JPH0259615A - High resolution absolute type encoder - Google Patents

Abstract

PURPOSE:To eliminate the need of fining a code pattern by taking a difference of storage data of a buffer register in a position detecting circuit of a movable head and a fixed head, and detecting absolutely a rotation angle of a shaft against a fixed case. CONSTITUTION:Both a movable magnetic head 3 fixed to a shaft 1 and held thereby and a fixed magnetic head 2 fixed to a fixed case face a magnetic disk 5. In such a state, signals (a), (b) from the fixed magnetic head 2, and signals (c), (d) from the movable magnetic head 3 are inputted to an input terminal 22 of a block 20, and an input terminal 23 of a block 21, respectively. Subsequently, a rotation angle theta of the movable head 3 against a zero point of the magnetic disk 5 is transferred to an output port 25 from a buffer register 36. Also, from an output port of the block 20, a position of the fixed head 2 against the zero point of the magnetic disk 5 is transferred. A microcomputer 28 detects absolutely the rotation angle theta of the movable head 3 against the fixed head 2.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to a high-resolution absolute encoder that detects a mechanical rotation angle in absolute value.

(Prior Art) Rotary encoders are classified into absolute type and incremental type depending on the output format, and further classified into optical type and magnetic type depending on the detection means.

A conventional optical absolute encoder has a recording surface of multiple tracks on a rotary code disk, and slits of code code patterns corresponding to each rotation angle are recorded on these multiple tracks by auto-etching technology using phosphorography. An optical head for detecting a code pattern is provided for each track, and a system is adopted in which a plurality of these heads read out the track code pattern in parallel.

Further, in a conventional magnetic absolute encoder, a rotating code disk is magnetized with a plurality of magnetization patterns in a plurality of tracks, and a plurality of magnetic sensors detect the pattern in parallel.

On the other hand, the incremental type encoder has only one track instead of the plurality of tracks of the above-mentioned absolute neat type encoder.

(Problems to be Solved by the Invention) In the above-mentioned absolute type encoder, since the magnetic type requires detecting the magnetization pattern of a stationary rotating code disk, an inductive magnetic head cannot be used, and a magnetic flux responsive type encoder cannot be used. Must use a magnetic head. The magnetic flux responsive head is a magnetoresistive type.

There is a Hall effect type, but in order to obtain high resolution,
When the magnetization pattern is made finer, the signal level becomes smaller, and the performance of the optical method is not as high as that of the optical method in terms of high resolution.

Even in optical systems, in order to achieve higher resolution, the slit pitch is made smaller, and as the pitch approaches the wavelength of light, the S/N of the signal decreases due to the effect of light diffraction, making detection impossible. becomes.

The biggest problem with absolute encoders is that when multiple tracks on a code plate are read in parallel, reading skew between tracks can cause reading errors, and there is a risk of erroneously outputting an abnormal code. Therefore, it was necessary to adopt an alternating binary code or to move the reading head to the center of the code when it came near the boundary between adjacent codes.

For this reason, it has been difficult to increase the resolution of the absolute encoder.

In addition, conventional absolute encoders needed to transmit multiple digits of data.6 However, in recent years, microcomputers have been applied to position control devices, and encoders that output in digital format are now being used exclusively as position detectors. It started to be done.

Further, when a speed detector is required in terms of control technology, it is necessary to use the amount of change in position of the position detection encoder every fixed time as speed information, so a high-resolution encoder is desired.

In addition, an absolute type encoder that can accurately detect the position of a machine even in the event of a power outage has a great advantage over an incremental type. It is an object of the present invention to provide a highly reliable, ultra-high resolution absolute encoder that requires only two systems of output signals, does not require its transmission frequency band to be so high, and allows for easy failure diagnosis.

[Structure of the invention]

(Means for Solving the Problems) The present invention includes the following means to achieve the above object.

a fixed case of the encoder; a shaft that is rotatably held around the axis of the fixed case and detects the mechanical rotation angle; and a shaft that is rotated at a constant speed around the axis of the fixed case and that is a rotary code disk on which index patterns of a constant pitch are recorded at equal intervals and a zero point pattern for specifying one of the plurality of large index patterns; The index pattern and zero point pattern of the disk are read out, and these are used as the first index input signal.

A movable head that produces a first zero-point pulse signal, and a fixed device that is fixed to the fixed case and reads out the index pattern and zero-point pattern of the rotary code disk and makes these a second index pulse signal and a second zero-point pulse signal. and a first counter that multiplies the frequency of the first index pulse signal of the encoder main body and increments at this multiplied frequency, and the overflow pulse of the first counter is a phase-locked loop (PLL) circuit synchronized with a first index input signal; a phase-locked loop (PLL) circuit that is incremented by the overflow pulse of the first counter and reset by the first zero point pulse signal; the position of a movable head consisting of a second counter having a capacity equal to the number of index patterns in one rotation of the plate; and a buffer register in which the contents of the first counter and the second counter are stored in response to a sample command; with detection circuit.

The fixed head position detection circuit has the same configuration as the movable head position detection circuit, and includes a fixed head position detection circuit which inputs the second index pulse signal and the second zero point pulse signal.

(Function) In the above configuration, the movable head fixed to the shaft generates a first index pulse signal every time it reads the index pattern on the rotary code disk. Further, each time the code disk rotates once, the zero point pattern is read out and a first zero point pulse signal is generated.

The rotational position of the movable head with respect to the zero point of the rotary code disk at the moment when the movable head generates the index pulse can accurately and absolutely detect how many index positions it is from the zero point pulse. In order to detect the rotational position of the movable head not only at the moment when this index pulse occurs but also at any arbitrary time, the position detection circuit of the movable head repeatedly generates the index pulse. A phase-locked loop having a first counter that multiplies the frequency of a first index pulse signal from the movable head and increments at the multiplied frequency, and the overflow pulse of the counter is synchronized with the first index pulse signal. to act. Further, the second counter is incremented every time the first counter overflows, and is reset by the first zero point pulse signal.

In this way, the position data obtained by interpolating between the indexes of the intermediate rotating code disk of the digital data representing the rotational position of the movable head with respect to the zero point of the rotating code disk that changes from moment to moment is stored in the first counter at which index position. The position data indicating the position is stored in the second counter as an absolute value.

The first thing that changes from moment to moment. The contents of the second counter are stored in a buffer register in the position detection circuit of the movable head by the sample command, and are determined by the position of the movable head relative to the zero point of the rotating code disk at the time of the sample command, that is,
The rotation angle of the shaft is stored in a buffer register.

On the other hand, the buffer register of the fixed head position detection circuit that inputs the second index input signal and the second zero point pulse signal contains the position of the fixed head relative to the zero point of the rotating code disk when the sample command is generated, that is, the fixed head position. The rotation angle of the case is stored in absolute form.

By taking the difference between the data stored in the buffer registers in the position detection circuits of the movable head and the fixed head, the rotation angle of the shaft with respect to the fixed case can be detected in absolute terms.

(Example) structure The present invention will be described in detail below with reference to the drawings.

FIG. 1 shows a cross-sectional view of a high-resolution magnetic absolute encoder with a resolution of 262,144 P/R (pulses/rotation) according to the present invention, and FIG. 2 is a perspective view of the main parts of the encoder shown in FIG. .

In FIG. 1, reference numeral 1 denotes a shaft for inputting the detected rotation angle θ, and this shaft 1 is held in a fixed case 6 via a bearing 7 and can freely rotate about one rotation axis 2-2.

This shaft 1 rotatably holds a ring 9 via a second bearing 8, and this ring 9 supports a magnetic disk 5.
and a multipolar magnetized ferrite magnet 10 is fixedly held.

Further, the shaft 1 fixedly holds a movable magnetic head 3, and this magnetic head 3 faces the magnetic recording surface of the magnetic disk 5. The shaft 1 fixes and holds the rotor side of the rotary transformer 4. The detection coil winding output of the movable magnetic head 3 is connected to the rotor side coil winding of the rotary transformer 4. The stator side of this rotary transformer 4 is fixed onto a printed board 13 held by a stationary case 6.

The stator side coil winding output of this rotary transformer 4 is input to an amplifier 15 on a printed board. The output of this amplifier is also connected to terminal 17.

Further, the fixed case 6 holds a printed board 12, and the motor coil 11 is fixedly held on this printed board 12.

This motor coil 11 is connected to the ferrite magnet l.
Place it opposite to O.

Further, the fixed case 6 holds a fixed magnetic head 2, and this magnetic head 2 faces the magnetic recording surface of the magnetic disk 5. The detection coil winding output of this fixed magnetic head 2 is input to an amplifier 14 on a printed board 13. The output of this amplifier 14 is connected to a terminal 16.

Function> The multi-pole magnetized ferrite magnet 10 and the motor coil 11 constitute a brushless motor, and the stator coil 11
applies rotational force to the rotor magnet 10, and the ring 9. A magnetic disk 5 is rotated around a shaft 1 at a constant rotation speed of 3600 rpm relative to a fixed case 6. The brushless motor described above can be easily implemented using a method currently widely used in floppy disk drives and hard disk drives.

Now, if the maximum permissible rotation speed of the shaft 1 is set to 80 Orpm or less, a high relative speed is obtained between the magnetic heads 2 and 3 and the magnetic disk recording surface, and an air layer is generated between them, and the air bearing action causes the recording surface to Heads face each other without contact.

Next, the operation of this embodiment will be explained with reference to FIG.

Tracks 18 on the magnetic recording surface of the magnetic disk 5 face the fixed magnetic head 2, and tracks 19 face the movable magnetic head 3.
This truck ta, 19 has 1 each.
An index pattern in which the N/S magnetization changes at a pitch that accurately divides the circumference into 2048 equal parts is recorded.

Here, in order to explain the action of the encoder shown in FIG.
This will be explained using the timing diagram shown in the figure. still.

FIG. 3 is a timing diagram when the shaft 1 is stopped at a constant rotation angle θ.

Now, when the magnetic disk 5 is rotating at 360 rpm, the fixed magnetic head 2 outputs the index pulse voltage shown in FIG. 3a every 1/2048 rotation of the disk 5.
An index pulse voltage shown at C is induced from the movable magnetic head 3. The pulse frequency of these index pulses a and a is 2048 x 3600/60 = 122.88K
It is Hz. In order to be able to identify a specific index pulse among the 2048 index pulses recorded in tracks 18 and 19 of the magnetic disk 5, the zero point pulse shown in d in FIG. Only one place was recorded. These zero point pulses b and d are approximately 1/8192 (1/
4 index) recorded as occurring after rotation.

Now, when the rotation angle of encoder shaft 1 is θ,
The time phase difference between the zero point pulse b generated from the fixed magnetic head 2 and the zero point pulse d generated from the movable magnetic head 3 is θ as shown in FIG.

Now, since the accuracy of the 2048 index magnetization patterns recorded in one rotation on the magnetic disk 5 determines the accuracy of the encoder of the present invention, the magnetic disk 5 is manufactured using the next master magnetic disk copying device.

In other words, a high-precision master magnetic disk on which 262,144 clock pulses are recorded at equal pitches on one track and a magnetic disk 5 used in this embodiment are combined into 360 disks.
It is coaxially fixed to the shaft of a copying machine that rotates with an Orpm, reads the clock of the master magnetic disk with a master magnetic head, and writes an index pulse to the magnetic disk 5 with a magnetic head dedicated to copying every time this clock generates 128 pulses. . Thus, 2048 index pulses are written with high precision per track revolution. Also, one zero point pattern is written at the same time during one round. With such a master magnetic disk copying device, the magnetic disk 5 is copied with the precision of the master magnetic disk, and a large number of magnetic disks 5 can be manufactured in a short time.

Now, the length of one lap of track 18.19 in Figure 2 is 150.
If m, to divide this one rotation into 262,144 equal parts, 1 pitch = 150 m/262,144 = 0
．． Although it is necessary to record on the recording surface at a wavelength of about 6 m, this can be sufficiently achieved since the recording wavelength of the floppy disk devices and hard disk devices currently in practical use is 0.5-.

Output signals a and b from the fixed magnetic head 2 enter an amplifier 14, where the power is amplified and outputted to a terminal 16, and transmitted to a position control device to be described next through a cable (not shown).

On the other hand, the output signals c and d from the movable magnetic head 3 pass through the rotary transformer 4 and enter the amplifier 15, where the power is amplified and outputted to the terminal 17, and similarly transmitted to the position control device.

A position detection circuit is provided inside the position control device, and processes the pulse trains of signals a, b, and signals c, d to convert the rotation angle O of the shaft 1 into an absolute digital quantity.

FIG. 4 is a block diagram for explaining the operation of this position detection circuit.

In FIG. 4, two blocks 20, 2 indicated by broken lines
1 has exactly the same circuit configuration, and the input terminal 22 of the block 20 is connected to the output terminal 16 of the encoder, and signals a and b from the fixed magnetic head 2 are input thereto. The output port 24 of the block 20 is then connected to the input of the microcomputer 28.

Further, the input terminal 23 of the block 21 is connected to the output terminal 17 of the encoder, and the input terminal 23 of the block 21 is connected to the output terminal 17 of the encoder, and receives the signal c from the movable magnetic head 3.
, d are input. Then, the output port 25 of the block 21 is connected to the input of the microcomputer 28. The sample command outputs from the microcomputer 28 are also connected to input terminals 26, 27 of blocks 20, 21, respectively.

Since the operations of blocks 20 and 21 are exactly the same, only block 21 will be described in conjunction with the timing diagram of FIG.

Output signals c and d from the movable magnetic head 3 are input to the input terminal 23.
through MM (mono step multi) 29, AND
Circuit 30. The signal is input to an INH (in pit) circuit 31. When this index pulse triggers MM29,
The output of MM29 is approximately 50% after the trigger as shown in signal e.
%Duty period 1 (state of 11#.This MM2
The output of the AND circuit 30 serves as a window for the zero point pulse d, and only the zero point pulse d is extracted from the output of the AND circuit 30 as shown in the signal f. On the other hand, the INH circuit 31 blocks only the zero point pulse d, and only the index pulse
appears in its output. The output from the circuit 31 enters the reference phase input terminal (R) of the phase comparator 32. This phase comparator 32, a VCO 33 whose frequency is controlled by its output voltage, and a binary 7 that divides the frequency of this VCO 33 into 1/12g.
The digit frequency division counter 34 constitutes a PLL (phase locked loop), and an OVF (overflow) occurs when the content of the frequency division counter 34 changes from 127 to 0.
The pulse is input to the feedback phase input terminal (FB) of the phase comparator 32.

By this PLL, not only the frequency of the index pulse and the ovF pulse but also the phase difference between them is set to 0.The output frequency of the VCO 33 is adjusted to 128 x 122.88 (7=15. 728.64
The contents of the frequency division counter 34, which increments at this frequency, change as shown in the signal diagram in Fig.
Synchronize so that the content changes from 127 to 0.

Note that when the rotation angle θ of the encoder shaft 1 changes, the phase of the index pulse also changes, so it is necessary to make the frequency of the VCO 33 follow this change, but the phase comparison sampling frequency of this PLL is 12
It is extremely high at 2.88 KIIz, and the maximum long frequency of the shaft l of a normal encoder is at most about 7 in 1, so the PLL can be followed with sufficient accuracy.

With this PLL, the contents of the frequency division counter 34 are changed from 0 to 127 during one index pulse as shown in the signal.
The index pulse period is interpolated into 128 equal parts.

The OVF pulse from the frequency divider counter 34 is connected to the clock input of an 11-digit binary counter 35, and each time an OVF pulse occurs, the contents of the counter 35 are incremented as shown in FIG. 3j. And the 2048th OVF
It is cleared to zero by a pulse and the increment operation is repeated again.

Further, the output of the AND circuit 30 is connected to the reset terminal R of the counter 35, and the zero point pulse d resets the counter 35 and makes its contents zero.

This operation is performed as initialization and backup.

The counters 34 and 35 correspond to lower digits and upper digits, and the 18 binary digit digital amount that is the concatenation of the contents is the movable head 3.
This means that the relative rotational position between the zero point and the rotating magnetic disk 5 is shown moment by moment with a high resolution of 2''=262,144.

Similarly, the rotational position between the zero points of the fixed magnetic head 2 and the rotating magnetic disk 5 is detected by the circuit of block 20.

Now, when the microcomputer 28 in the position control device reads the rotational position of the encoder, it generates a sample pulse from the sample command output and sends it via the terminal 27.
The LD (load) terminal of the 18-digit binary buffer register 36 is activated, and the contents of the counters 34 and 35 are stored in the buffer register 36. Thus, the rotation angle θ of the movable head 3 with respect to the zero point of the magnetic disk 5 at the time of the sample command is outputted from the buffer register 36 with a resolution of 211 and transferred to the output port 25. The terminal 26 is also activated at the same time, and the position of the fixed head 2 relative to the zero point of the magnetic disk 5 at the time of the sample command is transferred from the output port 24 of the block 20.

The microcomputer 28 reads each 18-bit data from the output ports 25 and 24, and subtracts the latter data from the former data, and the result is the rotation angle θ of the movable head 3 with respect to the fixed head 2, which is 21 in absolute terms.
This results in indexing and detection with a resolution of 0. This is absolute data of the rotation angle θ of the shaft 1 relative to the fixed case 6 to be detected.

<Effects> In this embodiment, if the maximum allowable rotation speed of the encoder shaft is 80 (lrp- or less), a high relative speed is generated between the magnetic disk 5 and the movable magnetic head 3 and the fixed magnetic head 2. An air bearing air layer can be easily obtained, and the space between them can be made non-contact, and the magnetic recording surface will not be damaged.Also, as a countermeasure when an air bearing air layer cannot be obtained when the power is turned off, air bearing surfaces on both sides of the magnetic head can be used. The recording track can be protected by making it higher than the air bearing surface of the magnetization detection gap at the center of the head to prevent the center of the head from contacting the magnetic recording track.

According to the embodiment, a highly accurate magnetic disk 5 can be manufactured by copying from a master magnetic disk in a short time. The master magnetic disk copying device used for this copying can also be easily manufactured with a simple mechanism. The code disks used in conventional encoders were manufactured using methods such as optical phosphorography, but the manufacturing equipment used in this manufacturing method was extremely expensive.

Due to the large scale, production of this code disk conventionally required a large amount of cost.

In this embodiment, rotation unevenness of the magnetic disk 5.

Wow and flutter cause jitter in the phase of the index pulse and cause a phase error in the PLL loop, but when the sampling frequency of the PLL loop is 122.
Because of the high speed of 88KHz, the response speed of the loop can also be increased.

On the other hand, it is easily possible to eliminate such high frequency wow and flutter due to the inertia of the magnetic disk 5. Furthermore, since both the PLL loops in block 20 and block 2■ in Fig. 4 exhibit completely the same phase error characteristics with respect to wow and flutter, the influence of disk rotation unevenness is canceled and the accuracy of position detection is reduced. do not have.

Further, even if the rotational speed of the magnetic disk 5 changes, the PLL follows it, so the accuracy of position detection does not deteriorate.

Further, according to this embodiment, since a high relative speed is generated between the rotating magnetic disk 5 and the magnetic heads 2 and 3, there is an advantage that an inductive magnetic head can be used.

(Other Embodiments) Configuration As another embodiment of the present invention, an optical absolute encoder with a resolution of 262,144 pulses/R will be described next. FIG. 5 shows a sectional view of the optical encoder of the present invention, and FIG. 6 is a perspective view of the main parts of the encoder shown in FIG.

In FIG. 5, reference numeral 37 denotes a shaft for inputting the detected rotation angle θ, and this shaft 37 is held in a fixed case 38 via a bearing 56 and rotates around a rotation axis Z-Z.

This shaft 37 has discs 58, 59. The printed board 60 is held fixed. Further, the shaft 37 is connected to the second bearing 5
7, a ring 61 is rotatably held, and this ring 61 fixedly holds a code plate 39, and the code plate 39 holds a magnet 48 and a multipolar magnetized ferrite magnet 49.

A disk 58 holds an LED (light emitting diode) 43 and a Hall element 47, and a disk 59 holds a slit plate 41 and a PD (photodiode) 45. The light from the LED 43 is transmitted to the code plate 39. Through the slit plate 41, P
Arrange it so that it is incident on D45. In addition, the Hall element 47
is arranged so as to detect the magnetic flux of the magnet 48.

The printed board 60 fixedly holds the electronic circuit 51 and the rotor side of the rotary transformer 52. The 0-rotation transformer 52 has two channels, and both of its rotor side coil ends are wired to the electronic circuit 51.

Also LED43. Hall element 47. Each lead terminal of the PD45 is connected to an electronic circuit 51 by a zero disk 58.59.
, the printed board 60 rotates together with the shaft 37, so the LEDs 43. Hall element 47° slit plate 41. PD45. Each component of the rotor of the electronic circuit 512 rotating transformer 52 also rotates together with the shaft.

A printed board 62 held in the fixed case 38 holds the stator side of the rotary transformer 52 and the second electronic circuit 53.

The two stator side coil ends of this rotary transformer 52 are wired to an electronic circuit 53.

The fixed case 38 is L E D42. Hall element 46. Slit plate 40. PD44. The motor coil 50 is held fixed. The light from the LED 42 is transmitted to the code plate 39.
The light is arranged so that it passes through the slit plate 40 and enters the PD 44. Further, the Hall element 46 is arranged to detect the magnetic flux of the magnet 48. Also, the motor coil 50.

It is arranged so as to face the ferrite magnet 49 on the code plate 39. And L E D42. Hall element 46. Each lead terminal of the PD 44 is wired to a second electronic circuit 53, respectively. The electronic circuit has output terminals 54,55 and power terminals 63.

<Function> Figure 5 shows the multi-pole magnetized ferrite magnet 49 and the motor.
The coil 50 constitutes a brushless motor, and the code plate 39
is rotated around the shaft 37 at a constant rotation speed of 3600 rpm with respect to the fixed case 38.

As shown in the perspective view of FIG.
The slit plates 40, 41 have slits of equal pitch.
A slit with the same pitch as this is cut. When the code plate 39 rotates at 3600 rpm+s, the LEDs 42 and 4
The amount of light incident on PD44, 45 from code plate 39
It is modulated by the action of the slit plates 40 and 41.

A signal a+J having a waveform shown in FIG. 7 appears from PD44.45.

FIG. 7 is a timing diagram when the shaft 37 is stopped at a constant rotation angle θ, where a is the output signal from the PD 44, and j
indicates the output signals of Pd45. These output signals repeat the same waveform every 1/2048 rotation of the code plate 39, and the frequency is 2048 x 3600/60 = 12
2.88 to seven.

On the other hand, as shown in FIG. 6, since the magnet 48 is attached to the code plate 39 rotating at 360 rpm, each time the magnet approaches the Hall element 46, 47, the Hall element 46
A signal with a waveform shown in d appears from the Hall element 47, and a signal with a waveform shown in m appears from the Hall element 47. These outputs are output when the code plate 39
Appears only once per rotation.

FIG. 8 is a functional block diagram of the electronic circuit 51, 53 shown in FIG.

The DC power supply voltage of the encoder of this embodiment is supplied through the power supply terminal 63 of the electronic circuit 53. This DC voltage causes the LED 42 to emit light via the terminal 64.

Further, this DC power supply voltage is converted into an AC voltage by an inverter 72. This AC voltage is applied to the first
is supplied to the rectifier 73 in the electronic circuit 51 through the channel. The rectifier 73 reconverts the DC voltage to create the necessary DC power for the electronic circuit 51. This DC voltage is applied to terminal 62.
The LED 43 is made to emit light via the .

Next, the operation of the electronic circuit will be explained with reference to the timing diagram of FIG.

The output of PD44 enters the input terminal 66 in FIG.
It is converted into a rectangular wave by a P (comparator) 74 and then converted into a pulse C by an MM (mono-stepple multi) 76.

This pulse C is applied to a clock input of an FF (flip-flop) 78 and an OR circuit 84. On the other hand, Hall element 4
The output d of 6 is inputted to the D terminal of the FF 78 via the input terminal 68, and a signal having the waveform shown in e appears from the Q output of the FF 78. The MM80 is triggered by the rising edge of the signal e, and its output changes as shown by the signal f. The signal f
The output waveform of MM82 triggered by the falling edge of is converted into a pulse indicated by g. The OR circuit 84 receives the Qtg pulse and outputs a pulse indicated by h, and this pulse is power amplified by the amplifier 86 and output to the encoder output terminal 54.

The output j of the PD 45 enters the input terminal 67 and is converted into a rectangular wave of k by the GOMP 75, and converted into a pulse Ω by the MM 77. This pulse α is ORed with the clock input of FF79.
added to circuit 85. On the other hand, the output m of the Hall element 47
is input to the D terminal of FF 79 via input terminal 69,
A signal having a waveform shown in n is obtained from the Q output of the FF 79. The MM81 is triggered by the rise of the signal n, and its output changes as shown by the signal p. A signal having a waveform shown in q is generated from the output of the MM83 triggered by the fall of the signal p. The OR circuit 85 receives pulses α and q as input and outputs a pulse shown as r. This pulse r is transmitted through the second channel of the rotary transformer 52 to the input of the amplifier 87 of the electronic circuit 53, power-amplified by the amplifier, and then output to the encoder output terminal 55.

Now, if the rotation angle of the shaft 37 is θ, the time phase difference between the rise of the output signal e of the FF 78 and the rise of the output signal n of the FF 79 is θ as shown in FIG.

Now, if we compare the timing diagrams in Figure 7 and Figure 3, we get
The pulse waveform shown in Figure 7 and Figure 3a.

The relationship is the same with the pulse waveform shown in b, and the pulse waveform shown in FIG. 7r and the pulse waveform shown in FIGS. 3c and d are the same.

Therefore, the same position detection circuit shown in FIG. 4 used in the embodiment shown in FIG. 1 can also be used in the optical encoder of this embodiment shown in FIG. 55 to the input terminals 22 and 23 in FIG. 4, the rotation angle θ of the shaft 37 of the microcomputer 28 in FIG. 4 can be determined and detected in absolute terms with a resolution of 2111.

Effect> In this embodiment, the accuracy of the slits in the code plate 39 determines the position detection accuracy of the encoder, so it is necessary to make them accurately, but it is sufficient to print 2048 identical slits at an equal pitch. It can be manufactured with high precision using methods such as lithography.

Furthermore, since the code plate 39 rotates without contacting other parts, it is highly reliable.

Furthermore, the detection timing precision required for the Hall element 46 is such that even if the rise timing of the output of the Hall element 46 varies between the pulse output signals C of the MM76, as shown in FIG.
Since the rise timing of the output signal e of FF78 does not change,
The detection timing accuracy of the Hall element 46 is 1/1 of the code plate.
It is sufficient as long as it is 2048 rotations or less. The same can be said about the Hall element 27.

〔Effect of the invention〕

According to the present invention, the signal transmission cable between the encoder, the position side, and the control device only needs to be capable of transmitting two systems of signals. This is a conventional absolute type encoder, and as in the present invention, 2
” = 2621.44 pulses/R resolution means that the cable interface is significantly simplified compared to the one that required 18 signal lines.

Incidentally, even with the currently widely used incremental encoder with a zero point, it is necessary to send three signals of A phase, B phase, and Z phase, but this is more advantageous than this.

The frequency of the signal transmitted through the two signal lines of the present invention is 122
．． The frequency is 88KHz, which is not a particularly high frequency, so it can be transmitted using inexpensive cables. Further, since the signal is transmitted at a substantially constant frequency, the noise resistance can be easily improved by using a noise reduction filter or the like.

Furthermore, according to the present invention, the index pulse has a substantially constant frequency of 122.88 KHz, and the zero point pulse has a substantially constant frequency of 607, so if the two systems of signals do not come in around these frequencies, it is likely that the encoder is malfunctioning. This makes it possible to easily detect encoder failures.

In addition, the PLL of the position detection circuit of the present invention filters out high frequency noise and performs smooth detection.

Further, according to the present invention, like a code plate provided with a plurality of tracks like a conventional absolute encoder,
Complex code boards are no longer required, and multi-track skew problems are eliminated.

Furthermore, if 2048/R patterns, which are not so many, are recorded on a magnetic disk or an optical code disk, a resolution of 2 L9 can be easily obtained. It is no longer necessary to make the slit pitch so small that the diffraction phenomenon becomes a problem.

[Brief explanation of the drawing]

FIG. 1 is a sectional view showing a magnetic encoder according to an embodiment of the present invention, FIG. 2 is a perspective view of the main components of the encoder shown in FIG. 1, and FIG. 3 is a timing diagram of an embodiment of the present invention. 4 is a block diagram showing one embodiment of a position detection circuit according to the present invention, FIG. 5 is a sectional view showing an optical encoder according to another embodiment of the present invention, and FIG. 7 is a timing diagram of the encoder shown in FIG. 5, and FIG. 8 is a functional block diagram of an electronic circuit for the encoder shown in FIG. 5. DESCRIPTION OF SYMBOLS 1...Shaft, 2...Fixed magnetic head 3...Movable magnetic head, 4...Rotating transformer 5.
...Magnetic disk, 6...Fixed case 10...
Ferrite magnet 11...Motor coil 28...Microcomputer 29...MM, 30...AND circuit 31...INH circuit, 32...Phase comparator 3
3...VCO34, 35...Counter 36...Buffer register, 37...Shaft 38...Fixed case, 39...Code plate 40, 41...Slit plate, 42.43...・LED44, 45...P
D, 46.47... Hall element 48... Magnet 49... Ferrite magnet 50... Motor coil, 51.53... Electronic circuit 52... Rotating transformer, 72... Inverter 73 ... Rectifier, 74.75-GOMP7
6, 77, 80.81.82.83...MM78, 7
9...FF, 84.85...OR circuit 86
, 87... Amplifier agent Patent attorney Nori Ken Chika Yudo Ken Daishimaru Figure 1 Figure Figure □ Time Figure Figure Time

Claims (2)

[Claims] (1) A fixed case, a shaft that is rotatably held around the axis of the fixed case and detects the mechanical rotation angle, and a shaft that is rotated at a constant speed around the axis of the fixed case and arranged at a constant pitch on the circumference. Index patterns are recorded at equal intervals,
a rotary code disk on which a zero point pattern for specifying one of the plurality of index patterns is recorded; a movable head fixed to the axis of the shaft for reading out the index pattern and zero point pattern of the rotary code disk; A high-resolution absolute type encoder, characterized in that a fixed head is fixed to the fixed case and reads out an index pattern and a zero point pattern of the rotary code disk. (2) A first counter to which the index pulse and zero point pulse read by the movable head and the fixed head are respectively input, multiplies the frequency of the index pulse, and increments at the multiplied frequency; a phase-locked loop (PLL) circuit that synchronizes an overflow pulse of the rotary code disk to the index pulse;
A movable head detection circuit and a fixed head detection circuit each comprising a second counter that overflows at a number of pulses equal to the index pattern of the circumference, and a buffer register that stores the contents of the first counter and the second counter according to a sample command. A high-resolution absolute type encoder, characterized in that it detects the rotation angle of the shaft at the time of a sample command.