layer 52, and a free layer 51, stacked in that order, with the antiferromagnetic layer 54 being closest to the lower electrode 41. The antiferromagnetic layer 54 is formed from an antiferromagnetic material. The antiferromagnetic layer 54 is in exchange interaction with the fixed magnetization layer 53 and thereby determines the magnetization direction of the fixed magnetization layer 53. It should be noted that the layers 51 to 54 of each MR element 50 are arranged in the Fig. The 7 items shown can be stacked in reverse order.

[0060] The free layer 51 has a first surface in contact with the non-magnetic layer and a second surface opposite the first surface. In the Fig. In the example shown in Figure 7, the second surface of the free layer 51 represents the upper end face of the MR element 50. Fig. 2 represents the symbol 11 the shape of the second surface of the free layer 51 of each MR element 50 in the first detection circuit 10 and the symbol 21 represents the shape of the second surface of the free layer 51 of each MR element 50 in the second detection circuit 20.

[0061] The in Fig. The bridge circuit 14 shown in Figure 6 contains four connecting electrodes 431, 432, 433 and 434 for the electrical connection of two magnetic detection elements. The connecting electrodes 431 to 434 are arranged on the substrate (not shown).

[0062] The connecting electrode 431 establishes an electrical connection between the MR element 50 located at one end of the magnetic sensing element R11 and the MR element 50 located at one end of the magnetic sensing element R13, and is electrically connected to the power supply terminal V1. The connecting electrode 432 establishes an electrical connection between the MR element 50 located at the other end of the magnetic sensing element R11 and the MR element 50 located at one end of the magnetic sensing element R12, and is electrically connected to the output terminal E11. The connecting electrode 433 establishes an electrical connection between the MR element 50 located at the other end of the magnetic sensing element R13 and the MR element 50 located at one end of the magnetic sensing element R14, and is electrically connected to the output terminal E12.The connecting electrode 434 establishes an electrical connection between the MR element 50, which is located at the other end of the magnetic sensing element R12, and the MR element 50, which is located at the other end of the magnetic sensing element R14, and is electrically connected to the ground terminal G1.

[0063] In the magnetic detection element R11, for example, the layers with fixed magnetization of the first to fourth MR element 50 from the power supply connection V1 have the same magnetization direction as that of the layer with fixed magnetization of the in Fig. 3 MR elements R111 shown, whereas the fixed magnetization layers of the fifth to eighth MR element 50 from the power supply terminal V1 have the same magnetization direction as that of the fixed magnetization layer of the in Fig. 3 MR elements R112 shown. In this case, the first MR element 50 is paired with the fifth MR element 50; the second MR element 50 is paired with the sixth MR element 50; the third MR element 50 is paired with the seventh MR element 50; and the fourth MR element 50 is paired with the eighth MR element 50.

[0064] In the magnetic detection element R12, for example, the layers with fixed magnetization of the first to fourth MR element 50 from the output terminal E11 have the same magnetization direction as that of the layer with fixed magnetization of the in Fig. 3 MR elements R121 shown, whereas the fixed magnetization layers of the fifth to eighth MR element 50 from the output terminal E11 have the same magnetization direction as that of the fixed magnetization layer of the in Fig. 3 MR elements R122 shown. In this case, the first MR element 50 is paired with the fifth MR element 50; the second MR element 50 is paired with the sixth MR element 50; the third MR element 50 is paired with the seventh MR element 50; and the fourth MR element 50 is paired with the eighth MR element 50.

[0065] In the magnetic detection element R13, for example, the layers with fixed magnetization of the first to fourth MR element 50 from the power supply connection V1 have the same magnetization direction as that of the layer with fixed magnetization of the in Fig. 3 MR elements R131 shown, whereas the fixed magnetization layers of the fifth to eighth MR element 50 from the power supply terminal V1 have the same magnetization direction as that of the fixed magnetization layer of the in Fig. 3 MR elements R132 shown. In this case, the first MR element 50 is paired with the fifth MR element 50; the second MR element 50 is paired with the sixth MR element 50; the third MR element 50 is paired with the seventh MR element 50; and the fourth MR element 50 is paired with the eighth MR element 50.

[0066] In the magnetic detection element R14, for example, the layers with fixed magnetization of the first to fourth MR element 50 from the output terminal E12 have the same magnetization direction as that of the layer with fixed magnetization of the in Fig. 3 MR elements R141 shown, whereas the fixed magnetization layers of the fifth to eighth MR element 50 from the output terminal E12 have the same magnetization direction as that of the fixed magnetization layer of the in Fig. 3 MR elements R142 shown. In this case, the first MR element 50 is paired with the fifth MR element 50; the second MR element 50 is paired with the sixth MR element 50; the third MR element 50 is paired with the seventh MR element 50; and the fourth MR element 50 is paired with the eighth MR element 50.

[0067] It should now refer to Fig. 3 and Fig. Section 4 describes how the value of the detected angle θs is calculated from the angle detection value calculation section 31. In the Fig. In the example shown, the first signal S1 should ideally have a cosine waveform that depends on the angle θ, and the second signal S2 a sine waveform that also depends on the angle θ. In this case, the second signal S2 differs in phase from the first signal S1 by 1 / 4 of the signal period T, i.e., by π / 2 (90°).

[0068] If the angle θ is greater than or equal to 0° and less than 90°, and if the angle θ is greater than 270° and less than or equal to 360°, the first signal S1 has a positive value. If the angle θ is greater than 90° and less than 270°, the first signal S1 has a negative value. If the angle θ is greater than 0° and less than 180°, the second signal S2 has a positive value. If the angle θ is greater than 180° and less than 360°, the second signal S2 has a negative value.

[0069] Based on the first and second signals S1 and S2, the angle detection value calculation section 31 calculates the value of the detected angle θs, which has a correspondence relationship to the angle θ. More precisely, the angle detection value calculation section 31 calculates θs according to the following equation (1). Note that "atan" stands for arctangent. θs=atan(E2 / S1)

[0070] The expression "atan (S2 / S1)" in equation (1) represents the arctangent calculation for determining θs. For θs in the range from 0° to less than 360°, equation (1) provides two solutions for θs with a difference of 180°. Which of the two solutions for θs from equation (1) is the true value of θs can be determined from the combination of positive and negative signs for S1 and S2. More precisely, if S1 has a positive value, θs falls within the range of greater than or equal to 0° to less than 90° and from greater than 270° to less than or equal to 360°. If S1 has a negative value, θs is greater than 90° and less than 270°. If S2 has a positive value, θs is greater than 0° and less than 180°. If S2 has a negative value, then θs is greater than 180° and less than 360°.Using equation (1) and based on the above determination from the combination of positive and negative signs at S1 and S2, the angle detection value calculation section 31 determines θs in the range from 0° to less than 360°.

[0071] If the direction DM of the target magnetic field MF rotates with the specified period, the first and second signals S1 and S2 change periodically with a signal period T equal to the specified period, as described previously. Ideally, the waveform of each of the signals S1 and S2 should describe a sinusoidal curve (including sine and cosine waveforms). In reality, both the first signal S1 and the second signal S2 have a waveform that is distorted relative to a sinusoidal curve if, for example, the magnetization direction of the fixed magnetization layer 53 of the MR element 50 changes under the influence of the target magnetic field MF or similar factors, or if the magnetization direction of the free layer 51 of the MR element 50 deviates from the direction DM of the target magnetic field MF due to effects such as the shape anisotropy of the free layer 51.

[0072] The distortion of the waveform of each of the signals S1 and S2 from a sinusoidal curve means that signals S1 and S2 each contain not only an ideal component that changes periodically in such a way that the curve has an ideal sinusoidal shape, but also one or more error components equivalent to one or more harmonics of the ideal component. The one or more error components contained in each of the signals S1 and S2 cause the value of the detected angle θs to contain an error. This error in the value of the detected angle θs is referred to below as the angular error. If the direction DM of the target magnetic field MF rotates with the specified period, the angular error may contain one or more angular error components that change periodically with one or more periods different from the specified period.

[0073] In the research conducted by the inventors of the present invention, it was found that magnetic sensors generally exhibit fluctuations in angular error depending on the strength of the target magnetic field, i.e., the angular error is dependent on the magnetic field strength. Furthermore, the research conducted by the inventors of the present invention revealed that this dependence of the angular error on the magnetic field strength is primarily attributable to one or more of the aforementioned angular error components, i.e., an angular error component that changes with a period of 1 / 4 of the predetermined period. This angular error component is hereinafter referred to as the fourth-order angular error component.The inventors have discovered that the fourth-order angular error component changes when the strength of the target magnetic field MF at the magnetic field detection unit 3 is varied, and the nature of this change corresponds to the nature of the change in the total angular error when the strength of the target magnetic field MF at the magnetic field detection unit 3 is varied. Hereinafter, the strength of the target magnetic field MF at the magnetic field detection unit 3 is referred to as the applied field strength. The inventors have further discovered that by reducing the fourth-order angular error component, the total angular error can be reduced.

[0074] We now consider a case where the dependence of the angular error on the magnetic field strength leads to a large fluctuation of the angular error within a predefined range, which is the expected range of the applied field strength during the use of magnetic sensor 1. In such a case, correction processing can be performed on the detected angle to reduce the angular error when the applied field strength assumes one of certain values ​​within the predefined range. However, this cannot sufficiently reduce the angular error when the applied field strength assumes a value that does not belong to the predefined range.

[0075] The maximum magnitude of the fourth-order angular error component for the target magnetic field MF rotating in direction DM with the specified period is denoted by the symbol E4. By reducing the rate of change of the maximum magnitude E4 of the fourth-order angular error component over a large range of applied field strengths, the rate of change of the angular error can be reduced over this large range of applied field strengths. If this is achieved, it becomes possible to reduce the angular error over this large range of applied field strengths even at large angular errors by performing a correction process that is independent of the applied field strength.

[0076] The magnetic sensor 1 according to this embodiment is configured such that the maximum magnitude E4 of the fourth-order angular error component changes by 0.1° or less when the applied field strength changes within a sub-range that is part of a range from 20 to 150 mT and has an upper and lower limit that differ from each other by 30 mT or more. This sub-range, whose upper and lower limits differ by 30 mT or more, is sufficiently large for the range of applied field strengths in the practical application of the magnetic sensor 1. The magnetic sensor 1 thus configured according to this embodiment is capable of reducing the magnitude of the change in angular error over a large range of the applied field strength.

[0077] It will now be described in detail how the properties of the magnetic sensor 1 described above can be realized. The research of the inventors of the present invention has shown that the fourth-order angular error component depends on two error components contained in each of the signals S1 and S2. This will be described first. If the direction DM of the target magnetic field MF rotates with a predetermined period, the first signal S1 contains a first ideal component V11, a first third harmonic component V13, and a first fifth harmonic component V15, and the second signal S2 contains a second ideal component V21, a second third harmonic component V23, and a second fifth harmonic component V25. The first ideal component V11 is a component that changes periodically such that it describes an ideal sine wave.The first third harmonic component V13 is an error component equivalent to a third harmonic of the first ideal component V11. The first fifth harmonic component V15 is an error component equivalent to a fifth harmonic of the first ideal component V11. The second ideal component V21 is a component that changes periodically to describe an ideal sine wave. The second third harmonic component V23 is an error component equivalent to a third harmonic of the second ideal component V21. The second fifth harmonic component V25 is an error component equivalent to a fifth harmonic of the second ideal component V21.

[0078] We now define a first ratio V13r, a second ratio V15r, a third ratio V23r, and a fourth ratio V25r as follows. The first ratio V13r is the ratio of the first third harmonic component V13 to the first ideal component V11 when the first ideal component V11 reaches a maximum value. The second ratio V15r is the ratio of the first fifth harmonic component V15 to the first ideal component V11 when the first ideal component V11 reaches its maximum value. The third ratio V23r is the ratio of the second third harmonic component V23 to the second ideal component V21 when the second ideal component V21 reaches its maximum value. The fourth ratio V25r is the ratio of the second fifth harmonic component V25 to the second ideal component V21 when the second ideal component V21 reaches its maximum value.The mean of the first ratio V13r and the third ratio V23r is hereinafter referred to as the ratio of the third harmonic component and denoted by the symbol V3r. The mean of the second ratio V15r and the fourth ratio V25r is hereinafter referred to as the ratio of the fifth harmonic component and denoted by the symbol V5r.

[0079] The research conducted by the inventors of the present invention has shown that the maximum magnitude E4 of the fourth-order angular error component is proportional to the magnitude of the difference between the ratio of the third harmonic component V3r and the ratio of the fifth harmonic component V5r. E4 is approximately expressed by equation (2) below. In equation (2), E4 is given in degrees, and V3r and V5r are given as percentages. E4=|V3r−V5r|×0.56

[0080] According to equation (2), E4 is in principle zero if V3r and V5r are equal. Furthermore, according to equation (2), E4 is equal to 0.1° if the magnitude of the difference between V3r and V5r is 0.18%; and E4 is 0.05° if the magnitude of the difference between V3r and V5r is 0.09%.

[0081] The amount by which the magnitude of the difference between V3r and V5r varies when the applied field strength changes within the sub-range is also referred to as the change in the magnitude of the difference between V3r and V5r within the sub-range. The amount by which E4 varies when the applied field strength changes within the sub-range is also referred to as the change in E4 within the sub-range. According to equation (2), the change in E4 within the sub-range is 0.1° or less if the change in the magnitude of the difference between V3r and V5r within the sub-range is 0.18% or less. Furthermore, the change in E4 within the sub-range is 0.05° or less if the change in the magnitude of the difference between V3r and V5r within the sub-range is 0.09% or less.

[0082] Examples one to four of the relationship between the fourth-order angular error component and a multitude of error components determined by simulation will now be presented with reference to Fig. 8A to Fig. 11C described.

[0083] Fig. 8A, Fig. 9A, Fig. 10A and Fig. Figure 11A shows each of the waveforms V11, V13r and V15r. Fig. 8B, Fig. 9B, Fig. 10B and Fig. Figure 11B shows each of the waveforms V21, V23r and V25r. Fig. 8C, Fig. 9C, Fig. 10C and Fig. Figures 11C each show the waveform of the fourth-order angular error component. In each of the Fig. 8A to Fig. 11C shows the angle θ plotted on the horizontal axis. Fig. 8A, Fig. 9A, Fig. 10A and Fig. 11A shows the values ​​of V11, V13r, and V15r plotted on the vertical axis. V11 is expressed by its maximum value at 100%. In Fig. 8B, Fig. 9B, Fig. 10B and Fig. In 11B, the values ​​of V21, V23r, and V25r are plotted on the vertical axis. V21 is expressed by the fact that it reaches its maximum value at 100%. Fig. 8C, Fig. 9C, Fig. 10C and Fig. 11C shows the value of the fourth-order angular error component plotted on the vertical axis.

[0084] Fig. 8A, Fig. 8B and Fig. 8C represents the first example. In the first example, V13r, V15r, V23r, and V25r each have a value of 0.1%. In this case, both V3r and V5r are 0.1%. Since V3r and V5r are equal in this case, the fourth-order angular error component is zero, regardless of the angle θ, so E4 is zero.

[0085] Fig. 9A, Fig. 9B and Fig. Example 9C represents the second example. In this example, both V13r and V23r are zero, and both V15r and V25r are 0.1%. In this case, V3r is zero and V5r is 0.1%. Furthermore, the absolute value of the difference between V3r and V5r is 0.1%, so E4 is 0.056°.

[0086] Fig. 10A, Fig. 10B and Fig. Example 10C represents the third example. In this example, both V13r and V23r are equal to 0.1%, and both V15r and V25r are zero. In this case, V3r is equal to 0.1% and V5r is zero. Furthermore, the absolute value of the difference between V3r and V5r is 0.1%, so E4 is equal to 0.056°.

[0087] Fig. 11A, Fig. 11B and Fig. Example 11C represents the fourth example. In this example, both V13r and V23r are equal to -0.1%, and both V15r and V25r are equal to 0.1%. In this case, V3r is equal to -0.1% and V5r is equal to 0.1%. Furthermore, the absolute value of the difference between V3r and V5r is 0.2%, so E4 is 0.11°.

[0088] The results of a simulation carried out to determine the relationship between E4 and the magnitude of the difference between V3r and V5r, as expressed in equation (2), are now discussed with reference to Fig. 12 described. In the simulation, the relationship between E4 and the magnitude of the difference between V3r and V5r was determined using various combinations of values ​​for V3r and V5r. Fig. The results will be shown in section 12. Fig. 12 represents the magnitude of the difference between V3r and V5r on the horizontal axis and E4 on the vertical axis. The filled squares in Fig. 12 each represent a combination of a value of E4 and the magnitude of the difference between V3r and V5r according to the simulation. The straight line in Fig. Equation 12 is an approximately straight line that represents the relationship between E4 and the magnitude of the difference between V3r and V5r. Equation (2) describes this approximately straight line.

[0089] The change in V3r, V5r, and E4 as a function of the applied field strength is referred to below as the dependence of V3r, V5r, and E4, respectively, on the magnetic field strength. As previously described, E4 is proportional to the magnitude of the difference between V3r and V5r. Consequently, it is clear that a reduction in the magnitude of the change in E4 within the sub-range can be achieved by reducing the magnitude of the difference between V3r and V5r for any applied field strength within the sub-range. This can be achieved by adjusting the dependencies of V3r and V5r on the magnetic field strength such that the rate of change of V3r with respect to an increase in the applied field strength and the rate of change of V5r with respect to an increase in the applied field strength converge within the sub-range.

[0090] In this embodiment, a first means and a second means, described below, are used to control the dependencies of V3r and V5r on the magnetic field strength such that the rate of change of V3r with respect to an increase in the applied field strength and the rate of change of V5r with respect to an increase in the applied field strength approximate each other in a desired sub-region. The first means is the adaptation of the shape of the MR element 50, in particular the shape of the free layer 51. The second means is the adaptation of the offset angle ϕ, which is determined with respect to Fig. 5 was described.

[0091] The first method will now be described. First, the adaptation of the shape of the free layer 51 in this embodiment will be described with reference to Fig. 13 to Fig. 15. In this embodiment, the shape of the free layer 51 is defined by the shape of the second surface of the free layer 51. The MR element 50 is produced, for example, by first forming a stack from a plurality of films, which become the layers constituting the MR element 50, and then etching the stack to a desired shape using an etching mask. The shape of the second surface of the free layer 51 is defined by the shape of the etching mask. The first surface of the free layer 51, as well as an upper and lower surface of each of the layers comprising the MR element 50, except for the free layer 51, have the same or a similar shape to the second surface of the free layer 51.

[0092] Fig. Figure 13 is an explanatory diagram illustrating the shape of the second surface of the free layer 51. Fig. Reference numeral 51a (reference numeral 13) denotes the outer edge of the second surface of the free layer 51. To facilitate understanding, the shape of the outer edge 51a of the second surface is shown in Fig. Figure 13 is exaggerated. The reference symbol 51C designates a reference circle used to define the shape of the second surface of the free layer 51. The reference circle 51C is an exact circle.

[0093] Here, the centroid of the second surface of the free layer 51 is assumed to be the center of the second surface. The symbol θa denotes the angle formed with respect to the X-direction by a straight line connecting any point on the outer boundary 51a of the second surface to the center of the second surface (see Fig. 2) The angle θa indicates the counterclockwise rotation relative to the X-direction. The angle θa is greater than or equal to 0° and less than 360°. The distance from the center of the second surface to a given point on the outer boundary 51a of the second surface is denoted by R(θa).

[0094] In this embodiment, the shape of the second surface is adapted during the fabrication of the MR element 50 as follows. Specifically, in this embodiment, the second surface is given a rotationally symmetric shape of order four, not order five or higher. Furthermore, a distortion ratio and distortion direction, described below, are adapted. The rotationally symmetric shape of order four, not order five or higher, does not include a shape with rotational symmetry of order 4×n, where n is an integer greater than or equal to 2. The rotationally symmetric shape of order four, not order five or higher, is referred to below as the order four symmetric shape.

[0095] If the second surface has the symmetrical shape of order four and the angle θa is varied within the range from 0° to less than 360°, R(θa) takes on a maximum value for four values ​​of the angle θa that are 90° apart from each other, and a minimum value for four other values ​​of the angle θa that are 90° apart from each other.

[0096] Reference circle 51C is defined here by the mean of the maximum and minimum values ​​of R(θa) as the radius Rc of reference circle 51C. The center of reference circle 51C coincides with the center of the second surface. The maximum value of R(θa) minus the radius Rc of reference circle 51C is the distortion d. The distortion ratio dr is defined as the ratio of the distortion d to Rc (expressed as a percentage). The ratio of R(θa) to Rc, expressed as a percentage, is denoted as r(θa). The distortion direction α is defined as an angle θa for which R(θa) assumes a maximum value in the range of the angle θa from 0° to less than 90°. In this embodiment, r(θa) is expressed, for example, by the following equation (3). r(θa)=100+dr⋅cos(4(θa−α))

[0097] Fig. Figure 14 shows an example of the relationship between the angle θa and r(θa). In this example, the distortion ratio dr is 1% and the distortion direction α is 0°.

[0098] Fig. Figure 15 shows a part of the reference circle 51C and a part of the outer boundary 51a of the second surface in the region of the angle θa between 0° and inclusive 90°, when the angle θa and r(θa) are the Fig. The relationship shown in section 14 is evident. Fig. Figure 15 plots the position in the X direction on the horizontal axis and the position in the Y direction on the vertical axis. The positions in the X and Y directions are expressed as relative values, with the radius Rc of the reference circle 51C set to 100 and the center of the reference circle 51C defined as the origin.

[0099] In this embodiment, the adjustment range for the distortion ratio dr is preferably 0% to 10%. If the shape of the outer edge 51a of the second surface is square, the distortion ratio dr exceeds 10%. In this embodiment, the shape of the outer edge 51a of the second surface cannot be a square. If the distortion ratio dr is on the order of 1%, the shape of the outer edge 51a of the second surface will not be as described in Fig. Figure 13 is not clearly distorted compared to a perfect circle, but rather represents an almost perfect circle, as in Fig. 15 shown.

[0100] With reference to the results of a first experiment, it is now described that the dependencies of V3r, V5r, and E4 on the magnetic field strength can change by modifying the shape of the second surface of the free layer 51. In the first experiment, the dependencies of V3r, V5r, and E4 on the magnetic field strength were determined by setting the offset angle ϕ to 0° and changing the shape of the second surface of the free layer 51 in three ways, from the first to the third shape. Fig. Figure 16 is a characteristic curve diagram showing how the dependencies of V3r and V5r on the magnetic field strength change with the change in the shape of the second surface of the free layer 51. Fig. Figure 17 is a characteristic curve diagram showing how the dependence of E4 on the magnetic field strength changes with the change in the shape of the second surface of the free layer 51.

[0101] The shape of the second surface of the free layer 51 is defined by the distortion ratio dr and the distortion direction α. ​​The first shape of the second surface is such that dr equals 0.5% and α equals 45°. The second shape of the second surface is such that dr is zero. The third shape of the second surface is such that dr equals 0.5% and α equals 0°. In Fig. In the first form, V3r and V5r are expressed as V3r (dr = 0.5%, α = 45°) and V5r (dr = 0.5%, α = 45°), respectively; in the second form, V3r and V5r are expressed as V3r (dr = 0) and V5r (dr = 0), respectively; and in the third form, V3r and V5r are expressed as V3r (dr = 0.5%, α = 0°) and V5r (dr = 0.5%, α = 0°), respectively. Fig. 17 In the case of the first form, E4 is expressed as E4 (dr = 0.5%, α = 45°); in the case of the second form, E4 is expressed as E4 (dr = 0); and in the case of the third form, E4 is expressed as E4 (dr = 0.5%, α = 0°).

[0102] Fig. 16 and Fig. Figure 17 shows an example to describe trends in how the dependencies of V3r, V5r, and E4 on the magnetic field strength change when at least one of the distortion ratio dr and distortion direction α is changed. Although not in Fig. 16 and Fig. As shown in Figure 17, the change trends described below in the dependencies of V3r, V5r and E4 on the magnetic field strength when dr is varied while α is fixed become more pronounced with increasing dr; with decreasing dr, the dependencies of V3r, V5r and E4 on the magnetic field strength approach those of the case where dr equals zero.

[0103] According to Fig. 16. The following change trends in the dependencies of V3r and V5r on the magnetic field strength can be observed. When the distortion ratio dr is set to a value greater than zero, the dependencies of V3r and V5r on the magnetic field strength differ from those in the case of a distortion ratio dr of zero. The dependencies of V3r and V5r on the magnetic field strength also change depending on the distortion direction α. ​​If at least one of the distortion ratio dr and distortion direction α is changed, the dependencies of V3r and V5r on the magnetic field strength change in different ways. In the Fig. In the example shown in Figure 16, setting dr to 0.5% and α to 0°, compared to the case dr = zero, changes the dependence of V3r on the magnetic field strength such that V3r is increased, and changes the dependence of V5r on the magnetic field strength such that V5r is decreased. Conversely, setting dr to 0.5% and α to 45°, compared to the case dr = zero, changes the dependence of V3r on the magnetic field strength such that V3r is decreased, and changes the dependence of V5r on the magnetic field strength such that V5r is increased.

[0104] The magnitude of the change in V3r when at least one of the distortion ratio dr and distortion direction α is changed, and the magnitude of the change in V5r when at least one of the distortion ratio dr and distortion direction α is changed, each depend on the applied field strength.

[0105] As described above, the dependencies of V3r and V5r on the magnetic field strength change in different ways when at least one of the distortion ratio dr and distortion direction α is changed. Therefore, if at least one of the distortion ratio dr and distortion direction α is changed, the dependence of E4 on the magnetic field strength can be changed, as shown in Fig. 17 shown.

[0106] The second method is now described with reference to the results of a second experiment. In the second experiment, the dependencies of V3r, V5r, and E4 on the magnetic field strength were determined by shaping the second surface of the free layer 51 such that dr is zero and setting the offset angle ϕ in three ways, i.e., to 0°, 8°, and 17°. Fig. Figure 18 is a characteristic curve diagram showing how the dependencies of V3r and V5r on the magnetic field strength change with the change of the offset angle ϕ. Fig. Figure 19 is a characteristic curve diagram showing how the dependence of E4 on the magnetic field strength changes with the change of the offset angle ϕ.

[0107] In Fig. In section 18, V3r and V5r are expressed as V3r (ϕ = 0°) and V5r (ϕ = 0°) respectively for an offset angle ϕ of 0°; for an offset angle ϕ of 8°, V3r and V5r are expressed as V3r (ϕ = 8°) and V5r (ϕ = 8°) respectively; and for an offset angle ϕ of 17°, V3r and V5r are expressed as V3r (ϕ = 17°) and V5r (ϕ = 17°) respectively. It should be noted that in Fig. 18 the three lines for V5r (ϕ = 0°), V5r (ϕ = 8°) and V5r (ϕ = 17°) coincide. In Fig. 19. E4 is expressed as E4 (ϕ = 0°) for an offset angle ϕ of 0°; for an offset angle ϕ of 8°, E4 is expressed as E4 (ϕ = 8°); and for an offset angle ϕ of 17°, E4 is expressed as E4 (ϕ = 17°).

[0108] As in Fig. As can be seen in Figure 18, a change in the offset angle ϕ leads to a slight or no change in the dependence of V5r on the magnetic field strength, but changes the dependence of V3r on the magnetic field strength. As shown in Figure 18, a change in the offset angle ϕ results in a slight or no change in the dependence of V5r on the magnetic field strength. Fig. As shown in Figure 19, a change in the offset angle ϕ therefore leads to a change in the dependence of E4 on the magnetic field strength.

[0109] As in Fig. 17 and Fig. As can be seen in Figure 19, the dependence of E4 on the magnetic field strength can be modified by using either the first or the second means alone. However, by using either the first or the second means alone, it is difficult to achieve a sufficient reduction in the magnitude of the change in E4 over a large range of the applied field strength. Therefore, when using either the first or the second means alone, it can sometimes be difficult to realize the characteristics of the magnetic sensor 1 according to this embodiment, i.e., the characteristics that E4 changes by 0.1° or less when the applied field strength changes within a sub-range that is part of the range from 20 to 150 mT and has an upper and lower limit that differ from each other by 30 mT or more.

[0110] In contrast to the case of using either the first means or the second means alone, the use of the first means and the second means in combination makes it possible to sufficiently reduce the change in E4 over a large range of the applied field strength, so that the above-described properties of the magnetic sensor 1 according to this embodiment can be easily realized.

[0111] A first to third example of the magnetic sensor 1 according to this embodiment, as well as a magnetic sensor of a comparative example, are now described. The first to third examples each use a first means and a second means in combination. The magnetic sensor of the comparative example is an example where the second surface of the free layer 51 is shaped such that dr is equal to zero, and the offset angle ϕ is set to 0°. [First example]

[0112] The magnetic sensor 1 of the first example is an example where the second surface of the free layer 51 is shaped such that dr equals 1% and α equals 0°, and the offset angle ϕ is set to 25°. Fig. Figure 20 is a characteristic curve diagram showing the dependencies of V3r and V5r on the magnetic field strength for the magnetic sensor 1 of the first example and the magnetic sensor of the comparison example. Fig. Figure 21 is a characteristic curve diagram showing the dependence of E4 on the magnetic field strength for both the magnetic sensor 1 of the first example and the magnetic sensor of the comparison example. Fig. In the case of magnetic sensor 1 of the first example, V3r and V5r are expressed as V3r (dr = 1%, ϕ = 25°) and V5r (dr = 1%, ϕ = 25°), respectively; in the case of the magnetic sensor of the comparison example, V3r and V5r are expressed as V3r (dr = 0, ϕ = 0°) and V5r (dr = 0, ϕ = 0°), respectively. Fig. 21 In the case of the magnetic sensor 1 of the first example, E4 is expressed as E4 (dr = 1%, ϕ = 25°), while in the case of the magnetic sensor of the comparison example, E4 is expressed as E4 (dr = 0, ϕ = 0°).

[0113] Fig. 20 and Fig. Figure 21 shows that, according to the magnetic sensor 1 of the first example, the magnitude of the difference between V3r and V5r changes by 0.18% or less and E4 by 0.1° or less in the range of 40 to 150 mT of the applied field strength. Furthermore, according to the magnetic sensor 1 of the first example, the magnitude of the difference between V3r and V5r changes by 0.09% or less and E4 by 0.05° or less in the range of 50 to 120 mT of the applied field strength. The magnetic sensor 1 of the first example meets the requirements for the characteristics of the magnetic sensor 1 according to this embodiment. In contrast, the magnetic sensor of the comparison example does not meet the requirements for the characteristics of the magnetic sensor 1 according to this embodiment.

[0114] A first example and a second example of correction processing, as described in correction processing section 32 of the [document / section], will now be presented. Fig. The calculation unit 30 shown in section 4 is described.

[0115] According to the first example of correction processing, the following test of the magnetic sensor 1 is performed before delivery to generate the correction information stored in the correction information memory section 33. In the test, the applied field strength is set to a specific value within the sub-range. Then, the angle θ formed by the direction DM of the target magnetic field MF with respect to the reference direction DR is varied at regular intervals in a range from 0° to 360° to determine the value of the detected angle θs, calculated by the angle detection value calculation section 31, and the angular error, which is an error in the value of the detected angle θs, for each of the plurality of angles θ. The angular error is subsequently denoted by the symbol AE.

[0116] The angular error AE changes with a period of 1 / 4 the period of the angle θ. The angular error AE consists predominantly of the fourth-order angular error component. Here, it is assumed that the value of the detected angle θs is to be determined for each of a number m of angles θ. To generate the correction information, m must be an integer greater than or equal to 8, according to the sampling theorem. The plurality of angles θ used to determine the values ​​of the detected angle θs preferably includes the angle 0°, and in this case, m is an integer greater than or equal to 16.

[0117] An angle θ for which the value of the measured angle θs is to be determined is denoted by θn, where n is an integer between 1 and inclusive of m. The value of the measured angle θs corresponding to θn is denoted by θsn. The angular error AE corresponding to θn is denoted by AEn. A combination of θsn and AEn is denoted by the expression (θsn, AEn). During testing, (θsn, AEn) is stored as correction information in correction information memory section 33.

[0118] When using the magnetic sensor 1, the correction processing section 32 determines the corrected value of the detected angle θt for the input value of the detected angle θs using the correction information (θsn, AEn) by linear interpolation. More precisely, the correction processing section 32 determines an approximate angular error AEs, corresponding to the input value of the detected angle θs, using two sets of correction information (θsn, AEn), corresponding to θsn before and after the input value of the detected angle θs, by linear interpolation, and determines θs minus AEs as the corrected value of the detected angle θt. The approximate angular error AEs is equivalent to the angular error AEn corresponding to a specific θsn when the input value of the detected angle θs falls within the specified θsn.If the entered value of the detected angle θs is a different value than θsn, the approximate angular error AEs is an angular error estimated by linear interpolation.

[0119] The first example of the correction processing is described in more detail below with reference to the magnetic sensor 1 of the first example. Here, it is assumed that the sub-range is the range of 50 to 120 mT of the applied field strength. Fig. Figure 22 shows the relationship between the angle θ and the angular error AE of the value of the detected angle θs for the applied field strengths of 20 mT, 80 mT and 150 mT in the magnetic sensor 1 of the first example. Fig. Figure 22 shows the angle θ plotted on the horizontal axis and the angular error AE plotted on the vertical axis. Fig. In Figure 22, AE (20 mT), AE (80 mT), and AE (150 mT) denote the angular errors AE at applied field strengths of 20 mT, 80 mT, and 150 mT, respectively. It should be noted that the relationship between the angle θ and the angular error AE remains almost unchanged across the entire range of applied field strengths from 50 to 120 mT. Therefore, it can be said that the relationship between the angle θ and the angular error AE at an applied field strength of 80 mT, as shown in Figure 22, is approximately 1. Fig. Figure 22 shows that the relationship between the angle θ and the angular error AE is representative in the range of 50 to 120 mT of the applied field strength.

[0120] For example, in the above-described test of magnetic sensor 1 before delivery, the applied field strength can be set to 50 mT and the number m of the plurality of angles θn for which the values ​​of the detected angle θs are determined can be set to 32. θn is 0° when n = 1. A combination of θn and AEn is denoted by (θn, AEn). The test provides information on (θn, AEn). Fig. Figure 23 is a diagram of the relationship between the angle θ and an estimated angular error AEe, which is generated from the information about (θn, AEn) by linear interpolation. Fig. Figure 23 shows the angle θ plotted on the horizontal axis and the estimated angular error AEe on the vertical axis. The estimated angular error AEe is equivalent to AEn if the angle θ equals θn. If the angle θ has a different value than θn, the estimated angular error AEe is an angular error estimated by linear interpolation. The characteristic curve in Fig. Equation 23, which represents the relationship between θ and AEe, is a polygon with vertices at (θn, AEn).

[0121] In the first example of correction processing, the correction information (θsn, AEn) described above is generated based on the information about (θn, AEn) obtained through the test and stored in the correction information memory section 33. When using magnetic sensor 1, the correction processing section 32 determines the corrected value of the detected angle θt for the input value of the detected angle θs, as previously described. An error in the corrected value of the detected angle θt is referred to here as the Corrected Angle Error (CAE).

[0122] For example, it shows Fig. 24 the relationship between the angle θ and the error of the corrected angle CAE when the applied field strength is 80 mT. In Fig. Figure 24 shows the angle θ on the horizontal axis and the error of the corrected angle CAE on the vertical axis. As in Fig. As shown in Figure 24, the maximum magnitude of the corrected angle error CAE is 0.05° or less when the applied field strength is 80 mT. This value is significantly smaller than the maximum magnitude of the angular error AE when the applied field strength is 80 mT, as shown in Figure 24. Fig. 22 shown.

[0123] The amplitude of the corrected angle error (CAE) is now defined as half the difference between a maximum and a minimum value of the CAE error. The amplitude of the CAE error is less than or equal to the maximum value of the CAE error. As in the example in Fig. As shown in Figure 24, the amplitude of the error of the corrected angle CAE is therefore 0.05° or less when the applied field strength is 80 mT.

[0124] Correction processing section 32 preferably performs the correction processing such that the amplitude of the error of the corrected angle CAE is 0.1° or less, regardless of the value of the applied field strength within the sub-area. Further preferably, correction processing section 32 performs the correction processing such that the amplitude of the error of the corrected angle CAE is 0.05° or less, regardless of the value of the applied field strength within the sub-area.

[0125] Fig. Figure 25 shows the relationship between the applied field strength and the error of the corrected angle CAE in the subrange of 50 to 120 mT. Fig. Figure 25 shows the applied field strength on the horizontal axis and the amplitude of the error of the corrected angle CAE on the vertical axis. As in Fig. As shown in Figure 25, the amplitude of the error of the corrected angle CAE in the subrange of 50 to 120 mT is less than or equal to 0.05°.

[0126] As mentioned previously, the angular error AE consists predominantly of the fourth-order angular error component. The approximate angular error AEs is very close to the fourth-order angular error component. Since the corrected value of the detected angle θt, according to the first correction processing example, is generated by subtracting the approximate angular error AEs from the value of the detected angle θs, the resulting corrected value of the detected angle θt contains hardly any error component with a period of 1 / 4 of the period of the angle θ. Therefore, the first correction processing example allows for an amplitude of the error of the corrected angle CAE in the subrange of 0.1° or less if the change in E4 in the subrange is 0.1° or less, and an amplitude of the error of the corrected angle CAE in the subrange of 0.05° or less if the change in E4 in the subrange is 0.05° or less.

[0127] The second example of correction processing is now described by correction processing section 32. In this second example, the same test of magnetic sensor 1 is performed before delivery as in the first example, whereby the angle θ is varied at regular intervals in the range of 0° to 360° to determine the angular error AE for each of the plurality of angles θ. Then, the amplitude F of the angular error AE, defined as 1 / 2 the difference between the maximum and minimum values ​​of the angular error AE, is determined. During testing, the amplitude F is stored as correction information in correction information memory section 33.

[0128] When using the magnetic sensor 1, the correction processing section 32 determines the corrected value of the detected angle θt for the input value of the detected angle θs using equation (4) including the correction information F. θt=θs−Fsin4θs

[0129] “Fsin4θs” in equation (4) is equivalent to the approximate angular error AEs in the first correction processing example. Like the first correction processing example, the second correction processing example also allows the amplitude of the error of the corrected angle CAE to be 0.1° or less in the subrange if the change in E4 is 0.1° or less, and the amplitude of the error of the corrected angle CAE to be 0.05° or less in the subrange if the change in E4 is 0.05° or less.

[0130] The examples of correction processing are not limited to the first and second examples described above. Correction processing can be any processing operation that enables the generation of a corrected value of the measured angle θt, which contains hardly any error component with a period of 1 / 4 of the period of the angle θ. [Second example]

[0131] The magnetic sensor 1 of the second example is an example where the second surface of the free layer 51 is shaped such that dr is equal to 0.1% and α is equal to 0°, and the offset angle ϕ is set to 10°. Fig. Figure 26 is a characteristic curve diagram showing the dependencies of V3r and V5r on the magnetic field strength for the magnetic sensor 1 of the second example and the magnetic sensor of the comparison example. Fig. Figure 27 is a characteristic curve diagram showing the dependence of E4 on the magnetic field strength for both the magnetic sensor 1 of the second example and the magnetic sensor of the comparison example. Fig. 26. In the case of magnetic sensor 1 of the second example, V3r and V5r are expressed as V3r (dr = 0.1%, ϕ = 10°) and V5r (dr = 0.1%, ϕ = 10°), respectively; in the case of the magnetic sensor of the comparison example, V3r and V5r are expressed as V3r (dr = 0, ϕ = 0°) and V5r (dr = 0, ϕ = 0°), respectively. Fig. 27 In the case of the magnetic sensor 1 of the second example, E4 is expressed as E4 (dr = 0.1%, ϕ = 10°), while in the case of the magnetic sensor of the comparison example, E4 is expressed as E4 (dr = 0, ϕ = 0°).

[0132] Fig. 26 and Fig. Figure 27 shows that, according to the magnetic sensor 1 of the second example, the magnitude of the difference between V3r and V5r changes by 0.18% or less and E4 by 0.1° or less in the range of 20 to 70 mT of the applied field strength. Furthermore, according to the magnetic sensor 1 of the second example, the magnitude of the difference between V3r and V5r changes by 0.09% or less and E4 by 0.05° or less in the range of 25 to 55 mT of the applied field strength. The magnetic sensor 1 of the second example meets the requirements for the characteristics of the magnetic sensor 1 according to this embodiment. [Third example]

[0133] The magnetic sensor 1 of the third example is an example where the second surface of the free layer 51 is shaped such that dr equals 1% and α equals 0°, and the offset angle ϕ is set to 30°. Fig. Figure 28 is a characteristic curve diagram showing the dependencies of V3r and V5r on the magnetic field strength for the magnetic sensor 1 of the third example and the magnetic sensor of the comparison example. Fig. Figure 29 is a characteristic curve diagram showing the dependence of E4 on the magnetic field strength for both the magnetic sensor 1 of the third example and the magnetic sensor of the comparison example. Fig. 28. In the case of magnetic sensor 1 of the third example, V3r and V5r are expressed as V3r (dr = 1%, ϕ = 30°) and V5r (dr = 1%, ϕ = 30°), respectively; in the case of the magnetic sensor of the comparison example, V3r and V5r are expressed as V3r (dr = 0, ϕ = 0°) and V5r (dr = 0, ϕ = 0°), respectively. Fig. 29 In the case of the magnetic sensor 1 of the third example, E4 is expressed as E4 (dr = 1%, ϕ = 30°), while in the case of the magnetic sensor of the comparison example, E4 is expressed as E4 (dr = 0, ϕ = 0°).

[0134] Fig. 28 and Fig.Figure 29 shows that, according to the magnetic sensor 1 of the third example, the magnitude of the difference between V3r and V5r changes by 0.18% or less and E4 by 0.1° or less in the range of 50 to 150 mT of the applied field strength. Furthermore, according to the magnetic sensor 1 of the third example, the magnitude of the difference between V3r and V5r changes by 0.09% or less and E4 by 0.05° or less in the range of 60 to 150 mT of the applied field strength. The magnetic sensor 1 of the third example meets the requirements for the characteristics of the magnetic sensor 1 according to this embodiment.

[0135] The first to third examples make it clear that the sub-area which meets the requirements that the difference between the upper and lower limits is 30 mT or more and the change amount of E4 is 0.1° or less, can be shifted within the range of 20 to 150 mT by changing the value of the offset angle ϕ and the values ​​of dr and α, which define the shape of the second surface of the free layer 51.

[0136] As described above, the magnetic sensor 1 according to this embodiment fulfills the requirement that the change in angular error E4 is 0.1° or less when the applied field strength changes within a sub-range that is part of the range from 20 to 150 mT and has an upper and lower limit that differ from each other by 30 mT or more. The magnetic sensor 1 according to this embodiment thus achieves a reduced change in angular error over a large range of the applied field strength.

[0137] If the angular error AE in this embodiment has a sufficiently small maximum value in the partial range, such as 0.1° or less or 0.05° or less, the processing unit 30 can output the value of the detected angle θs as the value of the angle θ detected by the magnetic sensor 1. In this case, the processing unit 30 does not need to include either the correction processing section 32 or the correction information storage section 33.

[0138] According to this embodiment, the correction processing section 32 also performs the correction processing when the maximum amount of the angular error AE is so large that, for example, 0.1° is exceeded in the sub-area, whereby the correction processing does not depend on the applied field strength in order to make it possible for the error of the corrected angle CAE, which is an error in the corrected value of the detected angle θt, to have a small maximum amount in the sub-area.

[0139] The values ​​of dr, α, and ϕ required to meet the requirements for the properties of the magnetic sensor 1 according to this embodiment can differ depending on the structure of the MR element 50 and other factors. Therefore, the values ​​of dr, α, and ϕ can be freely selected according to the structure of the MR element 50 and the upper and lower limits of the applied field strength that define the sub-range.

[0140] The larger the sub-area, the better. The difference between the upper and lower bounds of the sub-area, which define its width, is preferably 40 mT or more, and more preferably 50 mT or more. A minimum requirement for the sub-area is that the difference between its upper and lower bounds be 30 mT or more. A preferred requirement for the sub-area is that the difference between its upper and lower bounds be 40 mT or more. A further preferred requirement for the sub-area is that the difference between its upper and lower bounds be 50 mT or more.

[0141] The smaller the change in E4 within the subrange, the better. The change in E4 within the subrange is preferably 0.05° or less. A minimum requirement for E4 is that the change in E4 within the subrange is 0.1° or less. A preferred requirement for E4 is that the change in E4 within the subrange is 0.05° or less.

[0142] As described above, the change in E4 is 0.1° or less if the change in the amount of the difference between V3r and V5r in the sub-range is 0.18% or less. Therefore, to meet the minimum requirement for E4, the change in the amount of the difference between V3r and V5r in the sub-range is preferably 0.18% or less. Furthermore, the change in E4 in the sub-range is 0.05° or less if the change in the amount of the difference between V3r and V5r in the sub-range is 0.09% or less. Therefore, to meet the preferred requirement for E4, the change in the amount of the difference between V3r and V5r in the sub-range is preferably 0.09% or less.

[0143] The magnetic sensor 1 of the first example and the magnetic sensor 1 of the third example satisfy all conditions resulting from any combination of one of the requirements for the sub-area (minimum requirement, preferred requirement, and further preferred requirement) and one of the requirements for E4 (minimum requirement and preferred requirement). The magnetic sensor 1 of the second example satisfies the conditions resulting from a combination of one of the requirements for the sub-area (minimum requirement, preferred requirement, and further preferred requirement) and the minimum requirement for E4, as well as the conditions resulting from a combination of the minimum requirement for the sub-area and the preferred requirement for E4.

[0144] The present invention is not limited to the preceding embodiment, and various modifications can be made to it. For example, if the second surface of the free layer 51 is given a symmetrical shape of order four, the outer edge 51a of the second surface need not necessarily have a shape such that r(θa) is expressed by equation (3). For example, the outer edge 51a of the second surface can have a shape such that r(θa) changes into the form of a triangular wave.

[0145] In light of the foregoing explanations, it is clear that numerous modifications and variations of the present invention are possible. Accordingly, it is understood that, within the scope of the appended claims and their equivalents, the invention can be implemented in forms other than the most preferred embodiment described above.

Claims

[1] Magnetic sensor (1) comprising a magnetic field detection unit (3) and a computation unit (30), wherein the magnetic sensor (1) is configured to generate a value of a detected angle (θs) which has a correspondence to an angle (θ) formed by the direction (DM) of a detected magnetic field (MF) at a reference position (PR) with respect to a reference direction (DR), characterized by , that: the magnetic field detection unit (3) comprises a plurality of magnetoresistive elements (50) for detecting the magnetic field (MF) and outputs a first signal (S1) and a second signal (S2), wherein the first signal (S1) has a correspondence to the angle formed by the direction (DM) of the magnetic field (MF) with respect to a first direction (D1), and the second signal (S2) has a correspondence to the angle formed by the direction (DM) of the magnetic field (MF) with respect to a second direction (D2); Each of the plurality of magnetoresistive elements (50) comprises: a fixed magnetization layer (53) whose magnetization direction is fixed; a free layer (51) whose magnetization direction changes depending on the direction (DM) of the magnetic field (MF); and a non-magnetic layer (52) arranged between the fixed magnetization layer (53) and the free layer (51); the calculation unit (30) calculates the value of the detected angle (θs) based on the first signal (S1) and the second signal (S2); the value of the detected angle (θs), when the direction (DM) of the magnetic field (MF) rotates with a predetermined period, contains an angular error component that changes with a period of 1 / 4 of the predetermined period; and the distortion ratio dr and the offset angle ϕ are chosen such that a maximum amount of the angular error component changes by 0.1° or less when the strength of the magnetic field (MF) at the magnetic field detection unit (3) changes within a sub-range which is part of a range of 20 to 150 mT and has an upper limit and a lower limit which differ from each other by 30 mT or more, wherein the magnetic field detection unit (3) comprises a first detection circuit (10) for outputting the first signal (S1) and a second detection circuit (20) for outputting the second signal (S2); both the first detection circuit (10) and the second detection circuit (20) comprise a chain of magnetoresistive elements consisting of two or more of the plurality of magnetoresistive elements (50) connected in series; the free layer (51) of each of the plurality of magnetoresistive elements (50) has a first surface in contact with the non-magnetic layer (52) and a second surface opposite the first surface, wherein the second surface has a rotationally symmetric shape of order four and not of order five or higher; the number of two or more magnetoresistive elements (50) that form the chain of magnetoresistive elements is even; the two or more magnetoresistive elements (50) that form the chain of magnetoresistive elements contain one or more pairs of magnetoresistive elements (50); the magnetization directions of the layers with fixed magnetization (53) of two magnetoresistive elements (50) paired together form a predetermined relative angle that is not 0° or 180°; in the first detection circuit (10) the first direction (D1) is either an intermediate direction between the magnetization directions of the layers with fixed magnetization (53) of two paired magnetoresistive elements (50) or a direction opposite to the intermediate direction; and in the second detection circuit (20) the second direction (D2) is either an intermediate direction between the magnetization directions of the layers with fixed magnetization (53) of two paired magnetoresistive elements (50) or a direction opposite to the intermediate direction, • dr is the ratio between the offset distance d and the radius of the reference circle Rc, • Rc is the mean of the minimum and maximum values ​​of R(θa), • R(θa) is the distance from the center of the second surface to a given point on the outer boundary (51a) of the second surface and • The offset angle ϕ is the angle between the magnetization direction (D111) of the layer with fixed magnetization and the first direction D1, Distortion ratio and offset angle (dr, ϕ) selected view from (1%, 25°), (0.1%, 10°), (1%, 30°). [2] Magnetic sensor according to claim 1, wherein If the direction (DM) of the magnetic field (MF) rotates with a predetermined period, the first signal (S1) contains a first ideal component (V11), a first third harmonic component (V13), and a first fifth harmonic component (V15), and the second signal (S2) contains a second ideal component (V21), a second third harmonic component (V23), and a second fifth harmonic component (V25), where the first ideal component (V11) is a component that changes periodically to describe an ideal sinusoidal curve, the first third harmonic component (V13) is an error component equivalent to a third harmonic of the first ideal component (V11), the first fifth harmonic component (V15) is an error component equivalent to a fifth harmonic of the first ideal component (V11), and the second ideal component (V21) is a component that changes periodically to describe an ideal sinusoidal curve describesthe second third harmonic component (V23) is an error component equivalent to a third harmonic of the second ideal component (V21), the second fifth harmonic component (V25) is an error component equivalent to a fifth harmonic of the second ideal component (V21); and , a change in the magnitude of a difference between a ratio of the third harmonic component (V3r) and a ratio of the fifth harmonic component (V5r) by 0.18% or less when the strength of the magnetic field (MF) at the magnetic field detection unit (3) changes within the sub-area, wherein the ratio of the third harmonic component (V3r) is the mean of a first ratio and a third ratio, the ratio of the fifth harmonic component (V5r) is the mean of a second ratio and a fourth ratio, the first ratio is the ratio of the first third harmonic component (V13) to the first ideal component (V11) when the first ideal component (V11) assumes a maximum value, the second ratio is the ratio of the first fifth harmonic component (V15) to the first ideal component (V11) when the first ideal component (V11) assumes a maximum value,The third ratio is the ratio of the second third harmonic part (V23) to the second ideal part (V21) when the second ideal part (V21) reaches a maximum value, and the fourth ratio is the ratio of the second fifth harmonic part (V25) to the second ideal part (V21) when the second ideal part (V21) reaches a maximum value. [3] Magnetic sensor according to claim 1 or 2, wherein the calculation unit (30) comprises an angle detection value calculation section (31) for calculating the value of the detected angle (θs) and a correction processing section (32) to perform correction processing for the value of the detected angle (θs) and thereby generate a corrected value of the detected angle (θt), wherein; a maximum amount of error in the corrected value of the detected angle (θt) is less than a maximum amount of error in the value of the detected angle (θs). [4] Magnetic sensor according to claim 3, wherein the correction processing section (32) performs the correction processing such that 1 / 2 of a difference between a maximum value and a minimum value of the error in the corrected value of the detected angle (θt) is 0.1° or less, regardless of the value of the strength of the magnetic field (MF) at the magnetic field detection unit (3) within the sub-area. [5] Magnetic sensor according to claim 1, wherein the maximum amount of the angular error fraction changes by 0.05° or less when the strength of the magnetic field (MF) at the magnetic field detection unit (3) changes within the sub-area. [6] Magnetic sensor according to claim 5, wherein If the direction (DM) of the magnetic field (MF) rotates with a predetermined period, the first signal (S1) contains a first ideal component (V11), a first third harmonic component (V13), and a first fifth harmonic component (V15), and the second signal (S2) contains a second ideal component (V21), a second third harmonic component (V23), and a second fifth harmonic component (V25), where the first ideal component (V11) is a component that changes periodically to describe an ideal sinusoidal curve, the first third harmonic component (V13) is an error component equivalent to a third harmonic of the first ideal component (V11), the first fifth harmonic component (V15) is an error component equivalent to a fifth harmonic of the first ideal component (V11), and the second ideal component (V21) is a component that changes periodically to describe an ideal sinusoidal curve describesthe second third harmonic component (V23) is an error component equivalent to a third harmonic of the second ideal component (V21), the second fifth harmonic component (V25) is an error component equivalent to a fifth harmonic of the second ideal component (V21); and, a change in the magnitude of a difference between a ratio of the third harmonic component (V3r) and a ratio of the fifth harmonic component (V5r) by 0.09% or less when the strength of the magnetic field (MF) at the magnetic field detection unit (3) changes within the sub-area, wherein the ratio of the third harmonic component (V3r) is the mean of a first ratio and a third ratio, the ratio of the fifth harmonic component (V5r) is the mean of a second ratio and a fourth ratio, the first ratio is the ratio of the first third harmonic component (V13) to the first ideal component (V11) when the first ideal component (V11) assumes a maximum value, the second ratio is the ratio of the first fifth harmonic component (V15) to the first ideal component (V11) when the first ideal component (V11) assumes a maximum value,the third ratio is the ratio of the second third harmonic part (V23) to the second ideal part (V21) when the second ideal part (V21) assumes a maximum value, and the fourth ratio is the ratio of the , The second fifth harmonic part (V25) to the second ideal part (V21) is when the second ideal part (V21) takes on the maximum value. [7] Magnetic sensor according to claim 5 or 6, wherein the calculation unit (30) comprises an angle acquisition value calculation section (31) for calculating the value of the acquired angle (θs) and a correction processing section (32) to perform correction processing for the value of the acquired angle (θs) and thereby generate a corrected value of the acquired angle (θt), wherein a maximum amount of error in the corrected value of the detected angle (θt) is less than a maximum amount of error in the value of the detected angle (θs). [8] Magnetic sensor according to claim 7, wherein the correction processing section (32) performs the correction processing such that 1 / 2 of a difference between a maximum value and a minimum value of the error in the corrected value of the detected angle (θt) is 0.05° or less, regardless of the value of the strength of the magnetic field (MF) at the magnetic field detection unit (3) within the sub-area. [9] Magnetic sensor according to any one of claims 1 to 8, wherein the second direction (D2) is orthogonal to the first direction (D1).

Images