JP2009041983A - Method for detecting variation of zero-point error of multi-point probe - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for detecting variations of zero-point errors of multi-point probes, capable of easily compensating for zero-point errors, after drifts have occurred in a plurality of sensors. <P>SOLUTION: In the method for detecting variations of zero-point errors of multi-point probes, a sensor unit 220 is arranged, in such a way as to be opposed to the surface to be measured of a sample, in a state of rotating about a spindle, and a CPU 310 reads shape values along a concentric circle in the surface to be measured of the sample and measured values that include zero-point errors outputted from a plurality of displacement sensors A, B, C whenever the sample is rotated by a prescribed angle of rotation during one rotation of the sample and stores them in a storage device 340. On the basis of the new measured values and old measured values stored in the storage device 340 relating to the same sample, the amount of variations in the zero-point errors is computed. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a zero point error variation detection method for a multipoint probe.

  In a surface shape measuring apparatus that uses three displacement sensors to measure the surface shape of a measurement object sequentially using the three-point method, there is a zero point error as a measurement error factor that occurs between the displacement sensors. As shown in FIG. 1, when any one of the plurality of displacement sensors A to C is used as a reference, the zero point error corresponds to the zero point of the reference displacement sensor (displacement sensor B in FIG. 1) and the other displacement sensors A and C. This is the deviation of the zero point. Usually, in order to adjust the zero points of a plurality of displacement sensors, adjustment is made so that the reading becomes zero by applying to a plane which is a linear reference surface as a reference.

Alternatively, the zero point error is calculated using an inversion method or the like, and the zero point error is compensated.
However, once the zero point error is resolved or the zero point error is compensated by using an inversion method or the like, if the time has elapsed, the mounting position of the displacement sensor may drift between the three displacement sensors due to drift or the like. Zero error may fluctuate.

  Conventionally, when the zero point error between the sensors of the multipoint probe fluctuates due to the passage of time in this way, the zero point error is adjusted again, or the zero point error is compensated by performing an inversion method or the like. Therefore, there is a problem that work for eliminating the zero error is complicated.

  SUMMARY OF THE INVENTION An object of the present invention is to provide a zero point error variation detection method for a multipoint probe that can easily compensate for zero point errors after a plurality of sensor drifts have been solved. .

  In order to solve the above-mentioned problems, the invention of claim 1 is configured such that a multipoint probe for measuring a straight shape including a plurality of sensors is opposed to a measurement surface of an object to be measured being rotated by a rotating means. The measurement object is arranged by the multipoint probe support means and is output from the plurality of sensors each time the measurement object rotates by a predetermined rotation angle during at least one rotation of the measurement object. The measurement value including the shape value along the concentric circle on the measurement surface and the error based on the zero point error of the probe is read by the reading means and stored in the storage means, and the new measurement value relating to the same measurement object stored in the storage means Calculating the amount of variation of the zero error based on (hereinafter referred to as the first measurement value) and the measurement value (hereinafter referred to as the second measurement value) whose acquisition time is older than the acquisition time of the first measurement value. It is an gist variation detecting method in the zero point error of the multi-point probe according to symptoms.

  According to the invention of claim 1, since the fluctuation amount of the zero point error is calculated based on the first measurement value and the second measurement value, the zero point after the drift of the plurality of sensors is generated based on the fluctuation amount. Error compensation can be easily performed.

  According to a second aspect of the present invention, in the first aspect, the multipoint probe support means is provided so as to be movable in a radial direction of the rotating means, and the measurement value includes the multipoint probe support means in the radial direction. It includes an average value of values obtained when the measurement object is moved to a plurality of positions and the measurement object is rotated at least once or more by the rotation means at each position.

  According to the invention of claim 2, when the multipoint probe moves in the radial direction of the rotating means and measures the measurement surface at a plurality of positions, and the average of the measured values is taken, a chance error included in the measured values Less.

  Further, since the measured value is an average value, the shape value of the measurement surface remains unchanged. For example, even if the displacement is measured by a three-point probe or a two-point probe, if the displacement can be measured again at the same position, the zero-point error can be stored and compensated. However, for example, if the first and second circumferential measurement start positions cannot be measured at the same position, it becomes a memory / compensation error. However, in the second aspect of the invention, even if the measurement start position in the circumferential direction is deviated, the measurement value is an average value, and the shape value of the measurement surface remains unchanged. As a result, the zero point error can be stored and compensated, and the fluctuation amount of the zero point error can be calculated based on the old measured value (average value) and the new measured value (average value).

  According to a third aspect of the present invention, in the second aspect of the present invention, the plurality of sensors supported by the multipoint probe support means are three displacement sensors installed apart in the moving direction of the multipoint probe support means. .

According to the third aspect of the present invention, the amount of fluctuation of the zero point error in the three displacement sensors of the multipoint probe can be easily obtained.
According to a fourth aspect of the present invention, in the second aspect of the present invention, the plurality of sensors supported by the multipoint probe support means are two two-dimensional angle sensors installed apart in the moving direction of the multipoint probe support means. And

  According to the invention of claim 4, the amount of fluctuation of the zero point error in the two-dimensional angle sensor having two multi-point probes can be easily obtained.

  According to the first aspect of the present invention, the fluctuation amount of the zero point error of the multipoint probe can be easily obtained. As a result, the zero point error after the drift of the plurality of sensors is generated based on the fluctuation amount. It can be done easily.

  According to the invention of claim 2, when the multipoint probe moves in the radial direction of the rotating means and measures the measurement surface at a plurality of positions, and the average of the measured values is taken, a chance error included in the measured values Less. Further, since the measurement value is an average value, the shape value of the measurement surface can be made unchanged. That is, even if the measurement start position in the circumferential direction is deviated, the measurement value is an average value, so the shape value of the measurement surface remains unchanged. As a result, the zero point error can be stored and compensated, and the fluctuation amount of the zero point error can be calculated based on the old measured value (average value) and the new measured value (average value).

According to the invention of claim 3, the multipoint probe can easily obtain the variation amount of the zero point error in the three displacement sensors.
According to the invention of claim 4, the multipoint probe can easily determine the fluctuation amount of the zero point error in the two two-dimensional angle sensors.

(First embodiment)
A multipoint probe zero-point error related value recording apparatus (hereinafter simply referred to as a zero-point error related value recording apparatus) that can realize the zero-point error variation detection method of the multipoint probe of the present invention will be described below with reference to FIGS. explain.

  As shown in FIG. 2, a sample 200 as an object to be measured is mounted on a spindle 210 and is rotatable around the rotation axis (that is, the z axis) of the spindle 210. In the present embodiment, the upper surface of the sample 200 is the measurement surface 202. The sample 200 preferably has a sufficient thickness such that the shape of the measurement surface 202 does not change, and the measurement surface 202 is preferably a flat surface. Moreover, although the sample 200 is formed in a disk shape in this embodiment, it may be a square plate shape, and the shape is not limited. In short, it is only necessary to provide an area that can be measured by a multipoint probe for measuring a straight shape when the sample 200 is rotated by the spindle 210.

  A sensor unit 220 as a multipoint probe for measuring a straight shape is movable in the x direction, that is, the radial direction of the spindle 210 by a linear sensor carriage (hereinafter referred to as sensor carriage 230), and the sample 200 serves as a rotating means. When rotating by the spindle 210, the sample surface is scanned. The sensor carriage 230 corresponds to multipoint probe support means.

In the sensor unit 220, three displacement sensors A, B, and C are installed via the intervals d ₁ and d ₂ along the x direction, that is, along the moving direction of the sensor carriage 230.

Note that d ₁ = r _F −r ₀ and d ₂ = r ₀ −r _R (see FIG. 2).
In the present embodiment, the displacement sensors A, B, and C are non-contact sensors, and are composed of capacitive sensors. The displacement sensor is not limited to a non-contact type capacitive sensor, and may be a contact type. Alternatively, in the case of a non-contact type, for example, an optical sensor may be used instead of the capacitive type sensor. Also good.

  In the zero error related value recording device, the output from each displacement sensor A, B, C is converted into a digital value by A / D conversion or the like (not shown) and input to a computer system 300 such as a personal computer. The computer system 300 includes a CPU 310, a RAM 320, and a ROM 330. The CPU 310 corresponds to a reading unit and a calculating unit. A storage device 340 composed of a hard disk or the like is connected to the computer system 300, and measured values from the displacement sensors A, B, and C are stored in a readable manner.

In this specification, the zero-point error related value includes a value for calculating the zero-point error, a zero-point error fluctuation amount, and a zero-point error.
(Operation of the first embodiment)
Now, the shape on the circumference of the radius r _F , r ₀ , r _R of the measurement surface 202 in the sample 200 is f (r _F , θ), f (r ₀ , θ), f (r _R , θ), each The outputs (that is, measured values) of the displacement sensors A, B, and C are S _F , S ₀ , and S _R. Further, assuming that e _z (θ) and Φ _y (θ) are movement errors in the z-axis direction of the sample 200 and movement errors caused by the inclination of the rotation axis of the spindle 210 around the y-axis, the displacement sensors A, B, and C S _F , S ₀ , S _R can be expressed by the following equations (1) to (3). Here, α _F in the equation (1) is a zero point error of the displacement sensor A with the displacement sensor B as a reference. In addition, α _R in the equation (3) is a zero point error of the displacement sensor C with respect to the displacement sensor B. In addition, f (r _F , θ), f (r ₀ , θ), and f (r _R , θ) in the equations (1) to (3) are the shape values along the concentric circles on the measurement surface of the measurement object. Equivalent to.

変位センサＡ，Ｂ，Ｃがｙ軸上において、半径ｒ_F，r₀，r_Rにそれぞれ位置するときに、スピンドル２１０により試料２００を回転中心（すなわち、ｚ軸）の周りで１回転させた場合、この１回転中において、零点誤差関連値記録装置は所定回転角度経過する毎に、変位センサＡ，Ｂ，Ｃの出力Ｓ_F，Ｓ₀，Ｓ_R（すなわちデータ）を入力する。そして、ＣＰＵ３１０はこの測定値を記憶装置３４０に格納する。なお、所定回転角度は、例えば１度であってよく、０．５度でもよく、限定されるものではない。θは、原点位置からの回転角度である。

When the displacement sensors A, B, and C are located at radii r _F , r ₀ , and r _R on the y-axis, the sample 200 is rotated once around the rotation center (ie, the z-axis) by the spindle 210. In this case, during this one rotation, the zero point error related value recording apparatus inputs the outputs S _F , S ₀ , S _R (that is, data) of the displacement sensors A, B, C every time a predetermined rotation angle elapses. Then, the CPU 310 stores this measurement value in the storage device 340. The predetermined rotation angle may be, for example, 1 degree or 0.5 degree, and is not limited. θ is the rotation angle from the origin position.

The rotation angle θ and the position of the sensor carriage 230 on the x-axis can be detected by a position sensor such as a linear encoder or a rotary encoder (not shown).
For example, the position of the displacement sensor B on the x-axis is detected by the linear encoder, and the positions of the displacement sensors A and C are calculated by the CPU 310 based on the detected position of the displacement sensor B and d ₁ and d _2. .

(First measurement and recording)
In the first recording, the sample 200 is rotated once, the subscript “1” is added to the displacement output from each of the displacement sensors A, B, and C, and the output of the displacement sensors A, B, and C (that is, measured values). , S _F , S ₀ , S _R can be expressed by equations (4) to (6).

Here, the first zero error is first determined by a known method and compensated, and the zero errors of the displacement sensors A to C that can measure the surface shape of the sample 200 are α _F1 and α _R1. Let it be known.

試料２００が１回転された後、コンピュータシステム３００のＣＰＵ３１０は、１回転している際に入力された変位センサＡ，Ｂ，Ｃから得られた各測定値の平均値を算出する。

After the sample 200 has been rotated once, the CPU 310 of the computer system 300 calculates an average value of the respective measurement values obtained from the displacement sensors A, B, and C that are input during the one rotation.

  That is, if the predetermined rotation angle is 1 degree, the average value of the surface shape obtained by the displacement sensor A is obtained by summing up the data of the displacement sensor A and dividing by 360. Similarly, the average value of the surface shapes obtained by the displacement sensor B is obtained for the displacement sensor B.

The average values FF1, F01, FR1 of f (r _F , θ), f (r ₀ , θ), f (r _R , θ) are unchanged as long as the shape of the sample 200 is not changed. The average values of the motion errors e _z (θ) and Φ _y (θ) are EZ1 and ΦY1, respectively, and the average values of S _F1 (θ), S ₀₁ (θ), and S _R1 (θ) are SF1, S01, and SR1, respectively. Then, Expressions (4) to (6) become Expressions (7) to (9).

次に、ＣＰＵ３１０は式（７）−式（８）を演算すると、ｚ方向の運動誤差が消去されて式（１０）となる。

Next, if CPU310 calculates Formula (7)-Formula (8), the motion error of az direction will be erase | eliminated and it will become Formula (10).

次に、ＣＰＵ３１０は式（８）−式（９）を演算すると、同じくｚ方向の運動誤差が消去されて式（１１）となる。

Next, if CPU310 calculates Formula (8) -Formula (9), the motion error of az direction will be erase | eliminated similarly to Formula (11).

又、ＣＰＵ３１０は、式（１０）×（r₀−r_R）/( r_F−r0)を演算すると、式（１２）が得られる。

Further, when the CPU 310 calculates the equation (10) × (r ₀ −r _R ) / (r _F −r0), the equation (12) is obtained.

次に、ＣＰＵ３１０は式（１２）−式（１１）を演算すると、運動誤差解消値を表わす式（１３）を得る。

Next, CPU 310 calculates equation (12) -equation (11) to obtain equation (13) representing the motion error cancellation value.

ここで、式（１３）では、回転軸のｙ軸周りの傾きによる運動誤差も消去されている。又、式（１３）において、左辺は表面形状の２回差分に関するものであり、右辺は「測定して得られた値＋零点誤差」に関するものとなる。

Here, in Equation (13), the motion error due to the inclination of the rotation axis around the y-axis is also eliminated. In equation (13), the left side relates to the difference between the surface shapes twice, and the right side relates to “value obtained by measurement + zero error”.

In this way, the CPU 310 performs the first measurement value in the storage device 340, performs the above-described calculation for measurement, and stores the calculation result in the storage device 340.
This first measurement value corresponds to the second measurement value.

(Second measurement and recording)
After the first measurement, the zero point error of the displacement sensors A, B, C varies due to temperature drift or the like. Accordingly, when the surface shape of the workpiece is newly measured after the first measurement and recording, in the second measurement, the CPU 310 of the computer system 300 detects the displacement sensor input during one rotation of the sample 200. The average value of each measured value obtained from A, B, and C is calculated in the same manner as “first measurement and recording”. Then, the CPU 310 stores the calculation result and the like in the storage device 340. This second measurement value corresponds to the first measurement value.

  Here, for the convenience of explanation, detailed explanation is omitted, but in the equations (7) to (12), the subscript “2” is added to each variable explained in the first measurement instead of the subscript “1”. If expressed, the result of “second measurement” is obtained similarly. Also, the equation (13) obtained in the “first measurement” column represents the motion error elimination value in the second measurement as the equation (14).

１回目の運動誤差解消値を表わす式（１３）と２回目の運動誤差解消値を表わす式（１４）を比較すると、同一の試料２００を測定しているため、その表面形状は変化しないとしてもよいから、ＦＦ２＝ＦＦ１，Ｆ０２＝Ｆ０１，ＦＲ２＝ＦＲ１であり、右辺第１項は同じになる。従って、ＣＰＵ３１０は、式（１４）−式（１３）をＣＰＵ３１０は演算することにより、零点誤差の変動量を算出する。式（１５）は式（１４）−式（１３）で演算された零点誤差の変動量を求めるための式である。

Comparing equation (13) representing the first motion error cancellation value and equation (14) representing the second motion error cancellation value, since the same sample 200 is measured, the surface shape may not change. Therefore, FF2 = FF1, F02 = F01, FR2 = FR1, and the first term on the right side is the same. Therefore, the CPU 310 calculates the fluctuation amount of the zero point error by calculating the expression (14) -expression (13). Expression (15) is an expression for obtaining the fluctuation amount of the zero point error calculated by Expression (14) -Expression (13).

Note that d ₁ = r _F −r ₀ and d ₂ = r ₀ −r _R.

ここで、式（１５）中、ｒ_F，r₀，r_Rは、前述したように図示しないリニアエンコーダの検出値及び、ｄ_１，ｄ_２に基づいて得られる既知の値であり、演算結果として記憶装置３４０に格納されている。又、ＳＦ１，ＳＦ２，Ｓ０１，Ｓ０２，ＳＲ１，ＳＲ２は第１回目と第２回目の算出値であり、演算結果として記憶装置３４０に格納されている。

Here, in Expression (15), r _F , r ₀ , and r _R are detection values of a linear encoder (not shown) and known values obtained based on d ₁ and d ₂ as described above, and the calculation result Is stored in the storage device 340. SF1, SF2, S01, S02, SR1, SR2 are calculated values for the first time and the second time, and are stored in the storage device 340 as calculation results.

  When the fluctuation amount of the zero point error is obtained in this way, the first zero point error is already known. Therefore, the zero point error fluctuation amount is simply added to or subtracted from the first zero point error which is a known value. Thus, a new compensation amount for the zero point error after the sensor drift is obtained. In this way, it is possible to compensate for the zero point error based on the fluctuation amount of the zero point error.

When the distances d ₁ and d ₂ are the same, that is, when the displacement sensors A, B, and C are equally spaced, the equation (15) is simply expressed by the equation (16).

このように、変位センサＡ，Ｂ，Ｃが等距離に離間している場合は、零点誤差の変動量を容易に求めることができる。

As described above, when the displacement sensors A, B, and C are separated by an equal distance, the fluctuation amount of the zero point error can be easily obtained.

  As described above, when there is a variation in the zero point error of the displacement sensor due to the drift, the amount of variation of the zero point error can be obtained by the above-described zero point error related value recording device, and high-precision zero compensation can be performed.

  Therefore, for example, a zero point error related value recording device is provided adjacent to the shape measuring device, and the shape measurement device can measure the shape of the workpiece and the shape measurement of the measurement surface 202 of the sample 200 on the spindle 210. It is preferable that the sensor unit 220 can be moved by the sensor carriage 230 in a possible area.

The first embodiment has the following features.
(1) In the multipoint probe zero point error variation detection method of the present embodiment, a sample 200 (measurement object) having a measurement surface 202 is rotated by a spindle 210 (rotating means), and a plurality of displacement sensors A, B are rotated. , C including a straight shape measurement sensor unit 220 (multi-point probe) is disposed so as to face the measurement surface of the sample 200 being rotated by the spindle 210. The shape value and the zero point along the concentric circle on the measurement surface 202 of the sample 200 that are output from the plurality of displacement sensors A, B, and C each time the sample 200 rotates by a predetermined rotation angle during one rotation of the sample 200. The measurement value including the error is read by the CPU 310 (reading means) and stored in the storage device 340 (storage means). Then, based on the new measurement value (first measurement value) related to the same sample 200 stored in the storage device 340 and the measurement value (second measurement value) whose acquisition time is older than the acquisition time of the first measurement value. The variation of the zero point error is calculated.

  As a result, the fluctuation amount of the zero point error of the displacement sensors A, B, and C in the sensor unit 220 can be easily calculated, and the fluctuation amount of the zero point error of the displacement sensors A, B, and C of the sensor unit 220 can be easily obtained. Therefore, it is possible to easily compensate for the zero point error after the drift of the plurality of sensors occurs based on the fluctuation amount.

  (2) In the zero point error variation detection method of the multipoint probe of the present embodiment, the zero point error related value recording device rotates the sample 200 once by the spindle 210 for the first measurement value and the second measurement value. It was made to include the average value of the values obtained at the time. As a result, in the present embodiment, since the first measurement value and the second measurement value obtained by measuring the same sample 200 are average values, the shape value of the measurement surface 202 is not changed. This is because the measurement surface 202 is an average value of the measurement values obtained when the measurement start position (that is, the measurement start point) in the circumferential direction is shifted from the measurement start point by one round. It is used that the shape value of is invariant. As a result, there is no influence of the shape value when calculating the fluctuation amount of the zero point error.

  (3) In the zero point error variation detection method of the present embodiment, a plurality of displacement sensors A, B, and C supported by the sensor carriage 230 (multi-point probe support means) are installed apart in the movement direction of the sensor carriage 230. Yes. As a result, the sensor unit 220 (multi-point probe) can easily obtain the amount of variation of the zero point error in the three displacement sensors, and the amount of variation with respect to the zero point error before the known drift occurs can be easily obtained. The zero point error after the sensor drift has occurred can be calculated and the zero point error can be easily compensated.

(Second Embodiment)
Next, a zero point error related value recording apparatus according to a second embodiment capable of realizing a zero point error variation detection method of a multipoint probe will be described with reference to FIGS. In addition, the same code | symbol is attached | subjected about the structure which is the same as that of 1st Embodiment, or corresponds. The second embodiment differs from the first embodiment in that the sensor unit 220, which is a multipoint probe, includes two angle sensors P and Q.

  As shown in FIG. 4, the sample 200 is mounted on the spindle 210 and is rotatable around the rotation center (that is, the z axis). The sensor unit 220 is moved in the x direction by the sensor carriage 230 and scans the measurement surface 202 of the sample 200 when the sample 200 is rotated by the spindle 210. In the sensor unit 220, two two-dimensional angle sensors (hereinafter referred to as angle sensors) P and Q are installed with a gap d along the x direction, that is, along the moving direction of the sensor carriage 230. ing. The configuration of the sensor unit 220 will be described in detail later. In FIG. 3, the circular sample 200 is taken as an example, but the shape of the sample 200 is not limited to a circular shape, and may be another shape.

  In the zero error related value recording apparatus, the outputs from the angle sensors P and Q are converted into digital values by A / D conversion or the like (not shown) and input to a computer system 300 such as a personal computer.

  Since the angle sensors P and Q have the same configuration, the angle sensor Q will be described. The angle sensor Q includes a laser light source 231, a beam splitter 232, and a PSD (two-dimensional semiconductor position detection element) 233. The output of the PSD 233 is output to the CPU 310 of the computer system 300 for processing. In the angle sensor Q, the laser beam emitted from the laser light source 231 is reflected by the beam splitter 232 and travels toward the sampling point of the sample 200 via the objective lens 234. The laser light reflected at the sampling point passes through the objective lens 234 and the beam splitter 232 and reaches the PSD 233.

  Here, when the sampling point is orthogonal to the incident laser light, the reflected light is incident on the center of the PSD 233. However, if the sampling point is inclined, the laser light reflected at the sampling point is incident on a position shifted from the center of the PSD 233, so that an electrical signal corresponding to the difference is output as a measurement value to the CPU 310, thereby sampling. The tilt angle at the point, that is, the local two-dimensional tilt angle can be measured.

Further, details of the measurement principle of the two-dimensional angle sensor will be described. FIG. 6 shows a principle diagram of a two-dimensional angle sensor.
The two-dimensional angle sensor is basically for detecting a two-dimensional inclination of the surface of the sample 200 or a workpiece (not shown). By narrowing the incident beam, it can be used to detect a two-dimensional local slope of the surface. In order to detect a two-dimensional inclination, a method called a light lever shown in FIG. 6 is known as the simplest method.

  In FIG. 6, a laser beam 250 is incident on the surface of the sample 200 along the Z-axis direction. When the surface of the sample 200 is inclined, the position of the light spot of the reflected light beam on the PSD 233 changes according to the inclination angle. By detecting the coordinate change amounts Δx and Δy of the light spot, the inclination angle changes Δα and Δβ around the x and y axes can be obtained from the following equations.

なお、Ｌは試料２００の面からＰＳＤ２３３までの距離である。

Note that L is the distance from the surface of the sample 200 to the PSD 233.

This method has a problem that an error occurs in the measurement result when the distance L from the surface of the sample 200 to the PSD 233 changes.
Therefore, as shown in FIG. 7, a collimating lens (that is, the objective lens 234) is inserted between the sample 200 and the PSD 233, and angle detection is performed by the principle of autocollimation. As shown in FIG. 7, when the objective lens 234 is placed between the sample 200 and the PSD 233, the relationship between the sample surface inclination and the light spot coordinates on the PSD 233 depends only on the focal length F of the objective lens 234. The inclination detection accuracy can be improved. The inclination angle in this case is expressed as the following equation.

本実施形態の零点誤差関連値記録装置に使用する２次元角度センサでは、このオートコリメーション方式が採用されている。

The autocollimation method is employed in the two-dimensional angle sensor used in the zero error related value recording apparatus of the present embodiment.

Here, how to measure the height shape of the surface of the workpiece using the angle sensors P and Q will be described below. In the description here, the sample 200 is treated as a work for convenience of explanation.
As shown in FIG. 4, the sample 200 is mounted on a spindle 210 and is rotatable. The sensor unit 220 can be moved in the x direction by the sensor carriage 230 as shown in FIG. 5, and scans the surface of the sample 200 when the sample 200 is rotated by the spindle 210. In the sensor unit 220, two angle sensors P and Q) are disposed with a distance d along the x direction.

When measuring the surface shape, the outputs from the angle sensors P and Q are input to the computer system, and the calculation described below is performed.
FIG. 5 is a diagram showing sampling points of sample surface data measured by the sensor unit 220. As shown in FIG. 5, scanning in the x direction starts from the rotation center of the sample 200, and each sampling point (ie, sampling position) is x _i (i = 1, 2,..., M). The interval d between the two angle sensors P and Q and the sampling interval s are the same in the radial direction. In the sampling point x _i, two concentric circles on the sample 200 has two angle sensors P in the sensor unit 220 is scanned by Q. The sampling position on the circumference is θ _j (j = 1, 2,..., N). Outputs in the x direction of sensor unit 220 (corresponding to local slopes around the y axis) μ _Py (x _i , θ _j ) and μ _Qy (x _i , θ _j ) are expressed as follows. The sampling position (x _i , θ _j ) can be detected by a position sensor such as a linear encoder or a rotary encoder (not shown).

なお、ｉ＝１，２，…，Ｍ−１，ｊ＝１，２，…，Ｎである。又、ｅ_cＸ（ｘ_ｉ）は、センサキャリッジ２３０のロール誤差、ｅ_ＳＸ（ｘ_ｉ，θ_ｊ）はスピンドル２１０のｘ軸回りのアンギュラ誤差であり、ともに運動誤差である。α_Ｐとα_Ｑは角度センサＰ，Ｑのオフセットであり、真平らな平面を測定したときに角度センサＰ，Ｑに現れる出力である。ｆ'ｙ（ｘ，θ）は試料表面のｙ軸局部スロープであり、次のように定義される。

Note that i = 1, 2,..., M−1, j = 1, 2,. Further, e _cX (x _i ) is a roll error of the sensor carriage 230, and e _SX (x _i , θ _j ) is an angular error around the x axis of the spindle 210, both of which are motion errors. α _P and α _Q are offsets of the angle sensors P and Q, and are outputs that appear in the angle sensors P and Q when a completely flat plane is measured. f′y (x, θ) is the y-axis local slope of the sample surface and is defined as follows.

そして、センサキャリッジ２３０及びスピンドル２１０の運動誤差を取り除いた差動出力Δμｙ（ｘ_ｉ，θ_ｊ）は式（２２）から式（２１）を引くことにより次のようになる。

Then, the differential output Δμy (x _i , θ _j ) obtained by removing the motion error of the sensor carriage 230 and the spindle 210 is as follows by subtracting the equation (21) from the equation (22).

ここで、ｉ＝２，３，…，Ｍ−１，ｊ＝１，２，…，Ｎである。

Here, i = 2, 3,..., M−1, j = 1, 2,.

Then, for a fixed θj (j = 1, 2,..., N), f ′ _y (x _i , θ _j ) is obtained from the integration of Δμ _y (x _i , θ _j ) as follows.

ただし、ｉ＝２，３，…，Ｍ−１，ｊ＝１である。

However, i = 2, 3,..., M−1, j = 1.

In the fixed x _i (i = 2, 3,..., M), the height shape f (x _i , θ _j ) on the i-th concentric circle of the sample 200 is f ′ _y (x _i , θ _j ). It can be obtained from integration. Since the method for obtaining the integral of f ′ _y (x _i , θ _j ) is known, the description thereof is omitted here.

(Operation of Second Embodiment)
Now, the operation of the second embodiment will be described.
(First measurement and recording)
First, when the angle sensors P and Q of the sensor unit 220 are positioned at the sampling points x _i and x _{i + 1} , the sample 200 is rotated once around the rotation center (that is, the z axis) by the spindle 210. During this one rotation, the dew point recording device inputs the measured values of the angle sensors P and Q each time a predetermined rotation angle elapses, and stores them in the storage device 340 as storage means. The predetermined rotation angle may be, for example, 1 degree or 0.5 degree, and is not limited.

These angles and sampling positions can be detected by a position sensor such as a linear encoder or a rotary encoder (not shown) as in the first embodiment.
After one rotation, the CPU 310 of the computer system 300 calculates an average value of each measurement value obtained from the angle sensors P and Q input during one rotation.

  That is, if the predetermined rotation angle is 1 degree, the average value of the surface shape obtained by the angle sensor P is obtained by adding the data of the angle sensor P and dividing by 360. Similarly, for the angle sensor Q, an average value of measured values related to the surface shape obtained by the angle sensor Q is obtained. The measured value at this time is stored in the storage device 340 as the second measured value.

  By calculating this average value, Equation (21) and Equation (22) can be modified as follows.

なお、式（２６）、式（２７）は、下記のように、式（２１）、式（２２）中で記載されたものを平均化することにより得られたものである。「数１８」で示されている各式の右辺は平均化された値である。

In addition, Formula (26) and Formula (27) are obtained by averaging what was described in Formula (21) and Formula (22) as follows. The right side of each equation shown in “Equation 18” is an averaged value.

そして、ＣＰＵ３１０は式（２７）−式（２６）の算出を行うことにより、零点誤差αの算出ができる。

Then, the CPU 310 can calculate the zero point error α by calculating Expression (27) -Expression (26).

そして、このようにして得られた零点誤差αは、他のワークの表面形状を測定して算出する際、零点誤差補償に使用される。又、このときの零点誤差αはＣＰＵ３１０により記憶装置３４０に記憶される。

The zero error α thus obtained is used for zero error compensation when measuring and calculating the surface shape of another workpiece. Further, the zero point error α at this time is stored in the storage device 340 by the CPU 310.

(Second measurement and recording)
After the first measurement, the zero point error of the angle sensors P and Q varies due to temperature drift or the like. Therefore, when the surface shape of the workpiece is newly measured after the first measurement and recording, in the second measurement, the CPU 310 of the computer system 300 uses the angle sensor input during one rotation of the sample 200. The average value of each measurement value obtained from P and Q is calculated in the same manner as “first measurement and recording”. Then, the CPU 310 stores the calculation result and the like in the storage device 340. This second measurement value corresponds to the first measurement value.

Further, the CPU 310 calculates the zero point error α1 in the same manner as described in the “First measurement and recording” column. Then, α1−α is calculated to obtain the fluctuation amount of the zero point error.
As described above, when the zero point error of the angle sensors P and Q is fluctuated due to temperature drift or the like, the zero point error-related value recording device obtains the fluctuation amount of the zero point error, and performs high-precision zero point compensation. It can be carried out.

  Also in the second embodiment, for example, a zero point error-related value recording device is provided adjacent to the NC processing device, and the shape measurement of the workpiece of the NC processing device and the measurement of the sample 200 on the spindle 210 are performed. It is preferable that the sensor unit 220 can be moved by the sensor carriage 230 in an area where the shape of the surface 202 can be measured.

The second embodiment has the following characteristics in addition to the same functions and effects as the functions and effects (1) to (3) of the first embodiment.
(1) In the multipoint probe zero point error variation detection method of the present embodiment, the zero point error related value recording apparatus includes a plurality of angle sensors P and Q supported by the sensor carriage 230 (multipoint probe support means). 230 is set apart in the moving direction. As a result, the sensor unit 220 (multi-point probe) can easily determine the amount of fluctuation of the zero point error in the two angle sensors P and Q, and a plurality of sensors easily drift based on the amount of fluctuation. The subsequent zero error can be easily compensated.

In addition, embodiment of this invention is not limited to said each embodiment, You may change as follows.
In the first embodiment, the average value of the outputs of the displacement sensors A, B, and C when the sample 200 is rotated once is calculated. An average output value may be calculated. In the present specification, “one or more rotations” means integer rotation.

  In the first embodiment, one rotation of the sample 200 is performed twice at intervals with the displacement sensors A, B, and C fixed at specific positions at positions in the x direction (ie, radial direction). Based on the outputs of the displacement sensors A, B, and C each time, the average value of the measured values of each displacement sensor was calculated. Then, based on the average value of each displacement sensor A to C at each time, the motion error at the time of measurement is erased, the motion error cancellation value at each time is calculated, the difference between both motion error cancellation values is calculated, The variation of the zero error between intervals was obtained.

  Instead of fixing the displacement sensors A, B, and C at specific positions, the sensor carriage 230 is moved to a plurality of positions in the radial direction, and the sample 200 is rotated once or twice by the spindle 210 at each position. You may make it calculate the average value of the measured value of A, B, C of each displacement sensor based on each output of the displacement sensors A, B, C obtained at the time of making it above. In this case, each measurement value is stored in the storage device 340.

  In this way, the measurement values at all measurement positions of the displacement sensors A to C are obtained, and the average value is calculated. Therefore, the error is reduced by chance, and the fluctuation amount of the zero error with high accuracy is calculated. As a result, the zero-point error can be compensated with high accuracy.

  In the second embodiment, the sample 200 is rotated once in a state where the angle sensors P and Q are fixed at specific sampling points xi and xi + 1 at the position in the x direction (that is, in the radial direction), and the angle sensors P and Q are measured. The value was obtained and the zero error was calculated.

  Instead, the angle sensor P, obtained when the sensor carriage 230 is moved to a plurality of positions in the radial direction (that is, the x direction) and the sample 200 is rotated once by the spindle 210 at each position. You may make it calculate the average value of each output of Q. This calculation is performed in the same manner as in the second embodiment.

  Then, after all the average values of the outputs of the angle sensors P and Q at the respective positions are added for each output, the total value of the outputs of the angle sensors P and Q is divided by the added number. A value may be calculated, and a difference between the total average values of the two may be obtained as a zero point error.

  In this case, it is preferable to move the sensor unit 220 while making the interval d between the angle sensors P and Q and the sampling interval s the same in the radial direction (that is, the x direction), but it is not limited.

  In this way, since the total average value of each output is calculated based on the average value of each output at each position and the difference between the total average values is obtained as a zero point error, the chance error can be reduced. Since the zero error with high accuracy can be obtained, the calculated fluctuation amount of the zero error also has high accuracy. As a result, the zero error can be compensated with high accuracy.

  Although not described in the first and second embodiments, if the bearing that rotatably supports the spindle is a hydrostatic bearing, the accidental error included in the shape of the sample 200 can be reduced. Good. In this way, the calculation of the variation amount of the zero error becomes highly accurate, and the compensation accuracy can be improved.

In addition, when measuring the shape of the sample 200 on the circumference, if the spindle is first driven and then rotated by inertia, the chance error can be further reduced.
In the second embodiment, a two-dimensional angle sensor is used, but it is not necessary to be two-dimensional. Any line can be used as long as it can measure the angle around the Y axis when the line connecting the two sensors is the X axis in FIG. Further, the method and structure of the angle sensor are not limited to the two-dimensional angle sensor described in the second embodiment, and may be changed to other structures.

The technical ideas other than the claims that can be grasped from the above embodiment are listed below.
(1) When rotating the measurement object at least one section or more, it is performed in a state where the rotating means is rotatably supported by a hydrostatic bearing. 2. A method for detecting a variation in zero point error of a multipoint probe according to item 1. In this way, the calculation of the variation amount of the zero point error becomes highly accurate, and the compensation accuracy can be improved.

1 is a schematic diagram of a zero error related value recording apparatus according to a first embodiment embodying the present invention. Explanatory drawing which shows the arrangement | positioning relationship of a sample, a spindle, and a sensor unit. The schematic diagram of the zero point error arithmetic unit in the zero point error related value recording device of a 2nd embodiment which materialized the present invention. Explanatory drawing which shows the arrangement | positioning relationship of a sample, a spindle, and a sensor unit. Explanatory drawing which shows the sampling position in the scanning of a sensor unit. Explanatory drawing which shows the principle of an angle sensor. Explanatory drawing which shows the improvement of the principle of an angle sensor.

Explanation of symbols

200 ... sample (object to be measured), 210 ... spindle (rotating means),
220 ... sensor unit (multi-point probe),
230 ... sensor carriage (multi-point probe support means),
300 ... Computer system, 310 ... CPU (reading means, calculating means),
340 ... Storage device (storage means), A, B, C ... Displacement sensor,
P, Q ... An angle sensor.

Claims (4)

A multi-point probe for measuring a straight shape including a plurality of sensors is arranged at the multi-point probe support means so as to be opposed to the measurement surface of the measurement object being rotated by the rotation means,
The shape value along the concentric circle on the measurement surface of the measurement object and the zero point of the probe output from the plurality of sensors each time the measurement object rotates a predetermined rotation angle during at least one rotation of the measurement object A measurement value including an error based on the error is read by the reading means and stored in the storage means;
A new measurement value (hereinafter referred to as the first measurement value) related to the same measurement object stored in the storage means and a measurement value (hereinafter referred to as the second measurement value) whose acquisition time is older than the acquisition time of the first measurement value. And a zero-point error fluctuation detection method for a multi-point probe. The multipoint probe support means is provided to be movable in a radial direction of the rotation means;
The measurement value is obtained when the multipoint probe support means is moved to a plurality of positions in the radial direction and the measurement object is rotated at least one rotation or more by the rotation means at each position. 2. The method for detecting a variation in zero point error of a multipoint probe according to claim 1, wherein an average value of the obtained values is included.   3. The zero point error of the multipoint probe according to claim 2, wherein the plurality of sensors supported by the multipoint probe support means are three displacement sensors installed apart in the moving direction of the multipoint probe support means. Fluctuation detection method.   3. The multipoint probe according to claim 2, wherein the plurality of sensors supported by the multipoint probe support means are two two-dimensional angle sensors installed apart in the moving direction of the multipoint probe support means. Zero error variation detection method.

Images