JPH04158213A - High-resolution absolute value encoder - Google Patents

Abstract

PURPOSE:To enable the execution of high-resolution absolute value inspection having arbitrary resolution by determining absolute value data on each of rough, medium and precise angles and by obtaining continuous absolute value data by combining these absolute value data. CONSTITUTION:A first pattern 11 of which an opening length increases gradually till 180 degrees and decreases gradually in a half cycle of residual 180 degrees in accordance with a rotational angle, a second pattern 12 of which one cycle is four degrees and wherein transmitting and non-transmitting regions are repeated every two degrees, and a third pattern 13 which has a 1/3 cycle, are formed in a rotary slit plate 10. Besides, slits 21a to 23b and photosensors 31a to 33b corresponding to the patterns respectively are disposed in a fixed slit plate 20 and electric signals which the sensors 31a to 33b emit in accordance with the quantities of sensed light are subjected to voltage conversion 34. Pseudo sine signals 41a and 41b of which one cycle is one rotation, and of which phases are shifted by 90 degrees are obtained from the sensors 31a and 31b, while similar signals 42a and 42b/43a and 43b of which the respective brightness patterns are of one cycle are obtained from the sensors 32a and 32b/33a and 33b. These signals are subjected to an interpolation processing 35 at the same timing, absolute value data on each of rough, medium and precise angles thus obtained are combined 36, and thereby continuous absolute data are obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a high-resolution absolute value encoder suitable for use in detecting the angle and position of an object to be measured.

[Conventional technology]

FIG. 6 is a schematic diagram showing an example of a conventional absolute value type encoder. In the figure, 61 is a slit plate on the rotating side, and a plurality of tracks 62 are provided concentrically on the rotating slit plate 61.

Each track 62 is provided with an absolute value pattern, for example, a pattern in which a gray code is made up of a plurality of slits 63 consisting of areas through which light passes and areas through which light does not pass. 65 is a fixed slit plate. A fixed slit array 66 is provided on the fixed slit plate 65 and has a plurality of openings for selectively passing the light beam from the slit pattern of the steam can truck 62.

68 is a light emitting section having a plurality of light emitting elements, and 69 is a light receiving section having a plurality of light receiving elements.

In the configuration shown in the figure, the light flux emitted from each light emitting element of the light emitting section 68 passes through the slit 63 and the fixed slit row 66 and enters each light receiving element of the light receiving section 69 .

Each light receiving element generates an output signal corresponding to the amount of incident light, and the absolute rotation angle of the rotating slit plate 61 is determined from a combination of the output signals.

[Problem to be solved by the invention]

However, if a high-resolution absolute value encoder is constructed using only binary codes such as Gray codes, all code patterns corresponding to the resolution must be created. For example, in order to realize an encoder with 1 million divisions per rotation, 2 = 1,048,576, so 20 tracks of patterns are required. the same number as 2
This requires 0 sets of light emitting elements, light receiving elements, and their processing circuits, making them expensive. Furthermore, there is a limit to the slit width that can be realized due to the phenomenon of light diffraction, and the minimum diameter of the encoder is determined by the resolution. For example, the fixed slit row 66
To realize an encoder with a slit width of 10 μm and one million divisions per rotation, the circumference would be 10 m and the diameter would be approximately 3 m, far exceeding the practical dimensions from the calculation 10 μm x 1 million = 10 m. Furthermore, when the slit width is narrowed in this manner, the amount of light obtained by the light receiving element becomes weak, which necessitates increasing the amplification factor of the processing circuit section, which also has the disadvantage of making it more susceptible to the effects of noise.

Looking at typical examples of how rotary encoders are used, most rotary encoders are used for angle measuring instruments and have a minimum resolution of well-separated numerical values in degrees, minutes, and seconds, which is the unit of angle. In addition, for position detection of linear feed mechanisms using screws and the like, which are often used in FA equipment, the resolution of the encoder is usually selected so that the minimum resolution of the linear movement amount is a well-defined unit. For these reasons, the resolution per revolution is 60 or 100.
It can be seen that it is desirable to be a multiple of 0.

A method using Gray code as a binary code, 2n
In order to obtain an arbitrary resolution instead of
As disclosed in Publication No. 17, by removing an arbitrary code pair K from the 2n number of divisions, it becomes possible to detect the absolute value of any even number of divisions. However, in this method, when converting the detected Gray code to binary code, a table reference method is used instead of the general method.
The table becomes enormous if the resolution is high.

In a method that uses a binary code as a binary code, gray-binary conversion is unnecessary, but since multiple bits change at the same time, if the position of the fixed slit plate 65 is even slightly tilted, erroneous data may be generated. This makes assembly and adjustment difficult.

It is difficult.

The purpose of the present invention is to solve these problems, and to provide an absolute value encoder with high resolution and arbitrary resolution without increasing the size of the entire device.

[Means to solve the problem]

For this reason, the high-resolution absolute value encoder of the present invention has a movable slit plate for one period in which the amount of transmitted light gradually increases in the range of 1/2 period, and the amount of transmitted light gradually decreases in the range of the remaining 1/2 period. A first pattern consisting of codes and a second pattern consisting of striped codes obtained by dividing the range in which the first pattern is arranged at equal intervals are arranged.

From the first pattern, one rotation (360') has one period,
By detecting two pseudo sine wave signals with a phase shift of 90 degrees and electrically interpolating them, absolute value data of the rough angle is obtained. From the second pattern, two pseudo sine waves with a phase shift of 90 degrees are detected at a predetermined period, and these are electrically interpolated to obtain absolute value data of medium and fine angles, and absolute value data of coarse angles are obtained. to obtain continuous absolute value data. Any resolution can be achieved by appropriately selecting the period of the second pattern and the interpolation magnification.

By the above method, the number of tracks of the code pattern can be reduced, miniaturization is possible, and a high-resolution absolute value encoder having arbitrary resolution can be obtained.

〔Example〕

Hereinafter, as an example of the present invention with reference to the drawings, a resolution of 1 second (
An absolute value encoder (1,296.000 divisions per rotation) will be explained.

Fig. 1 is a diagram showing a pattern of a rotary slit plate which is a movable slit plate of the encoder, Fig. 2 is a diagram showing the fixed slit pattern of the encoder developed in a straight line, and Fig. 3 is a diagram showing the pattern of the rotary slit plate, which is a movable slit plate of the encoder. Figure 4 is a block diagram of the signal processing system of the absolute value encoder. Figure 4 is a diagram showing the relationship between coarse angle data, medium angle data, and fine angle data. Figure 5 is a diagram showing the relationship between coarse angle data, medium angle data, and fine angle data. 3 is a flowchart representing a processing method for combining.

<Basic idea> Record three types of butter on the rotating slit plate 10,
A signal of one rotation and one period is detected from the pattern, and it is divided into 180 by interpolation. From the second pattern as a subsequent stage of the first pattern, a signal of one period is detected at twice the number of divisions of the first pattern, that is, 2/180 rotations, and it is divided into 160 by interpolation. In this way, a resolution of 90×160 per revolution is obtained, and from the third pattern as a subsequent stage of the second pattern, the resolution is twice the number of divisions of the second pattern, that is, 2/(90×1.60).
Detect one period of rotation signal and interpolate it to process 8o
By dividing, per revolution: 90x80x180=360x60x60=1,296
．． OOO [division], that is, obtain a resolution of 1 second.

Angle detection> On the rotating slit plate 10, as shown in FIG.
As a pattern, the opening length is proportional to the rotation angle.
Increasing gradually from 180° to 3600”
A gradually decreasing pattern 11 is formed in the range of . The second pattern is 1/90 of the first pattern. (360@
/90=4'') as one period, forming a pattern 12 in which light passing areas and non-light passing areas are repeated every 2 degrees,
As the third pattern, 1/80° (
4×60/80=3) as one period, and a pattern 16 is formed in which a light passage area and a light non-passage area are repeated every 1.5.

As shown in FIG. 2, on the fixed slit plate 2o, corresponding to the first pattern 11 on the rotating slit, INKs are provided with a width narrower than the resolution by interpolation, for example, l, at a distance of 90'.
Light passing portions 21&, 21b having a width of o are formed. A slit group 2 having the same width and pitch as each pattern corresponds to the second and third patterns 12 and 13 on the rotating slit plate 10.
2a, 22b, 26a, and 26b are formed. The slit groups 22b and 23b are the slit group 22a1.
231L, and are formed at positions shifted by 1/4 period in phase.

Six noslit groups 21a, 21b, 22a.

22b, 21, 23b, 1 each, total 6 pieces (7
) 7f To sensor 318% 3 l b, 32a% 32
b, 33a, and 33b are arranged.

A current signal proportional to the amount of light received is obtained from the photosensor groups 31 to 33b, and is converted into a voltage by a current/voltage conversion circuit 64 as shown in FIG.

Photo sensors 31a, 3 corresponding to the first pattern 11
From 1b, corresponding to the rotating slit plate 100 rotation angle,
Pseudo sine wave signal 4 with a phase shift of 90 degrees in one rotation and one period
1a and 41b are obtained, and the second and third patterns 12.1
Photo sensors 32a, 32b/33a13 compatible with 6
From 3b, each light and dark pattern is one cycle, 90
Pseudo sine wave signal 42&, 42b/438 with degree phase shift
．． 43b is obtained.

<Data combination> These three sets of pseudo sine wave signals 41a, 41b/421k
, 42b/43&, 43b are interpolated at the same timing by the interpolation processing unit 35, and as a result, absolute value data of three coarse angles, medium angles, and fine angles are obtained, and these are sent to the data combining unit 36. The three angle data are combined into continuous absolute value data.

FIG. 5 is a flowchart showing the contents of the data combination process, which will be explained using FIG. 4.

In Fig. 5, fin is fine angle data, mid is medium angle data, crs is coarse angle data, F, mag is interpolation magnification of fine angle data, Mmag is interpolation magnification of medium angle data, MID is correction The latter intermediate angle data, CR3, represents the coarse angle data after correction.

From FIG. 4, it can be seen that the medium angle data changes every 1.5, and the fine angle data changes every 3.

Therefore, if the absolute value data of the fine angle increases monotonically, the absolute value data of the medium angle will always be either an odd number or an even number at the point where it changes from the maximum value to 0. This is determined by design. Incidentally, when the absolute value data of the fine angle is monotonically decreasing, the same thing can be said in that it changes from 0 to the maximum value. Also, the change point is 22 &, 2
2b, 23a, and 25b (FIG. 2) can be shifted by adjusting the phases.

Here, for example, if the intermediate angle data is an odd number at the point where the fine angle data changes from the maximum value to 0 as shown in FIG. 4, then if the value of the intermediate angle data is m i d, then ■ mid If is an even number, the corrected medium angle data MID
m id / 2, ■ If mid is an odd number and the fine angle data fin is more than half the interpolation magnification Fmgf of the fine angle data, the medium angle data MID after correction is (m id −1) / 2. , ■ If m i d is an odd number and the fine angle data fin is less than half of the interpolation magnification Fmgf of the fine angle data, the corrected medium angle data MID is (m i d + 1).
/ 2, however, when this MID becomes more than half of the interpolation magnification Fmgf, it is set to 0.

On the other hand, although not shown in the flowchart in Figure 5,
If the intermediate angle data is an even number at the point where the fine angle data changes from the maximum value to O, then if the value of the intermediate angle data is m i d, if +'pmid is an odd number, the intermediate angle data after correction is M
I D is m id / 2, ■ If m i d is an even number and the fine angle data fin is more than half the interpolation magnification Fmgf of the fine angle data, the medium angle data MID after correction is (m id − 1) /
2. However, when the medium angle data m i d is 0, it is calculated as the interpolation magnification value Mmgf of the medium angle data. ■ When m i d is an even number and the fine angle data fin is the interpolation magnification value Fmgf of the fine angle data. If it is less than half, the corrected medium angle data MID is (m i d + 1)
/ 2, and take the quotient.

By the above calculation, the solution becomes a value in which one period 3 of the precision angle data is one unit.

By setting one cycle of the fine angle data to be twice the minimum resolution of the medium angle data in this way, at the point where the fine angle data changes from the maximum value to 0, the corresponding medium angle data changes. Even if there is a deviation in the range, the error can be absorbed.

Further, the coarse angle data changes every 2 degrees, and the medium angle data changes every 4 degrees. These relationships are exactly the same as the relationship between medium angle data and coarse angle data mentioned earlier, except for the angle values being divided and the interpolation magnification. It is a value with 4 degrees as one unit.

Continuous absolute value data can be obtained by using the solution obtained through the above two calculation processes as the minimum unit of the rough angle. Specifically, it is based on the following formula.

Absolute value data − Rough angle data + Medium angle data after correction × Interpolation magnification of coarse angle data + Rough angle data after correction × Interpolation magnification of medium angle data × Interpolation magnification of rough angle data ÷ 2 The obtained angle data can be directly output in parallel, but as shown in Figure 3, if the P/S converter 67 performs P/S conversion processing and transmits it as serial data, the number of signal lines can be reduced.

The method for realizing 1,296,000 divisions per rotation has been explained above, but the number of divisions is determined by the × third interpolation magnification, and any number of divisions can be obtained by appropriately combining the interpolation magnifications at each stage. is obtained. Also, if a larger number of divisions is required, conversely, by increasing the number of interpolation stages,
If such high resolution is not required, by using two stages of interpolation, division into about 20,000 can be achieved in a similar manner with a simpler configuration. Furthermore, the number of divisions can be increased by increasing the interpolation magnification.

In the above explanation, an example in which the aperture length gradually increases and decreases is shown as the first pattern, but as another example, there are parts with transmittance of 0% and transmittance of 100% at 180° opposing positions, and in between. It may be a pattern in which the transmittance changes continuously.

In addition, in the above embodiment, the first pattern is one stitched 90
By detecting at two points separated by 90 degrees, a pseudo sine wave with a phase shift of 90 degrees is obtained. However, by forming two first patterns with a phase shift of 90 degrees and detecting them at the same point,
Since it is possible to obtain a low-pressure sinusoidal wave with a phase shift of 0 degrees, it is possible to make the fixed slit smaller.

Further, although the rotary encoder has been described above, it goes without saying that this can be extended in a straight line and applied to a linear encoder. Note that when applied to a linear encoder, the first pattern can be formed within a range of one period, so it can be considered that the first pattern is formed for one period in this case as well.

〔Effect of the invention〕

As detailed above, according to the high-resolution absolute value encoder that adopts the detection method of the present invention, high-resolution detection is possible with a small number of tracks, so there is no need to increase the size of the entire equipment and purchase. , absolute value detection with arbitrary resolution becomes possible.

Furthermore, since the slit width of the fixed mask can be set sufficiently wide, the amplification factor of the current/voltage conversion circuit can be kept low, making it less susceptible to noise.

[Brief explanation of the drawing]

FIG. 1 is a diagram showing a pattern of a rotary slit plate, which is a movable slit plate, of an encoder according to an embodiment of the present invention, and FIG. 2 is a diagram showing a linear development of the fixed slit pattern of an encoder according to an embodiment of the present invention. , FIG. 3 is a block diagram of the signal processing system of the absolute value encoder to which the present invention is applied, FIG. 4 is a diagram showing the relationship between coarse angle data, medium angle data, and coarse angle data, and FIG. FIG. 6 is a flowchart showing a processing method for combining angle data, intermediate angle data, and coarse angle data, and is a perspective view of a conventional absolute value detection type encoder. 10.61... Rotating slit plate, 11...
...First pattern, 12...Second pattern, 16...Third pattern, 20.65...Fixed slit plate, 64...
... Current/voltage converter, 65 ... Interpolation processing section, 36 ... Data combination section. Figure 1 Figure 6

Claims (3)

[Claims] (1) A movable slit plate on which first and second patterns are arranged, a first detection means for detecting two waveform signals having a phase shift of 90 degrees from the first pattern, and a second pattern. A second waveform signal that detects two waveform signals with a phase shift of 90 degrees from
and 2 detected by the first detection means.
a first interpolation processing means for interpolating one waveform signal; a second interpolation processing means for interpolating two waveform signals detected by the second detection means; 2
data combining means for combining the interpolated data obtained by the interpolation processing means of the first pattern, the amount of transmitted light gradually increases in the range of 1/2 cycle, and the amount of transmitted light gradually increases in the range of 1/2 period, and The second pattern consists of a code for one period in which the amount of transmitted light gradually decreases in the range of one period, and the signal obtained from this amount of transmitted light is output in a pseudo-sine wave shape, and the second pattern 1. A high-resolution absolute value encoder comprising: a striped code that divides a range in which a pattern No. 1 is arranged at equal intervals. (2) The second pattern is composed of a plurality of tracks in which the interval between the striped codes becomes finer in sequence according to a predetermined integer ratio, and a plurality of second interpolation processing means are provided corresponding to each of the tracks. The high-resolution absolute value encoder according to claim 1, characterized in that: (3) The later pattern among the adjacent patterns has a 90-degree phase, with one cycle being twice the resolution of the interpolated data obtained by the interpolation processing means that interpolates the signal obtained from the previous pattern. 3. The high-resolution absolute value encoder according to claim 1, wherein the high-resolution absolute value encoder is configured to obtain two pseudo-sinusoidal output signals that are shifted from each other.