JPH11325971A - Absolute linear scale - Google Patents

Abstract

PROBLEM TO BE SOLVED: To output absolute value information, which is changed by moving a linear scale, at the time point of request. SOLUTION: Concerning data D1, D2, D3 and D4 which are changed for every sampling, when a request signal is inputted at a time point Q, for example, corresponding to the fixed data D3 at a stored sampling time point S3, data D4 at a sampling time point S4 to next fix data, sampling term T4 and time ΔR from the last sampling time point S3 to the time point Q when the request signal is inputted, data DQ3 at a time point, when the request signal comes, are outputted by D3+(D4-D3)*ΔR/T4.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a linear scale for measuring an absolute amount of movement between two objects, and in particular, in an optical scale, outputs an actual moving distance of the linear scale as an absolute value. Thus, for example, the present invention relates to an optical absolute trinier scale capable of knowing a moving amount of a workpiece such as a machine tool by an absolute value.

[0002]

2. Description of the Related Art In a machine tool or the like, it is extremely important to accurately measure the amount of movement of a tool relative to a workpiece with high accuracy in performing precision machining. There is a demand for quick output as serial data. As one example, an outline of an optical scale using moiré fringes obtained by superposing two optical gratings will be described below. As shown in FIG. 5, the optical scale is a main scale 101 in which a grid (notched line) is provided on one surface of a transparent glass scale 100 such that light transmitting portions and non-light transmitting portions are arranged at a predetermined pitch.
And an index scale 103 having a grid (notched line) on one surface of a transparent glass scale 102 such that light-transmitting portions and non-light-transmitting portions are arranged at a predetermined pitch, as shown in FIG. In addition, the index scale 103 is opposed to the main scale 101 at a small interval,
As shown in FIG. 2B, the grid of the index scale 103 is arranged so as to be tilted by a small angle with respect to the grid of the main scale 101.

With this arrangement, moire fringes shown in FIG. 6 are generated according to the movement of the scale. The interval between the moire fringes is W, and a dark portion or a bright portion occurs at every interval W / 2. The dark portion or the bright portion moves from top to bottom or from bottom to top when the index scale 103 moves left and right relative to the main scale 101. In this case, assuming that the pitch of the grids of the main scale 101 and the index scale 103 is P, and the mutual inclination angle is θ [rad], the interval W of the moire fringes is expressed as W = P / θ, and the pitch P is Then, the interval W between the moiré fringes can be detected by enlarging it by θ times. That is, the grid is 1
When the pitch moves, the moiré fringes are displaced by W, but the pitch P becomes θ times the W. Therefore, by detecting the change in the moiré fringes, the amount of movement in the pitch P can be measured with high accuracy. Become.

Therefore, as shown in FIG. 7, a photoelectric conversion element 110 for optically detecting a change in moiré fringes is provided on an index scale, and a light source is provided on the opposite side of the main scale. The change in the current flowing through the photoelectric conversion element 110 is read while the scale 103 is relatively moved. Index scale 103 is A for main scale 101
In this state, the amount of light applied to the photoelectric conversion element 110 becomes the largest, and the current flowing through the photoelectric conversion element 110 becomes the maximum value I ₁ . Next, when relatively moved to the state of B, the amount of light applied to the photoelectric conversion element 110 slightly decreases, and the current becomes I ₂ . Is irradiated with the least amount of light, and its current is also the smallest I ₃ . Then, when further moved to the state of D, the amount of light applied to the photoelectric conversion element 110 slightly increases, the current becomes I ₂ , and when the state moves to the state of E, it becomes the position where the amount of light is largest again. current becomes maximum I _1. As described above, the current flowing through the photoelectric conversion element 110 changes sinusoidally, and the main scale 101 and the index scale 103 relatively move by the lattice interval P when one cycle of the change has elapsed.

As shown in FIG. 8, when two photoelectric conversion elements 111 and 112 are provided at 90 ° or one cycle (interval W) and 90 ° apart, an A-phase photoelectric conversion element 111 is provided.
The current flowing through the B-phase photoelectric conversion element 112 with respect to the current flowing through is a current deviated by 90 ° as shown in FIG. That is, when the current flowing through the A-phase photoelectric conversion element 111 is a sine wave, the current flowing through the B-phase photoelectric conversion element 112 is a cosine wave. In this case, the phase of the current flowing in the B-phase photoelectric conversion element 112 with respect to the current flowing in the A-phase photoelectric conversion element 111 is 90 ° advanced or 90 ° depending on the relative movement direction of the main scale 101 and the index scale 103. If two photoelectric conversion elements are arranged shifted by 90 °, the relative movement direction can be detected by detecting the phase between the two.

The optical scale constructed as described above is mounted on an NC machine tool and measures the relative movement amount between the workpiece and the tool. In general, when performing numerical control, the optical scale is always used. Or, it is required to grasp the absolute position of the work table by the scale. Hereinafter, a method of detecting the movement amount of the scale as an absolute value from the A-phase signal and the B-phase signal will be described with reference to FIG.

In this figure, 21 is a carrier signal CK
A balanced modulator (hereinafter also simply referred to as a modulator) to which (a, b, c) is input, 22 is a low-pass filter, 23 is a waveform shaping circuit, and 24 is a waveform-shaped binary signal as will be described later. A phase digital signal interpolating between one pitch to
And a digital signal processing unit for forming a B-phase digital signal. The modulator 21 has been disclosed by the present applicant in
As disclosed in Japanese Unexamined Patent Publication No. -132104, a carrier signal is balanced-modulated using a sinusoidal signal level as a modulation signal, and the signal level is output as carrier phase information. As shown in FIG. 11, the input A-phase signal is supplied to the resistor network RT via the operational amplifier OP1 operating as a buffer, and is also inverted and supplied to the resistor network RT by the operational amplifier OP2. The B-phase signal is supplied to the resistor network RT via the operational amplifier OP3 operating as a buffer, and is also inverted by the operational amplifier OP4 and supplied to the resistor network RT.

That is, the A-phase signal, the inverted A-phase signal, the B-phase signal, and the inverted B-phase signal are mixed and added by the resistor network RT to generate a mixed signal having the opposite voltage and divided into eight equal voltages. 8 input terminals (0) ~
(7). This multiplexer A
Select signals A, B, and C shown in FIG. 12C are input to the input terminals C1, C2, and C3 of M.
C, the input terminals (0) to (7) of the multiplexer AM
Are sequentially selected, and a step-like output signal S shown in FIG. 12A is output from the output terminal to. The frequency of the signal S output from the multiplexer AM is the same as the period of the selection signal C as shown in FIG. 12, and after all, the selection signal C is used as a carrier and its phase is changed to the A-phase signal (B
The output signal S balanced-modulated by the level of the phase signal) is output from the multiplexer AM. That is,
The carrier wave whose phase is shifted according to the level of the A-phase signal (B-phase signal) is output.

The modulated wave thus balanced is applied to the low-pass filter 22 to form a smooth sinusoidal wave as shown in FIG.

The signal converted into a sinusoidal wave has the following formula: S = K · Cos (ωt−2π · x) where ω is the angular velocity of the frequency of the carrier, p is the interval of the grid (notched line) of the scale, and x is the amount of movement. / P), which is an AC signal in which the ratio x / p of the scale movement amount x to the pitch p is represented as a phase displacement.

The phase-modulated AC signal is converted by the next waveform shaping circuit 23 into a binary signal having a zero-level point as an edge. FIG. 13 shows the relationship between the phase of the binary signal output from the waveform shaping circuit 23 and the levels of the A-phase signal and the B-phase signal output from the optical means. The left side of the figure shows the A-phase signal and the B-phase signal output from a scale of a sine wave, and the pulse waveform shown on the right side is a waveform shaping circuit 23 which has undergone a phase shift. Is a binary signal of the carrier, and the position of the broken line is the carrier CK (a,
b, c).

As shown in FIG. 1A, when the scale is stopped, for example, when the A-phase signal is at the maximum positive level and the B-phase signal is at the zero level, the binary signal is shifted by 90 ° in phase. If the A-phase signal is at zero level and the B-phase signal is at the maximum positive level, the binary signal is shifted by 180 °, and the A-phase signal is at the negative maximum level. Is a binary signal having a phase shift of 270 ° when the signal is zero level, and a phase shift of 360 ° when the signal of the phase A is zero level and the maximum level of the phase B signal is negative. Then, the binary signal is returned to the original state without phase shift.

The digital signal processing unit 24 uses the capture function to detect the rising point of the binary signal output from the co-waveform shaping circuit 23 and the interval T between the reference phase indicated by the dotted line. Since the interval T indicates information on a position obtained by dividing the grid pitch P, the period T is counted by a predetermined clock to form an interpolation pulse signal for interpolating one pitch of the scale. At the same time, by counting the interpolation pulse signal, it is possible to obtain data on the absolute position of the scale obtained by dividing one pitch. For example,
When the counting of the counter is started by the edge of the carrier wave output from the digital signal processing unit 24 and the counting of the counter is stopped by the rising edge of the binary output of the waveform shaping circuit 23, the counter of the digital signal processing unit 24 An interpolation absolute value obtained by dividing the pitch P can be detected. In the case of this embodiment, one pitch of the moire fringes output from the optical means indicates 40 μm, but FIG. 13 shows a clock having a frequency that is 40 times the carrier by the counter function in the digital signal processing unit 24. When the period T shown is counted, 1 /
A counting pulse is obtained with an accuracy of 40 pitches, and a movement of 1 μm can be detected from the moire fringe output signal of the optical means.

[0014]

The absolute value of the interpolation is obtained by, for example, setting the zero phase point in FIG.
At the end of the sampling period, and the value is output as interpolation data Dn at the end point of the sampling period t. When moving, as shown in FIG. 14, each sampling point s1, s
.., Data D1, D2, D3,.
Can be read. When this data is output when a request signal Sr is input from the outside, the digital signal processing unit 24 outputs absolute value data for interpolating within the pitch indicated by the scale corresponding to the request signal Sr. Can be. Also, during the movement of the scale, the variation of the interpolated pulse can be output as an A / B phase signal.

However, the request signal Sr is
As shown in FIG. 4, since the request signal Sr generally arrives at a period longer than the sampling period tn and the request signal Sr is output from the machine tool side, the interval differs depending on the target machine tool. Even if the signal is input at regular intervals, the signal is asynchronous with the sampling timing. Therefore, for example, when the request signal Sr is input at the timing shown in the drawing with respect to the sampling period, at the time when this signal is input,
Since the sampling has not yet been completed at that time, in such a case, the data (D1) held after the completion of the sampling one cycle before is output, and the absolute value at the point Q at which the request was made is output. There is a problem that the insertion signal cannot be output.

[0016]

SUMMARY OF THE INVENTION In order to solve the above-mentioned problems, the present invention relates to a main scale having at least a ruled line graduated at equal intervals in the length direction and a main scale having the same. An index scale that is movably disposed and has an engraved line intersecting the engraved line, and detects moiré fringes generated by the engraved line between the two scales. Photoelectric conversion means for generating a signal which changes periodically, and a binary signal pulse train corresponding to the ruled line pitch outputted from the photoelectric conversion means, and a higher-order detection for outputting the moving position of the scale as the number of pulses (G) Means, modulation means for performing balanced phase modulation of the signal fetched by the photoelectric conversion means with a carrier signal of a predetermined frequency, and a modulation signal phase-modulated by the modulation means and the carrier. Sampling the phase difference of the signal period t
And lower-order detection means for generating interpolation data for outputting data (Dn) for interpolating within one pitch, and measuring an edge of the sampling period and an input timing period of a request signal supplied from the outside. The ratio (ΔR / tn) between the calculated value (ΔR) and the sampling period (tn) is measured, and NG + Dn−
1+ (Dn−Dn−1) * (ΔR / tn) is output as the absolute value of the scale.

[0017]

When the scale is stopped or when the scale is moving in one direction, each sampling period tn has a substantially constant value. At least during the period tn of each sampling period, a large fluctuation does not occur. can not see.
Therefore, by performing the interpolation processing at the time of arrival of the request signal, the data of the absolute value interpolated at that time can be output. Further, by monitoring the tendency of the past sampling period, the accuracy of the output data can be further improved by providing a correction means for estimating the current sampling period from the sampling period one period before.

[0018]

FIG. 1 shows an outline of an embodiment of an absolute scale according to the present invention. In this figure, 1 is a light source lamp, 2 is a light receiving element for receiving the moire fringe light of the A-phase and B-phase signals transmitted through the scale, and 3 is the received A / B
This is an amplifier that amplifies the phase signal to a predetermined level.

In this embodiment, the optical means is, for example, 4
The resolution of an output pulse constituted by a scale having a pitch of 0 μm and interpolated by phase division is 1 μm.
The case where the length can be measured in units of will be described.

A phase amplified by the amplifier 3 or B
The phase reference signal is input to the waveform shaping circuit 4,
The signal is converted into a binary signal by inverting the signal that changes in a sinusoidal manner at a zero cross. Then, the signal converted into the binary signal is detected in the gate array circuit 5, for example, by detecting the period between the rising edges thereof.
One pulse signal train is formed at one pitch of the scale, and the pulse signal train is counted by the microcomputer 6 to store the upper absolute value (in this case, a unit of 40 μm) that changes with the movement of the scale. I am trying to do it.

The A-phase signal and the B-phase signal amplified by the amplifier 3 are input to the balanced modulator 7 and are subjected to signal processing for interpolating between pitches as shown in the above equation. Then, the level change of the A-phase and B-phase moiré fringes is converted into a signal that changes the phase of the alternating signal, and the resulting signal is supplied to a low-pass filter 8 that makes the phase-modulated signal a sine wave. The signal converted into a sine wave is input to the comparator 9 so as to be inverted at the zero cross point,
The digital output converted to the value signal is supplied to the microcomputer 6.

The microcomputer 6 divides the clock signal by 1/4 (（
0), the digitized carrier signal CK (a, b, c) is supplied to the balanced modulator 7 as described above, and the output phase of the comparator 9 receiving the phase modulation and the carrier wave The clock is counted using the reference edge of the signal CK (a, b, c) as a sampling period, and an interpolation pulse signal for further dividing the interval between 1-pitch lines is formed from the readout information of the scale. Is counted, and the value is detected as lower-order absolute position data obtained by dividing one pitch.

That is, as shown in FIG. 2, when the displacement within one pitch is temporarily increased due to the movement of the scale, a displacement that rises to the right occurs at each of the sampling times S1, S2, S3,. .., Which are displaced for each sampling, are added to the higher-order absolute position data G which is output from the gate array 5 and incremented for each pitch, and output. Can output an absolute value over the entire length of the scale.

In the case of the present invention, as shown in FIG. 2, for example, the request signal Sr is applied to the data D1, D2, D3, and D4 interpolating within one pitch, which changes every sampling, at the time Q. , The stored data D3 at the sampling time point S3 and the data D at the sampling time point S4 at which the next data is determined.
4, and the sampling period t4, and the time ΔR from the previous sampling time S3 and the time Q when the request signal Sr was input, the data DQ3 at the time when the request signal came is expressed as “DQ3 = D3 + (D4−D
3) * (ΔR / t4) ”.

FIG. 3 shows a flow of the microcomputer for performing such an operation. In accordance with this flow, the microcomputer is always in a standby state in which a request signal is input in step 100, and when a request signal is input, first, at that time, a higher-order absolute value data (output of a gate array output) The measured value G is latched (S1).
01). Then, the sampling data (Dn-1) at the previous time point which has already been completed is read out at step 102, and the previous sampling time (Sn-
1) and the time ΔR until the request signal Sr is input
Is measured and stored.

Next, it is determined whether or not the interpolated data currently being sampled and measured is determined (S1).
04), the period (T
n) and the determined data Dn is read (105). Then, the difference (Dn-Dn-D) between the data (Dn-1) sampled last time and the data (Dn) sampled this time is obtained.
A value obtained by multiplying 1) by the sampling period Tn and the ratio (ΔR / Tn) of the time ΔR during which the request signal was input is calculated in step (106), and when the interpolation number is N, the measured value is calculated. Interpolation processing for adding G to the value of N times is performed (107), and GN + Dn-1 + (Dn-Dn-1) *
(ΔR / tn) is output as the absolute value of the arrival time of the request signal.

When such a calculation is output by the microcomputer 6 every time a request signal arrives, the absolute value at the time when the request signal Sr is input is latched, latched at least until the next sample time, and transmitted to the machine tool side. can do.

During the interpolation process, the measured value G is taken in such a manner that it is incremented in synchronization with the sampling period, so that even when the upper absolute value changes, the upper + lower absolute value can be accurately obtained. Output at an appropriate timing.

Further, when the scale is moved in the direction in which the sampling value is temporarily increased as shown in FIG. 4 during the interpolation processing, immediately after the upper absolute value G is incremented to G + 1, the lower absolute value G becomes lower. When the interpolation absolute value (D4) becomes zero, 0 is set to 40 and the interpolation calculation (40-39) is performed. Similarly, when the scale moves in the direction in which the sampling value monotonously decreases, the timing occurs in which the upper absolute value G decreases to G-1, but immediately thereafter, the lower absolute value changes to 39. A flow routine is provided so as to output -1 when it becomes, and (0-1) is interpolated.

According to the present invention, the data at the time of the arrival of the request can be calculated by the interpolation processing. Therefore, the sampling period can be set arbitrarily in accordance with the processing speed of the microcomputer regardless of the timing shown in the above embodiment. Can be determined.

[0031]

As described above, according to the present invention, the lower absolute value is output in response to the request signal by using the conventional interpolation pulse output means by phase division which measures the relative moving distance. In this case, since the ratio between the time when the request signal is input and the sampling period is detected and the lower absolute value at the request time is interpolated and output, an absolute value that matches the request timing is output. There is an effect that can be.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an example for detecting an absolute value of a linear scale according to the present invention.

FIG. 2 is an explanatory diagram of interpolated absolute value data, a request signal, and a sampling cycle.

FIG. 3 is a flowchart showing an operation of a microcomputer for outputting an absolute value.

FIG. 4 is an explanatory diagram of lower-order data when higher-order data is incremented;

FIG. 5 is a principle diagram of an optical scale.

FIG. 6 is a diagram showing moire fringes.

FIG. 7 is a diagram showing movement of moire fringes.

FIG. 8 is a diagram showing a position where a photoelectric conversion element is installed.

FIG. 9 is a waveform diagram of an A-phase signal and a B-phase signal.

FIG. 10 is an explanatory circuit diagram for obtaining an absolute value from A-phase and B-phase signals.

FIG. 11 is a circuit diagram of a phase modulation circuit.

FIG. 12 is a timing chart of the phase modulation circuit.

FIG. 13 is an explanatory waveform chart for obtaining lower interpolation data.

FIG. 14 is an explanatory diagram when lower interpolation data is output by a request signal.

[Explanation of symbols]

1 light source 2 pho and diode 3 amplifier
4 Waveform shaping circuit 5 Gate array 6 Microcomputer 7 Modulator
8 Low-pass filter 9 Comparator 101
Main scale 103 indexes

Claims (2)

[Claims] 1. A main scale having a score line graduated at least at equal intervals in a length direction, and a score line movably disposed with respect to the main scale and intersecting the score line. Index scale, a photoelectric conversion unit that detects a moiré fringe generated by a score line between the two scales, and generates a signal that periodically changes each time the unit moves relatively unit length; and Higher-order detection means for forming a binary signal pulse train corresponding to the ruled line pitch to be output and outputting the moving position of the scale as the number of pulses (G), and a carrier signal of a predetermined frequency which is a signal fetched by the photoelectric conversion means. A phase difference between the modulation signal phase-modulated by the modulation means and the carrier signal is detected at a sampling period tn. And a lower detection means for generating interpolation data among the output interpolated data (Dn), the sampling period of the edge and the value obtained by measuring the input timing period of the external request signal supplied from the (delta
R) and the ratio (ΔR / t) of the sampling period (tn).
n), and when the number of interpolated data is N, NG + Dn-1 + (Dn-Dn-1) .times. (. DELTA.R / Tn) is output as an absolute value of the scale. . 2. The absolute trinier scale according to claim 1, wherein said upper detecting means latches the number of pulses (G) in synchronization with said sampling period.