JPH04329312A - Absolute encoder - Google Patents

Abstract

PURPOSE:To provide high resolution without need for decreasing the pitch of a sensor to lambda/2. CONSTITUTION:There are three patterns consisting of two one-track type absolute patterns with phase difference of lambda/2 from each other, and one incremental pattern 4 on a symbol plate 1. An output means 9 switches two sets of one-track type absolute patterns 2, 3 alternately at intervals of lambda/2 at the timing with the boundary range of minimum read-out unit avoided preferably, according to binary signals 0, 1 read out from the incremental pattern 4.

Description

[Detailed description of the invention]

0001

[Industrial Application Field] The present invention reads a one-track absolute pattern by alternately switching two sets of sensors to avoid the boundary area of the minimum reading unit.
This invention relates to an absolute encoder that reduces pattern reading errors caused by sensors.

0002

2. Description of the Related Art An absolute encoder is a measuring instrument that outputs information on the position of a detection unit with respect to a code plate using an absolute position signal specific to each relative position. The code plate and the detection unit are combined so that they can be moved relative to each other, and the code plate has an absolute pattern in which the numbers of the absolute position signal are replaced with physical information and is continuously arranged, and the detection unit has the physical information of the pattern. It has a sensor that discriminates. Absolute encoders are divided into linear types in which the detection part moves linearly along a band-shaped code plate, and rotary types in which the detection part moves angularly relative to a disc or cylindrical code plate. Although it is broadly divided into two types, in any case, scale information recorded as physical information on the absolute pattern (
Absolute position signal) is directly read by the sensor in the detection section,
It is a measuring instrument that reassembles the numbers into numbers and outputs them.

[0003] Conventionally, a multi-track type absolute encoder has been common. This is called a multi-track absolute pattern, which uses multiple parallel incremental tracks, each with a different pitch.
This is an absolute encoder in which an absolute pattern is constructed from patterns. The detection unit for reading the multi-track absolute pattern is basically equipped with multiple sensors, one for each incremental pattern, and converts the output of the multiple sensors into a binary code. An absolute position signal is assembled from ``ordered binary numbers'' such as

For example, in a 4-track absolute encoder that reads a 4-digit binary code,
The code board has 4 parallel incremental patterns, each with a different pitch, 23 tracks...000000001
1111111 22 Truck …00
00111100001111 21 Track …0011001100110011
20 Tracks…0101010101010
101 Absolute position
abcdefghijklmnop is formed. In the detection section, four sensors are arranged vertically, and the same phase positions of the four tracks are read in parallel, and a=0000, b
From =0001 to o=1110 and p=1111, 16 absolute position signals a to p in binary numbers are obtained in order.

On the other hand, in recent years, instead of the multi-track type,
One-track type devices are being actively researched and put into practical use. This is an absolute encoder in which the absolute pattern is composed of one irregular pattern called a one-track absolute pattern. A one-track absolute pattern is a special binary number sequence such as a full period sequence or an M sequence, in which 1 and 0 are replaced with two types of minimum read units having different physical properties and arranged in one line. On the other hand, in the detection section, a plurality of sensors are basically arranged in a row at a pitch of the minimum reading unit length along a one-track absolute pattern, and the outputs of the plurality of sensors are An absolute position signal is assembled consisting of "binary numbers of different but completely random order."

[0006] The one-track absolute encoder is: (1) Only one absolute pattern is required; (2) Many sensors are arranged in a row at equal intervals, so many sensors can be formed all at once on one board. The absolute encoder can be made smaller and the overall structure including wiring can be simplified due to the fact that a sensor array can be used, and there is no need to adjust the phase of the sensor arrangement between tracks as in the case of a multi-track type. This is extremely advantageous.

For example, in a one-track absolute encoder using an M sequence consisting of 15 codes per cycle: 000100110101111, the 0s and 1s of the M sequence are encoded on the code plate in two types with different physical properties. 1-track type absolute by replacing the minimum reading unit
The pattern 000100110101111 abcdefghijklmno is formed. However, after the last o, the pattern returns to the first a, and each relative position on the pattern is represented by the leftmost alphabetic character. On the other hand, in the detection section, four sensors are arranged corresponding to four consecutive minimum reading units, and a=0001
, in order from b=0010, n=1100, o=1000
Up to 15 absolute position signals are obtained, each consisting of binary numbers that are different but in random order. A random absolute position signal can also be converted into a binary code using a semiconductor memory device.

By the way, one track type absolute
The encoder is highly likely to malfunction (misread the pattern) when the sensor passes through the boundary area of the minimum reading unit. In other words, in the boundary area of the minimum reading unit where the output is reversed, the sensor output is inherently unstable, and due to pattern and sensor pitch errors, variations in sensor characteristics, etc. The timing of output reversal may be delayed. At this time, an abnormal absolute position signal in which at least one digit of the binary number constituting the absolute position signal is inverted is output.

[0009] Such a one-track type absolute
In view of the problem of reading errors in encoders, the applicant of the present invention previously proposed in Japanese Patent Application Laid-Open No. 63-322189 a one-track type system that avoids the boundary area of the minimum reading unit by alternately switching two sets of sensor groups. Absolute
We proposed a technology to read patterns.

FIG. 2 is a schematic diagram of a conventional absolute encoder disclosed in Japanese Patent Application Laid-Open No. 63-322189.

In FIG. 2, the code plate 11 has a number sequence: 10110011110100 starting from the triangular mark.
A one-track absolute pattern 12 formed by replacing the 0's and 1's of . Ru. On the other hand, code plate 11
The detecting unit 15 combined so as to be movable relative to the
Opposed to the absolute pattern 12, the pitch λ
A sensor array 16 in which eight photosensors are collectively formed on one substrate with a ratio of 1/2 is arranged, and a sensor 18 is arranged to face the incremental pattern 14. The sensor array 16 includes a first sensor group consisting of four sensors P1 to P4 with a pitch λ, and a second sensor group consisting of sensors Q1 to Q4.
The sensors are divided into sensor groups.

In the code plate 11, the patterns 12 and 14 have a mutual phase difference of λ/4. Detector 15
, the sensors Q1 to Q4 and the sensor 18 have the same phase and have a phase difference of λ/2 with respect to the sensors P1 to P4. Therefore, when the sensor 18 faces one code (shaded area) of the pattern 14, the sensors Q1 to Q4 face the boundary area of the minimum reading unit, while the sensor 18 faces the other code (white area) of the pattern 14. , the sensors Q1 to Q4 face the center of the minimum reading unit. Furthermore, when the sensors Q1 to Q4 face the boundary area of the minimum reading unit, the sensors P1 to P4 always face the center of the minimum reading unit.

In the absolute encoder configured as described above, when the sensor 18 detects the shaded portion of the pattern 14, the first sensor group consisting of sensors P1 to P4 is detected, while the sensor 18 detects the white portion of the pattern 14. When detecting the
The sensor group is selected and the pattern 12 is read. In this way, on the pattern 12, the sensor group located at the center of the minimum reading unit is always selected, and no matter the relative position of the code plate 11 and the detection unit 15, the sensor group is continuously selected without any break. An absolute position signal can be obtained, and the absolute position signal read from the boundary area of the minimum reading unit of the pattern 12 does not need to be output to the outside.

[0014]

[Problems to be Solved by the Invention] In order to meet the recent demands for smaller absolute encoders and higher resolution, code plates (one-track absolute pattern) are required to be further miniaturized. . However, compared to a one-track absolute pattern that is just a structural element, it is much more difficult to manufacture a sensor that is a functional element that reads it.
In reality, the reduction in sensor pitch is a technical constraint for increasing the resolution of absolute encoders. In this respect, the pitch of the sensors on the sensor array must be reduced to 1/2 of the minimum reading unit length λ, and an absolute encoder of the type that alternates between two sets of sensors every λ/2 as described above. However, it was difficult to achieve high resolution.

For example, in the absolute encoder shown in FIG. 2, the sensor array 16 consists of eight photosensors P1 to P4 and Q1 to Q4 with a pitch of λ/2, and the aperture length of each sensor is λ/2 or less. be. Here, reducing the pitch of the sensors P1 to P4 and Q1 to Q4 not only increases the difficulty in manufacturing, but also causes a decrease in output and noise resistance. Therefore, in the current design, sensors P1 to P
4. Due to the performance and manufacturing limitations of Q1 to Q4, the pitch in the sensor array 16 is first determined to be 40 to 50 μm, and in accordance with this, the minimum reading unit length λ of the 1-track absolute pattern 12 is as follows: 80-10
We are forced to adopt an extraordinarily large value of 0 μm.

[0016]

[Means for Solving the Problems] The present invention is characterized by: (1) being able to obtain an absolute position signal continuously without any break at any relative position between the code plate and the detection unit, and (2) achieving the minimum reading unit of a 1-track absolute pattern. Absolute position signals read from the boundary area are not output to the outside, and the pitch of the sensor does not have to be reduced to 1/2 of the minimum reading unit length of a 1-track absolute pattern, making it easy to achieve high resolution.・The purpose is to provide an encoder.

The absolute encoder of the present invention has the following features:
1 track type absolute with minimum reading unit length λ
In an absolute encoder comprising a code plate on which a pattern is formed and a detecting section movable relative to the code plate, a second encoder having the same length λ and arrangement but different in phase by λ/2 is provided in the pattern. A track-type absolute pattern and an incremental pattern with a pitch λ are arranged in parallel with the pattern, and the pattern and the second 1-track type absolute pattern are arranged in parallel with the above-mentioned pattern, and the pattern and the second 1-track type absolute pattern are arranged in parallel with the above-mentioned pattern. An output means for selecting and reading one of the two is provided.

[0018]

[Operation] In the absolute encoder of the present invention, two one-track absolute patterns are
Every time the incremental pattern is reversed, it is read alternately and an absolute position signal is output. That is, when the binary signal obtained by reading the incremental pattern is one value, the output means connects the sensor group provided for the first one-track absolute pattern, and when the binary signal is the other value, the output means connects the sensor group to the second one when the binary signal obtained by reading the incremental pattern is the other value. The sensor group provided for the one-track absolute pattern is selected and the read absolute position signal is output to the outside.

[0019]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Examples of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic diagram of an absolute encoder according to an embodiment. Here, two one-track absolute patterns having a phase difference of λ/2 on the code plate are alternately read by two sets of sensors provided on the detection unit in the same phase.

In FIG. 1, a first one-track absolute pattern 2, a second one-track absolute pattern 3, and an incremental pattern 4 are arranged on a code plate 1. Pattern 2 and pattern 3 are each a sequence of numbers starting from the triangular mark: 101
It is formed by replacing the 0s and 1s of 10011110100... with two types of minimum reading units, the white part and the shaded part, whose length is λ, and both have the same minimum reading unit length λ and arrangement. However, the phase differs by λ/2. Further, incremental pattern 4 has a phase difference of λ/4 from patterns 2 and 3.

On the other hand, the detection section 5, which is movable relative to the code plate 1, includes a sensor array 6 facing the pattern 2, a sensor array 7 facing the pattern 3, and a sensor array 7 facing the pattern 3.
Sensors 8 are arranged to face the patterns 4, respectively. The sensor arrays 6 and 7 are formed by collectively forming two sets of four photosensors R1 to R4 and S1 to S4 with a pitch λ on the same substrate, and the photosensors R1 to R4 and the photosensors S1 to S4 are The phases are made equal. Also,
Sensor 8 occupies the same phase with respect to sensors R1-R4, S1-S4.

Therefore, when sensor 8 faces one symbol (shaded area) of pattern 4, sensors R1 to R4 face the boundary area of the minimum reading unit of pattern 2, but sensors S1 to S4 face the boundary area of pattern 3. Opposed to the center of the minimum reading unit. On the other hand, sensor 8 detects the other code of pattern 4 (
When facing the white part), the sensors S1 to S4 are in pattern 3.
is opposite to the boundary area of the minimum reading unit of sensor R1
~R4 faces the center of the minimum reading unit of pattern 2.

The output circuit 9 outputs a sensor array 7 consisting of sensors S1 to S4 when the sensor 8 detects the shaded area of the pattern 4, and outputs signals from the sensors R1 to R4 when the sensor 8 detects the white area of the pattern 4. The sensor arrays 6 are selected and connected to the output 9a.

In the absolute encoder configured as described above, the output circuit 9 alternately switches between the sensor array 6 and the sensor array 7 in accordance with the binary signal 0 and 1 obtained from the pattern 4. 2 and pattern 3 are read alternately to assemble an absolute position signal. In this way, the sensor array located at the center of the minimum reading unit is always selected, and the absolute position is continuously maintained without any break in all the phase positions that the detection unit 5 can take with respect to the code plate 1. Even though the signal is obtained, the absolute position signal obtained by reading the boundary area of the minimum reading unit of patterns 2 and 3 does not need to be output to the outside.

Furthermore, looking at the photocurrent output from one photosensor, in the case of the conventional example, the pitch is 40 μm and the width is 30 μm.
Whereas an output of 80 nA was obtained from a 00 μm photosensor, the same resolution (minimum reading unit length 8
Despite using a code plate of 0 μm), the pitch is 80
160 μm, using a photo sensor with a width of 3000 μm
An output of nA is obtained.

In this embodiment, the phase difference between the two one-track absolute patterns on the code plate is λ.
/2. The phase difference between the two sensor groups in the detection section is assumed to be 0, but basically, each combination of pattern and sensor group has the same function as long as the mutual phase relationship is shifted by λ/2. You can expect. In other words, the phase difference between the two 1-track absolute patterns is 0, the phase difference between the two sensor groups is λ/2, and the phase difference between the two 1-track absolute patterns is λ/4, 2. A similar function can be expected even when the phase difference between a set of sensor groups is set to λ/4.

[0028]

Effects of the Invention In the absolute encoder of claim 1, (1) an absolute position signal can be obtained continuously without any break in all the relative positions that the detecting section can take with respect to the code plate; (2) a one-track absolute encoder; Absolute position signals read from the boundary area of the minimum reading unit of the pattern are not output to the outside, and (2) the pitch of the sensor does not have to be set to 1/2 of the minimum reading unit length of the 1-track type absolute pattern.

Therefore, the output of the sensor is increased, and the circuit design of the absolute encoder is facilitated. Also, 1
Even if the minimum reading unit length of the track-type absolute pattern is reduced to 1/2, the sensor output can be obtained at the same level as in the past, so it is possible to further increase the resolution of the code plate.

[Brief explanation of the drawing]

FIG. 1 is a schematic diagram of an absolute encoder according to an embodiment.

FIG. 2 is a schematic diagram of a conventional absolute encoder.

[Explanation of symbols]

1 Code plate 2 1 track type absolute pattern 3 1
Track type absolute pattern 4 Incremental pattern 5 Detection section 6 Sensor array 7 Sensor array 8 Sensor 9 Output circuit

Claims (1)

[Claims] 1. An absolute encoder comprising a code plate on which a one-track absolute pattern with a minimum reading unit length λ is formed, and a detection section movable relative to the code plate, wherein the pattern has a length λ. A second one-track absolute pattern with the same λ and arrangement but with a phase difference of λ/2 and an incremental pattern with a pitch λ are arranged in parallel to the pattern, and a binary signal obtained from the incremental pattern is obtained. An absolute encoder characterized in that an output means is provided for selecting and reading one of the pattern and a second one-track type absolute pattern in correspondence with the pattern.