JP5249158B2 - Encoder offset correction method - Google Patents

Description

  The present invention relates to an offset value correction technique for correcting an error in the detection result of an angular position in an encoder that detects the angular position of a rotating body.

  2. Description of the Related Art A magnetic encoder or the like is known as a device that detects the amount of displacement of a movable detection object and the absolute value of the displacement. In such a magnetic encoder, for example, a Sin signal obtained by rotating a single pole pair of permanent magnets (rotating bodies) in conjunction with a movable object to be detected and detecting a change in the magnetic field by a sensor element such as an MR element. After the AD conversion of the Cos signal and the Cos signal, an arctangent signal of the Sin signal and the Cos signal is calculated, and the absolute value of the angular position of the permanent magnet is detected by using the arctangent signal.

  Here, when the phase of the arctangent signal is used as a parameter, the Sin signal is plotted as the Y coordinate of the orthogonal coordinate system, and the Cos signal is plotted as the X coordinate of the orthogonal coordinate system, a so-called Lissajous waveform is obtained (see, for example, Patent Document 1).

  If such a Lissajous waveform is assumed that the Sin signal and the Cos signal are ideal signals without noise or the like, the Lissajous waveform has a circular shape with no center shift or distortion. However, in practice, the distance from the intersection of the Y-coordinate and the X-coordinate to the circumference of the Lissajous waveform may differ due to variations in sensor elements or the like. Therefore, when the encoder is shipped from the factory, it is common to apply an offset value correction width.

JP 2008-2904 A

  However, even when an offset value correction width is applied when the encoder is shipped from the factory, the angle depends on the temperature characteristics of the MR element, the temperature characteristics of the amplifier circuit until the MR element detection signal is input to the AD converter, and the like. There is a problem that an error occurs in the position detection result. More specifically, when the encoder is used under a temperature condition different from the offset adjustment environment at the time of shipment, an offset error occurs in the Sin signal and the Cos signal. Difficult to predict and deal with.

  Here, the present inventor proposes to sequentially correct the offset value while actually using the encoder after shipment. However, when the offset value is corrected while the encoder is actually used, there is a problem that if the offset value is sharply corrected, the encoder output will fluctuate greatly and the device with the encoder will malfunction. If the correction of (2) is performed slowly, there is a problem that a state in which a detection error occurs continues indefinitely.

  In view of the above problems, an object of the present invention is to quickly correct an offset value to an appropriate value without greatly changing the encoder output even when the offset value is corrected every time the encoder is actually used. Another object of the present invention is to provide a method for correcting an offset value of an encoder.

  In order to solve the above problems, in an encoder offset value correction method to which the present invention is applied, an encoder that detects an angular position of the rotating body based on an output signal of a sensor element that changes in conjunction with the rotation of the rotating body. In order to optimize the offset value for correcting the output signal in order to eliminate the error in the detection result of the angular position, a correction process for correcting the offset value is performed every time power supply to the encoder is started, In the correction step, a first correction step for detecting a deviation amount of the output signal from an ideal state and correcting the offset value based on the deviation amount; and after the first correction step, the first correction step And a second correction step of correcting the offset value with a smaller gain.

  In the present invention, after the encoder is shipped from the factory, the offset value is corrected while the encoder is actually used. For this reason, even when the encoder is used under a temperature environment different from the offset adjustment environment at the time of shipment, the offset value is updated to an optimum value according to the temperature environment. For this reason, even when the encoder is used under a temperature environment different from the offset adjustment environment at the time of shipment, it is possible to prevent an offset error from occurring. Further, when updating the offset value, the amount of deviation of the output signal from the ideal state when the detection operation at the encoder is started is detected, and after performing the first correction step based on the amount of deviation, A second correction step for correcting the offset value with a smaller gain than the correction step is performed. For this reason, immediately after the operation of the encoder is started, the offset value is sharply corrected in the first correction step, and thereafter, the offset value is gently corrected in the second correction step. For this reason, the offset value can be quickly corrected to an appropriate value, and a situation in which the output of the encoder greatly fluctuates does not occur.

  In the present invention, in the second correction step, each time the second correction step is performed, a deviation amount from the ideal state of the output signal is detected, and the offset value is corrected based on the deviation amount. preferable.

  In the present invention, in the second correction step, the total sum of the shift amounts is calculated every time the shift amount from the ideal state of the output signal is detected, and a correction width for the offset value is determined based on the total sum. It is preferable. If comprised in this way, an offset value can be gradually approached to an optimal value.

  In the present invention, the output signals are, for example, an A-phase signal and a B-phase signal having a phase difference of π / 2 or a phase difference of approximately π / 2, and the first correction step and the second correction step. Then, when a Lissajous waveform is formed on the XY coordinate axis based on the A phase signal and the B phase signal, the distance from the X axis at the two intersections where the Lissajous waveform and the Y axis intersect is equal, and The offset value is corrected in a direction in which the distances from the Y axis at the two intersections where the Lissajous waveform and the X axis intersect are equal. With this configuration, the point where the Lissajous waveform intersects the Y axis and the Y axis can be obtained only by rotating the rotating body by an electrical angle of 270 °, and the deviation amount can be calculated from the detection result.

  In the present invention, the A phase signal is preferably a Sin wave, and the B phase signal is preferably a Cos wave. With this configuration, the peak value of the A phase signal of the Sin wave and the peak value of the B phase signal of the Cos wave respectively correspond to the points where the Lissajous waveform intersects the Y axis and the Y axis. Therefore, it is possible to calculate the amount of deviation only by obtaining the peak value of the S-phase A-phase signal and the peak value of the Cos-wave B-phase signal.

  In the present invention, after the encoder is shipped from the factory, the offset value is corrected while the encoder is actually used. For this reason, even when the encoder is used under a temperature environment different from the offset adjustment environment at the time of shipment, the offset value is updated to an optimum value according to the temperature environment. For this reason, even when the encoder is used under a temperature environment different from the offset adjustment environment at the time of shipment, it is possible to prevent an offset error from occurring. Further, when updating the offset value, the amount of deviation of the output signal from the ideal state when the detection operation at the encoder is started is detected, and after performing the first correction step based on the amount of deviation, A second correction step for correcting the offset value with a smaller gain than the correction step is performed. For this reason, immediately after the operation of the encoder is started, the offset value is sharply corrected in the first correction step, and thereafter, the offset value is gently corrected in the second correction step. For this reason, the offset value can be quickly corrected to an appropriate value, and a situation in which the output of the encoder greatly fluctuates does not occur.

It is a block diagram which shows the hardware constitutions of the encoder with which the offset value correction method which concerns on embodiment of this invention is performed. It is a block diagram which shows the function of the microcomputer used with the encoder to which this invention is applied. It is explanatory drawing of the Lissajous waveform used when detecting the deviation | shift amount with the deviation | shift amount detection part used with the encoder to which this invention is applied. It is a flowchart which shows the content of the deviation amount detection process performed in the deviation amount detection part used with the encoder to which this invention is applied. It is explanatory drawing which shows the specific example at the time of correct | amending an offset value in the encoder to which this invention is applied. It is explanatory drawing which shows the simulation result of the detection error of the permanent magnet in the encoder to which this invention is applied.

  An encoder offset value correction method to which the present invention is applied will be described with reference to the drawings.

[Hardware configuration]
FIG. 1 is a block diagram showing a hardware configuration of an encoder in which an offset value correction method according to an embodiment of the present invention is performed, and FIGS. 1A and 1B are diagrams showing an outline of the hardware configuration. It is a figure and an image figure of hardware constitutions.

  As shown in FIGS. 1A and 1B, an encoder 1 to which the present invention is applied includes an MR element 10 (MR Element / sensor element) whose output signal changes in conjunction with the rotation of a rotating body, a microcomputer, and the like. In this embodiment, a permanent magnet 11 in which a pair of S poles and N poles are magnetized is used as the “rotating body”. For this reason, in the encoder 1, the Sin signal (A phase signal) and the Cos signal (B phase signal) having a phase difference of π / 2 or substantially π / 2 from the MR element 10 toward the microcomputer 20. Are entered. More specifically, the MR element 10 includes an A phase sensor that outputs a sinusoidal Sin signal corresponding to the rotation of the rotating body, and a B phase that outputs a sinusoidal Cos signal corresponding to the rotation of the rotating body. And a sensor. In FIG. 1A, only the MR element 10 which is one component of the A-phase sensor and the B-phase sensor is shown, but for example, a rectifier circuit, a low-pass filter, a differential amplifier, and an MR element 10 A phase sensor and a B phase sensor are constituted by various electric elements such as a driver for supplying an exciting current to the. The microcomputer 20 has an interpolator and electrical elements such as an RS422 driver and an open collector, but may have any other elements.

  As shown in FIG. 1A, the microcomputer 20 has an A / D conversion unit 21, and the analog signal is digitized by the A / D conversion unit 21. Although not particularly shown in FIG. 1, the microcomputer 20 has various electric elements such as a CPU, a ROM, and a RAM, and has a function for calculating the angular position of the permanent magnet 11 and a function for calculating a Lissajous waveform. In addition, it has an offset value correction function to be described later. That is, in the microcomputer 20, various programs for executing the above functions (calculation of the angular position of the permanent magnet 11, calculation of the Lissajous waveform, correction of the offset value) are stored in a storage medium such as a RAM / ROM. Reads out the program as appropriate and executes various processes while using the RAM as a working area.

[Configuration of Microcomputer 20]
FIG. 2 is a block diagram showing functions of the microcomputer 20 used in the encoder to which the present invention is applied.

  In the encoder 1 shown in FIG. 1, a deviation may occur in the relationship between the rotation of the permanent magnet 11 (rotating body) and the waveform of the output signal from the MR element 10 due to variations in the MR element 10. When the 1 is shipped from the factory, the encoder 1 is set with an offset value (fixed offset value). However, when the encoder 1 is used under a temperature environment different from the offset adjustment environment at the time of shipment, an error occurs with the fixed offset value set at the time of shipment. Therefore, in the encoder 1 of the present embodiment, the microcomputer 20 is configured as follows so that the offset value can be sequentially corrected while the encoder 1 is actually used.

  First, the microcomputer 20 includes an A / D converter 21 that converts an output signal (analog signal / A phase signal and B phase signal) from the MR element 10 into a digital signal, and an A / D converter 21 that converts the latest offset value. And an offset correction unit 50 that subtracts from the output of the signal, and the signal calculated by the offset correction unit 50 is output to the angular position detection unit 30. Here, the A / D conversion unit 21 is 10-bit data having no decimal part, whereas the offset value used in the offset correction unit 50 is 14-bit data having a decimal part. For this reason, 4 bits 22 are added to the output signal from the MR element 10 and converted into 14-bit data. This 4 bits 22 has a resolution of 0.0625 (1/16). In addition, the angular position detection unit 30 includes an angle calculation unit 33 based on a CORDIC (Coordinate Rotation Digital Computor) algorithm. For this reason, the angular position detection unit 30 includes a conversion unit 31 that adds 1 bit to 14-bit data and converts it to 15 bits, thereby increasing the accuracy of angle calculation. Therefore, the magnetic field change when the permanent magnet 11 rotates is detected by the MR element 10, and the angle calculation unit 33 calculates the arctangent signal of the Sin signal and the Cos signal based on the detection result (Sin signal and Cos signal). Then, the absolute value of the angular position of the permanent magnet 11 can be detected. Therefore, as an offset value used in the offset correction unit 50, an offset value for correcting a shift in the relationship between the rotation of the permanent magnet 11 (rotating body) and the waveform of the output signal from the MR element 10 before shipment from the factory. By using (fixed offset values COfsA, COfsB), the absolute value of the angular position of the permanent magnet 11 can be detected with high accuracy as long as the encoder 1 is used under the same temperature conditions as the offset adjusted before factory shipment. Can do.

[Configuration of offset value update unit]
In this embodiment, even when the encoder 1 is used under a temperature environment different from the offset adjustment environment at the time of shipment, every time power supply to the encoder 1 is started for the purpose of setting the offset value to an optimum value. An offset value update unit 100 that corrects the offset value is configured. The offset value update unit 100 corrects each error by adding or subtracting an offset value so that data a1 and a2 and data b1 and b2, which will be described later, are approximately equidistant from the intersection of the X axis and the Y axis, respectively. is there.

  In the present embodiment, the offset value update unit 100 generally includes a shift amount detection unit 60 that detects the shift amounts ΔA and ΔB of the relationship between the rotation of the permanent magnet 11 (rotary body) and the waveform of the output signal from the MR element 10. The latest offset value determining unit 70 for determining the latest offset values NOfsA and NOfsB based on the amount of deviation detected by the deviation amount detecting unit 60 and the fixed offset values COfsA and COfsB, and the latest determined by the latest offset value determining unit 70 A latest offset value output unit 80 for outputting the offset values NOfsA and NOfsB to the offset correction unit 50;

  First, the deviation amount detection unit 60 includes a data acquisition unit 61 that acquires data a1, a2, b1, and b2, which will be described later, and an error determination unit 62 that determines whether the data a1, a2, b1, and b2 are obtained. The acquisition result confirmation unit 63 for confirming whether or not the data a1, a2, b1, b2 are at appropriate positions, and the deviation amounts ΔA, ΔB from the ideal positions of the data a1, a2, b1, b2 are calculated. And a deviation amount calculation unit 64.

  The latest offset value determination unit 70 includes an offset value correction width determination unit 78 that determines offset value correction widths ΔOfsA and ΔOfsB at the time of offset update performed based on the shift amounts ΔA and ΔB, and an offset value correction width determination unit 78. A latest offset value calculation unit 79 that calculates the latest offset values NOfsA and NOfsB based on the determined offset value correction widths ΔOfsA and ΔOfsB. In the present embodiment, the latest offset value calculation unit 79 adds the offset value correction widths ΔOfsA and ΔOfsB determined by the offset value correction width determination unit 78 and the fixed offset values COfsA and COfsB to obtain the latest offset values NOfsA and NOfsB. calculate.

  In this embodiment, the offset value correction width determination unit 78 includes a correction number monitoring unit 75 that monitors the number of corrections n after the power is turned on, and an offset value correction width with a predetermined gain depending on the number of corrections this time. And an offset value correction width calculation unit 72 that calculates ΔOfsA and ΔOfsB. Further, in this embodiment, the offset value correction width determination unit 78 includes a deviation amount integration unit 71 that adds the second and subsequent deviation amounts ΔA and ΔB to obtain the integrated values SA and SB of the deviation amounts ΔA and ΔB. .

Here, the offset value correction width calculation unit 72 and the latest offset value calculation unit 79 indicate that the current correction is the first correction after the power is turned on in the monitoring result of the correction count monitoring unit 75, and 2 In the case of correction after the first time, the processing conditions are switched. Specifically, when the current correction is the first correction after the power is turned on, the offset value correction width calculation unit 72 sets x of the gain 2- ^x multiplied by the shift amounts ΔA and ΔB as 1, and the offset value. The correction widths ΔOfsA and ΔOfsB are calculated, and the latest offset value calculator 79 adds the offset value correction widths ΔOfsA and ΔOfsB and the fixed offset values COfsA and COfsB to calculate the latest offset values NOfsA and NOfsB.

On the other hand, when the current correction is the second or subsequent correction after the power is turned on, the offset value correction width calculation unit 72 multiplies the value calculated by the deviation amount integration unit 71 by the gain 2- ^x . At the same time, offset value correction widths ΔOfsA and ΔOfsB are calculated by setting x of gain 2 ^−x to 2. Further, the latest offset value calculation unit 79 sets the fixed offset values COfsA and COfsB offset adjusted before shipment from the factory in the first correction, and the latest offset value NOfsA obtained in the first correction in the second and subsequent corrections. , NOfsB is set as fixed offset values COfsA and COfsB, and the offset values correction widths ΔOfsA and ΔOfsB are added to the fixed offset values COfsA and COfsB to calculate the latest offset values NOfsA and NOfsB.

(Specific Configuration of Deviation Amount Detection Unit 60)
FIG. 3 is an explanatory diagram of a Lissajous waveform used when the deviation amounts ΔA and ΔB are detected by the deviation amount detection unit 60 used in the encoder 1 to which the present invention is applied. FIG. 4 is a flowchart showing the contents of the shift amount detection process performed by the shift amount detection unit 60 used in the encoder 1 to which the present invention is applied.

  In the encoder 1 of this embodiment, in order to detect the amount of deviation of the relationship between the rotation of the permanent magnet 11 and the waveform of the output signal from the MR element 10, as described below, the A phase signal and the B phase signal are detected. Based on the sum of the distances from the X-axis of the two intersections where the Lissajous waveform and the Y-axis intersect when the Lissajous waveform is formed on the XY coordinate axis based on the A-phase signal deviation amount ΔA, Based on the sum of the distances from the Y axis at the two intersections where the Lissajous waveform and the X axis intersect, the amount of deviation ΔB of the B phase signal is detected.

  More specifically, first, in the deviation amount detection unit 60, the data acquisition unit 61 calculates a Lissajous waveform shown in FIGS. This calculation is performed by the microcomputer 20 taking in the Sin signal (Sin value) and the Cos signal (Cos value) digitized at a predetermined sampling period by the A / D converter 21. Next, as shown in FIGS. 3B and 3C, the data acquisition unit 61 uses the Lissajous waveform on the XY plane calculated in the first step on the circumference of the Lissajous waveform and the X axis. Two points of data (b1, b2) in the vicinity of the two intersecting points are acquired. Similarly, data (a1, a2) of two points in the vicinity of each of the two points intersecting the Lissajous waveform on the Y axis are acquired.

  More specifically, in FIG. 4A, first, it is determined whether or not the Cos value satisfies −L <Cos value <L with respect to a certain reference value L (step S1). Here, the reference value L is a value for specifying the vicinity of the Cos value 0 shown in FIG. 3A, and the range of the reference value L is set based on the following idea. In FIG. 3A, the number of data sampling within the range of the reference value L is determined by the rotational speed of the permanent magnet 11 and the sampling period by the A / D conversion unit 21, but the data a1, a2, b1, In order to specify b2, at least one sampled data within the range of the reference value L is required in the range corresponding to each of the data a1, a2, b1, and b2. Here, when the range of the reference value L is set narrow, it is possible to calculate a range closer to the vicinity of “0” on the X axis and the Y axis. However, since the range of the reference value L is set to be narrow, the sampled data does not fall within the range of the reference value L, that is, there is a risk of data loss, and the data a1, a2, b1, b2 are specified. I can't. When data loss occurs in this way, correction calculation is not performed, that is, the offset value is not newly updated, and the correction accuracy of the offset value is deteriorated. On the other hand, when the range of the reference value L is set wide, the sampled data easily enters the range of the reference value L, and there is no fear of data loss. However, since the range in the vicinity of “0” on the X axis and the Y axis is widened, the data a1, a2, b1, b2 cannot be specified with high accuracy. Therefore, the accuracy of offset value correction also deteriorates. Therefore, the setting of the range of the reference value L in the present embodiment ensures an appropriate number of samplings within the range of the reference value L from the relationship between the rotational speed of the permanent magnet 11 and the sampling period by the A / D converter 21. It is set to be possible.

  In this embodiment, if the conditional expression in step S1 shown in FIG. 4A is not satisfied (step S1: NO), the process ends. On the other hand, if this condition is satisfied (step S1: YES), the data acquisition unit 61 acquires two points of data (a1, a2). Next, it is determined whether or not the Sin value is 0 or more (step S2). If it is 0 or more (step S2: YES), the Sin value is specified as data a1 (step S3). If it is less than 0 (step S2: NO), the Sin value is used as data a2. Specify (step S4). At that time, in the deviation amount detection unit 60 shown in FIG. 2, the error determination unit 62 determines whether or not the data a1 and a2 are properly obtained, and the acquisition result confirmation unit 63 determines that the polarity of the data a1 and a2 is Check if the opposite is true. Then, when the error determination unit 62 appropriately obtains the data a1 and a2, and the acquisition result confirmation unit 63 confirms that the polarities of the data a1 and a2 are opposite, the data a1 and a2 are determined.

  For example, as shown in FIGS. 3B and 3C, the data a1 and a2 obtained in this way are located on the circumference of a substantially circular Lissajous waveform. Here, when there is an offset, the data a1 and a2 have different distances from the intersection D of the X axis and the Y axis. That is, the distance from the position corresponding to the data a1 to the X axis is different from the distance from the position corresponding to the data a2 to the X axis. Therefore, in the deviation amount detection unit 60 shown in FIG. 2, the deviation amount calculation unit 64 calculates the deviation amount ΔA of the A phase from the value obtained by adding the Y coordinates of the data a1 and a2 (= a1 + a2).

  On the other hand, as shown in FIG. 4B, the CPU of the microcomputer 20 determines whether or not the Sin value satisfies −L <Sin value <L (step S11). If this conditional expression is not satisfied (step S11: NO), the process ends. On the other hand, if this condition is satisfied (step S11: YES), the data acquisition unit 61 acquires two points of data (b1, b2). Next, it is determined whether the Cos value is 0 or more (step S12). If it is 0 or more (step S12: YES), the Cos value is specified as data b1 (step S13). If it is smaller than 0 (step S12: NO), the Cos value is used as data b2. Specify (step S14). At that time, in the deviation amount detection unit 60 shown in FIG. 2, the error determination unit 62 determines whether or not the data b1 and b2 are obtained, and the acquisition result confirmation unit 63 reverses the polarities of the data b1 and b2. Check if it exists. Then, when the error discriminating unit 62 appropriately obtains the data b1 and b2, and the acquisition result confirming unit 63 confirms that the polarities of the data b1 and b2 are opposite, the data b1 and b2 are determined.

  For example, as shown in FIGS. 3B and 3C, the data b1 and b2 obtained in this way are located on the circumference of the substantially circular Lissajous waveform calculated in the first step. Here, when there is an offset, the data b1 and b2 have different distances from the intersection D of the X axis and the Y axis. That is, the distance from the position corresponding to the data b1 to the Y axis is different from the distance from the position corresponding to the data b2 to the Y axis. Accordingly, in the shift amount detection unit 60 shown in FIG. 2, the shift amount calculation unit 64 calculates the shift amount ΔB of the B phase from the value (= b1 + b2) obtained by adding the X coordinates of the data b1 and b2.

[Offset value correction method]
FIG. 5 is an explanatory diagram showing a specific example when the offset value is corrected in the encoder to which the present invention is applied. FIG. 6 is an explanatory diagram showing a simulation result of a detection error of the permanent magnet 11 (rotating body) in the encoder to which the present invention is applied.

  In this embodiment, the offset value update unit 100 shown in FIG. 2 adjusts the offset values so that the data a1, a2 and the data b1, b2 are approximately equidistant from the intersection of the X axis and the Y axis, respectively. The error is corrected.

  After the encoder 1 of this embodiment is mounted on a device, a correction process for correcting the offset value is performed every time power supply to the encoder 1 is started. At that time, in the correction step, as described below, the deviation amounts ΔA and ΔB from the ideal state of the output signal when the detection operation in the encoder 1 is started are detected, and the deviation amounts ΔA and ΔB are canceled out. First, a first correction step for correcting the offset value with the first gain, and a second correction step for correcting the offset value with a second gain smaller than the first correction step are performed after the first correction step. Since the offset value correction method for the A phase signal and the B phase signal is the same, in the following description, only the offset value correction method for the A phase signal will be described, and the offset value correction method for the B phase signal will be described. Description is omitted.

(First correction process)
First, power supply to the encoder 1 is started, and when the permanent magnet 11 rotates at least 135 °, the A-phase signal is input to the angular position calculation unit 30 after being corrected by the offset correction unit 50. In the first step, the offset value used in the offset correction unit 50 is the fixed offset value COfsA.

At that time, first, the first offset value update process (first correction process) is performed. Specifically, in the offset value update unit 100, the deviation amount detection unit 60 detects the deviation amount ΔA described with reference to FIG. Next, in the latest offset value determination unit 70, when the correction count monitoring unit 75 determines that the current correction is the first correction since the start of power supply (n = 1), the offset value correction width determination Each of the unit 78 and the latest offset value calculating unit 79 calculates the offset value correction width ΔOfsA and the latest offset value NOfsA in the following procedure. Specifically, the offset value correction width calculation unit 72 calculates the offset value correction width ΔOfsA by setting x of the gain 2- ^x multiplied by the deviation amount ΔA to 1, and the latest offset value calculation unit 79 calculates the offset value correction width. The latest offset value NOfsA is calculated by adding ΔOfsA and the fixed offset value COfsA.

Here, as shown in FIG. 5, the fixed offset value COfsA at the time of factory shipment is set to “512”, and the data a1 and a2 are the following conditions a1 + a2 = 0
An offset value (optimum offset value) for offset value correction width so as to satisfy the condition is “516”. In such a condition, the deviation amount ΔA (= (a1 + a2)) is “8”.

Accordingly, in the latest offset value determination unit 70, the offset value correction width calculation unit 72 multiplies the deviation amount ΔA (= 8) by the gain 2 ^−x (x = 1) to obtain the offset value correction width ΔOfsA (= 4) is calculated.

  Next, in the latest offset value determination unit 70, the latest offset value calculation unit 79 sets the fixed offset value COfsA (= 512) with respect to the offset value correction width ΔOfsA (= 4) determined by the offset value correction width determination unit 78. To obtain the latest offset value NOfsA (= 516), and outputs it to the latest offset value output unit 80. As a result, the latest offset value output unit 80 outputs the latest offset value NOfsA (= 516) to the offset correction unit 50.

Each value in the first correction step is as follows: Fixed offset value COfsA at factory shipment = 512
Optimal offset value = 516
Amount of deviation ΔA = 8
Offset value correction width ΔOfsA = ΔA / gain 2 ^−x = ΔA / 2 = 4
Latest offset value NOfsA
= Fixed offset value COfsA + offset value correction width ΔOfsA
= 516
Therefore, the latest offset value NOfsA matches the optimum offset value (= 516). In this way, the offset value correction width in the first offset value update process is completed.

  Next, when the permanent magnet 11 performs the next rotation, the A-phase signal is input to the angular position calculation unit 30 after being corrected by the offset correction unit 50. In this process, the offset value used in the offset value correction width unit 50 is the latest offset value NOfsA (= 516).

(Second correction step)
Next, the second and subsequent offset value update steps (second correction step) will be described. In the offset value update process (second correction process) after the second time, in the offset value update unit 100, the shift amount detection unit 60 detects the shift amount ΔA described with reference to FIG. 4 as in the first time.

Next, in the latest offset value determining unit 70, when the correction count monitoring unit 75 determines that the current correction is the second correction after starting the power supply (n> 2), the offset value correction width determining unit 78 and the latest offset value calculation unit 79 respectively calculate the offset value correction width ΔOfsA and the latest offset value NOfsA in the following procedure. Specifically, the deviation amount integration unit 71 adds the second and subsequent deviation amounts ΔA to obtain an integral value SA of the deviation amount ΔA, and the offset value correction width calculation unit 72 gains 2 ⁻ to multiply the deviation amount ΔA. ^The offset value correction width ΔOfsA is calculated by setting ^x of x to 2, and the latest offset value calculation unit 79 adds the offset value correction width ΔOfsA and the fixed offset value COfsA to calculate the latest offset value NOfsA. Here, the latest offset value calculation unit 79 calculates the latest offset value NOfsA using the latest offset value NOfsA (= 516) obtained by the first correction as the fixed offset value COfsA.

In the second correction, since the first offset value correction width has already been performed and the optimum offset value has not changed, the integral value SA is zero. Therefore, each value has the following condition Fixed offset value COfsA = 516
Optimal offset value = 516
Integral value of deviation amount ΔA = 0
Offset value correction width ΔOfsA = ΔA / gain 2 ^−x = ΔA / 4 = 0
Latest offset value NOfsA
= Fixed offset value COfsA + offset value correction width ΔOfsA
= 516
It becomes. Therefore, the latest offset value calculation unit 79 outputs the latest offset value NOfsA (= 516) to the latest offset value output unit 80, and the latest offset value output unit 80 outputs the latest offset value NOfsA (= 516) to the offset correction unit 50. Output to. As a result, the A phase signal is corrected by the offset value correction width unit 50 with the latest offset value NOfsA (= 516) and then input to the angular position calculation unit 30.

Next, in the latest offset value determining unit 70, when the correction count monitoring unit 75 determines that the current correction is the third correction since the start of power supply, the offset value correction width determining unit 78 and the latest offset value are determined. The calculation unit 79 calculates the offset value correction width ΔOfsA and the latest offset value NOfsA in the same procedure as in the second correction. In the third correction, since the first offset value correction width has already been performed and the optimum offset value has not changed, the integral value SA is zero. Therefore, each value has the following condition Fixed offset value COfsA = 516
Optimal offset value = 516
Integral value of deviation amount ΔA = 0
Offset value correction width ΔOfsA = ΔA / gain 2 ^−x = ΔA / 4 = 0
Latest offset value NOfsA
= Fixed offset value COfsA + offset value correction width ΔOfsA
= 516
It becomes. Therefore, the latest offset value calculation unit 79 outputs the latest offset value NOfsA (= 516) to the latest offset value output unit 80, and the latest offset value output unit 80 outputs the latest offset value NOfsA (= 516) to the offset correction unit 50. Output to. As a result, the A phase signal is corrected by the offset value correction width unit 50 with the latest offset value NOfsA (= 516) and then input to the angular position calculation unit 30.

Next, by continuously operating the device, when the temperature inside the device rises and the optimum offset in the encoder 1 changes, the optimum offset is changed from “516” in the fourth offset update process (second correction process), for example. The case where it changes to “520” will be described. Under such conditions, since the deviation amount ΔA is “8”, the integral value SA of the deviation amounts ΔA obtained by adding the second and subsequent deviation amounts ΔA in the deviation amount integration unit 71 of the offset value correction width determination unit 78. Is “8”. Further, the offset value correction width calculation unit 72 of the offset value correction width determination unit 78 calculates the offset value correction width ΔOfsA by setting x of the gain 2- ^x multiplied by the deviation amount ΔA to 2, and the latest offset value calculation unit 79 The latest offset value NOfsA is calculated by adding the offset value correction width ΔOfsA and the fixed offset value COfsA. Here, the latest offset value calculation unit 79 calculates the latest offset value NOfsA using the latest offset value NOfsA (= 516) obtained by the first correction as the fixed offset value COfsA. Therefore, each value is as follows: Fixed offset value COfsA = 516
Optimal offset value = 520
Integral value of deviation amount ΔA = 8
Offset value correction width ΔOfsA = ΔA / gain 2 ^−x = ΔA / 4 = 2
Latest offset value NOfsA
= Fixed offset value COfsA + offset value correction width ΔOfsA
= 518
It becomes. Therefore, the latest offset value calculation unit 79 outputs the latest offset value NOfsA (= 518) to the latest offset value output unit 80, and the latest offset value output unit 80 outputs the latest offset value NOfsA (= 518) to the offset correction unit. Output to 50. As a result, the A-phase signal is corrected to the latest offset value NOfsA (= 518) by the offset value correction width unit 50 and then input to the angular position calculation unit 30.

Next, in the fifth offset update process (second correction process), the deviation amount ΔA is “4”, and thus the deviation amount obtained by adding the second and subsequent deviation amounts ΔA in the deviation amount integration unit 71. The integrated value SA of ΔA is “12”. Further, the offset value correction width calculation unit 72 calculates the offset value correction width ΔOfsA by setting x of the gain 2- ^x multiplied by the deviation amount ΔA to 2, and the latest offset value calculation unit 79 is fixed to the offset value correction width ΔOfsA. The latest offset value NOfsA is calculated by adding the offset value COfsA. Here, the latest offset value calculation unit 79 calculates the latest offset value NOfsA using the latest offset value NOfsA (= 516) obtained by the first correction as the fixed offset value COfsA. Therefore, each value is as follows: Fixed offset value COfsA = 516
Optimal offset value = 520
Integral value of deviation amount ΔA = 12
Offset value correction width ΔOfsA = ΔA / gain 2 ^−x = ΔA / 4 = 3
Latest offset value NOfsA
= Fixed offset value COfsA + offset value correction width ΔOfsA
= 519
It becomes. Therefore, the latest offset value calculation unit 79 outputs the latest offset value NOfsA (= 519) to the latest offset value output unit 80, and the latest offset value output unit 80 outputs the latest offset value NOfsA (= 519) to the offset correction unit. Output to 50. As a result, the A-phase signal is corrected to the latest offset value NOfsA (= 518) by the offset value correction width unit 50 and then input to the angular position calculation unit 30.

Next, in the sixth offset update process (second correction process), the shift amount ΔA is “2”, and thus the shift amount obtained by adding the second and subsequent shift amounts ΔA in the shift amount integration unit 71. The integrated value SA of ΔA is “14”. Further, the offset value correction width calculation unit 72 calculates the offset value correction width ΔOfsA by setting x of the gain 2- ^x multiplied by the deviation amount ΔA to 2, and the latest offset value calculation unit 79 is fixed to the offset value correction width ΔOfsA. The latest offset value NOfsA is calculated by adding the offset value COfsA. Here, the latest offset value calculation unit 79 calculates the latest offset value NOfsA using the latest offset value NOfsA (= 516) obtained by the first correction as the fixed offset value COfsA. Therefore, each value is as follows: Fixed offset value COfsA = 516
Optimal offset value = 520
Integral value of deviation amount ΔA = 14
Offset value correction width ΔOfsA = ΔA / gain 2 ^−x = ΔA / 4 = 3.5
Latest offset value NOfsA
= Fixed offset value COfsA + offset value correction width ΔOfsA
= 519.5
It becomes. Therefore, the latest offset value calculation unit 79 outputs the latest offset value NOfsA (= 519.5) to the latest offset value output unit 80, and the latest offset value output unit 80 offsets the latest offset value NOfsA (= 519). Output to the correction unit 50. As a result, the A-phase signal is corrected to the latest offset value NOfsA (= 518) by the offset value correction width unit 50 and then input to the angular position calculation unit 30.

  The correction contents thereafter are as shown in FIG. Therefore, as long as the encoder 1 is operating, the second correction process of the offset value is performed, and the offset value is sequentially updated.

  According to such a correction method, as shown in FIG. 6, the offset value correction width is completed in the first correction (first correction step). Therefore, according to this embodiment, the rotation angle of the permanent magnet 11 Immediately after passing 135 °, the offset value approaches the optimum value greatly. In the second and subsequent corrections (second correction step), correction is performed with a second gain (= 1/4) smaller than the first gain (= 1/2) in the first correction. Accordingly, in the second correction process after the fourth time, the deviation amount ΔA is reduced to 8, 4, 2, 1, 0.5, 0.25,..., While the offset value based on the fixed offset value COfsA. The correction width ΔOfsA is gradually increased to 2, 3, 3.5, 3.75, 3.875, 3.9375,. For this reason, the error of the result of the angular position of the permanent magnet 11 gradually becomes a small value.

  The correction process is stopped when power supply to the encoder 1 is stopped. When power is supplied to the encoder 1, all values other than the fixed offset value COfsA are initialized, and the above correction process is performed again.

[Main effects of this embodiment]
As described above, in this embodiment, after the encoder 1 is shipped from the factory, the offset value is corrected during the period in which the encoder 1 is actually used. For this reason, even when the encoder 1 is used under a temperature environment different from the offset adjustment environment at the time of shipment, the offset value is updated by the offset value updating unit 100 to an optimum value according to the temperature environment. For this reason, even when the encoder 1 is used under a temperature environment different from the offset adjustment environment at the time of shipment, it is possible to prevent the occurrence of an offset error, so that high detection accuracy can be obtained.

  Further, when updating the offset value, the offset values ΔA and ΔB from the ideal state of the output signal when the detection operation in the encoder 1 is started are detected, and the offset values are set so as to cancel the deviation amounts ΔA and ΔB. After performing the first first correction step with a large first gain, in the second correction step, a second correction step for correcting the offset value with a second gain smaller than the first correction step is performed. For this reason, immediately after the operation of the encoder 1 is started, the offset value is sharply corrected in the first correction step, and thereafter, the offset value is gently corrected in the second correction step. For this reason, the offset value can be quickly corrected to an appropriate value, and a situation in which the output of the encoder 1 fluctuates greatly does not occur.

  In this embodiment, since the A-phase signal is a Sin wave and the B-phase signal is a Cos wave, the peak value of the A-phase signal of the Sin wave and the peak value of the B-phase signal of the Cos wave are respectively Y and the Lissajous waveform is Y. This corresponds to the point where the axis and the Y axis intersect. Therefore, the shift amounts ΔA and ΔB can be calculated only by obtaining the peak value of the A phase signal of the Sin wave and the peak value of the B phase signal of the Cos wave. Further, since the A phase signal and the B phase signal have a phase difference of π / 2 or a phase difference of approximately π / 2, the Lissajous waveform can be converted into the Y axis and the Y axis only by rotating the permanent magnet 11 by 135 °. The points where the axes intersect can be obtained, and the deviation amounts ΔA and ΔB can be calculated from the detection result. Therefore, the first correction process can be performed only by rotating the permanent magnet 11 by an electrical angle of 270 °.

1 Encoder 10 MR element 11 Permanent magnet (rotating body)
20 Microcomputer 21 A / D conversion unit 30 Angular position detection unit 50 Offset correction unit 60 Deviation amount detection unit 61 Data acquisition unit 62 Error discrimination unit 63 Acquisition result confirmation unit 64 Deviation amount calculation unit 70 Latest offset value determination unit 71 Deviation amount Integration unit 72 Offset value correction width calculation unit 75 Correction frequency monitoring unit 78 Offset value correction width determination unit 79 Latest offset value calculation unit 80 Latest offset value output unit 100 Offset value update unit

Claims (5)

In an encoder that detects the angular position of the rotating body based on the output signal of the sensor element that changes in conjunction with the rotation of the rotating body, an offset that corrects the output signal to eliminate an error in the detection result of the angular position In optimizing the value,
A correction process for correcting the offset value every time power supply to the encoder is started,
In the correction step, a first correction step for detecting a deviation amount of the output signal from an ideal state and correcting the offset value based on the deviation amount; and after the first correction step, the first correction step And a second correction step of correcting the offset value with a smaller gain than the encoder offset value correction method.   In the second correction step, a deviation amount from an ideal state of the output signal is detected every time the second correction step is performed, and the offset value is corrected based on the deviation amount. Item 4. The encoder offset value correction method according to Item 1.   In the second correction step, every time the amount of deviation from the ideal state of the output signal is detected, the sum of the amount of deviation is calculated, and a correction width for the offset value is determined based on the sum. The encoder offset value correction method according to claim 2. The output signals are an A-phase signal and a B-phase signal having a phase difference of π / 2 or a phase difference of approximately π / 2.
In the first correction step and the second correction step, when a Lissajous waveform is formed on the XY coordinate axes based on the A phase signal and the B phase signal, the Lissajous waveform and the Y axis intersect with each other. The offset value is corrected in a direction in which the distance from the X-axis of the intersection is equal and the distance from the Y-axis of two intersections where the Lissajous waveform and the X-axis intersect is equal. The encoder offset value correction method according to any one of claims 1 to 3.   5. The encoder offset value correction method according to claim 4, wherein the A-phase signal is a Sin wave and the B-phase signal is a Cos wave.

Images