JP2020027027A - Calculation device of eccentricity of detection rotor - Google Patents

Abstract

To provide a detection rotor eccentricity calculation device capable of further precisely calculating eccentricity of a detection rotor.SOLUTION: A calculation device of eccentricity of a detection rotor comprises a centering center detection sensor 30, a turntable 2 for rotating the centering center detection sensor 30, and a computing unit 72 for calculating the eccentricity of the detection rotor with respect to a rotary shaft. The centering center detection sensor 30 includes two X-direction detection sections disposed to face each other at 180°, and two Y-direction detection sections disposed with their phases respectively offset by 90° from the X-direction detection sections, and outputs a differential value between output signals of the two X-direction detection sections as an X-direction displacement and a differential value between output signals of the two Y-direction detection sections as a Y-direction displacement. When a divisor of the number of teeth of the detection rotor and a natural number equal to or more than 5 is defined as n, the computing unit 72 rotates the turntable by [360/n]° and calculates the eccentricity of the detection rotor on the basis of the X-direction displacement and the Y-direction displacement read out for every angle of [360/n]°.SELECTED DRAWING: Figure 3

Description

  This specification calculates the amount of eccentricity of the detection rotor with respect to the rotation shaft in a manufacturing process of the rotation position detector that detects the rotation position based on the teeth (irregularities) of the detection rotor fixed to the end face of the rotation shaft. An eccentricity calculating device is disclosed.

  Patent Documents 1 and 2 disclose well-known centering devices for centering a detection rotor with respect to a rotating shaft in a manufacturing process of a rotational position detector. In Patent Document 1, at least one detection winding that outputs a signal proportional to the distance from the rotor is wound around one or more teeth of a rotor displacement detection core that includes a plurality of teeth on an inner peripheral portion and is made of a magnetic material. It constitutes a detection sensor. In Patent Document 1, when the center detection sensor is relatively rotated with respect to the detection rotor, the position of the detection rotor with respect to the rotation center is changed based on an output signal obtained from the detection winding. Further, in Patent Literature 2, indirectly, a detection rotor is pressed against a rotating shaft portion via a drum, and an impact applying mechanism using a piezoelectric element is brought into contact with the drum to drive the impact applying mechanism. Discloses a technique for moving a detection rotor.

JP 2017-83357 A JP-A-2017-32487

  Patent Literature 2 discloses only a technique for moving a detection rotor, and does not sufficiently describe calculation of an eccentric amount. Further, according to the technique of Patent Document 1, a signal proportional to the distance from the detection rotor is output from each winding. This output signal fluctuates with the relative rotation of the center detection sensor with respect to the detection rotor. The output signal usually includes a fluctuation component caused by the eccentricity of the detection rotor, a fluctuation component caused by a roundness error of the detection rotor, and a fluctuation component caused by unevenness (teeth) of the detection rotor. include. In Patent Literature 1, a fluctuation component caused by a roundness error of the detection rotor and a fluctuation component caused by unevenness (teeth) of the detection rotor have not been sufficiently studied. As a result, in Patent Literature 1, the eccentric amount of the detection rotor cannot be accurately calculated.

  Therefore, the present specification discloses a detection rotor eccentricity calculation device that can more accurately calculate the eccentricity of the detection rotor.

  The eccentricity calculating apparatus of the present specification includes a detection rotor having a plurality of teeth provided on a peripheral surface thereof, a rotation shaft that rotates together with the detection rotor, and a sensor unit that detects a rotation angle of the detection rotor. An eccentricity calculation device for a detection rotor for calculating an eccentricity amount of the detection rotor with respect to the rotation shaft portion during a manufacturing process of the rotation position detector having the eccentricity amount output signal according to a distance to a peripheral surface of the detection rotor. A detecting unit configured to detect four center detection sensors provided at intervals of 90 degrees, a rotation table for rotating the center detection sensor relative to the detection rotor, and an output signal from the center detection sensor. An arithmetic unit for calculating the amount of eccentricity of the detection rotor with respect to the rotation shaft portion, wherein the center detection sensor comprises two X-direction detection units which are arranged to face each other by 180 degrees; And two Y-direction detectors arranged with a phase shift of 90 degrees with respect to the two Y-direction detectors, wherein a difference value between output signals of the two X-direction detectors is defined as an X-direction displacement. Is output as a Y-direction displacement, and the arithmetic unit sets the rotation table to [360 / n] °, where n is a natural number that is a divisor of the number of teeth of the detection rotor and is 5 or more. And the eccentricity of the detection rotor is calculated based on at least one of the X-direction displacement and the Y-direction displacement read at every [360 / n] ° angle.

  The X-direction displacement and the Y-direction displacement, which are the difference values of the output signals of the two detectors, include not only the fluctuation component E due to the eccentricity, but also the fluctuation component e1 due to the circular error of the detection rotor and the fluctuation component e1 of the detection rotor. A fluctuation component e2 caused by teeth (irregularities) provided on the peripheral surface is also included. Since the value of the variation component e2 included in the X-direction displacement and the Y-direction displacement read for each angle of [360 / n] ° is the same, the variation component e1 can be treated as a simple offset. Also, since the fluctuation components e1 included in the n X-direction displacements and the Y-direction displacements read out at every [360 / n] ° angle have symmetry, they must be canceled out by performing a predetermined operation. Can be. As a result, only the fluctuation component E due to eccentricity can be extracted from at least one of the obtained X-direction displacement and Y-direction displacement, and the eccentricity of the detection rotor can be calculated more accurately.

In this case, when the X-direction displacement and the Y-direction displacement read out at the k-th time are X _k and Y _k , the arithmetic unit calculates the eccentric amount D (θ) = A × sin which fluctuates sinusoidally with rotation. The amplitude A and the phase α of (θ + α) may be calculated based on at least one of Expressions 1-4 and 5-8.

  By performing the calculations of Expressions 1 and 2 or Expressions 5 and 6, the fluctuation components e1 and e2 included in the X-direction displacement or the Y-direction displacement can be canceled. Then, by obtaining the amplitude A and the phase α of the eccentric amount based on XSS, XCS or YSS, YCS in which the fluctuation components e1, e2 have been canceled, the eccentric amount of the detection rotor can be calculated more accurately.

In this case, the computing unit may calculate the amount of eccentricity in the X direction DX (θ) and the amount of eccentricity in the Y direction DY (θ) at the rotation angle θ based on Equation 9-10.

  The X-direction displacement and the Y-direction displacement are calculated by using both the amplitude A = RX and the phase α = PX obtained from the X-direction displacement and the amplitude A = RY and the phase α = PY obtained from the Y-direction displacement. , Errors can be averaged, and the amount of eccentricity of the detection rotor can be calculated more accurately.

  According to the eccentricity calculating device disclosed in the present specification, the eccentricity of the detection rotor can be calculated more accurately.

FIG. 2 is a schematic configuration diagram of a rotor centering device incorporating an eccentricity amount calculating device. FIG. 2 is a schematic sectional view taken along line AA in FIG. 1. FIG. 2 is a schematic block diagram of an eccentricity calculating device. FIG. 4 is a diagram illustrating a fluctuation component included in an X-direction displacement. FIG. 5 is a diagram in which a fluctuation component e1 (θ) caused by a perfect circle error is extracted from FIG. FIG. 5 is a diagram in which a fluctuation component e2 (θ) caused by unevenness is extracted from FIG.

  FIG. 1 shows a schematic configuration diagram of a detection rotor centering device in which the eccentricity amount calculating device is incorporated. FIG. 2 is a schematic sectional view taken along line AA of FIG. FIG. 3 is a schematic block diagram of the eccentricity calculating device. The detection rotor centering device centers the detection rotor W1 with respect to the rotation shaft W3 in the manufacturing process of the rotational position detector. The rotation position detector includes: a rotation shaft portion W3 that rotates together with a detection target; a detection rotor W1 fixed to an axial end face of the rotation shaft portion W3; bearings W4 and W5 press-fitted into the rotation shaft portion W3; It has a housing W2 fixed to the outer ring of W4 and W5 with an adhesive, and a sensor unit (not shown) fixed to the housing W2. The rotational position detector of this example is of a reluctance type resolver type. Therefore, the detection rotor W1 is made of a magnetic material, and has a plurality of teeth (irregularities) formed on the outer peripheral surface. The sensor unit is a so-called resolver stator, and includes a core having a plurality of teeth formed on an inner peripheral portion, and a winding wound around the core.

  In the manufacturing process, the detection rotor W1 is placed on the rotation shaft W3 via an adhesive. This adhesive is, for example, a highly viscous elastomer-modified acrylate adhesive and maintains fluidity for a while after being applied. Until the adhesive is cured, the frictional resistance between the detection rotor W1 and the rotation shaft W3 is extremely low, and the detection rotor W1 may move in the plane direction with respect to the rotation shaft W3. it can. When centering the detection rotor W1, the rotating shaft portion W3 is fixed to the base 7 via the fixed shaft 8 by screws or the like. In other words, the rotation shaft portion W3 and the detection rotor W1 are set in the detection rotor centering device in a state where they cannot rotate.

  A centering center detection sensor 30 is fixed to the housing W2 with screws or the like instead of the sensor unit. The centering center detection sensor 30 has substantially the same configuration as the resolver stator (sensor unit) of the rotational position detector, and includes a core having a plurality of teeth formed on an inner peripheral portion thereof, and a winding wound around the core. And The resolver stator (sensor section) of the rotational position detector may be used as the centering center detection sensor 30.

  The housing W2 is supported by the turntable 2. The rotary table 2 is a table that rotates around the fixed shaft 8 (and thus the rotating shaft W3). As the rotary table 2 rotates, the phase (rotation angle) of the centering center detection sensor 30 fixed to the housing W2 with respect to the detection rotor W1 changes.

  An air cylinder 51 is provided above the detection rotor W1 so as to be able to advance and retreat (up and down) in the axial direction. The air cylinder 51 is fixed to the base 7 via the support 28. A drum 50 is interposed between the air cylinder 51 and the detection rotor W1. The air cylinder 51 presses the detection rotor W1 in the axial direction via the drum 50 by the pressure of air (not shown). The contact surface between the air cylinder 51 and the drum 50 is in contact with a mirror-finished surface coated with grease, and the frictional force is sufficiently small. Therefore, the drum 50 can move in the radial direction with respect to the air cylinder 51. On the other hand, a thin rubber or the like having a large friction coefficient is adhered to the lower surface of the drum 50, and the drum 50 is frictionally engaged with the axial end surface of the detection rotor W1. Therefore, when the drum 50 receives an impact and moves, the detection rotor W1 moves together with the drum 50.

  Around the drum 50, four solenoid actuators 31, 32, 33, 34 are arranged at 90-degree intervals. Each of the solenoid actuators 31, 32, 33, and 34 functions as an impact force applying mechanism for moving the detection rotor W1 in the radial direction together with the drum 50 by moving in the radial direction and colliding with the drum 50. The solenoid actuators 31, 32, 33, and 34 are attached to the base 7 via advance and retreat cylinders 61, 62, 63, and 64, and columns 41, 42, 43, and 44, which will be described later. Cannot rotate.

  The solenoid actuators 31, 32, 33, and 34 take a state in which the movable portion advances radially (moves in a direction approaching the drum 50) and a state in which the movable portion retreats (moves in a direction away from the drum 50). Can be. The four solenoid actuators 31, 32, 33, and 34 are arranged in a standby state with a sufficient gap provided with the drum 50. This gap is set to be equal to or larger than the gap between the detection rotor W1 and the centering center detection sensor 30, and is, for example, about 2 mm. The strokes of the solenoid actuators 31, 32, 33, 34 are greater than this gap, for example, about 7 mm. Therefore, it can be said that the solenoid actuators 31, 32, 33, and 34 can advance and retreat from the standby position completely separated from the drum 50 to a position where the solenoid collides with the drum 50.

  A voltage is applied to the solenoids of the solenoid actuators 31, 32, 33, 34 via a variable power supply 73 and an insulating amplifier 74. When a voltage is applied, a current flows through a coil (solenoid) in a magnetic field, and a driving torque is generated. With this torque, the drive unit of the solenoid actuator advances in the radial direction (moves in a direction approaching the drum 50). The drive unit moves the drum 50 by colliding with the drum 50 after the gap section is accelerated. When the solenoid actuators 31, 32, 33, and 34 are retracted in the radial direction (moved in a direction away from the drum 50), the voltage application to the solenoid is released. As a result, the drive unit returns to the original position by the spring installed on the solenoid actuator.

  As described above, the drum 50 can slide in the horizontal direction with respect to the air cylinder 51, while being frictionally engaged with the detection rotor W1. Therefore, when an impact force is applied to the drum 50, the drum 50 moves in the horizontal direction while dragging the detection rotor W1.

  In addition, when attaching and detaching the rotation position detector from the centering device, a larger gap (for example, several tens of cm) is required between each of the solenoid actuators 31, 32, 33, and 34 and the drum 50. Therefore, in this example, the solenoid actuators 31, 32, 33, and 34 are mounted on the forward and backward cylinders 61, 62, 63, and 64 in order to secure a space for attaching and detaching the rotational position detector. Each of the forward and backward cylinders 61, 62, 63, and 64 is fixed to the base 7, which is a fixed member, by screws or the like via the columns 41, 42, 43, and 44 instead of the rotary table 2. Each of the forward and backward cylinders 61, 62, 63, 64 retreats in the radial direction when the rotational position detector is attached or detached, so that the entire solenoid actuators 31, 32, 33, 34 are largely separated from the drum 50. On the other hand, when performing the centering process, each of the forward and backward cylinders 61, 62, 63, 64 advances in the radial direction to bring the entire solenoid actuators 31, 32, 33, 34 close to the drum 50.

  Such a detection rotor centering device incorporates an eccentricity calculating device for calculating the eccentricity of the detecting rotor. As shown in FIG. 3, the eccentricity amount calculating device includes a centering center detection sensor 30, a rotation table 2 for rotating the centering center detection sensor 30 relative to the detection rotor W1, and a detection value of the centering center detection sensor 30. And a calculator 72 for calculating the amount of eccentricity based on

  In this example, the centering center detection sensor 30 has the same configuration as the resolver stator of the rotational position detector. Therefore, the centering center detection sensor 30 is provided with, for example, four detection units that output a signal corresponding to the distance to the radial end surface of the detection rotor W1 at 90-degree intervals. Each detection unit is configured by winding a winding around a pole tooth of a stator core made of a magnetic material. Hereinafter, of the four detection units, two detection units that are 180 degrees opposite to each other are referred to as “X direction detection units”, and two detection units that are 90 degrees out of phase with the X direction detection units are “Y direction detection units”. Call. The centering center detection sensor 30 outputs a difference signal between the two X-direction detectors as an X-direction displacement, and outputs a difference signal between the two Y-direction detectors as a Y-direction displacement. In the following, the rotation angle of the rotary table 2 (and, by extension, the centering center detection sensor 30) is “θ”, and the angle at which the phases of the two X-direction detectors and the solenoid actuators 31, 32 match θ = 0 °. And The displacement in the X direction at the rotation angle θ is represented by X (θ), and the displacement in the Y direction is represented by Y (θ).

  The output signal of the centering center detection sensor 30, that is, the displacement in the X direction and the displacement in the Y direction are input to the computing unit 72. The calculator 72 calculates the amount of eccentricity of the detection rotor W1 based on the X-direction displacement and the Y-direction displacement. The computing unit 72 also controls the variable power supply 73 and the insulating amplifier 74 and controls the rotation drive of the turntable 2 based on the calculated eccentricity. In other words, the computing unit 72 also functions as a control unit of the detection rotor centering device.

  Before describing the calculation of the amount of eccentricity by the calculator 72, the displacement in the X direction and the displacement in the Y direction will be described with reference to FIGS. FIG. 4 is a diagram showing a fluctuation component included in the X-direction displacement X (θ). In FIG. 4, the horizontal axis is the rotation angle θ.

  In general, the X-direction displacement X (θ) changes with the change of the rotation angle θ. The X-direction displacement X (θ) includes a variation component E (θ) due to the eccentricity of the detection rotor W1, a variation component e1 (θ) due to the roundness error of the detection rotor W1, and a variation component E1 (θ) of the detection rotor W1. And a fluctuation component e2 (θ) caused by unevenness (teeth). In FIG. 4, the two-dot chain line indicates the fluctuation component E (θ) due to eccentricity, the one-dot chain line indicates the fluctuation component e1 (θ) due to the perfect circular error, and the broken line indicates the fluctuation component e2 ( θ) are shown. FIG. 5 is a graph illustrating a variation component e1 (θ) due to a perfect circular error, and FIG. 6 is a graph illustrating a variation component e2 (θ) due to unevenness.

  The X direction displacement (θ) is the sum of these three fluctuation components, and X (θ) = E (θ) + e1 (θ) + e2 (θ). In FIG. 4, the solid line indicates the displacement (θ) in the X direction. Note that the Y-direction displacement Y (θ) is a signal that is 90 degrees out of phase with the X-direction displacement (θ), that is, X (θ) = Y (θ + 90 °).

  Each variable component will be specifically described. When the detection rotor W1 is not a perfect circle, even if the detection rotor W1 is not eccentric, the distance from the X-direction detection unit to the outer peripheral surface of the detection rotor W1 changes depending on the rotation angle θ, and the X-direction displacement X (θ) Fluctuates. The fluctuation component e1 (θ) is a fluctuation component generated due to the true circular error. The fluctuation component e1 (θ) is a component that fluctuates sinusoidally at the same cycle as the true circular error. The perfect circle error often has mainly two periods and three periods. FIG. 4 exemplifies a fluctuation component e1 (θ) that has a perfect circle error of three cycles and fluctuates in a cycle of 120 degrees.

  Further, a plurality of irregularities (teeth) are formed on the outer peripheral surface of the detection rotor W1. Due to the unevenness, even if the detection rotor W1 is not eccentric, the distance from the X-direction detection unit to the outer peripheral surface of the detection rotor W1 changes depending on the rotation angle, and the X-direction displacement X (θ) changes. . The fluctuation component e2 (θ) is a fluctuation component generated due to the unevenness of the detection rotor W1. The fluctuation component e2 (θ) fluctuates sinusoidally with the same cycle as the number of teeth of the detection rotor W1. FIG. 4 shows a case where the number of teeth of the detection rotor W1 is 5, and the fluctuation component e2 (θ) fluctuates in a cycle of 360 ° / 5 = 72 °.

  Here, the displacement information of the centering center detection sensor 30 should have a pole tooth and winding structure that cancels out the unevenness component (e2 (θ)) of the detection rotor W1. I want to detect only the difference in the X and Y directions of the gap between the detection rotors. However, the unevenness component of the detection rotor, that is, the fluctuation component e2 (θ) slightly remains due to the shape error of the pole teeth and the like.

  The fluctuation component E (θ) due to the eccentricity is a component that fluctuates in a sinusoidal waveform at a cycle of 360 °. Therefore, the fluctuation component E (θ) can be expressed as E (θ) = A × sin (θ + α) + B. The distance of the fluctuation component E (θ) from the center of the vibration, that is, D (θ) = A × sin (θ + α) is determined by the amount of eccentricity to be calculated by the calculator 72 (that is, the detection rotor is moved for centering). Amount to be performed). Therefore, the amount of eccentricity D (θ) can be calculated by determining the amplitude A and the phase difference α of the fluctuation component E (θ) due to eccentricity. The arithmetic unit 72 removes the fluctuation component e1 (θ) due to the perfect circular error and the fluctuation component e2 (θ) due to the irregularity from the X-direction displacement X (θ) and the Y-direction displacement Y (θ). Then, the amplitude A and the phase difference α of the fluctuation component E (θ) due to the eccentricity are obtained, and the eccentricity D (θ) is calculated.

  Next, a specific calculation procedure of the eccentric amount D (θ) by the arithmetic unit 72 will be described. Prior to calculating the amount of eccentricity D (θ), the calculator 72 instructs the turntable 2 to rotate by [360 / n] °. Here, the numerical value n is a divisor of the number of teeth of the detection rotor and a natural number of 5 or more. This numerical value n may be selected by the user and input to the calculator 72. As another form, the arithmetic unit 72 may automatically specify the numerical value n from the number of teeth of the detection rotor. For example, as described later, the smaller the numerical value n, the smaller the number of calculation processes. Therefore, the arithmetic unit 72 may specify the smallest numerical value as n, which is a divisor of the number of teeth of the detection rotor and a natural number of 5 or more. Therefore, for example, when the number of teeth of the detection rotor W1 is 35 (divisors 1, 5, 7, 35), n may be set to 5. When the number of teeth is 28 (divisors 1, 2, 4, 7, 14, 28), n may be set to 7.

In any case, the computing unit 72 rotates the rotary table 2 by [360 / n] ° to acquire the X-direction displacement X (θ) and the Y-direction displacement Y (θ) n times. For example, when the number of teeth of the detection rotor W1 is 5 and n = 5, the computing unit 72 rotates the rotary table by 72 ° and changes the X-direction displacement X (θ) and the Y-direction displacement Y (θ) by 5 degrees. Get it twice. In the following, the X-direction displacement X (θ) and the Y-direction displacement acquired at the k-th time (k = 1, 2,..., N), that is, at the rotation angle θ = k × (360 / n) ° Y (θ) is denoted as X _k and Y _k , respectively.

If n X-direction displacements _Xk and Y-direction displacements _Yk can be obtained, the arithmetic unit 72 calculates the amplitude A and the phase α of the fluctuation component E (θ) due to the eccentricity based on Expression 1-8. . In Expressions 4 and 8, Atan2 (x, y) means an angle between the x-axis and a line passing through the origin (0, 0) and coordinates (x, y).

Here, the calculation of Expressions 1 and 2 cancels out the fluctuation component e1 (θ) caused by the perfect circular error and the fluctuation component e2 (θ) caused by the unevenness. That is, when the variation component E (θ) included in X _k is Ek, the value of Expression 1 is equivalent to Σ (Ek × sin (360 / n) × k). The value of Expression 2 is equivalent to Σ (Ek × cos (360 / n) × k). Similarly, in the calculations of Equations 3 and 4, the fluctuation component e1 (θ) caused by the perfect circular error and the fluctuation component e2 (θ) caused by the unevenness are canceled.

  The arithmetic unit 72 substitutes XSS and XCS, from which the effects of the fluctuation component e1 (θ) and the fluctuation component e2 (θ) have been removed, into Expressions 3 and 4, thereby obtaining the fluctuation included in the X-direction displacement X (θ). The amplitude A and the phase difference α of the component E are calculated. Similarly, by substituting YSS and YCS from which the effects of the fluctuation component e1 (θ) and the fluctuation component e2 (θ) have been removed into Equations 7 and 8, the fluctuation component E included in the Y-direction displacement Y (θ) is obtained. Is obtained.

  The fluctuation component E (θ) included in the Y-direction displacement Y (θ) is a signal having a phase shifted by 90 degrees from the fluctuation component E (θ) included in the X-direction displacement (θ). Are the same. Therefore, the value of Expression 3 indicating the amplitude A of the fluctuation component E (θ) included in the X-direction displacement X (θ) and the amplitude A of the fluctuation component E (θ) included in the Y-direction displacement Y (θ) are shown. The value of equation 7 is theoretically the same. Further, the fluctuation component E (θ) included in the X-direction displacement X (θ) and the fluctuation component E (θ) included in the Y-direction displacement (θ) are out of phase by 90 degrees, so that their phase differences are different. PX and PY are shifted by 90 degrees in theory, and PX = (PY−90 °) is satisfied.

The eccentric amount D (θ) is a combination of the amplitude A = 2 · RX calculated from the X-direction displacement X _k and the phase difference α = PX, or the amplitude A = 2 · RY and the phase difference calculated from the Y-direction displacement Y _k It can be obtained from the combination of α = PY. In this example, in order to cancel the measurement error and the like included in the X-direction displacement and the Y-direction displacement, all of these four numerical values RX, PX, RY, and PY are used, and the eccentric amounts DX (θ) and DY (θ) are used. Is calculated. Specifically, the amount of eccentricity DX (θ) in the X direction at the angle θ is obtained by the following equation 9, and the amount of eccentricity DY (θ) in the Y direction is obtained by the following equation 10.

However, in theory, RY = RX and PY = PX + 90 °. Therefore, at least one set of RX and PX and RY and PY may be calculated, and the eccentricities DX (θ) and DY (θ) may be calculated from the one set. For example, DX (θ) and DY (θ) may be calculated by the following Expressions 11 and 12 without calculating RX and PY.
DX (θ) = (RX * sin (RY + θ)) Equation 11
DY (θ) = DX (θ + 90 °) Equation 12

  In the actual control, the computing unit 72 obtains the X-direction eccentricity DX (0) and the Y-direction eccentricity DY (0) at the rotation angle θ = 0. When the amount of eccentricity DX (θ) and DY (θ) can be calculated, the arithmetic unit 72 sends a command to the solenoid actuators 31, 32, 33, and 34 via the insulating amplifier 74 to cause the drum 50 to operate the solenoid actuators 31, 32, and. The detection rotor is indirectly moved by DX (0) in the X direction and DY (0) in the Y direction by hitting the movable parts 33 and 34.

  After hitting the drum 50, the rotary table 2 is rotated to the position of θ = 0, and the distance to the detection rotor W1 is measured again using the centering center detection sensor 30. As a result of the measurement, if the amount of movement of the detection rotor W1 has not reached or exceeded the amount of eccentricity DX (θ), DY (θ), a command is issued to the solenoid actuators 31, 32, 33, 34 again. The drum 50 is fed, and the drum 50 is hit with the movable portions of the solenoid actuators 31, 32, 33, 34. On the other hand, if the detection rotor W1 has moved by the eccentricities DX (θ) and DY (θ) described above, the process ends.

Here, in this example, the X-direction displacement X (θ) and the Y-direction displacement Y (θ) are acquired n times for each rotation angle of [360 / n] °. In this example, the numerical value n is a divisor of the number of teeth of the detection rotor W1 and a natural number of 5 or more. The reason for this configuration will be described with reference to FIGS. As described above, FIGS. 4 to 6 show the X-direction displacement X (θ) when the detection rotor W1 having five teeth is measured, and the fluctuation component E (θ) included in the X-direction displacement X (θ). , E1 (θ), and e2 (θ). A point _X 1 to X in FIG. _{4. 5,} respectively, represents the value of the obtained X-direction displacement _{X k} for each 72 °, Figure 5, the point _e1 1 to E1 ₅ in FIG. 6, the point e2 ₁ to e2 _5, the variation component included in the X-direction displacement _{X k} e1 (theta), which indicates the value of e2 (theta).

As shown in FIG. 6, the variation component e2 (theta), in order to vary with the same period as the teeth of the detection rotor W1, included in the X-direction displacement X _k to be read for each angle having equal intervals about the number of the number of teeth The values of the fluctuation components e2 _k are all the same. Therefore, in this case, the fluctuation component e2 _k can be treated as a mere offset, and in the calculation of the amplitude A and the phase α, the influence of the unevenness formed on the outer peripheral end of the detection rotor W1 can be ignored.

Further, the X-direction displacement X (θ) includes not only the fluctuation component e2 (θ) but also the fluctuation component e1 (θ) resulting from the roundness error of the detection rotor W1. The variation component e1 (θ) differs depending on the detection rotor W1, but often includes only two-cycle components and three-cycle components. Such two-period components and three-period components have symmetry when sampled at angles obtained by dividing 360 ° by a natural number of 5 or more. For example, in the example of FIG. 4, a _{e1 1 = -1 × E1 4,} e1 2 = -1 × e1 3 next, e1 _{5 is, (e1 1 -e1 4) /} 2 _{and,, (e1} 2 -e1 ₄ ) / 2. As described above, since the fluctuation component e1 _k has symmetry, it is possible to cancel the fluctuation component e1 _k by performing the operations of Expressions 1 and 2 or Expressions 5 and 6. Such cancellation is possible only when n is set to 5 or more. Therefore, in this example, n is set to 5 or more.

  As is apparent from the above description, according to the eccentricity calculating apparatus of the present example, the influence of the roundness error of the detection rotor W1 and the irregularities (teeth) of the detection rotor W1 can be offset in the process of calculating the eccentricity. As a result, the amount of eccentricity of the detection rotor can be calculated more accurately.

  In the above description, the rotary table 2 may have another configuration as long as the centering center detection sensor 30 can be rotated relatively to the detection rotor W1. For example, a rotation table that disables the centering center detection sensor 30 and rotates the rotation position detector (detection rotor W1) and the solenoid actuators 31, 32, 33, and 34 may be provided.

W1 detection rotor, W2 housing, W3 rotary shaft, W4, W5 bearing, 2 rotary table, 7 base, 8 fixed shaft, 28 columns, 30 centering center detection sensor, 31, 32, 33, 34 solenoid actuator, 41, 42, 43, 44 props, 51 air cylinders, 61, 62, 63, 64 forward and backward cylinders, 50 drums, 72 computing units, 73 variable power supply, 74 isolation amplifier.

Claims (3)

In the manufacturing process of a rotation position detector having a detection rotor provided with a plurality of teeth on its peripheral surface, a rotation shaft that rotates together with the detection rotor, and a sensor unit that detects a rotation angle of the detection rotor, An eccentricity calculation device for a detection rotor that calculates an eccentricity of the detection rotor with respect to the rotation shaft portion,
A center detection sensor provided with four detection units that output a signal corresponding to the distance to the peripheral surface of the detection rotor at 90-degree intervals;
A rotation table for rotating the center detection sensor relative to the detection rotor,
A calculator that calculates an eccentric amount of the detection rotor with respect to the rotation shaft portion based on an output signal from the center detection sensor;
With
The center detection sensor has two X-direction detectors arranged 180 degrees opposite to each other, and two Y-direction detectors arranged with a phase shift of 90 degrees with respect to the X-direction detector, The difference between the output signals of the two X-direction detectors is set as the X-direction displacement, and the difference between the output signals of the two Y-direction detectors is set as the Y-direction displacement. Where n is a natural number that is a divisor of 5 or more, the X-direction displacement and the Y-direction displacement read out at every [360 / n] ° angle by rotating the turntable by [360 / n] °. Calculating the amount of eccentricity of the detection rotor based on at least one of
An eccentricity calculation device for a detection rotor, characterized in that: The eccentricity calculation device for a detection rotor according to claim 1,
When the X-direction displacement and the Y-direction displacement read out at the k-th time are represented by X _k and Y _k , the arithmetic unit D (θ) = A × sin (θ + α) which varies sinusoidally with rotation. And calculating the amplitude A and phase α of the detection rotor based on at least one of Expressions 1-4 and 5-8.
An eccentricity calculation device for a detection rotor according to claim 2,
The arithmetic unit calculates the amount of eccentricity in the X direction DX (θ) and the amount of eccentricity in the Y direction DY (θ) at the rotation angle θ based on Equation 9-10.