JP2679207B2 - Code plate for absolute encoder - Google Patents

Description

TECHNICAL FIELD The present invention relates to a code plate for an absolute encoder.

[Prior Art] Japanese Patent Application Laid-Open No. 54-118259 discloses a code plate having tracks of two or more magnetic patterns, and a magnetoresistive element (Magn.
A magnetic absolute encoder combined with a detector using an etoresistive element (hereinafter referred to as MR element) is shown.

Generally in order to obtain a resolution of 2 ^N in this type of absolute encoder requires minimum N of tracks as absolute pattern, the code disc if the disc type, a plurality of tracks are arranged in concentric circles. For example, when N = 4, four circular tracks are concentrically provided on the disk-shaped code plate, and a binary code absolute pattern depending on the direction of magnetization is formed on these tracks. A sensor consisting of an MR element is assigned to each track, and as the code plate rotates relative to the sensor around its center, the output of each sensor obtained at any rotation position of the code plate The combination is such that it gives a code signal giving the absolute rotational position of the code plate.

On the other hand, JP-A-57-175211, or JP-A-60-1529
In Japanese Patent No. 16 publication, two kinds of minimum reading units corresponding to a "0" signal and a "1" signal, that is, a minimum reading unit having magnetism and a minimum reading unit having no magnetism are arranged on one track in a predetermined arrangement. And a read head composed of a plurality of detectors arranged in the track length direction and movable relative to the code plates, and the signal train output from each detector is absolute as a binary code. A magnetic or optical absolute encoder for position detection is shown.

[Problems to be Solved by the Invention] When the above-mentioned conventional absolute encoder is a code plate, particularly when it is a magnetic type, as shown in FIG. Therefore, there is a problem in that one of them adversely affects the other, which becomes noise, and eventually the reading accuracy deteriorates.

[Means for Solving the Problems] Therefore, in the present invention, in at least one of the two types of minimum reading units in such a code plate, the signal determining unit is not a whole unit but a part of the unit.

[Operation] In the present invention, the signal determining unit has a unit of “0”.
It indicates the area (range) that determines whether it represents a signal or a "1" signal. For example, the magnetic type will be explained in an easy-to-understand manner. As shown in FIG. 1 corresponding to the embodiment, one unit representing the "1" signal among a plurality of minimum reading units arranged at a constant pitch is defined as one unit. Has a magnetized region (hatched portion) in only a part thereof, whereas the other unit representing the "0" signal has no magnetized portion at all. The part of the magnetized area in one of the units is the signal determining unit that determines the unit as the "1" signal.

On the other hand, in the conventional code plate, as shown in FIG. 2, the entire region of one unit representing the "1" signal is occupied by the magnetized portion, and the entire unit is the signal determining unit.

Since the code plate of the present invention has the signal fixing portion only in a partial area of the unit, it is necessary to read in the magnetized portion of the partial area when reading the unit. Therefore, it is not preferable for the same type of reading unit to have the signal determining unit at different positions for each unit, and for any reading unit, it is desirable that the signal determining unit be present at a substantially central position in the unit area.

Further, in the code plate of the present invention, the reading unit is read by the signal determining unit, so that a clock signal for timing the reading is separately required. As this clock signal, there is a method of obtaining it by using another clock, but on the code plate of the present invention or another code plate arranged in parallel with it, the same unit pitch, that is, the length, along the absolute pattern is obtained. It is advisable to provide an incremental pattern of equal reading units and read it with another detector to obtain the clock signal. In this case, the clock signal obtained by reading the incremental pattern has a slight variation in synchronism (this variation occurs when the positional relationship between the pitch of the incremental pattern and the pitch of the absolute pattern does not match). The variation range is calculated in advance from the phase shift between the code plate and the read head to obtain the variation range in the track length direction of the reading position for each unit in advance, and the minimum length of the signal fixing portion is included so as to include this variation range. It is preferable to make the range wider than the fluctuation range.

By adding a track of an incremental pattern along a 1-track type absolute pattern on the same or an adjacent code plate, and operating the latch circuit by using the clock signal generated by using this incremental track as a strobe signal, A method to obtain the absolute output by simultaneously reading the output pulses of the detector at the optimum timing (Japanese Patent Application No. Sho 63-170782), or a serial data format from each detector by a clock signal created from an incremental pattern using a similar code plate. A method (Japanese Patent Application No. 63-171922) for obtaining an absolute output by fetching the absolute signal of the above into a shift register or the like at the optimum timing and performing serial / parallel data conversion has been proposed. Any of these can be used as an application example of the code plate of the present invention. As an example, the outline of the former method is shown in FIG.

In the example of FIG. 12 (a), the scale scale number is 2 ⁴ = 16 (N
= 4) In the optical example of pulse, a track 102 having an absolute pattern of one track and an incremental track 103 are manufactured on a disc type code plate 101. Truck 1
The absolute pattern of 02 is "0000110101111001". Reference numerals 104-1 to 104-4 denote optical sensors (an example of detectors) for reading the track 102.
Obtain an absolute signal consisting of a binary code of bits. Incremental track 103 is another optical sensor 105
Read by. The reference numeral 106 designates the axis of rotation of the code plate 1.

FIG. 12 (b) shows an example of an output circuit for processing the reading result by these sensors. It shows the incremental signals from the sensor 105 and the sensors 104-1 to 104-4.
And the absolute signal from the pulse shaping circuit 11
0, 110-1 to 110-4 waveform shaping into a rectangular wave, then
The absolute signal is directly input to the latch circuit 114, and the incremental pulse is input to the one-shot circuit 112 to generate a clock pulse at both the rising and falling points of the incremental pulse, which is used as a strobe signal for the latch circuit 114. In this case, the clock pulse rises at a time point approximately at the center of the pulse width of the unit pulse forming the absolute signal. Latch circuit 114
Each time a clock pulse arrives, it reads the high and low level of the pulse of the absolute signal at that time (the high and low of this level correspond to one and the other of the "0,1" signal), and the value is read until the next clock comes in. Latch the output terminal 11
Continue outputting to 6-1 to 116-4.

Although the above-mentioned error generation of the encoder output was avoided by such a latch system, the Japanese Patent Application No. 63-
In the method of 170782, the signal determination unit for determining the signal content of the minimum reading unit in the absolute pattern of the code plate as "0" or "1" has a length of the minimum reading unit as shown in FIG. Since it is formed to have the same size as the length dimension λ (dimension in the longitudinal direction of the track), for example, in the magnetic system, the magnetized portion serving as a signal determining portion must be formed with accurate length dimension accuracy.

Further, in the case of the optical type, if one minimum reading unit is represented by, for example, a punched hole formed in the code plate, if the units are continuous, the mechanical strength of the code plate will be reduced, and reinforcement will be necessary.

Further, in either method, since the continuous unit having a length up to N times the unit pitch λ must be formed in the absolute pattern of the code plate, the larger the number N of bits of the absolute code, It must be formed with high dimensional accuracy.

However, if the signal fixing portion is formed in a part of the area of the reading unit as in the present invention, these requirements are eliminated, so that the absolute encoder can be easily assembled and adjusted.

In the absolute encoder using the code plate of the present invention, for example, a plurality of detectors that can move relative to the code plate of the present invention in the longitudinal direction of the track are provided, and the high and low levels of the pulse of the absolute signal are detected by these detectors. At the time of reading, only the signal fixing portion shorter than the length λ of the minimum reading unit is synchronized with another clock pulse signal and is read at a time point within the width of the corresponding pulse.

In the absolute encoder using the code plate of the present invention, when the absolute pattern of the track of the code plate is read by each detector, the pulse train extracted from each detector corresponds to the digital code array according to the constant unit pitch of the absolute pattern. Therefore, the high and low levels of the unit pulse constituting the pulse train are read in synchronization with another clock pulse signal for each unit. Thus, the output of each detector is read at a stable time point apart from the rising and falling edges of each unit pulse of the pulse train based on the clock signal.
It is possible to extremely reduce the occurrence of errors in the encoder output.

Further, in the present invention, since the signal determining unit for determining whether the minimum reading unit in the absolute pattern is “0” or “1” is formed smaller than the length λ of the unit, the adjacent signal determining unit is formed. There will be a space between and.

The clock pulse is preferably obtained from the reading result of the incremental pattern provided together with the code plate on which the absolute pattern is formed.
This may be obtained from the incremental pattern of another code plate connected by the shaft, and in this case, the dimension of the signal determining portion in the track length direction is determined in consideration of the twist of the shaft.

Embodiments of the present invention will be described below with reference to the drawings.

[Embodiment] FIGS. 3 (a) and 3 (b) show an embodiment of the present invention in the case of a magnetic absolute encoder using an MR element. FIG. 3 (a) shows a disk-shaped code plate. FIG. 2B is a schematic plan view showing a sensor of an MR element for reading an absolute pattern signal magnetized on the disc type code plate.

In FIG. 3 (a), the code plate 1 has its rotational axis 3
Three circular tracks 2a, 2b, 2c centering on the center are set concentrically.

Here, the tracks 2a and 2b are tracks having an absolute pattern, and in this embodiment, the case where the S / N ratio of the reading of the detector is improved by using the complementary double track is illustrated. It may be by an absolute pattern track of a book.

On the first circular track 2a, an absolute pattern of 16 divisions (N = 4 bits) with the circumference divided into 16 (1 division unit corresponds to π / 8 radian) is magnetized in the magnetization direction indicated by the arrow. Magnetized graduations 6a to 6d made up of graduation units (minimum reading unit) having a section (corresponding to the signal determining section) M and unmagnetized graduations 7a made up of graduated units (minimum reading unit) not having the magnetized section M. ~ 7d and formed. The actual state of the magnetized portion M is as shown in FIG. 3 (c) here.

In the first circular track 2a of FIG. 3 (a), the graduation structure will be described in order from the 12 o'clock position in the clockwise direction. The unmagnetized graduation 7a is a continuous four "0" graduation and the magnetized graduation 6a is continuous. Two "1" scales, a non-magnetized scale 7b is a single "0" scale, a magnetized scale 6b is a single "1" scale, a non-magnetized scale 7c is a single "0" scale, and a magnetized The scale 6c can be represented as four consecutive "1" scales, the non-magnetized scale 7d can be represented as two consecutive "0" scales, and the magnetized scale 6d can be represented as a single "1" scale, thus The absolute code is "0000110101111001".

Also on the second circular track 2b, the absolute pattern of 16 divisions (N = 4 bits) with the circumference divided into 16 divisions (one division unit corresponds to π / 8 radians) is the same as the magnetized division.
It is formed by 6e-6h and unmagnetized scales 7e-7h,
This absolute pattern is just the reverse pattern of the absolute pattern of the first track, and therefore the absolute code of this pattern is "1111001010000110".

Each of the graduation units constituting each of the magnetized graduations 6a to 6h has a length dimension (angle range) λ of one graduation, and a magnetized portion M having a dimension α <λ in the longitudinal direction of the track approximately at the center thereof.
Have respectively.

Here, λ is the length dimension of one graduation unit and the arrangement pitch of the graduation unit, and since N = 4 in FIG. 3 (a), λ is in the angular range on the circular track of the code plate 1. Equivalent to 360/16 = 22.5 degrees (π / 8 radians).

The magnetizing portion M is a signal determining portion, and the dimension α in the track longitudinal direction is set to be less than the dimension λ in increments of one scale, and
As will be described later, the size is sufficient to cover the relative error with the clock signal so that the clock signal as the synchronization signal always appears in the unit pulse width of the absolute signal corresponding to the size α of the magnetized portion M. Has been done.

The innermost third track 2c has an incremental pattern for obtaining the aforementioned clock signal, and 16 tracks corresponding to the length dimension (angle range) λ of exactly one scale are provided on this track 2c. The magnetized sections N and S are alternately arranged with different polarities, and the entire circumference is divided into 16 parts to form an incremental pattern.

As shown in FIG. 3B, the detector 10 has N = 4 MR element sensors 11a to 14a and 11b to 14b for each absolute pattern track 2a, 2b of the code plate 1 and an incremental pattern. It has one MR element sensor 30 for reading the track 2c. This detector 10
As shown by the chain line in FIG. 3 (b), the sensors 11a to 14a are placed opposite to the code plate 1 and the relative rotation between them about the rotation axis 3 causes the sensors 11a to 14a to attach to the first circular track 2a. At the same time as reading the absolute pattern by the magnetic scale, the sensor
11b to 14b read the inverted absolute pattern of the second circular track 2b, and the sensor 30 reads the incremental pattern on the third circular track 2c so as to obtain a synchronized clock signal at that time.

When reading the absolute pattern, the arrangement interval between MR element sensors on the same track is the unit size λ.
Alternatively, it may be an integral multiple thereof, and in FIG. 3 (b), this interval is just λ. However, in the case of an integral multiple, the absolute pattern is different from the above. For example, for an absolute encoder with 10 scales, the interval is λ
In the case of, the absolute pattern is as shown in FIG. 8 (d), and when the interval is 2λ, the absolute pattern is as shown in FIG.

FIG. 4 shows the configuration of the absolute pattern reading sensors 11a and 11b for each MR element sensor together with the configuration related to the incremental pattern reading sensor 30. In addition, another sensor for absolute pattern reading 12
Since the configurations of a, b to 14a, b are the same as the configurations of the both sensors 11a, 11b, the description thereof will be omitted.

As shown in FIG. 4, the MR element 15a that constitutes the MR element sensor 11a for the first track 2a and the second track 2b.
The MR element 15b that constitutes the MR element sensor 11b for is arranged in a line in a direction orthogonal to the tracks 2a and 2b in a plane parallel to the track plane of the code plate. The MR element sensor 30 for the third track 2c is composed of two thin MR elements 15c and 15d spaced at a distance of λ / 2 in the track longitudinal direction. In this case, the MR elements of the sensors 11a and 11b are aligned with each other in the longitudinal direction of the track without any displacement. This is because both tracks 2a and 2b have a zero phase difference in the circumferential direction as shown in FIG. 3 (a). When these tracks are provided with a certain phase difference in the circumferential direction, the MR elements 15a and 15b of the sensors 11a and 11b are also displaced by the corresponding phase difference. Good.

MR elements 15a and 15b for reading the absolute pattern
As shown in FIG. 4, 15a and resistors 32a and 15b and resistor 32b
Are connected in series and are connected in series, and DC power supply terminals 17 and 20 are connected so that the current directions are opposite to each other in both series connection paths.
A bridge circuit is formed between them and a detection output is generated between the signal output terminals 18 and 19. In FIG. 4, magnetized scales 6a and 6f for both tracks are added as an example.

In addition, the sensor 30 for reading the incremental pattern
The MR elements 15c and 15d are connected in series between the DC power supply terminals 17 and 20 as shown in FIG. 4, and are similarly connected to the bridge circuit together with fixed resistors 32c and 32d connected in series between the DC power supply terminals. And the intermediate connection point of each serial connection bus is connected to the signal output terminals 33 and 34. When a horizontal magnetic field is applied to the MR element, the MR element lowers its electric resistance value according to its strength regardless of the polarity of the magnetic field. Therefore, the signals generated at the output terminals 18 and 19 of the absolute pattern reading sensor shown in FIG. 4 by the relative movement of the detector 10 and the code plate 1 are as follows.

For example, when a horizontal magnetic field from the magnetizing scale is applied to the MR element 15a, the resistance value of the MR element 15a decreases, so the potential of the output terminal 18 rises, and when a magnetic field is applied to the MR element 15b, the output terminal 18
Potential drops. The potential of output terminal 19 is fixed resistance 32a, 32
It is fixed to a predetermined potential in the middle between the DC power supply terminals 17 and 20 by b. The patterns on both tracks are complementary, and since the MR elements in both sensors are not displaced in the track length direction, the voltage generated between the output terminals 18 and 19 has an amplitude waveform just above and below in both cases. Be symmetrical.

The signals generated at the output terminals 33 and 34 of the incremental pattern reading sensor 30 are as follows. That is, when a horizontal magnetic field is applied to one MR element 15c, its resistance value decreases, so the potential of the output terminal 33 rises, and when a horizontal magnetic field is applied to the other MR element 15d, its resistance value decreases. , The potential of the output terminal 33 decreases. On the other hand, the potential of the output terminal 34 is fixed to a predetermined potential in the middle between the DC power supply terminals 17 and 20 by the fixed resistors 32c and 32d.

The distance between the MR elements 15c and 15d is λ, which is just half the length λ of each magnetized scale unit of the incremental pattern.
/ 2, so when one of the MR elements 15c, 15d has the maximum resistance value, the other has the minimum resistance value. Therefore, the code plate 1 and the detector
Due to the relative movement with respect to 10, a signal output that changes corresponding to the horizontal magnetic field distribution along the track 2c of the incremental pattern is obtained between the output terminals 33 and 34.

FIG. 5 shows an example of a signal processing circuit for processing the detection output of the detector 10, and FIG. 6 shows complementary absolute patterns and tracks formed on the tracks 2a and 2b of the code plate 1. An example of the horizontal magnetic field pattern along the track by each magnetized section of the incremental pattern formed in 2c and the waveform of each part of the signal processing circuit is shown.

In the detector 10, the absolute pattern reading
One output terminal 18 of the MR element sensors 11a and 11b is connected to the input terminal 22 of the signal processing circuit 21, and the other output terminal 19
Is connected to the input terminal 23. Since the same applies to the other MR element sensors 12a, 12b to 14a, 14b for reading the absolute pattern, only the set of MR element sensors 11a, 11b will be described here.

The magnetic field pattern by the magnetizing graduation of the absolute pattern formed on the track 2a of the code plate 1 is as shown in FIG. 6 (a), and by the magnetic graduation of the reverse pattern formed on the track 2b complementary to this. The magnetic field pattern is as shown in FIG. 6 (b). When this is relatively scanned by the sensors 11a and 11b each composed of an MR element, a pulsating pulse signal as shown in FIG. 6 (d) appears at the output terminal 18, and the other output terminal of FIG. A voltage signal having a substantially constant level as shown in (e) is generated. When the signals appearing at both output terminals 18 and 19 are input to the differential amplifier 24 of the signal processing circuit 21 and differentially amplified, the output end of the differential amplifier 24 has a substantially rectangular wave shape as shown in FIG. 6 (f). Signal is obtained. The pulse rise and fall of this signal are constant and steep regardless of the pulse width. Therefore, when this signal is converted into a rectangular wave at a certain comparison level by the comparator 25, a rectangular wave signal as shown in FIG. 6 (g) is obtained at the output end of the comparator 25. This rectangular wave signal is input to the input end of a latch circuit 29 as a part of the reading means, and similarly, a pair of sensors 12a and 12b, a pair of sensors 13a and 13b, and a sensor 14a.
A rectangular wave signal based on the detection signal from each of the groups 14 and 14b is also input to the corresponding input terminal of the latch circuit.

On the other hand, the horizontal magnetic field pattern of the incremental pattern formed on the track 2c of the code plate 1 is as shown in FIG. 6 (c), and two MR elements with an interval of λ / 2 are provided.
When relative scanning is performed with the sensor 30 composed of 15c and 15d, the sensor 30
Since a pulsating pulse-shaped signal having a waveform corresponding to the horizontal magnetic field pattern appears between the output terminals 33 and 34 of, the waveform is shaped into a rectangular wave by the amplifier circuit 26 and the Schmitt trigger circuit 27 in the signal processing circuit 21. Then, it is input to the mono-multi circuit 28. The mono-multi circuit 28 outputs a scobe pulse as the clock signal as shown in FIG. 6 (h) at both the rising edge and the falling edge of the rectangular wave, and this is output to the latch circuit 29.
Has given to.

In this case, as shown in FIGS. 6 (g) and 6 (h), the scroll pulse is generated in synchronization with a time point approximately at the center of the pulse width of the minimum scale constituent unit of the rectangular wave signal which is the reading result of the absolute pattern. As a result, the timing of reading the high-low level of the rectangular wave signal from each comparator 25 is aligned with the strobe pulse time point by the latch circuit 29, and the rising and rising edges of the rectangular wave shown in FIG. Each rectangular wave signal is read at the same time at a stable point in the middle of the pulse width away from the downward direction, thereby preventing reading with wrong contents. The binary number obtained as a result of reading is shown in FIG. 6 (i).

In the present embodiment, such selection of the reading timing is performed by selecting the track 2c of the incremental pattern and its detector 30 for the combination of the tracks 2a, 2b of the absolute pattern and its detectors 11a, b to 14a, b. This is achieved by appropriately selecting the phase difference in the arrangement of the combination with.

The latch circuit 29 latches the high and low levels of the rectangular wave signal at each input end each time the strobe pulse as the clock signal arrives, and outputs it to its output terminals 27a to 27d. In this way, four-digit binary code signals with different combinations of "0, 1" are obtained from the four output terminals 27a to 27d for each rotation angle of π / 8 radians of the code plate 1. .

The rectangular wave signals appearing at the output terminals 25a, 25b, 25c, 25d of the signal processing circuit 25 by the detection signals from the four sets of sensors are shown in FIG. 7 corresponding to FIGS. 3 (a) and (b). (A) (b) (c) (d) shown in FIG.
Incidentally, FIG. 7 (e) shows a rectangular wave signal corresponding to the reading result of the incremental pattern output from the Schmitt trigger circuit 27, and the strobe pulse giving the reading timing at both the rising and falling edges of this signal is a strobe pulse. Is generated. In this case, for the fixed detector 10, the code plate 1
Is rotating counterclockwise as indicated by the arrow in FIG. 3 (b). In this embodiment, N =
Since it is 4, the absolute code has 2 ^N = 16 scales, and as shown in FIG. 3 (b), four sets of MR element sensors 11a-11b, which are arranged at intervals of λ in the circumferential direction of the code plate 1, are arranged. 12a−
On the track 2, the detection signals from 12b, 13a-13b, and 14a-14b prevent the same combination of "0, 1" code signals from being generated from the output terminals 27a to 27d for one rotation of the code plate 1. The absolute pattern array (absolute code) of is specified, which is "0000110101111001" as described above.

Therefore, output terminal 27a is 2 ⁰ , 27b is 2 ¹ , 27c is 2 ² and 27d is 2 ³
, The absolute signals of four scales with different contents are obtained for each relative rotation angle π / 8 radian. In FIG. 7, hexadecimal numbers corresponding to the absolute signals are added as (f). As can be seen, if the rectangular wave signal shown in FIG. 7 is digitized as it is, it becomes 16 hexadecimal numbers, and this does not become the same value at one place when the code plate 1 is rotated once. It can be seen that the encoder is configured.

The arrangement of the absolute pattern is determined as follows.

That is, when the number of graduations is small, it may be sequentially performed by trial and error, but when the number of graduations is large, it is necessary to calculate by a computer.

In the case of the above-described four scales, for example, when each scale is “0”, there is always a case, so a series of four “0” s “0,0,0,0” is considered first. Then, if five “0” s are consecutive, the same combination will occur, so “0”
It is assumed that "1" always comes after four consecutive characters. In this way, “0” or “1” is sequentially added so that the same combination of contents does not occur when shifting one division at a time by four divisions.

The result calculated by the computer in this way is
Figures (a), (b), (c) and (d) show.

FIG. 8 (a) is 32 scales, that is, an absolute code when N = 5 bits, and FIG. 8 (b) is 64 scales, that is, N =
Absolute code for 6 bits, Fig. 8 (c)
Is an absolute code for 256 scales, that is, N = 8 bits, and FIG. 8 (c) is an absolute code for 1024 scales, that is, N = 10 bits.

The absolute codes shown in FIGS. 8 (b), (c), and (d) are constructed as a series in which the scale at the end of a line is connected to the scale at the beginning of the next (lower) line.

When the absolute code shown in FIG. 8 is used for a rotary encoder, the last scale on the bottom line is the first scale.
It is connected to the scale at the beginning of the line so that it continues endlessly.

In the example of Fig.8, the code array is shown when the MR element sensor is continuously arranged at the interval (λ) corresponding to one graduation of the absolute pattern. If it is physically difficult to continuously arrange the MR elements, the MR element sensors can be arranged at intervals of 2λ, for example, every other scale of the code arrangement by devising the code arrangement of the absolute pattern. As such an example, FIG. 9 shows an absolute code when N = 10. In this case N = 10
Ten MR element sensors are arranged on a scale, that is, at intervals of 2λ.

Of course, it is possible to appropriately determine the absolute code for other intervals as well, and generally, an absolute code can be created for an interval that is an integral multiple of λ.

According to such an absolute code, since an absolute pattern can be realized in one track, it is possible to obtain an absolute encoder whose size is almost the same as that of a so-called incremental encoder.

The above is an example of a magnetic absolute encoder, but FIGS. 10 to 11 show an embodiment of an optical absolute encoder.

In FIG. 10 (a), a track on which the same absolute pattern as that described above is provided on the code plate 4 is composed of "0" scale units formed only of opaque portions and "0" scale units including transparent portions T. 5a and a track 5b that divides one round into 16 equal parts and alternately repeats each divided area into an opaque portion and a transparent portion are formed as an incremental pattern. The transparent portion T on the track 5a of the absolute pattern is a signal determining portion, and the dimension α in the track length direction is smaller than the dimension λ in the scale unit, as in the case of the magnetic type described above.

As shown in FIG. 10 (b), a pair of light sources 81a to 84a and photoelectric sensors 81b to 84b and a pair of light source 80a and photoelectric sensor 80b facing each other with the code plate sandwiched therebetween are combined on the code plate 4 as detectors. Then, the detector and the code plate 4 perform relative rotation about the rotation axis 9.

This detector has four photoelectric sensors 81a, 82a, 82a, 82a, 82a, which are arranged in the track longitudinal direction at intervals of λ to read the track 5a.
83a, 84a, and is composed of a single photoelectric sensor 80b for reading the track 5b, the relative arrangement relationship of these photoelectric sensors 81b ~ 84b mutually λ, as in the case of the embodiment of the magnetic absolute encoder described above. Tracks at intervals
Photoelectric sensors 80b are arranged along the track 5b and are arranged opposite the tracks 5b at positions angularly displaced from each other by λ / 2.

FIG. 11 shows an example of a signal processing circuit for processing the detection output of each of the photoelectric sensors 80b to 84b. Eleventh
In the signal processing circuit of the figure, the detection output of each of these photoelectric sensors is shaped by the pulse shaping circuits 90 to 94 to obtain rectangular wave signals as shown in FIG. In FIG. 7, (a) to (d) correspond to the outputs of the shaping circuits 91 to 94, and (e) corresponds to the output waveform of the shaping circuit 90.
As shown in FIG. 7, the rising and falling timings of the rectangular wave signal (e) obtained from the detection signal of the photoelectric sensor 80b corresponding to the reading result of the incremental pattern are as follows.
It is substantially at the center of the pulse width of the smallest scale unit of the rectangular wave signals (a) to (d). This rectangular wave signal (e) is input to the mono-multi circuit 95 as a trigger signal,
95 produces pulse outputs on both the rising and falling edges of the trigger pulse. 96 is a latch circuit, a pulse shaping circuit
The rectangular wave signals (a) to (d) output from 91 to 94 are latched at the arrival of the pulse (strobe pulse) output from the monomulti circuit 95 and taken out to the output terminals 97a to 97d. ing. This allows each output terminal 97a
Needless to say, a binary code signal of 4 scales similar to that described above can be obtained in .about.97d.

Although the rotary encoder for reading the rotational angle position has been mainly described in the above-described embodiments, the present invention can be applied to a linear encoder for reading a linear position. The above-mentioned absolute pattern may be linearly formed on the code plate.

Also, an example was given where the code plate was provided with an incremental pattern track to obtain the clock signal (strobe pulse). However, in the case of an encoder in which the relative movement of the code plate and the detector is always performed at a constant speed. Is
By providing a clock pulse oscillator with a constant frequency in the signal processing circuit, the track of the incremental pattern and its detection system can be omitted.

[Advantages of the Invention] As described above, according to the present invention, since the signal determining portion in the absolute turn of the code plate is formed smaller than the minimum reading unit, the influence of the adjacent reading units is reduced. The reading accuracy is improved. Further, it becomes easy to form the absolute pattern of the code plate and to assemble and adjust the encoder. In other words, the rectangular wave signal of the read result is
Since it is read based on another clock signal within a pulse width smaller than the minimum reading unit equivalent, it is possible to take in the pulse high and low levels of the detection signals from the individual detectors at the same time at a stable point, resulting in an error in the output. Thus, it is possible to obtain a high-resolution absolute encoder that is free from the possibility of noise.

[Brief description of the drawings]

FIG. 1 is a schematic partial plan view showing an example of a code plate of the present invention,
2 is a schematic partial plan view of a conventional example, FIG. 3 (a) is a schematic plan view of a code plate according to an embodiment of the present invention for a magnetic absolute encoder, and FIG. 3 (b) is a view of the code plate. FIG. 4C is a schematic plan view showing a detector composed of an MR element sensor for reading an absolute pattern by the magnetizing section, FIG. 6C is an explanatory view of the magnetizing section M, and FIG. 4 is an explanatory view showing the configuration of each MR element sensor. FIG. 5 is a circuit diagram showing an example of a signal processing circuit for processing the output signal of the detector, and FIG.
(I) is a diagram showing the horizontal magnetic field pattern by the magnetizing scales of the absolute pattern and the incremental pattern formed on the track of the code plate and the waveform of each part of the signal processing circuit, and FIGS. A diagram showing the final output waveform of the absolute encoder using the code plate of the invention, FIGS. 8 (a) to (d) are some of the absolute codes for determining the absolute patterns for obtaining the absolute signals of different scale numbers. Explanatory diagram showing an example,
FIG. 9 is an explanatory diagram showing another example of the absolute code,
FIG. 10 (a) is a schematic plan view showing a code plate for an optical absolute encoder according to another embodiment of the present invention, FIG. 10 (b) is a front view seen from the direction of arrow A in the previous figure, and FIG. Circuit diagram showing an example of a signal processing circuit for processing the output signal of the photoelectric sensor of the previous figure, FIG. 12 (a) is a schematic plan view of a code plate for an optical absolute encoder according to a conventional example,
FIG. 2B is a circuit diagram showing the signal processing circuit. (Explanation of symbols of main parts) 1: Code plate 2a, 2b: Track (absolute) 2c: Track (incremental) 3: Rotation axis 6a to 6h: Magnetized scale 7a to 7h: Not magnetized scale 10: Detector 11a, b to 14a, b: MR element sensor 15a to d: MR element 17: + side power supply terminal 18, 19: Output terminal (absolute) 20: -side power supply terminal 21: Signal processing circuit 24: Differential amplifier 25: Comparator 28: Mono-multi circuit 29: Latch circuit 30: MR element sensor 33, 34: Output terminal (incremental)

Claims (1)

(57) [Claims] 1. An absolute encoder code plate in which two kinds of minimum reading units corresponding to a "0" signal and a "1" signal are arranged in a predetermined array on one track, and at least one of the minimum reading units is A code plate for an absolute encoder, which has a signal fixing portion shorter than the length of the unit.