Position detection device and position detection method

Abstract

A position detection device includes a signal processing section. A period corresponding to one cycle of a mechanical angle of a rotor is divided into a plurality of quadrants. The period corresponding to one cycle of the mechanical angle of the rotor is divided into a plurality of zones. The signal processing section performs: first acquisition processing that acquires a quadrant number indicating a quadrant; second acquisition processing that acquires a zone number; mechanical angle calculation processing that calculates information related to the mechanical angle of the rotor; and abnormality sensing processing that senses an abnormality in the first detection signal and an abnormality in the second detection signal. The abnormality sensing processing includes: first abnormality determination processing that determines that there is an abnormal part in the first detection signal when a time-series order of the quadrant numbers is different from a time-series order of the quadrant numbers at normal times; and second abnormality determination processing that determines that there is an abnormal part in the second detection signal when a time-series order of the zone numbers is different from a time-series order of the zone numbers at normal times.

Description

Technical Field

[0001] This invention relates to a position detection device and a position detection method.

[0002] This application is based on Japanese Patent Application No. 2024-051177, filed on March 27, 2024. This application claims priority to that application, the entire contents of which are incorporated herein by reference. Background Technology

[0003] Previously, rotation detection devices were known that could continue to perform rotation detection even if a part of the circuit malfunctioned, by means of a circuit that detects the rotation of a motor with two systems (e.g., Patent Document 1).

[0004] Existing technical documents

[0005] Patent documents

[0006] Patent Document 1: Japanese Patent Application Publication No. 2017-191093 Summary of the Invention

[0007] The problem that the invention aims to solve

[0008] In the aforementioned rotation detection device, multiple identical detection circuits need to be installed, which increases the cost of the rotation detection device.

[0009] In view of the above, one of the objectives of this invention is to provide a position detection device and a position detection method that can suppress cost increases and improve the redundancy of motor rotational position detection.

[0010] Solution for solving the problem

[0011] One embodiment of the position detection device of the present invention is a position detection device that detects the rotational position of a rotor in a motor. The position detection device comprises: a first magnet having one pole pair that rotates with the rotor; a second magnet having P pole pairs that rotates with the rotor, where P is an integer greater than or equal to 2; N1 first magnetic sensors opposite the first magnet and arranged at intervals in the rotational direction of the first magnet, where N1 is an integer greater than or equal to 2; N2 second magnetic sensors opposite the second magnet and arranged at intervals in the rotational direction of the second magnet, where N2 is an integer greater than or equal to 3; and a signal processing unit that processes N1 first detection signals output from the N1 first magnetic sensors and N2 second detection signals output from the N2 second magnetic sensors. A period corresponding to one cycle of the rotor's mechanical angle is divided into multiple quadrants corresponding to the combination of signal strengths in the N1 first detection signals. A period corresponding to one cycle of the rotor's mechanical angle is divided into multiple partitions by the intersection points and zero-crossing points of the waveforms in the N2 second detection signals, where the zero-crossing point is the point where the N2 second detection signals intersect with a reference value. The signal processing unit performs: a first acquisition process, acquiring a quadrant number representing the quadrant based on the N1 first detection signals; a second acquisition process, acquiring a partition number representing the partition based on the N2 second detection signals; a mechanical angle calculation process, calculating information related to the rotor's mechanical angle based on at least one of the quadrant number and the partition number; and an anomaly sensing process, sensing anomalies in the first detection signals and anomalies in the second detection signals. The anomaly sensing process includes: a first anomaly determination process, which determines that an abnormal part exists in the first detection signal if the temporal order of the quadrant numbers obtained through the first acquisition process is different from the temporal order of the quadrant numbers under normal conditions; and a second anomaly determination process, which determines that an abnormal part exists in the second detection signal if the temporal order of the partition numbers obtained through the second acquisition process is different from the temporal order of the partition numbers under normal conditions.

[0012] One aspect of the position detection method of the present invention is a position detection method for detecting the rotational position of a rotor in a motor. This position detection method includes signal processing of N1 first detection signals output from N1 first magnetic sensors opposite a first magnet and N2 second detection signals output from N2 second magnetic sensors opposite a second magnet, where N1 is an integer greater than or equal to 2 and N2 is an integer greater than or equal to 3. The first magnet has one pole pair and rotates with the rotor. The second magnet has P pole pairs and rotates with the rotor, where P is an integer greater than or equal to 2. The N1 first magnetic sensors are arranged at intervals in the rotational direction of the first magnet. The N2 second magnetic sensors are arranged at intervals in the rotational direction of the second magnet. A period corresponding to one cycle of the rotor's mechanical angle is divided into multiple quadrants corresponding to the combination of signal intensities in the N1 first detection signals. A period corresponding to one cycle of the rotor's mechanical angle is divided into multiple partitions by the intersection points and zero-crossing points of the waveforms in the N2 second detection signals, where the zero-crossing point is the point where the N2 second detection signals intersect with a reference value. The signal processing includes: a first acquisition process, acquiring quadrant numbers representing the quadrants based on the N1 first detection signals; a second acquisition process, acquiring partition numbers representing the partitions based on the N2 second detection signals; a mechanical angle calculation process, calculating information related to the rotor's mechanical angle based on at least one of the quadrant numbers and the partition numbers; and an anomaly sensing process, sensing anomalies in the first detection signals and anomalies in the second detection signals. The anomaly sensing process includes: a first anomaly determination process, which determines that an abnormal part exists in the first detection signal if the temporal order of the quadrant numbers obtained through the first acquisition process is different from the temporal order of the quadrant numbers under normal conditions; and a second anomaly determination process, which determines that an abnormal part exists in the second detection signal if the temporal order of the partition numbers obtained through the second acquisition process is different from the temporal order of the partition numbers under normal conditions.

[0013] Invention Effects

[0014] According to one aspect of the present invention, the cost increase of the position detection device can be suppressed, and the redundancy of the rotational position detection of the motor can be improved. Attached Figure Description

[0015] Figure 1 This is a schematic diagram illustrating the position detection device of the first embodiment.

[0016] Figure 2This is a flowchart illustrating an example of the learning process in the first embodiment.

[0017] Figure 3 This is a diagram showing the first detection signal and the second detection signal in the first embodiment.

[0018] Figure 4 This is a diagram showing a portion of the second detection signal in the first embodiment.

[0019] Figure 5 This is a flowchart illustrating an example of the position detection process in the first embodiment.

[0020] Figure 6 This is a diagram showing the second detection signal that has undergone processing to convert it into a triangular wave in the first embodiment.

[0021] Figure 7 This is a diagram showing the changes in quadrant numbering in the first embodiment.

[0022] Figure 8 This is a diagram showing the changes in the partition numbering in the first embodiment.

[0023] Figure 9 This is a graph showing the changes in the first and second count values ​​in the first embodiment.

[0024] Figure 10 This is a flowchart illustrating an example of the signal processing procedure of the position detection unit in the first embodiment.

[0025] Figure 11 This is a diagram illustrating an example of the UVW signal in the first embodiment.

[0026] Figure 12 This is a diagram illustrating an example of the ABZ signal in the first embodiment.

[0027] Figure 13 This is a diagram showing the changes in the output signal in the first embodiment.

[0028] Figure 14 This is a flowchart illustrating an example of the signal processing procedure of the position detection unit in the second embodiment. Detailed Implementation

[0029] <First Implementation>

[0030] Figure 1The position detection device 100 shown in this embodiment is a device for detecting the rotational position of the rotor 20 in the motor 10. The position detection method of this embodiment is performed by the position detection device 100. In this embodiment, the motor 10 is a three-phase motor. The rotor 20 is a rotating body that can rotate about the central axis J. In the following description, the axial direction extending from the central axis J is sometimes simply referred to as the "axial direction", and the circumferential direction about the central axis J is sometimes simply referred to as the "circumferential direction".

[0031] The position detection device 100 includes a first magnet 31, a second magnet 32, a first magnetic sensor 41, a second magnetic sensor 42, and a signal processing unit 50. Furthermore, although not shown in the figures, the position detection device 100 includes a circuit board. The first magnetic sensor 41, the second magnetic sensor 42, and the signal processing unit 50 are, for example, mounted on this circuit board.

[0032] The first magnet 31 has a pair of magnetic poles. A pair of magnetic poles refers to a pair of N poles and S poles. In the first magnet 31, an N pole and a S pole are arranged across the central axis J. The first magnet 31 rotates together with the rotor 20. The first magnet 31 is, for example, fixed to the axial end of the shaft of the rotor 20. The first magnet 31 is, for example, in the shape of a circular plate centered on the central axis J.

[0033] The second magnet 32 ​​has multiple pole pairs. In this embodiment, the second magnet 32 ​​has five pole pairs. It should be noted that when P is set to an integer greater than or equal to 2, the number of pole pairs of the second magnet 32 ​​is only P, and there is no particular limitation. In the second magnet 32, five N poles and five S poles are arranged alternately in a circumferential direction. In this embodiment, the second magnet 32 ​​is disposed on the rotor 20. The second magnet 32 ​​is, for example, an annular shape surrounding the central axis J. In this embodiment, the second magnet 32 ​​is a rotor magnet whose magnetic flux flows between the stator (not shown) and the rotor 20 in the motor 10. The second magnet 32 ​​rotates together with the rotor 20.

[0034] The first magnetic sensor 41 is axially opposed to the first magnet 31. In this embodiment, the first magnetic sensor 41 includes three sensors: first magnetic sensor 41a, first magnetic sensor 41b, and first magnetic sensor 41c. It should be noted that when N1 is set to an integer greater than or equal to 2, the number of first magnetic sensors 41 is only N1 and is not particularly limited. The three first magnetic sensors 41a, 41b, and 41c are arranged at intervals in the rotation direction of the first magnet 31, that is, in the circumferential direction around the central axis J. The three first magnetic sensors 41a, 41b, and 41c are, for example, Hall ICs with built-in Hall elements and latching circuits. The three first magnetic sensors 41a, 41b, and 41c each output a digital signal whose level is reversed every time the first magnet 31 rotates 180°.

[0035] In this embodiment, three first magnetic sensors 41a, 41b, and 41c are arranged at 120° intervals along the rotation direction of the first magnet 31. Therefore, the digital signals output from the three first magnetic sensors 41a, 41b, and 41c have a 120° phase difference with each other at a mechanical angle θ. Hereinafter, the digital signals output from the three first magnetic sensors 41a, 41b, and 41c will be referred to as first detection signals. First magnetic sensor 41a outputs the first detection signal HA1 to the signal processing unit 50. First magnetic sensor 41b outputs the first detection signal HA2 to the signal processing unit 50. First magnetic sensor 41c outputs the first detection signal HA3 to the signal processing unit 50.

[0036] The second magnetic sensor 42 is axially opposed to the second magnet 32. In this embodiment, the second magnetic sensor 42 includes three types: second magnetic sensor 42a, second magnetic sensor 42b, and second magnetic sensor 42c. It should be noted that when N2 is set to an integer greater than or equal to 3, the number of second magnetic sensors 42 is only N2 and is not particularly limited. The three second magnetic sensors 42a, 42b, and 42c are arranged at intervals in the rotation direction of the second magnet 32, i.e., circumferentially around the central axis J. The three second magnetic sensors 42a, 42b, and 42c are Hall elements or linear Hall ICs, respectively. The second magnetic sensors 42a, 42b, and 42c output analog signals whose electrical signals vary according to the magnetic field strength. One cycle of the electrical angle of each analog signal output from each second magnetic sensor 42a, 42b, and 42c corresponds to 1 / P of one cycle of the mechanical angle. In this embodiment, the number of pole pairs P of rotor 20 is "5", so the electrical angle of each analog signal is equivalent to 1 / 5 of the mechanical angle of one cycle, i.e., the mechanical angle is 72°.

[0037] In this embodiment, three second magnetic sensors 42a, 42b, and 42c are arranged at 24° intervals along the rotation direction of the rotor 20. Therefore, the analog signals output from the second magnetic sensors 42a, 42b, and 42c have a phase difference of 120° electrical angle between them. Hereinafter, the analog signals output from the second magnetic sensors 42a, 42b, and 42c will be referred to as second detection signals. Second magnetic sensor 42a outputs the second detection signal Hu to the signal processing unit 50. Second magnetic sensor 42b outputs the second detection signal Hv to the signal processing unit 50. Second magnetic sensor 42c outputs the second detection signal Hw to the signal processing unit 50.

[0038] The signal processing unit 50 performs signal processing on three first detection signals HA1, HA2, and HA3 output from three first magnetic sensors 41a, 41b, and 41c, and three second detection signals Hu, Hv, and Hw output from three second magnetic sensors 42a, 42b, and 42c. Based on the three first detection signals HA1, HA2, and HA3, and the three second detection signals Hu, Hv, and Hw, the signal processing unit 50 detects the rotational position of the motor 10, i.e., the rotational position of the rotor 20. The signal processing unit 50 includes an arithmetic unit 51 and a storage unit 52.

[0039] The arithmetic unit 51 is, for example, a microprocessor such as an MCU (Microcontroller Unit). The arithmetic unit 51 is input with three first detection signals HA1, HA2, and HA3, and three second detection signals Hu, Hv, and Hw. The arithmetic unit 51 is connected to the storage unit 52 via a data bus for data communication.

[0040] It should be noted that the three second detection signals Hu, Hv, and Hw, which are analog signals, are converted into digital signals by an A / D converter inside the arithmetic unit 51. However, for ease of explanation, the digital signals output from the A / D converter are also referred to as the second detection signals Hu, Hv, and Hw. Furthermore, in the following description, the three first detection signals HA1, HA2, and HA3 and the three second detection signals Hu, Hv, and Hw input to the arithmetic unit 51 are sometimes collectively referred to as "input sensor signals".

[0041] The storage unit 52 includes non-volatile memory for storing programs, various values, and learning data required for the arithmetic unit 51 to perform various processes, and volatile memory used as a temporary storage destination for data when the arithmetic unit 51 performs various processes. Non-volatile memory may be, for example, EEPROM (Electrically Erasable Programmable Read-Only Memory) or flash memory. Volatile memory may be, for example, RAM (Random Access Memory).

[0042] The arithmetic unit 51 can perform various processes according to the program stored in the storage unit 52. The arithmetic unit 51 performs position detection processing to calculate the rotational position of the rotor 20, i.e., the mechanical angle θ. The position detection processing is performed based on the learning data stored in the storage unit 52.

[0043] Learning data for position detection processing is acquired through learning processing. In this embodiment, the arithmetic unit 51 can perform this learning processing. It should be noted that the arithmetic unit 51 may also not perform this learning processing. In this case, learning data acquired by a learning device different from the arithmetic unit 51 may also be stored in the storage unit 52.

[0044] The following explanation uses the example of the learning process performed by the arithmetic unit 51 to illustrate the learning process. Figure 2 As shown, when the arithmetic unit 51 begins learning processing, it first rotates the rotor 20 (step S1). Next, the arithmetic unit 51 acquires three first detection signals HA1, HA2, and HA3 via three first magnetic sensors 41a, 41b, and 41c (step S2). Figure 3 As shown, the three first detection signals HA1, HA2, and HA3 are digital signals whose levels reverse every 180° of rotation of the first magnet 31 and which have a phase difference of 120° electrical angle between them. Figure 3 In this context, the period from time t1 to time t11 corresponds to one cycle of the machine angle. In the learning process of this embodiment, one cycle of the machine angle corresponds to the learning period. Figure 3 In this context, the periods from time t1 to time t2, from time t2 to time t4, from time t4 to time t6, from time t6 to time t8, from time t8 to time t10, and from time t10 to time t11 are respectively equivalent to 1 / 6 of one cycle of the mechanical angle, i.e., the mechanical angle 60°.

[0045] Next, the arithmetic unit 51 acquires three second detection signals Hu, Hv, and Hw through three second magnetic sensors 42a, 42b, and 42c (step S3). Figure 3 As shown, the electrical angle of each of the three second detection signals Hu, Hv, and Hw in one cycle is equivalent to 1 / 5 of the mechanical angle in one cycle, i.e., the mechanical angle is 72°. Figure 3 In this context, the periods from time t1 to time t3, from time t3 to time t5, from time t5 to time t7, from time t7 to time t9, and from time t9 to time t11 correspond to mechanical angles of 72°. Furthermore, the three second detection signals Hu, Hv, and Hw have an electrical phase difference of 120°.

[0046] Next, the arithmetic unit 51 divides one cycle of the mechanical angle into multiple quadrants based on the three first detection signals HA1, HA2, and HA3 obtained during one cycle of the mechanical angle corresponding to the learning period (step S4). Each quadrant has a distinct "N1" bit digital value. "N1" is the number of first magnetic sensors 41. In this embodiment, there are three first magnetic sensors 41; therefore, in step S4, the arithmetic unit 51 divides one cycle of the mechanical angle into multiple quadrants with distinct 3-bit digital values. In this embodiment, the uppermost bit of the 3-bit digital value is the value of the first detection signal HA1, the middle bit is the value of the first detection signal HA2, and the lowermost bit is the value of the first detection signal HA3.

[0047] like Figure 3 As shown, the arithmetic unit 51 divides one cycle of the mechanical angle into six quadrants based on three first detection signals HA1, HA2, and HA3. The arithmetic unit 51 divides the period from time t1 to time t2 within one cycle of the mechanical angle into a first quadrant with a 3-bit digital value "101 (HLH)". The arithmetic unit 51 divides the period from time t2 to time t4 within one cycle of the mechanical angle into a second quadrant with a 3-bit digital value "100 (HLL)". The arithmetic unit 51 divides the period from time t4 to time t6 within one cycle of the mechanical angle into a third quadrant with a 3-bit digital value "110 (HHL)". The arithmetic unit 51 divides the period from time t6 to time t8 within one cycle of the mechanical angle into a fourth quadrant with a 3-bit digital value "010 (LHL)". The arithmetic unit 51 divides the period from time t8 to time t10 within one cycle of the mechanical angle into a fifth quadrant with a 3-bit digital value "011 (LHH)". The arithmetic unit 51 divides the period from time t10 to time t11 in one cycle of the mechanical angle into the sixth quadrant with a 3-bit digital value "001 (LLH)". The digital value "0" is the value when the first detection signals HA1, HA2, and HA3 are at a low level L, and the digital value "1" is the value when the first detection signals HA1, HA2, and HA3 are at a high level H.

[0048] exist Figure 3In this diagram, "L" represents a value of "0" and "H" represents a value of "1". When the signal strength of the first detection signal HA1 is high (H), the most significant bit of the 3-bit digital value in the quadrant becomes "1". When the signal strength of the first detection signal HA1 is low (L), the most significant bit of the 3-bit digital value in the quadrant becomes "0". When the signal strength of the first detection signal HA2 is high (H), the middle bit of the 3-bit digital value in the quadrant becomes "1". When the signal strength of the first detection signal HA2 is low (L), the middle bit of the 3-bit digital value in the quadrant becomes "0". When the signal strength of the first detection signal HA3 is high (H), the least significant bit of the 3-bit digital value in the quadrant becomes "1". When the signal strength of the first detection signal HA3 is low (L), the least significant bit of the 3-bit digital value in the quadrant becomes "0".

[0049] As described above, the period corresponding to one cycle of the mechanical angle of rotor 20 is divided into multiple quadrants corresponding to the combination of signal strengths in the three first detection signals HA1, HA2, and HA3. The first quadrant is where the signal strength of the first detection signal HA1 is high (H), the signal strength of the first detection signal HA2 is low (L), and the signal strength of the first detection signal HA3 is high (H). The second quadrant is where the signal strength of the first detection signal HA1 is high (H), the signal strength of the first detection signal HA2 is low (L), and the signal strength of the first detection signal HA3 is low (L). The third quadrant is where the signal strength of the first detection signal HA1 is high (H), the signal strength of the first detection signal HA2 is high (H), and the signal strength of the first detection signal HA3 is low (L). The fourth quadrant is where the signal strength of the first detection signal HA1 is low (L), the signal strength of the first detection signal HA2 is high (H), and the signal strength of the first detection signal HA3 is low (L). The fifth quadrant is the quadrant where the signal strength of the first detection signal HA1 is low (L), the signal strength of the first detection signal HA2 is high (H), and the signal strength of the first detection signal HA3 is high (H). The sixth quadrant is the quadrant where the signal strength of the first detection signal HA1 is low (L), the signal strength of the first detection signal HA2 is low (L), and the signal strength of the first detection signal HA3 is high (H).

[0050] like Figure 2As shown, next, the arithmetic unit 51, based on the three second detection signals Hu, Hv, and Hw obtained during one cycle of the mechanical angle corresponding to the learning period, divides one cycle of the mechanical angle into five pole pair regions associated with pole pair numbers representing the pole pair positions of the five pole pairs of the second magnet 32 ​​(step S5). In this embodiment, in order to detect the rotational position of the rotor 20, pole pair numbers representing pole pair positions are assigned to the five pole pairs of the second magnet 32. For example, as Figure 1 As shown, the five pole pairs of the second magnet 32 ​​are assigned pole pair numbers in a clockwise order of “0”, “1”, “2”, “3”, and “4”.

[0051] like Figure 3 As shown, in step S5, the arithmetic unit 51 divides one cycle of the mechanical angle into five pole pair regions based on the three second detection signals Hu, Hv, and Hw obtained in one cycle of the mechanical angle. The arithmetic unit 51 divides the period from time t1 to time t3 of one cycle of the mechanical angle into pole pair regions associated with pole pair number "0". The arithmetic unit 51 divides the period from time t3 to time t5 of one cycle of the mechanical angle into pole pair regions associated with pole pair number "1". The arithmetic unit 51 divides the period from time t5 to time t7 of one cycle of the mechanical angle into pole pair regions associated with pole pair number "2". The arithmetic unit 51 divides the period from time t7 to time t9 of one cycle of the mechanical angle into pole pair regions associated with pole pair number "3". The arithmetic unit 51 divides the period from time t9 to time t11 of one cycle of the mechanical angle into pole pair regions associated with pole pair number "4".

[0052] like Figure 2 As shown, next, the arithmetic unit 51 divides each pole pair region into multiple partitions and associates the segment number representing the rotational position of the rotor 20 with each partition (step S6). Figure 4 As shown, in step S6, the arithmetic unit 51 divides each of the five pole pair regions into twelve partitions based on the three second detection signals Hu, Hv, and Hw obtained in one cycle of the mechanical angle, and establishes an association between the segment number representing the rotational position of the rotor 20 and each of the twelve partitions.

[0053] Each of the five pole pair regions comprises twelve sub-regions, each assigned a sub-region number from "0" to "11". Conversely, consecutive numbers throughout a single cycle of the mechanical angle are associated with each sub-region as segment numbers. Specifically, in the pole pair region associated with pole pair number "0", segment numbers "0" to "11" are associated with sub-region numbers "0" to "11". In the pole pair region associated with pole pair number "1", segment numbers "12" to "23" are associated with sub-region numbers "0" to "11". In the pole pair region associated with pole pair number "2", segment numbers "24" to "35" are associated with sub-region numbers "0" to "11". In the pole pair region associated with pole pair number "3", segment numbers "36" to "47" are associated with sub-region numbers "0" to "11". In the polar pair region associated with polar pair number "4", the segment numbers "48" to "59" are associated with the partition numbers "0" to "11".

[0054] The following is a detailed explanation of the process in step S6 where the arithmetic unit 51 divides each of the five pole pair regions into twelve partitions. Figure 4 In this context, the baseline value for amplitude is "0". Figure 4 In this context, a positive amplitude value is used as an example to represent the magnetic field strength of the N pole. Conversely, a negative amplitude value is used as an example to represent the magnetic field strength of the S pole. The arithmetic unit 51 performs the processing of extracting the zero-crossing points, which are the points where the three second detection signals Hu, Hv, and Hw included in each of the five pole pair regions intersect with the reference value "0". For example... Figure 4 As shown, the arithmetic unit 51 extracts zero-crossing points P1, P3, P5, P7, P9, P11, and P13 for each pole pair region.

[0055] The arithmetic unit 51 performs the process of extracting the intersection points, i.e., the points where the three second detection signals Hu, Hv, and Hw, contained in each of the five pole pair regions intersect. For each pole pair region, the arithmetic unit 51 extracts intersection points P2, P4, P6, P8, P10, and P12 as intersection points. The arithmetic unit 51 performs the process of determining the intervals between adjacent zero-crossing points and intersection points as partitions. The arithmetic unit 51 determines the interval between zero-crossing point P1 and intersection point P2 as partition number "0". The arithmetic unit 51 determines the interval between intersection point P2 and zero-crossing point P3 as partition number "1". The arithmetic unit 51 determines the interval between zero-crossing point P3 and intersection point P4 as partition number "2". The arithmetic unit 51 determines the interval between intersection point P4 and zero-crossing point P5 as partition number "3". The arithmetic unit 51 determines the interval between the zero-crossing point P5 and the intersection point P6 as the partition assigned partition number "4". The arithmetic unit 51 determines the interval between the intersection point P6 and the zero-crossing point P7 as the partition assigned partition number "5". The arithmetic unit 51 determines the interval between the zero-crossing point P7 and the intersection point P8 as the partition assigned partition number "6". The arithmetic unit 51 determines the interval between the intersection point P8 and the zero-crossing point P9 as the partition assigned partition number "7". The arithmetic unit 51 determines the interval between the zero-crossing point P9 and the intersection point P10 as the partition assigned partition number "8". The arithmetic unit 51 determines the interval between the intersection point P10 and the zero-crossing point P11 as the partition assigned partition number "9". The arithmetic unit 51 determines the interval between the zero-crossing point P11 and the intersection point P12 as the partition assigned partition number "10". The arithmetic unit 51 determines the interval between the intersection point P12 and the zero-crossing point P13 as the partition assigned the partition number "11".

[0056] As described above, the period corresponding to one cycle of the mechanical angle θ of the rotor 20 is divided into multiple partitions by the intersection points and zero-crossing points of the waveforms of the three second detection signals Hu, Hv, and Hw. The zero-crossing point is the point where the three second detection signals Hu, Hv, and Hw intersect with the reference value. It should be noted that, in the following description, for example, the partition assigned partition number "0" is sometimes referred to as "partition 0", and the partition assigned partition number "11" is referred to as "partition 11", etc.

[0057] like Figure 2 As shown, next, the arithmetic unit 51 acquires learning data and stores it in the storage unit 52 (step S7). The learning data acquired in step S7 includes data representing the correspondence between segment numbers associated with the partitions included in each of the six quadrants and pole pair numbers representing pole pair positions. For example... Figure 4As shown, the first quadrant includes ten partitions, from partition 0 to partition 9. The learning data includes data representing the correspondence between the segment number "0" to the segment number "9" and the pole pair number "0" that are associated with the partitions 0 to 9 included in the first quadrant.

[0058] The second quadrant includes two partitions, 10 and 11, and eight partitions from 0 to 7. The learning data includes data representing the correspondence between segment numbers "10" and "11" associated with partitions 10 and 11 in the second quadrant and their corresponding pole pair number "0". Additionally, the learning data includes data representing the correspondence between segment numbers "12" to "19" associated with partitions 0 to 7 in the second quadrant and their corresponding pole pair number "1".

[0059] The third quadrant comprises four partitions (8 through 11) and six partitions (0 through 5). The learning data includes data representing the correspondence between segment numbers "20" to "23" and pole pair number "1" that establish associations with partitions 8 through 11 in the third quadrant. Additionally, the learning data includes data representing the correspondence between segment numbers "24" to "29" and pole pair number "2" that establish associations with partitions 0 through 5 in the third quadrant.

[0060] The fourth quadrant comprises six partitions, numbered 6 through 11, and four partitions, numbered 0 through 3. The learning data includes data representing the correspondence between segment numbers "30" to "35" and pole pair number "2" that establish associations with partitions 6 through 11 in the fourth quadrant. Additionally, the learning data includes data representing the correspondence between segment numbers "36" to "39" and pole pair number "3" that establish associations with partitions 0 through 3 in the fourth quadrant.

[0061] The fifth quadrant comprises eight partitions, from partition 4 to partition 11, as well as two partitions, 0 and 1. The learning data includes data representing the correspondence between segment numbers "40" to "47" and pole pair number "3" that establish associations with partitions 4 to 11 in the fifth quadrant. Additionally, the learning data includes data representing the correspondence between segment numbers "48" and "49" and pole pair number "4" that establish associations with partitions 0 and 1 in the fifth quadrant.

[0062] The sixth quadrant comprises ten partitions, from partition 2 to partition 11. The learning data includes data representing the correspondence between the segment numbers "50" to "59" and the pole pair number "4" that establish associations with the partitions 2 to 11 included in the sixth quadrant.

[0063] The learning data includes the values ​​of the mechanical angle θ at each zero-crossing point P1, P3, P5, P7, P9, P11, and P13 for each pole pair number, and the values ​​of the mechanical angle θ at each intersection point P2, P4, P6, P8, P10, and P12. Furthermore, the learning data also includes the normalization coefficient k[i] described in Japanese Patent No. 6233532 corresponding to each segment number. i is an integer representing the segment number. The normalization coefficient k[i] is a coefficient determined based on the slope of the portion of the waveform of the second detection signals Hu, Hv, and Hw that connects the intersection points and zero-crossing points located at the start and end points of each segment. In the following description, the portion of the waveform of the second detection signals Hu, Hv, and Hw that connects the intersection points and zero-crossing points located at the start and end points of each segment is referred to as the segmented signal. In the method described in Japanese Patent No. 6233532, three second detection signals Hu, Hv, and Hw are converted into triangular waves for correction. Based on the relationship between the digital values ​​of the three second detection signals Hu, Hv, and Hw (when the segmented signals of each segment of one cycle of the mechanical angle are extracted from the corrected three second detection signals Hu, Hv, and Hw and connected to each other in a roughly straight line) and the mechanical angle θ, the mechanical angle θ is estimated and detected. During the conversion to the triangular wave, a correction is also performed to ensure that the amplitudes of the three second detection signals Hu, Hv, and Hw are all of the same value.

[0064] exist Figure 5 The diagram illustrates an example of three second detection signals, Hut, Hvt, and Hwt, corrected to triangular waves as described above. The second detection signal Hut represents the second detection signal Hu converted to a triangular wave. The second detection signal Hvt represents the second detection signal Hv converted to a triangular wave. The second detection signal Hwt represents the second detection signal Hw converted to a triangular wave. Figure 5 The image shows the three second detection signals Hut, Hvt, and Hwt in pole pair number "0". Figure 5 The “No.” in the text represents the partition number and the segment number.

[0065] In segment number "0", zero-crossing point P1 and intersection point P2 are connected by the segmentation signal U2b in the waveform of the second detection signal Hut. In segment number "1", intersection point P2 and zero-crossing point P3 are connected by the segmentation signal W1a in the waveform of the second detection signal Hwt. In segment number "2", zero-crossing point P3 and intersection point P4 are connected by the segmentation signal W2a in the waveform of the second detection signal Hwt. In segment number "3", intersection point P4 and zero-crossing point P5 are connected by the segmentation signal V1b in the waveform of the second detection signal Hvt. In segment number "4", zero-crossing point P5 and intersection point P6 are connected by the segmentation signal V2b in the waveform of the second detection signal Hvt. In segment number "5", intersection point P6 and zero-crossing point P7 are connected by the segmentation signal U1a in the waveform of the second detection signal Hut. In segment "6", zero-crossing point P7 and intersection point P8 are connected by the segmentation signal U2a in the waveform of the second detection signal Hut. In segment "7", intersection point P8 and zero-crossing point P9 are connected by the segmentation signal W1b in the waveform of the second detection signal Hwt. In segment "8", zero-crossing point P9 and intersection point P10 are connected by the segmentation signal W2b in the waveform of the second detection signal Hwt. In segment "9", intersection point P10 and zero-crossing point P11 are connected by the segmentation signal V1a in the waveform of the second detection signal Hvt. In segment "10", zero-crossing point P11 and intersection point P12 are connected by the segmentation signal V2a in the waveform of the second detection signal Hvt. In segment "11", intersection point P12 and zero-crossing point P13 are connected by the segmentation signal U1b in the waveform of the second detection signal Hut.

[0066] The segmented signal U1a is the first portion between the intersection point P6 of the other two second detection signals Hvt and Hwt, and the zero-crossing point P7. The segmented signal U2a is the second portion between the zero-crossing point P7 and the intersection point P8 of the other two second detection signals Hvt and Hwt. The segmented signal U1b is the first portion between the intersection point P12 of the other two second detection signals Hvt and Hwt, and the zero-crossing point P13. The segmented signal U2b is the second portion between the zero-crossing point P1 and the intersection point P2 of the other two second detection signals Hvt and Hwt.

[0067] The segmented signal V1a is the first portion between the intersection point P10 of the second detection signal Hwt with the other two second detection signals Hut and Hwt, and the zero-crossing point P11. The segmented signal V2a is the second portion between the zero-crossing point P11 and the intersection point P12 of the other two second detection signals Hut and Hwt. The segmented signal V1b is the first portion between the intersection point P4 of the second detection signal Hwt with the other two second detection signals Hut and Hwt, and the zero-crossing point P5. The segmented signal V2b is the second portion between the zero-crossing point P5 and the intersection point P6 of the other two second detection signals Hut and Hwt.

[0068] The segmented signal W1a is the first portion between the intersection point P2 of the other two second detection signals Hut and Hvt, and the zero-crossing point P3. The segmented signal W2a is the second portion between the zero-crossing point P3 and the intersection point P4 of the other two second detection signals Hut and Hvt. The segmented signal W1b is the first portion between the intersection point P8 of the other two second detection signals Hut and Hvt, and the zero-crossing point P9. The segmented signal W2b is the second portion between the zero-crossing point P9 and the intersection point P10 of the other two second detection signals Hut and Hvt.

[0069] As described above, the three second detection signals Hu output from the three second magnetic sensors 42a, 42b, and 42c each have a first part and a second part.

[0070] The normalization coefficient k[i] is determined for each segment based on the slope of each segmented signal. When the absolute value of the deviation of the digital value from the start point of each segment to the current time is set as ΔX, and the deviation of the mechanical angle θ from the start point of each segment to the current time is set as Δθ, the normalization coefficient k[i] satisfies the relationship Δθ=k[i]×ΔX.

[0071] Next, the position detection processing performed by the arithmetic unit 51 will be explained. For example... Figure 6 As shown, during the initial position detection process, the arithmetic unit 51 first acquires three first detection signals HA1, HA2, and HA3 using three first magnetic sensors 41a, 41b, and 41c (step S8). Next, the arithmetic unit 51 acquires three second detection signals Hu, Hv, and Hw using three second magnetic sensors 42a, 42b, and 42c (step S9).

[0072] Next, the arithmetic unit 51 performs a first acquisition process (step S10) to obtain the quadrant number representing the quadrant based on the three first detection signals HA1, HA2, and HA3. That is, the signal processing in the position detection method of this embodiment includes the first acquisition process. In the first acquisition process (step S10), the arithmetic unit 51 determines the current quadrant from the six quadrants based on the 3-bit digital value represented by the three first detection signals HA1, HA2, and HA3, and obtains the quadrant number. As an example, when the 3-bit digital value is "100", the first detection signal HA1 is high level H, the first detection signal HA2 is low level L, and the first detection signal HA3 is low level L. Therefore, in the first acquisition process, the arithmetic unit 51 determines that the current mechanical angle θ of the rotor 20 is included in the second quadrant and obtains "2" as the quadrant number.

[0073] Next, the arithmetic unit 51 executes a second acquisition process (step S11) to obtain the partition number representing the partition based on the three second detection signals Hu, Hv, and Hw. That is, the signal processing in the position detection method of this embodiment includes the second acquisition process. In the second acquisition process (step S11), the arithmetic unit 51, for example, determines the current partition from the twelve partitions based on the magnitude relationship of the detection values ​​of the three second detection signals Hu, Hv, and Hw and the sign of each detection value, and obtains the partition number. Figure 4 As shown, for example, in partition 2, the detection value of the second detection signal Hu is the largest and has a positive sign. Furthermore, in partition 2, the detection value of the second detection signal Hw is the second largest and has a negative sign. Furthermore, in partition 2, the detection value of the second detection signal Hv is the smallest and has a negative sign. In the second acquisition process, if the magnitude relationship of the detection values ​​of the three second detection signals Hu, Hv, and Hw, and the sign of each detection value, meet the conditions for partition 2 as described above, the arithmetic unit 51 determines partition 2 as the current partition and acquires "2" as the partition number.

[0074] like Figure 6As shown, next, the arithmetic unit 51 performs a mechanical angle calculation process (step S12) to calculate information related to the mechanical angle θ of the rotor 20 based on the acquired quadrant number and the acquired partition number. That is, the signal processing in the position detection method of this embodiment includes mechanical angle calculation processing. The information related to the mechanical angle θ of the rotor 20 obtained in the mechanical angle calculation process (step S12) of this embodiment is an estimated value of the mechanical angle θ of the rotor 20. For example, as described above, let "2" be acquired as the current quadrant number, and "2" be acquired as the current partition number. The learning data includes data representing the correspondence between the segment numbers "12" to "19" associated with the partitions 0 to 7 included in the second quadrant and the pole pair number "1". Therefore, in the mechanical angle calculation process (step S12), when the quadrant number is "2" and the partition number is "2", the calculation unit 51 determines that the rotation position of the rotor 20 is in the position included by the pole pair number "1" and in the position included by the segment number "14" corresponding to the partition number 2 in the pole pair number "1".

[0075] Here, as described above, the storage unit 52 pre-stores the mechanical angle θ corresponding to the start and end points of each segment number, and the normalization coefficient k[i] corresponding to each segment number. Therefore, by adding the value of the deviation Δθ of the mechanical angle θ calculated according to Δθ=k[i]×ΔX to the value of the mechanical angle θ stored for the start point of the segment number corresponding to the rotational position of the rotor 20, the mechanical angle θ at the current moment can be calculated. A more specific method for calculating the mechanical angle θ using three second detection signals Hu, Hv, and Hw can be appropriately adopted from the calculation method described in Japanese Patent No. 6233532. In this embodiment, by using three first detection signals HA1, HA2, and HA3 and three second detection signals Hu, Hv, and Hw, it is possible to determine which segment of a mechanical angle cycle the current rotational position of the rotor 20 is in, as described above. Therefore, even immediately after the position detection device 100 is powered on, the rotational position of the rotor 20 can be detected.

[0076] The signal processing unit 50 performs anomaly sensing processing to detect anomalies in the three first detection signals HA1, HA2, and HA3, and anomalies in the three second detection signals Hu, Hv, and Hw. That is, the signal processing in the position detection method of this embodiment includes anomaly sensing processing. In this embodiment, the anomaly sensing processing is performed in the arithmetic unit 51. In this embodiment, the anomaly sensing processing includes a first anomaly determination process, a second anomaly determination process, and a third anomaly determination process.

[0077] The first anomaly detection process involves determining an abnormal location in the first detection signals HA1, HA2, and HA3 if the temporal order of the quadrant numbers obtained in the first acquisition process differs from the normal temporal order. Figure 7 The diagram shows the change in quadrant numbering based on time t as rotor 20 rotates. For example... Figure 7 As shown, when the rotor 20 is rotating, under normal conditions, the quadrant number obtained through the first acquisition process (step S10) described above repeats the following changes: from "1" to "6" in the order of increasing by 1 each time, when it becomes "6", it then returns to "1" and changes again in the order of increasing by 1 each time.

[0078] For example, in Figure 7 In the case of anomaly 1, the quadrant number becomes "3". Normally it should become "4", but it remains "3" and then changes to "5". In this situation, during the first anomaly determination process, the signal processing unit 50 determines that the timing order of the quadrant numbers is different from the normal timing order, and determines that there is an abnormal part in the first detection signals HA1, HA2, and HA3. Furthermore, for example, in... Figure 7 In the case of anomaly 2, after quadrant numbering becomes "3", it should normally become "4", but it does not become "4" and instead becomes a state where quadrant numbering cannot be obtained, then it becomes "5". In this case, during the first anomaly determination process, the signal processing unit 50 also determines that the timing order of the quadrant numbers is different from the timing order of the quadrant numbers in normal conditions, and determines that there is an abnormal part in the first detection signals HA1, HA2, and HA3. Figure 7 In the examples of anomalies 1 and 2 shown, since the quadrant number did not become "4" and the quadrant number changed from "3" to "5", the signal processing unit 50 determined that there was an anomaly in the part where the quadrant number became "4" in one cycle of the machine angle.

[0079] It should be noted that, as Figure 7 The anomaly shown in case 2, where the quadrant number cannot be obtained, is, for example, the absence of a quadrant corresponding to the combination of digital values ​​of the three first detection signals HA1, HA2, and HA3. Specifically, for example, it is the case where the 3-bit digital value represented by the three first detection signals HA1, HA2, and HA3 is "111 (HHH)". It should be noted that in this case, even if the rotor 20 does not rotate, the signal processing unit 50 can sense that an anomaly has occurred in the three first detection signals HA1, HA2, and HA3.

[0080] The second anomaly determination process involves determining that an anomaly exists in the second detection signals Hu, Hv, and Hw if the temporal order of the partition numbers obtained in the second acquisition process (step S11) differs from the normal temporal order of the partition numbers. Figure 8 The diagram shows the change in partition numbers based on time t as rotor 20 rotates. For example... Figure 8 As shown, when the rotor 20 is rotating, under normal conditions, the partition number obtained by the second acquisition process repeats the following changes: from "0" to "11" in the order of increasing by 1 each time. When it becomes "11", it returns to "0" and changes again in the order of increasing by 1 each time.

[0081] For example, in Figure 8 In the abnormal situation shown, after the partition number becomes "6", it should normally become "7", but it does not become "7" and remains "6", and then becomes "9" in the next change. In this case, in the second abnormality determination process, the signal processing unit 50 determines that the timing order of the partition numbers is different from the timing order of the partition numbers in the normal situation, and determines that there is an abnormal part in the second detection signals Hu, Hv, Hw. Figure 8 In the example of the abnormal situation shown, since the quadrant number did not become "7" or "8", the quadrant number changed from "6" to "9". Therefore, the signal processing unit 50 determined that there was an abnormality in the part where the partition number became "7" or "8" in one cycle of the machine angle.

[0082] In this embodiment, by comparing the changes in quadrant numbers with those under normal conditions during the first anomaly determination process, it is possible to detect which quadrant number is abnormal in the first detection signals HA1, HA2, and HA3. Furthermore, by comparing the changes in partition numbers with those under normal conditions during the second anomaly determination process, it is possible to detect which partition number is abnormal in the second detection signals Hu, Hv, and Hw. Therefore, when anomalies occur in the first detection signals HA1, HA2, HA3 and the second detection signals Hu, Hv, and Hw, the abnormal location can be determined, and the calculation of the mechanical angle θ at the abnormal location can be supplemented based on the non-abnormal locations in the first detection signals HA1, HA2, HA3 and the second detection signals Hu, Hv, and Hw. Therefore, without the need for multiple systems using the same circuit for calculating the mechanical angle θ, the position detection device 100, which has a single system with a first magnetic sensor 41 and a second magnetic sensor 42, can easily continue the detection of the rotor 20's rotational position even when anomalies occur in the detection signals. Therefore, the cost increase of the position detection device 100 can be suppressed, and the redundancy of the rotational position detection of the motor 10 can be improved. In addition, since anomalies can be detected simply by comparing the temporal order of the detected quadrant numbers and partition numbers with the normal state, anomaly detection is easy.

[0083] The third anomaly detection process involves determining that at least one of the first detection signals HA1, HA2, HA3 and the second detection signals Hu, Hv, Hw is abnormal if the difference between the count value obtained through the first counting process and the count value obtained through the second counting process is greater than or equal to a predetermined number. The first counting process counts each time the rotor 20 is determined to have rotated by a predetermined angle based on the three first detection signals HA1, HA2, HA3. The second counting process counts each time the rotor 20 is determined to have rotated by a predetermined angle based on the three second detection signals Hu, Hv, Hw. In this embodiment, the predetermined angle is 360°. It should be noted that the predetermined angle is not particularly limited and can be any angle.

[0084] The signal processing unit 50 performs the first and second counting processes described above. In the first counting process, the signal processing unit 50 counts each time it determines that the rotor 20 has rotated one revolution based on the three first detection signals HA1, HA2, and HA3, incrementing the first count value by 1 each time. In the second counting process, the signal processing unit 50 counts each time it determines that the rotor 20 has rotated one revolution based on the three second detection signals Hu, Hv, and Hw, incrementing the second count value by 1 each time. Therefore, as... Figure 9As shown, under normal conditions, the first and second count values ​​increase by 1 at approximately the same timing as time t.

[0085] In the third anomaly determination process of this embodiment, if the difference between the first count value and the second count value is greater than or equal to 2, the signal processing unit 50 determines that there is an abnormal part in at least one of the first detection signals HA1, HA2, HA3 and the second detection signals Hu, Hv, Hw. For example, in Figure 9 In the case of the anomaly shown, even if the first count value increases to 4, 5, or 6, the second count value remains unchanged at 3. In the third anomaly determination process, the signal processing unit 50 determines that an anomaly has occurred at the point when the first count value changes to 5 while the second count value remains at 3. In this case, the signal processing unit 50 determines that there is an abnormal portion in the second detection signals Hu, Hv, and Hw used to count the second count value which is smaller. Through the third anomaly determination process, the signal processing unit 50 can sense that at least one of the first detection signals HA1, HA2, HA3 and the second detection signals Hu, Hv, and Hw has an anomaly without comparing the changes in quadrant numbering and partition numbering with the normal situation.

[0086] For example, if the first magnetic sensors 41a, 41b, and 41c completely stop and the first detection signals HA1, HA2, and HA3 are interrupted, quadrant numbers will no longer be acquired. In this case, an anomaly cannot be determined using the first anomaly determination process described above. Furthermore, for example, if the second magnetic sensors 42a, 42b, and 42c completely stop and the second detection signals Hu, Hv, and Hw are interrupted, partition numbers will no longer be acquired. In this case, an anomaly cannot be determined using the second anomaly determination process described above. Moreover, if either the quadrant number or the partition number stops at a constant value, it is difficult to distinguish between a stop in the rotor 20 and an anomaly. Even in these cases, by comparing the first and second count values ​​in the third anomaly determination process, the signal processing unit 50 can sense that at least one of the first detection signals HA1, HA2, and HA3 and the second detection signals Hu, Hv, and Hw has an anomaly.

[0087] As described above, in this embodiment, the difference between the first count value and the second count value when the signal processing unit 50 determines that an anomaly has occurred in the third anomaly determination process, i.e., a predetermined number, is 2. The timing of the upward counting of the first count value and the second count value may sometimes be misaligned due to errors in the detection signals. Therefore, if this predetermined number is 1, an anomaly may be determined to have occurred in the third anomaly determination process even if no anomaly has occurred. In contrast, by setting this predetermined number to be greater than or equal to 2, false sensing of anomalies by the signal processing unit 50 can be suppressed.

[0088] In this embodiment, if the signal processing unit 50 determines during anomaly sensing that an abnormal portion exists in one of the first detection signals HA1, HA2, HA3 and the second detection signals Hu, Hv, Hw, it performs fail-safe processing. Fail-safe processing includes estimation processing, which estimates information obtained under normal conditions based on the abnormal portion in one of the first detection signals HA1, HA2, HA3 and the second detection signals Hu, Hv, Hw, based on at least a portion of the portions not determined to be abnormal in the first detection signals HA1, HA2, HA3 and the second detection signals Hu, Hv, Hw. Furthermore, fail-safe processing includes processing that calculates information related to the mechanical angle θ of the rotor 20 based on the information estimated through the estimation processing for the abnormal portion determined during anomaly sensing. Therefore, even if an abnormal portion exists in one of the first detection signals HA1, HA2, HA3 and the second detection signals Hu, Hv, Hw, the mechanical angle θ of the rotor 20 can be calculated at that abnormal portion. Therefore, even if an anomaly occurs in one of the first detection signals HA1, HA2, HA3 and the second detection signals Hu, Hv, Hw, the rotational position detection of the motor 10 can continue. It should be noted that the signal processing unit 50 can also use the information estimated through the above-described estimation process to calculate the mechanical angle θ of the determined anomaly location through the same processing as the above-described mechanical angle calculation process. Furthermore, the signal processing unit 50 can also use the information estimated through the above-described estimation process to calculate the mechanical angle θ of the determined anomaly location through a different processing than the above-described mechanical angle calculation process.

[0089] In this embodiment, the fault-safe processing includes a first fault-safe processing and a second fault-safe processing. The first fault-safe processing is performed when the first anomaly determination processing determines that there are abnormal parts in the first detection signals HA1, HA2, and HA3. The second fault-safe processing is performed when the second anomaly determination processing determines that there are abnormal parts in the second detection signals Hu, Hv, and Hw.

[0090] The first failsafe processing includes the following process: based on the sequential relationship of the first abnormal part in the quadrant numbers obtained in the first acquisition process (step S10), the quadrant number of the first abnormal part that is determined to be abnormal is estimated. Therefore, even if a first abnormal part is generated in the first detection signals HA1, HA2, HA3, the quadrant number can be obtained at the first abnormal part in the same way as in the normal case. As a result, the signal processing unit 50 can use the quadrant number estimated from the first abnormal part to calculate the machine angle θ using the same processing as the machine angle calculation process described above.

[0091] Specifically, for example, in Figure 7 In cases 1 and 2 of the anomaly shown, the first abnormal location is the area between quadrant number "3" and quadrant number "5". Based on the relationship between this first abnormal location, i.e., quadrant number "3" and quadrant number "5", the signal processing unit 50 estimates that the quadrant number of the first abnormal location is "4". Therefore, even if an abnormal location is determined to exist in the first detection signals HA1, HA2, and HA3 during the first anomaly determination process, the normal quadrant number "4" can be estimated and obtained at the first abnormal location where the anomaly is determined. Thus, even if an anomaly occurs in the first detection signals HA1, HA2, and HA3, the signal processing unit 50 can use the estimated quadrant number at the first abnormal location where the anomaly occurred to perform mechanical angle calculation processing, and can calculate the mechanical angle θ of the rotor 20.

[0092] When the signal processing unit 50 determines, through the first anomaly determination process, that there is a first anomalous location in the arrangement of quadrant numbers, and the rotational position of the rotor 20 becomes the first anomalous location in the next subsequent cycle, it uses a first fail-safe process to estimate the quadrant number of the first anomalous location and performs mechanical angle calculation based on this estimated quadrant number. Specifically, when the signal processing unit 50 determines that the quadrant preceding the first anomalous location based on the three second detection signals Hu, Hv, and Hw has ended, it does not use the three first detection signals HA1, HA2, and HA3, but instead sets the quadrant number of the next quadrant to the quadrant number estimated through the first fail-safe process and uses this quadrant number for mechanical angle calculation. Furthermore, when the signal processing unit 50 determines that the next quadrant following the first anomalous location based on the three second detection signals Hu, Hv, and Hw has begun, it reverts to the method of obtaining the quadrant number of the next quadrant using the first acquisition process obtained based on the three first detection signals HA1, HA2, and HA3.

[0093] For example, in such Figure 7 When the quadrant numbered "4" is the first abnormal location, as shown in cases 1 and 2, the signal processing unit 50 determines, based on the three second detection signals Hu, Hv, and Hw, that the quadrant numbered "3" has ended, and sets the quadrant number to "4" as the next quadrant number, which was deduced through the first failsafe processing. In this case, if the signal processing unit 50 determines that the quadrant numbered "5" has begun, it reverts to the method of obtaining the next quadrant number based on the three first detection signals HA1, HA2, and HA3.

[0094] The determination of the end of the quadrant preceding the quadrant that becomes the first abnormal part, and the determination of the start of the next quadrant of the quadrant that becomes the first abnormal part, are based on whether the mechanical angle θ calculated in the mechanical angle calculation process is the same as the mechanical angle θ at the point that is the boundary point of the quadrant preceding or following the quadrant that becomes the first abnormal part, among the intersection points and zero-crossing points of the second detection signals Hu, Hv, and Hw stored in the storage unit 52. For example, the boundary point between the third and fourth quadrants is the zero-crossing point P7 in pole pair number "2". When the mechanical angle θ calculated by the mechanical angle calculation process is the value of the mechanical angle θ stored for the zero-crossing point P7 of pole pair number "2", the signal processing unit 50 determines that the third quadrant has ended. For example, the boundary point between the fourth and fifth quadrants is the zero-crossing point P5 in pole pair number "3". When the mechanical angle θ calculated by the mechanical angle calculation process becomes the value of the mechanical angle θ stored at the zero-crossing point P5 for pole pair number "3", the signal processing unit 50 determines that the fourth quadrant has ended and the fifth quadrant has begun.

[0095] The second fault-safe processing includes the following steps: determining which quadrant number the second abnormal part, identified as abnormal in the partition number obtained in the second acquisition process (step S11), belongs to; and calculating information related to the mechanical angle θ of the rotor corresponding to the period of the quadrant number containing the second abnormal part, based on information predetermined for the period of the quadrant number containing the second abnormal part. Therefore, even if a second abnormal part occurs in the second detection signals Hu, Hv, and Hw, the signal processing unit 50 can calculate the mechanical angle θ of the rotor with high accuracy at that second abnormal part.

[0096] The process of “determining which quadrant number the second abnormal part identified in the partition number belongs to, and calculating the information related to the mechanical angle θ of the rotor corresponding to the period of the second abnormal part based on the information of the mechanical angle θ of the rotor that is predetermined for the period of the quadrant number that includes the second abnormal part” includes the estimation process in fail-safe processing and the process of calculating the information related to the mechanical angle θ of the rotor 20 based on the information estimated by the estimation process.

[0097] Specifically, for example, in Figure 8In the case of the abnormality shown, the second abnormal location is the location with partition numbers "7" and "8". During the second fail-safe processing, the signal processing unit 50 determines which quadrant number the second abnormal location belongs to, based on the quadrant number obtained from the first detection signals HA1, HA2, and HA3, within the partition numbers "7" and "8" that constitute the second abnormal location. For example, if the partition numbers "7" and "8" that constitute the second abnormal location belong to quadrant number "4", during the second fail-safe processing, the signal processing unit 50 calculates information related to the mechanical angle θ of the rotor 20 corresponding to the partition numbers "7" and "8" that constitute the second abnormal location, based on information about the mechanical angle θ of the rotor 20 predetermined for the period corresponding to quadrant number "4", i.e., the fourth quadrant. In this embodiment, the fourth quadrant is the period from the zero-crossing point P7 of pole pair number "2" to the zero-crossing point P5 of pole pair number "3". Therefore, the information regarding the mechanical angle θ of the rotor 20 predetermined for the fourth quadrant includes angle information from the mechanical angle θ stored corresponding to the zero-crossing point P7 of pole pair number "2" to the mechanical angle θ stored corresponding to the zero-crossing point P5 of pole pair number "3". In the second failsafe processing, the signal processing unit 50 calculates the mechanical angle θ at the partition numbers "7" and "8" that constitute the second abnormal location based on this angle information. It should be noted that "calculating information related to the mechanical angle θ of the rotor 20 corresponding to the period of the second abnormal location based on the information regarding the mechanical angle θ of the rotor predetermined for the period including the quadrant number of the second abnormal location" means that it is sufficient to calculate the information related to the mechanical angle θ of the rotor 20 corresponding to the period of the second abnormal location based on at least a portion of the information regarding the mechanical angle θ of the rotor predetermined for the period including the quadrant number of the second abnormal location.

[0098] In this embodiment, when the signal processing unit 50 determines, through the second anomaly determination process, that there is a second anomaly in the arrangement of partition numbers, and the rotational position of the rotor 20 becomes the second anomaly at a later time, it calculates the mechanical angle θ at the second anomaly location through the second fail-safe processing. Similar to the case where the first anomaly occurs, the signal processing unit 50 performs the calculation of the mechanical angle θ in the second fail-safe processing starting from the partition number determined to be the second anomaly. In the case where the second anomaly is partition number "7" or "8", the signal processing unit 50 begins calculating the mechanical angle θ based on the second fail-safe processing from the time point when the previous partition number "6" was determined to be the second anomaly, i.e., the time point when the partition number becomes the second anomaly.

[0099] In the second fail-safe processing, the signal processing unit 50 calculates the mechanical angle θ of the rotor 20 based on the angle information predetermined for the period of the quadrant number including the second abnormal location, the time elapsed since the partition number became the second abnormal location, and the rotational speed of the rotor 20. That is, in this embodiment, the second fail-safe processing includes the following process: calculating information related to the mechanical angle θ of the rotor corresponding to the period of the second abnormal location based on the information of the mechanical angle θ of the rotor predetermined for the period of the quadrant number including the second abnormal location, the time elapsed since the partition number became the second abnormal location, and the rotational speed of the rotor 20. Therefore, through the second fail-safe processing, the signal processing unit 50 can calculate the mechanical angle θ at the second abnormal location with high accuracy and ease. It should be noted that the method for detecting the rotational speed of the rotor 20 and the method for measuring the time elapsed since the partition number became the second abnormal location are not particularly limited.

[0100] "The process of calculating information related to the mechanical angle θ of the rotor corresponding to the period of the second abnormal part based on the information of the mechanical angle θ of the rotor predetermined for the period including the quadrant number of the second abnormal part, the time elapsed from the time point when the partition number becomes the second abnormal part, and the rotational speed of the rotor 20" includes the estimation process in fail-safe processing and the process of calculating information related to the mechanical angle θ of the rotor 20 based on the information estimated by the estimation process.

[0101] For example, if the second abnormal location is partition number "7" or "8", in the second fail-safe process, the signal processing unit 50 reads the mechanical angle θ stored corresponding to the intersection point P8 between partition number "6" and partition number "7" at the time point when partition number "7" is determined. The signal processing unit 50 calculates the mechanical angle θ of the rotor 20 at the second abnormal location by adding the value obtained by multiplying the time elapsed since the time point when partition number "7" is determined to be "7" by the rotational speed of the rotor 20 to the read mechanical angle θ. Similar to the case where the first abnormal location is determined to be the second abnormal location, the signal processing unit 50 ends the calculation of the mechanical angle θ in the second fail-safe process when the partition number that determines the second abnormal location has ended, and starts the calculation of the mechanical angle θ based on the normal mechanical angle calculation process again.

[0102] It should be noted that in the second fail-safe processing, the signal processing unit 50 can also calculate the mechanical angle θ of the rotor 20 based on the aforementioned angle information predetermined for the period of the quadrant number including the second abnormal location, the time elapsed since the quadrant number became the quadrant number including the second abnormal location, and the rotational speed of the rotor 20. That is, the second fail-safe processing can also include the following process: based on the information of the mechanical angle θ of the rotor 20 predetermined for the period of the quadrant number including the second abnormal location, the time elapsed since the quadrant number became the quadrant number including the second abnormal location, and the rotational speed of the rotor 20, to calculate information related to the mechanical angle θ of the rotor 20 corresponding to the period of the second abnormal location. Even in this case, the signal processing unit 50 can calculate the mechanical angle θ at the second abnormal location with high accuracy and ease through the second fail-safe processing. In this case, the signal processing unit 50 reads the mechanical angle θ stored at the intersection point or zero-crossing point starting with the quadrant number including the second abnormal location. The signal processing unit 50 calculates the mechanical angle θ of the rotor 20 within the quadrant that includes the quadrant number of the second abnormal part by adding the mechanical angle θ to the value obtained by multiplying the time elapsed since the time point of becoming the quadrant number by the rotational speed of the rotor 20. Furthermore, in this case, if the signal processing unit 50 determines that the quadrant number including the second abnormal part has ended, it terminates the calculation of the mechanical angle θ based on the second fail-safe processing and restarts the calculation of the mechanical angle θ based on the normal mechanical angle calculation processing.

[0103] Signal processing unit 50, for example Figure 10 As shown in the flowchart, the above-described anomaly detection processing and fault-safe processing are performed. Figure 10As shown, during the initial position detection process (step S13), the signal processing unit 50 performs a third anomaly determination process (step S14) to determine whether the difference between the first and second count values ​​is less than a predetermined number. If, in the third anomaly determination process (step S14), it is determined that the difference between the count values ​​is greater than or equal to the predetermined number (step S14: No), the signal processing unit 50 stops using the detection signal that generated the anomaly (step S17). In step S17, the signal processing unit 50, for example, compares the magnitude relationship between the first and second count values ​​to determine which of the first detection signals HA1, HA2, HA3 and the second detection signals Hu, Hv, Hw has generated an anomaly. In step S17, the signal processing unit 50, for example, determines that the detection signal used to count the smaller of the first and second count values ​​has generated an anomaly, and stops using that detection signal. If the signal processing unit 50 determines that the position detection of the rotor 20 can be performed using only a detection signal different from the detection signal that has been stopped from use, it continues the position detection process. If the signal processing unit 50 determines that the position detection of the rotor 20 cannot be performed using only a detection signal that is different from the detection signal that has been discontinued, the position detection processing is stopped.

[0104] If, in the third anomaly determination process (step S14), it is determined that the difference in the count values ​​is less than a predetermined number (step S14: Yes), the signal processing unit 50 executes the first anomaly determination process (step S15) to determine whether the timing order of the quadrant numbers is normal. If, in the first anomaly determination process (step S15), it is determined that the timing order of the quadrant numbers is abnormal (step S15: No), the signal processing unit 50 executes the first failsafe process (step S18).

[0105] If the timing sequence of the quadrant numbers is determined to be normal in the first anomaly determination process (step S15) (step S15: Yes), the signal processing unit 50 executes the second anomaly determination process (step S16) to determine whether the timing sequence of the partition numbers is normal. If the timing sequence of the partition numbers is determined to be abnormal in the second anomaly determination process (step S16) (step S16: No), the signal processing unit 50 executes the second failsafe process (step S19). If the timing sequence of the partition numbers is determined to be normal in the second anomaly determination process (step S16) (step S16: Yes), the signal processing unit 50 continues the position detection process as is.

[0106] In this embodiment, the signal processing unit 50 performs output processing. Output processing is the process of converting information related to the mechanical angle θ of the rotor 20, calculated through the aforementioned mechanical angle calculation processing, into different signals and outputting them externally. The signals output through output processing are, for example, output to a control device that controls the motor 10. The signal processing unit 50 converts the mechanical angle θ calculated in the mechanical angle calculation processing into U-phase, V-phase, and W-phase signals or A-phase, B-phase, and Z-phase signals in the output processing and outputs them.

[0107] exist Figure 11 The image shows an example of the U-phase signal PU as the U-phase signal, the V-phase signal PV as the V-phase signal, and the W-phase signal PW as the W-phase signal. For example... Figure 11 As shown, the U-phase signal PU, V-phase signal PV, and W-phase signal PW are rectangular wave pulse signals with phases offset from each other. In this embodiment, the phases of the U-phase signal PU, V-phase signal PV, and W-phase signal PW are offset from each other by an electrical angle of 120°. The U-phase signal PU, V-phase signal PV, and W-phase signal PW are rectified signals used to rectify the current supplied to the motor 10, equivalent to the "first phase signal". The control device, which receives the U-phase signal PU, V-phase signal PV, and W-phase signal PW as inputs, supplies the motor 10 with three-phase AC current of U-phase current, V-phase current, and W-phase current based on the U-phase signal PU, V-phase signal PV, and W-phase signal PW. In the following description, the U-phase signal PU, V-phase signal PV, and W-phase signal PW are collectively referred to as UVW signals.

[0108] exist Figure 12 The image shows an example of the A-phase signal PA as the A-phase signal, the B-phase signal PB as the B-phase signal, and the Z-phase signal PZ as the Z-phase signal. For example... Figure 12 As shown, phase A signal PA, phase B signal PB, and phase Z signal PZ are rectangular wave pulse signals. The phase difference between phase A signal PA and phase B signal PB is offset by a predetermined value. In this embodiment, this predetermined value is an electrical angle of 90°. Phase Z signal PZ is a pulse signal output for each revolution of rotor 20. Phase A signal PA, phase B signal PB, and phase Z signal PZ are signals corresponding to the incremental signals of phases A, B, and Z output from the rotary encoder, respectively, and are equivalent to "second phase signals". In the following description, phase A signal PA, phase B signal PB, and phase Z signal PZ are collectively referred to as ABZ signals.

[0109] like Figure 13As shown, in the output processing of this embodiment, the signal processing unit 50 outputs the UVW signal as an output signal at the power-on time of the motor 10. That is, the output processing includes: when the rotor 20 starts to rotate, generating a UVW signal for rectifying the current supplied to the motor 10 based on information related to the mechanical angle θ of the rotor 20 calculated in the mechanical angle calculation processing, and outputting the UVW signal as an output signal.

[0110] like Figure 13 As shown in the normal case, during output processing, if no abnormality is detected in the aforementioned abnormality sensing process during the period from the start of rotor 20 rotation until a predetermined time ta, the signal processing unit 50 switches the output signal from the UVW signal to the ABZ signal. That is, the output processing includes: if no abnormality is detected by the abnormality sensing process during the period from the start of rotor 20 rotation until the predetermined time ta, generating A-phase, B-phase, and Z-phase signals, i.e., the ABZ signal, based on information related to the mechanical angle θ of rotor 20 calculated in the mechanical angle calculation process, and outputting the ABZ signal as the output signal. Therefore, it can handle situations where it is desired to switch the output signal from the UVW signal to the ABZ signal after the predetermined time ta. Furthermore, if there are abnormalities in the three first detection signals HA1, HA2, HA3 and the three second detection signals Hu, Hv, Hw, the information related to the mechanical angle θ of rotor 20 calculated in the mechanical angle calculation process is prone to larger errors compared to the normal case. In the case of the ABZ signal, the mechanical angle θ is obtained by counting the pulses of the AB phase. However, when the ABZ signal is calculated based on information with large errors, the pulse count of the AB phase becomes abnormal. The resulting mechanical angle θ will have an error corresponding to the abnormal pulse count. In particular, when the pulse output width of the ABZ signal is shortened to improve resolution, the pulse count of the AB phase is easily affected by the magnitude of the error, and the error in the obtained mechanical angle θ is also easily amplified. Therefore, if no abnormality is detected through abnormality sensing processing from the start of rotor 20's rotation until a predetermined time ta has elapsed, the ABZ signal is output as the output signal. This suppresses the output of an ABZ signal with abnormal pulse counts and suppresses information about the mechanical angle θ with large errors. The length of the predetermined time ta is not particularly limited.

[0111] like Figure 13As shown in the case of an anomaly, during output processing, if the signal processing unit 50 senses an anomaly during the aforementioned anomaly sensing process from the start of rotor 20 rotation until a predetermined time ta has elapsed, the signal processing unit 50 will maintain the output signal as a UVW signal even after the predetermined time ta has elapsed. That is, the output processing includes: if an anomaly is sensed by the anomaly sensing process during the period from the start of rotor 20 rotation until the predetermined time ta has elapsed, the output signal will not be set to an ABZ signal, but will instead output a UVW signal as is. Compared to the ABZ signal, the UVW signal outputs fewer pulses per revolution of rotor 20. Therefore, even if there are abnormal parts in the three first detection signals HA1, HA2, HA3 and the three second detection signals Hu, Hv, Hw, the UVW signal generated based on information related to the machine angle θ calculated in the machine angle calculation process is less likely to be an abnormal signal compared to the ABZ signal. Therefore, by continuously outputting the UVW signal instead of the ABZ signal when an anomaly is sensed by the anomaly sensing process, the impact on the processing of the upper-level device that provides the output signal input can be reduced.

[0112] <Second Implementation>

[0113] In this embodiment, the fault-safe processing includes a third fault-safe processing. In this embodiment, the signal processing unit 50 performs the third fault-safe processing instead of the second fault-safe processing described in the first embodiment. The third fault-safe processing is performed when the second anomaly determination processing determines that an anomaly exists in a portion of the three second detection signals Hu, Hv, and Hw.

[0114] The third fail-safe processing includes the following steps: based on other second detection signals from parts not identified as having an anomaly, information related to the mechanical angle θ of the rotor 20, which was obtained under normal conditions based on a subset of second detection signals from parts identified as having an anomaly. Therefore, by using normal second detection signals to replace the abnormal second detection signal among the three second detection signals Hu, Hv, and Hw, it is easy to accurately estimate information related to the mechanical angle θ of the rotor 20.

[0115] For example, the case where partition numbers "7" and "8" are identified as second abnormal locations through the second anomaly detection process will be explained. Figure 5As shown, in partition numbers "7" and "8", if normal, the mechanical angle θ is calculated based on the segmented signals W1b and W2b in the second detection signal Hwt, which is converted into a triangular wave. Therefore, in the second anomaly determination process, if the signal processing unit 50 determines that an anomaly has occurred in partition numbers "7" and "8", it determines that an anomaly has occurred in the second detection signal Hw among the three second detection signals Hu, Hv, and Hw. In this case, the signal processing unit 50 estimates the mechanical angle θ in partition numbers "7" and "8" based on the two second detection signals Hu and Hv that are not determined to have an anomaly. In this embodiment, the signal processing unit 50 uses the digital values ​​of the portions of the two second detection signals Hu and Hv that are converted into triangular waves and located in partition numbers "7" and "8" to estimate the mechanical angle θ of partition numbers "7" and "8". That is, in this embodiment, the third failsafe processing includes the following process: based on the other second detection signals Hvt and Hwt that have undergone the triangular wave conversion process, information related to the mechanical angle θ of the rotor 20 obtained from a portion of the second detection signal Hu that is determined to have an anomaly is estimated. In the second detection signals Hut, Hvt, and Hwt, which are converted into triangular waves, the absolute values ​​of the slopes of the waveforms of each second detection signal Hut, Hvt, and Hwt in each partition become very close to each other. Therefore, the time variations of the absolute values ​​of the digital values ​​in each second detection signal Hut, Hvt, and Hwt are very close to each other. Therefore, by using the digital values ​​of other second detection signals Hut, Hvt, and Hwt and the normalization coefficient k[i] corresponding to the partition determined to have generated an anomaly, the mechanical angle θ can be calculated according to the above relationship Δθ=k[i]×ΔX, and the mechanical angle θ can be estimated with high accuracy.

[0116] The second detection signal Hwt located in partition number "7" is the segmented signal W1b. The segmented signal W1b is the first part from the intersection point P8 with the second detection signal Hut to the zero-crossing point P9. In the third fail-safe process, the signal processing unit 50, in partition number "7", replaces the segmented signal W1b, which is the first part, and uses the third part U3 of the waveform of the second detection signal Hut located in partition number "7" to estimate the mechanical angle θ. Specifically, in the third fail-safe process, the signal processing unit 50, in partition number "7", uses the digital value of the third part U3 and the normalization coefficient k[i] stored corresponding to the segment number of partition number "7", and estimates the mechanical angle θ according to the above-mentioned relationship Δθ=k[i]×ΔX.

[0117] The second detection signal Hw located in partition number "8" is the segmented signal W2b. The segmented signal W2b is the second part between the zero-crossing point P9 and the intersection point P10 with the other second detection signal Hvt. In the third fail-safe process, the signal processing unit 50 replaces the segmented signal W2b, which is the second part, in partition number "8" and estimates the mechanical angle θ based on the third part V3 located in partition number "8" of the waveform of the second detection signal Hvt. Specifically, in the third fail-safe process, the signal processing unit 50 estimates the mechanical angle θ in partition number "8" using the digital value of the third part V3 and the normalization coefficient k[i] stored corresponding to the segment number of partition number "8" according to the above-mentioned relationship Δθ=k[i]×ΔX.

[0118] As described above, in this embodiment, the third failsafe processing includes: if an abnormal part is determined to exist in a second detection signal Hwt, based on the second detection signal Hut that intersects with the segmented signal W1b, which is the first part, at the intersection P8, information related to the mechanical angle θ of the rotor 20 obtained under normal conditions based on the segmented signal W1b, which corresponds to the first part of a second detection signal Hwt, is estimated; and based on the other second detection signal Hvt that intersects with the segmented signal W2b, which is the second part, at the intersection P10, information related to the mechanical angle θ of the rotor 20 obtained under normal conditions based on the segmented signal W2b, which corresponds to the second part of a second detection signal Hwt, is estimated.

[0119] Here, in the second detection signals Hut, Hvt, and Hwt, which are converted into triangular waves, the waveform becomes curved in the portion that forms the peak of the waveform amplitude and in the portion near that peak, compared to the portion near the zero-crossing point that intersects with the reference value. Therefore, for example, the third part U3 of the second detection signal Hut, which intersects with the segmented signal W1b used to calculate the machine angle θ at intersection point P8 in partition number "7", is closer to a straight line than the portions of other second detection signals Hvt in partition number "7" that do not intersect with the segmented signal W1b. Thus, by replacing the segmented signal W1b, which corresponds to the first part of the second detection signal Hwt that is determined to have generated an anomaly, with the digital value of the third part U3 of the second detection signal Hut, the machine angle θ can be estimated with higher accuracy based on the relationship Δθ=k[i]×ΔX. Furthermore, for example, the third part V3 of the second detection signal Hvt, which intersects with the segmented signal W2b used to calculate the machine angle θ at intersection point P10 in partition number "8", is closer to a straight line than the parts of other second detection signals Hut in partition number "8" that do not intersect with the segmented signal W2b. Therefore, by replacing the segmented signal W2b, which corresponds to the second part of the second detection signal Hwt that is determined to have generated an anomaly, with the digital value of the third part V3 of the second detection signal Hvt, the machine angle θ can be estimated with higher accuracy based on the relationship Δθ=k[i]×ΔX.

[0120] The signal processing unit 50 in this embodiment is, for example, as follows: Figure 14 As shown in the flowchart, anomaly sensing and fault-safe processing are performed. Figure 14 The flowchart shown is identical to the one in which step S19 is replaced by step S20. Figure 10 The flowcharts shown are the same. For example... Figure 14 As shown, if the second anomaly determination process (step S16) determines that there is an anomaly in the timing order of the partition numbers (step S16: no), the signal processing unit 50 executes the third fault safety process described above (step S20).

[0121] The other configurations and methods in this embodiment are the same as those in the first embodiment. It should be noted that the third fail-safe processing in this embodiment can also be performed if two of the three second detection signals Hu, Hv, and Hw become abnormal. In this case, the signal processing unit 50 replaces the segmented signal from the two abnormal second detection signals and uses the digital value of the remaining segmented signal that did not become abnormal to estimate the mechanical angle θ in each partition based on the relationship Δθ = k[i] × ΔX.

[0122] This invention is not limited to the embodiments described above. Other configurations and methods can be employed within the scope of the technical concept of this invention. The first magnetic sensor can be any magnetic sensor. The first magnetic sensor can be a magnetic sensor that outputs an analog signal as a first detection signal. In this case, the multiple quadrants can be divided, for example, based on a combination of whether the signal strength of the first detection signal is large or small relative to a predetermined value. The multiple quadrants can be arbitrarily divided as long as they correspond to the combination of signal strengths in the first detection signal. The second magnetic sensor can be any magnetic sensor. The second magnetic sensor can be any magnetic sensor. The second magnetic sensor can be a magnetic sensor that outputs a digital signal as a second detection signal. The second magnet can also be a component different from the rotor magnet. In this case, the number of pole pairs of the second magnet can also be different from the number of pole pairs of the rotor magnet.

[0123] The mechanical angle calculation process can be performed based on at least one of the quadrant number and the partition number. That is, the mechanical angle calculation process can also be performed based on only one of the quadrant number and the partition number. For example, in the above embodiment, the signal processing unit 50 can also perform mechanical angle calculation processing based only on the second detection signals Hu, Hv, and Hw when the first detection signals HA1, HA2, and HA3 are interrupted. The motor that detects the rotational position of the rotor by the position detection device can be any motor.

[0124] It should be noted that this technology can be configured as described below.

[0125] (1) A position detection device for detecting the rotational position of a rotor in a motor, wherein the position detection device comprises: a first magnet having a pole pair and rotating together with the rotor; a second magnet having P pole pairs and rotating together with the rotor, wherein P is an integer greater than or equal to 2; N1 first magnetic sensors opposite to the first magnet and arranged at intervals in the rotational direction of the first magnet, wherein N1 is an integer greater than or equal to 2; and N2 second magnetic sensors opposite to the second magnet and... The second magnets are arranged at intervals along their rotational direction, where N2 is an integer greater than or equal to 3; and a signal processing unit processes N1 first detection signals output from N1 first magnetic sensors and N2 second detection signals output from N2 second magnetic sensors, dividing a period corresponding to one cycle of the rotor's mechanical angle into multiple quadrants corresponding to the combination of signal intensities in the N1 first detection signals, and dividing the period corresponding to one cycle of the rotor's mechanical angle into intervals formed by the intersections of waveforms in the N2 second detection signals. The system divides the system into multiple partitions based on points and zero-crossing points. The zero-crossing points are the points where N2 of the second detection signals intersect with a reference value. The signal processing unit performs the following: a first acquisition process, acquiring quadrant numbers representing the quadrants based on N1 of the first detection signals; a second acquisition process, acquiring partition numbers representing the partitions based on N2 of the second detection signals; a mechanical angle calculation process, calculating information related to the mechanical angles of the rotor based on at least one of the quadrant numbers and the partition numbers; and an anomaly sensing process, sensing anomalies in the first and second detection signals. The anomaly sensing process includes: a first anomaly determination process, determining that an abnormal part exists in the first detection signal if the temporal order of the quadrant numbers acquired through the first acquisition process differs from the normal temporal order of the quadrant numbers; and a second anomaly determination process, determining that an abnormal part exists in the second detection signal if the temporal order of the partition numbers acquired through the second acquisition process differs from the normal temporal order of the partition numbers.

[0126] (2) According to the position detection device of (1), wherein the signal processing unit performs: a first counting process, which counts whenever it is determined that the rotor has rotated by a predetermined angle based on N1 of the first detection signals; and a second counting process, which counts whenever it is determined that the rotor has rotated by the predetermined angle based on N2 of the second detection signals, wherein the abnormal sensing process includes a third abnormal determination process: if the difference between the count value obtained by the first counting process and the count value obtained by the second counting process is greater than or equal to a predetermined number, it is determined that there is an abnormal part in at least one of the first detection signal and the second detection signal.

[0127] (3) According to the position detection device of (1) or (2), wherein the signal processing unit determines in the abnormality sensing process that there is an abnormal part in one of the first detection signal and the second detection signal, and performs fault safety processing, the fault safety processing including: estimation processing, estimating information obtained under normal conditions based on the abnormal part in one of the first detection signal and the second detection signal based on at least a portion of the parts not determined to be abnormal in the first detection signal and the second detection signal; and processing to calculate information related to the mechanical angle of the rotor based on the information estimated by the estimation processing at the part determined to be abnormal.

[0128] (4) According to the position detection device described in (3), wherein the fault safety processing includes a first fault safety processing performed when it is determined in the first anomaly determination processing that there is an abnormal part in the first detection signal, the first fault safety processing includes the following processing: based on the sequential arrangement of the quadrant numbers obtained by the first acquisition processing and the preceding and following relationship of the first abnormal part, the quadrant number of the first abnormal part determined to be abnormal in the quadrant numbers obtained by the first acquisition processing is estimated.

[0129] (5) According to the position detection device of (3) or (4), wherein the fault safety processing includes a second fault safety processing performed when it is determined in the second anomaly determination processing that there is an abnormal part in the second detection signal, the second fault safety processing includes the following processing: determining which quadrant number the second abnormal part determined to be abnormal in the partition number obtained by the second acquisition processing is included in, and calculating information related to the mechanical angle of the rotor corresponding to the period of the second abnormal part based on the information of the mechanical angle of the rotor predetermined for the period of the quadrant number including the second abnormal part.

[0130] (6) According to the position detection device of (5), wherein the second fault safety processing includes the following processing: based on the information of the mechanical angle of the rotor predetermined for the period of the quadrant number including the second abnormal part, the time elapsed from the time point when the partition number becomes the second abnormal part or the time point when the quadrant number becomes the quadrant number including the second abnormal part, and the rotation speed of the rotor, information related to the mechanical angle of the rotor corresponding to the period of the second abnormal part is calculated.

[0131] (7) The position detection device according to (3) or (4), wherein the fault safety processing includes a third fault safety processing, which is performed when the second anomaly determination processing determines that a portion of the N2 second detection signals have an abnormal location, and the third fault safety processing includes the following processing: based on the other second detection signals that are not determined to have an abnormal location, presuming information related to the mechanical angle of the rotor obtained under normal conditions based on the portion of the second detection signals.

[0132] (8) According to the position detection device of (7), wherein the third failsafe processing includes the following processing: based on the other second detection signal that has been processed into a triangular wave, presuming information related to the mechanical angle of the rotor obtained based on the portion of the second detection signal.

[0133] (9) The position detection device according to (7) or (8), wherein the number of the second magnetic sensors is 3, and each of the three second detection signals output from the three second magnetic sensors has: a first portion from the intersection point of the second detection signal with one of the other two second detection signals to the zero-crossing point; and a second portion from the zero-crossing point to the intersection point of the second detection signal with the other of the other two second detection signals, wherein the third failsafe processing includes: if it is determined that there is an abnormal part in one of the second detection signals, estimating information related to the mechanical angle of the rotor obtained under normal conditions based on the first portion of the one second detection signal based on the one second detection signal, and estimating information related to the mechanical angle of the rotor under normal conditions based on the second portion of the one second detection signal based on the other second detection signal.

[0134] (10) The position detection device according to any one of (1) to (9), wherein the signal processing unit performs output processing to output an output signal, the output processing comprising: when the rotor starts to rotate, generating a first phase signal for rectifying current supplied to the motor based on information related to the mechanical angle of the rotor calculated in the mechanical angle calculation processing, and outputting the first phase signal as the output signal; and if no abnormality is detected by the abnormality sensing processing during the period from when the rotor starts to rotate until a predetermined time has elapsed, generating second phase signals of phase A, phase B and phase Z based on information related to the mechanical angle of the rotor calculated in the mechanical angle calculation processing, and outputting the second phase signal as the output signal.

[0135] (11) A position detection method for detecting the rotational position of a rotor in a motor, wherein the position detection method includes signal processing as follows: processing N1 first detection signals output from N1 first magnetic sensors opposite a first magnet and N2 second detection signals output from N2 second magnetic sensors opposite a second magnet, wherein N1 is an integer greater than or equal to 2 and N2 is an integer greater than or equal to 3, the first magnet has one pole pair and rotates together with the rotor, the second magnet has P pole pairs and rotates together with the rotor, wherein P is an integer greater than or equal to 2, the N1 first magnetic sensors are arranged at intervals in the rotational direction of the first magnet, the N2 second magnetic sensors are arranged at intervals in the rotational direction of the second magnet, a period corresponding to one period of the mechanical angle of the rotor is divided into multiple quadrants corresponding to the combination of signal intensities in the N1 first detection signals, and the period corresponding to one period of the mechanical angle of the rotor is divided by the intersection points and zero crossing points of the waveforms in the N2 second detection signals. The system comprises multiple partitions, wherein the zero-crossing point is the point where N2 of the second detection signals intersect with the reference value. The signal processing includes: a first acquisition process, which acquires a quadrant number representing the quadrant based on N1 of the first detection signals; a second acquisition process, which acquires a partition number representing the partition based on N2 of the second detection signals; a mechanical angle calculation process, which calculates information related to the mechanical angle of the rotor based on at least one of the quadrant number and the partition number; and an anomaly sensing process, which senses anomalies in the first detection signal and the second detection signal. The anomaly sensing process includes: a first anomaly determination process, which determines that there is an abnormal part in the first detection signal if the temporal order of the quadrant numbers acquired by the first acquisition process is different from the temporal order of the quadrant numbers under normal conditions; and a second anomaly determination process, which determines that there is an abnormal part in the second detection signal if the temporal order of the partition numbers acquired by the second acquisition process is different from the temporal order of the partition numbers under normal conditions.

[0136] The structures and methods described above can be appropriately combined within the scope of mutual non-contradiction.

[0137] Explanation of reference numerals in the attached figures

[0138] 10: Motor; 20: Rotor; 31: First magnet; 32: Second magnet; 41, 41a, 41b, 41c: First magnetic sensor; 42, 42a, 42b, 42c: Second magnetic sensor; 50: Signal processing unit; 100: Position detection device; HA1, HA2, HA3: First detection signal; Hu, Hut, Hv, Hvt, Hw, Hwt: Second detection signal; P1, P3, P5, P7, P9, P11, P13: Zero crossing point; ta: Specified time; U1a, U1b, V1a, V1b, W1a, W1b: Segmentation signal (first part); U2a, U2b, V2a, V2b, W2a, W2b: Segmentation signal (second part); θ: Mechanical angle.

Claims

1. A position detection device for detecting the rotational position of a rotor in a motor, wherein, This location detection device has the following features: A first magnet having a pair of magnetic poles rotates together with the rotor; A second magnet having P pole pairs rotates together with the rotor, where P is an integer greater than or equal to 2; N1 first magnetic sensors are positioned opposite the first magnet and are arranged at intervals in the rotational direction of the first magnet, wherein N1 is an integer greater than or equal to 2; N2 second magnetic sensors, which are opposite to the second magnet and arranged at intervals in the rotational direction of the second magnet, wherein N2 is an integer greater than or equal to 3; and The signal processing unit processes the N1 first detection signals output from the N1 first magnetic sensors and the N2 second detection signals output from the N2 second magnetic sensors. The period corresponding to one cycle of the rotor's mechanical angle is divided into multiple quadrants corresponding to the combination of signal intensities in N1 of the first detection signals. The period corresponding to one cycle of the rotor's mechanical angle is divided into multiple partitions by the intersection points and zero-crossing points of the waveforms in the N2 second detection signals, where the zero-crossing point is the point where the N2 second detection signals intersect with the reference value. The signal processing unit performs: The first acquisition process involves acquiring the quadrant number representing the quadrant based on N1 of the first detection signals; The second acquisition process involves acquiring a partition number representing the partition based on N2 of the second detection signals; The mechanical angle calculation process calculates information related to the mechanical angle of the rotor based on at least one of the quadrant number and the partition number; and Anomaly detection processing detects anomalies in both the first and second detection signals. The anomaly sensing process includes: The first anomaly detection process determines that an abnormal part exists in the first detection signal if the temporal order of the quadrant numbers obtained through the first acquisition process differs from the normal temporal order of the quadrant numbers; and The second anomaly determination process determines that if the temporal order of the partition numbers obtained through the second acquisition process is different from the normal temporal order of the partition numbers, an abnormal part exists in the second detection signal.

2. The position detection device according to claim 1, wherein, The signal processing unit performs: The first counting process involves counting whenever the rotor is determined to have rotated a predetermined angle based on N1 of the first detection signals; and The second counting process involves counting whenever, based on N2 of the second detection signals, it is determined that the rotor has rotated by the predetermined angle. The abnormal sensing process includes the following third abnormal determination process: if the difference between the count value obtained by the first counting process and the count value obtained by the second counting process is greater than or equal to a predetermined number, it is determined that there is an abnormal part in at least one of the first detection signal and the second detection signal.

3. The position detection device according to claim 1, wherein, If the signal processing unit determines during the anomaly sensing process that an abnormality exists in either the first detection signal or the second detection signal, it performs fault-safe processing. The fail-safe handling includes: The estimation process, based on at least a portion of the un-identified abnormal portions in the first and second detection signals, estimates information obtained under normal conditions based on the portions identified as abnormal in one of the detection signals; and The process of calculating information related to the mechanical angle of the rotor based on the information estimated through the estimation process at the location determined to be abnormal.

4. The position detection device according to claim 3, wherein, The fault-safe processing includes a first fault-safe process performed when the first anomaly determination process determines that there is an abnormal part in the first detection signal. The first fault safety processing includes the following processing: based on the sequential arrangement of the quadrant numbers obtained through the first acquisition process and the preceding and following relationships of the first abnormal part, the quadrant number of the first abnormal part that is determined to be abnormal in the quadrant numbers obtained through the first acquisition process is presumed.

5. The position detection device according to claim 3, wherein, The fail-safe processing includes a second fail-safe process performed when the second anomaly determination process determines that there is an abnormal part in the second detection signal. The second fault safety processing includes the following processing: determining which quadrant number the second abnormal part identified as abnormal in the partition number obtained through the second acquisition process belongs to, and calculating information related to the mechanical angle of the rotor corresponding to the period of the quadrant number including the second abnormal part based on the information of the rotor's mechanical angle predetermined for the period of the quadrant number including the second abnormal part.

6. The position detection device according to claim 5, wherein, The second fault-safe processing includes the following processing: based on the information of the rotor's mechanical angle, which is predetermined for the period of the quadrant number including the second abnormal part, the time elapsed from the time point when the partition number becomes the second abnormal part or the time point when the quadrant number becomes the quadrant number including the second abnormal part, and the rotor's rotational speed, calculating information related to the rotor's mechanical angle corresponding to the period of the second abnormal part.

7. The position detection device according to claim 3, wherein, The fault-safe processing includes a third fault-safe processing step, which is executed when the second anomaly determination process determines that a portion of the N2 second detection signals are abnormal. The third failsafe process includes the following processing: based on other second detection signals from parts not identified as having an abnormality, information related to the mechanical angle of the rotor obtained under normal conditions based on the portion of the second detection signals is presumed.

8. The position detection device according to claim 7, wherein, The third failsafe process includes the following processing: based on the other second detection signals that have been processed into triangular waves, information related to the mechanical angle of the rotor obtained based on the portion of the second detection signals is estimated.

9. The position detection device according to claim 7, wherein, The number of the second magnetic sensors is 3. The three second detection signals output from the three second magnetic sensors each have: The first portion between the intersection point of the second detection signal and one of the other two second detection signals and the zero crossing point; as well as The second portion between the zero-crossing point and the intersection point of the other two second detection signals. The third fault-safe processing includes: if an abnormal part is determined to exist in one of the second detection signals, inferring information related to the mechanical angle of the rotor obtained under normal conditions based on the first part of the second detection signal based on one of the second detection signals, and inferring information related to the mechanical angle of the rotor under normal conditions based on the second part of the second detection signal based on the other second detection signal.

10. The position detection device according to any one of claims 1 to 9, wherein, The signal processing unit performs output processing to output the output signal. The output processing includes: When the rotor begins to rotate, based on information related to the rotor's mechanical angle calculated in the mechanical angle calculation process, a first-phase signal for rectifying the current supplied to the motor is generated, and the first-phase signal is output as the output signal; and If no abnormality is detected by the abnormality sensing process during the period from when the rotor starts to rotate until a predetermined time has elapsed, second phase signals of phase A, phase B, and phase Z are generated based on information related to the mechanical angle of the rotor calculated in the mechanical angle calculation process, and the second phase signals are output as the output signals.

11. A position detection method for detecting the rotational position of a rotor in a motor, wherein, The position detection method includes the following signal processing: processing N1 first detection signals output from N1 first magnetic sensors opposite to the first magnet and N2 second detection signals output from N2 second magnetic sensors opposite to the second magnet, wherein N1 is an integer greater than or equal to 2 and N2 is an integer greater than or equal to 3. The first magnet has a pair of magnetic poles and rotates together with the rotor. The second magnet has P pole pairs and rotates together with the rotor, where P is an integer greater than or equal to 2. N1 of the first magnetic sensors are arranged at intervals in the rotational direction of the first magnet. N2 second magnetic sensors are arranged at intervals in the rotational direction of the second magnet. The period corresponding to one cycle of the rotor's mechanical angle is divided into multiple quadrants corresponding to the combination of signal intensities in N1 of the first detection signals. The period corresponding to one cycle of the rotor's mechanical angle is divided into multiple partitions by the intersection points and zero-crossing points of the waveforms in the N2 second detection signals, where the zero-crossing point is the point where the N2 second detection signals intersect with the reference value. The signal processing includes: The first acquisition process involves acquiring the quadrant number representing the quadrant based on N1 of the first detection signals; The second acquisition process involves acquiring a partition number representing the partition based on N2 of the second detection signals; The mechanical angle calculation process calculates information related to the mechanical angle of the rotor based on at least one of the quadrant number and the partition number; and Anomaly detection processing detects anomalies in both the first and second detection signals. The anomaly sensing process includes: The first anomaly detection process determines that an abnormal part exists in the first detection signal if the temporal order of the quadrant numbers obtained through the first acquisition process differs from the normal temporal order of the quadrant numbers; and The second anomaly determination process determines that if the temporal order of the partition numbers obtained through the second acquisition process is different from the normal temporal order of the partition numbers, an abnormal part exists in the second detection signal.

Images