JP3225547B2 - Encoder - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an encoder for generating an absolute position, and more particularly, it can be used for a DC servomotor, an AC servomotor or the like which requires a high-resolution absolute position.

[0002]

2. Description of the Related Art Conventionally, an absolute encoder has been used by Yaskawa Electric.
As shown in Vol. 51, No. 196, No. 3, 1987, p257, it is composed of a rotating disk on which position information per rotation is recorded, and a part for counting the number of rotations. Since the position information per rotation is recorded as an absolute value on the rotating disk, the resolution per rotation is 1
If it is 1 bit, the rotating disk is provided with at least 11 signal slits. These eleven signals are detected by eleven sensors and processed as binary signals. The absolute value that has been converted into a binary signal is combined with the count value of the number of rotations, and is output as an encoder output signal of a multi-rotation absolute value. In the case of Yaskawa Electric, the rotation signal of the multi-rotation absolute value is output as a serial signal in A phase,
Position information per rotation is output as an initial incremental pulse of the A phase and the B phase. For the multi-turn absolute value encoder, see Samtak Corporation, Rotary Encoder / Rotation Sensor General Catalog VOL.02 199
January 0 There is an AEM series at P106. This is 1
There is a sensor that detects the absolute value signal per rotation (11 bits and at least 11 slits are on the rotating disk) and a sensor that detects the number of rotations, detects the signal of the slit on the rotating disk and counts the number of rotations And the position per rotation. The counted number of rotations and the position during one rotation are combined by an absolute register, and become 24-bit multi-rotation absolute position information, which is output via a modem.

[0003]

In the above prior art, it was necessary to provide a slit on the rotating disk in accordance with the required position resolution to obtain position information during one rotation. That is, assuming that the required resolution is 2048 pulses / rotation (2 11 pulses / rotation), at least 11 slits are required. If two phase signals (RA phase and RB phase) for detecting the A phase, the B phase, the Z phase, and the number of revolutions are provided, five more slits are required. For this reason, there is a problem that the rotating disk becomes large and a large number of sensors are required, so that it becomes expensive and cannot be downsized. In addition, the number of parts increases, and there is a problem that the number of manufacturing steps increases, the reliability decreases, and the yield is poor. Further, when the number of A-phase and B-phase pulses is increased to increase the resolution, it is necessary to increase the absolute value per rotation. Therefore, it becomes impossible to provide a slit on a rotating disk of the same size, or There is also a problem that must be increased. As described above, the resolution improvement and the encoder size are contradictory issues, and the size of the encoder attached to the motor is a major restriction on the product.

An object of the present invention is to provide a small and inexpensive high-resolution absolute value encoder using a small encoder disk such as a current incremental encoder.

[0005]

In order to achieve the above-mentioned object, many uses of a high-resolution multi-rotation absolute value encoder are mainly intended to eliminate the conventional home position adjustment time and home position adjustment operation. The rotation of the motor is often acceptable. Therefore, instead of detecting all positions during one rotation as absolute values, a counter that presets the absolute value during one rotation at this reference position based on a plurality of positions during one rotation is provided in the encoder. I did it. The counter counts incremental signals after presetting, and constantly detects the absolute value of an arbitrary position during one rotation. Furthermore, if the detected value and the count value of the number of revolutions are stored in an absolute data register and output via a modem, the same function as the above-described multi-turn absolute value encoder can be achieved.

[0006]

It is necessary to provide at least one reference position during one rotation, and a signal indicating this position is provided on the slit. Reference position signal obtained from slit or
Since the reference position signal generated based on the signal obtained from the slit indicates a specific position during one rotation,
An accurate absolute position during one rotation is obtained. By presetting the absolute position in the counter, the absolute value of the reference signal generation position becomes the value of the counter, and the value of the counter indicates the accurate absolute position during one rotation. By counting the amount of rotation obtained by the incremental signal with the counter based on the preset position, the absolute position of an arbitrary position during one rotation can be obtained as the value of the counter.

This reference position signal can be generated by the edges of the U, V, and W phases when an incremental encoder with U, V, and W phases for an AC servo is used, for example. On the other hand, an absolute value encoder having a low resolution
When using a gray code signal, it can be generated by a changing edge of a gray code signal often used as a slit pattern. In the case of this absolute value encoder, a signal for operating the rotation speed counter is obtained from the upper signal of the gray code signal.

As described above, it is not necessary to provide a large number of absolute value slits on the rotating disk, and a high-resolution absolute value encoder can be manufactured within the technical range of a conventional incremental encoder. Never be.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, U, V, W
An embodiment in which an incremental encoder is used will be described with reference to FIGS. In FIG. 1, a rotation speed correction output unit (85) and a rotation speed correction signal generation unit (89)
, Except that the output of the circuit 83 is directly input to the circuit 41, and will be described using the symbols RA, RB, U, V, and W in the figure. Thus, the figure shows a 4-pole, 2048 pulse / rev AC
Servo encoder [Commitment signal (CS signal)
U, V, and W-phase signal-equipped encoder] to which the signals RA and RB-phase signals for detecting the rotational speed are added.

First, the quadrupole motor shown in FIG. 2 will be described. The rotating disk 11 (desk) of the encoder is
A slit pattern as shown in FIG. 9 is formed, and each pattern generates the next target signal. .RA and RB with 90-phase difference to count the number of rotations
Two-phase incremental signal consisting of two phases ・ CS signal consisting of U, V, W phases for detecting magnetic pole position ・ IA, IB with 90-phase difference to detect rotation
A two-phase incremental signal consisting of two phases, where the edge of the RA phase is 0-, 18 for the rotation angle of the motor of 0-360-
It is assumed that the number of rotations is counted by the edge change of the RA phase when the RB phase is High at the 0-, 360-position. Also,
The edge of the U phase of the CS signal U, V, W phase is 0-, 90-, 180-, 2
70-, 360-; V-phase edges are 60-, 150-, 240-, 330-; W-phase edges are 30-, 120-, 210-, 300-. By the combination of the three-phase CS signals, the number of rotations of the motor becomes
The rotation angle region of the motor can be divided into twelve divisions by combining them with the RA phase or the RB phase.
D12 can. On the other hand, IA, IB indicating the rotation amount
Assume that there are 2048 phase incremental pulses per revolution and that the edges of the IA phase are at motor positions 0- and 360-.

In the above encoder, RA,
The number of rotations is counted as −1, 0, +1 by the RB phase, and I
The rotation amount is counted from 0 to 2047 (multiple amount of the incremental pulse × 1) according to the A and IB phases. By combining the result of counting the number of revolutions and the count amount of the incremental pulse, the function of the absolute value encoder is achieved.

The present embodiment is to simply configure an absolute encoder by permitting an operation during one division among the 12 divided areas and updating the absolute position at the boundary of the divided areas.

Next, the configuration of an embodiment of the present invention will be described with reference to FIG. The light of the LED (LD1, LD2) passed through the slit on the rotating disk 11 (desk) is received by the photodiode (PD), the waveform is shaped by the comparator (CMP), and the signals RA, RB, U, V, W are output. , IA, and IB are obtained. The signals RA and RB are counted by the rotation number counting means 15 to count the rotation number. The LED (LD1), the photodiode (PD), the comparator (CMP) and the rotation number counting means 15 are designed to be able to count and store the rotation number even when the power is off by driving the battery back. . The other LEDs, photodiodes, and circuits operate when the main power supply Vcc of the encoder is applied. The range of the battery backup is set to 21, 23, 25, 27, 31, 3 in addition to the above.
If it is expanded to 3,41 elements, the absolute position can be stored even during a power failure.

At least one of the rotation speed detection signals RA, RB
Then, the CS signals U, V, W are input to the rotational position detecting means 21 (moving position detecting means). Rotational position detecting means 21
Outputs a sampling pulse SP every time each edge of the U, V, and W phases is detected. Further, the rotational position detecting means 21 outputs to the absolute value setting means 27 a signal AD (position specifying signal) indicating the rotational position of the encoder divided into 12 at the time of outputting the sampling pulse SP. Upon receiving the AD signal, the absolute value setting means 27 determines an accurate absolute position per one rotation when the sampling pulse sp (position reference signal) is output by the rotation amount counting means 33 (movement amount counting means) [SN74193. Such a presettable counter: hereinafter referred to as a counter 33]. The absolute value setting means 27 may be a storage element such as a ROM that outputs constant information according to an input.

On the other hand, when Vcc power is turned on, a reset signal is output from the reset signal generating means 23, and a reset sampling pulse rsp is output. An element such as Mitsubishi M51967BL is used to generate the reset signal.
The preceding rsp and the sampling pulse sp are ORed by the OR circuit 25 (preset means), and become the load signal SP of the counter 33. Further, the rotation amount detecting means 31 (movement amount detecting means) receives the incremental signals IA and IB and receives the signals from Japanese Patent Publication No. 60-492.
4. As shown in Japanese Patent Application No. 58-109812, a pulse indicating the determination of the rotation direction and the rotation amount, an up pulse, and a down pulse are output. this
The up pulse and the down pulse are input to the counter 33 and update the value of the counter. As described above, the counter 33 always holds the absolute value of the rotational position by counting the incremental signal. Although the pulse is not doubled in the description of the present embodiment, it is natural that the contents of the preset data only need to be changed when the pulse is doubled and used, ie, × 2 or × 4.

The rotation amount counting means 33 and the rotation number counting means
The data stored in 15 is input to the multi-rotation absolute value generation means 41 and becomes multi-rotation absolute value data. The multi-turn absolute value data is transmitted to the higher order via the modem 43 and the line driver. Here, 15, 41 and 31 are AEM made by Samtaku
This can be achieved by using an encoder. As the modem, DN1811 manufactured by Mitsubishi Cable Industries, or the like can be considered.

Next, the function of this embodiment will be described with reference to FIG. In this example, the rotation angle of the motor is from 120 ° to 150 °
It is assumed that the Vcc power supply is turned on at the position. Vc
c When the power is turned on, the reset sampling pulse rsp indicates the motor rotation angle 135 ° (HEX data: 2F
F) is loaded into the counter 33. Next, the motor rotates (in the case of normal rotation, the sampling pulse sp +
A sampling pulse sp + 2 is generated at a position of 1,180 °, and in the case of reverse rotation, a sampling pulse sp-1,90 is generated at a position of 120 °.
When the sampling pulse sp-2 is generated at the position of °), the sampling pulse sp is generated from the rotational position detecting means 21. This pulse is generated by the edge of the CS signal, becomes an accurate position signal of the motor rotation angle, and presets the counter value to a normal value. Thereafter, by counting the incremental signals IA and IB, the value of the counter indicates the correct rotation angle. FIG. 4 shows a rotational position detecting means.
21 shows a specific example of the OR circuit 25 and the reset signal generating means 23. However, the circuits 22, 34, and 36 will be described later, and the basic operation without this circuit will be described.

The operation of the rotational position detecting means 21 will be described. C
The S signal U, V, and W phase signals are sampled in synchronization with the clock signal by two-stage D-type flip-flops (hereinafter abbreviated as D-FFs). The output of each D-FF becomes the A0 and A1 inputs (A2 and A3 for the V phase and A4 and A5 for the W phase) of the absolute value setting means 27 in the case of the U phase, and serves as a signal for specifying the absolute value. Further, at least one of the rotation speed detection signals RA and RB is sampled by the clock signal, and becomes an A6 input of the absolute value setting means 27. Absolute value setting means 27
Outputs the absolute value per one rotation of the motor as preset data of the counter 33 according to the inputs A0 to A6. The two-stage D-FF output is input to an EX-OR,
When both signals (A0 and A1, A2 and A3, A4 and A5) do not match (when the edge of each phase is detected), a sampling pulse sp of 1 clock width is output.

On the other hand, the reset signal generated by the reset signal generating means 23 is sampled by two stages of D-FFs, and a reset sampling pulse rsp by the reset signal is output by the EX-OR. When the sampling pulse sp and the reset sampling pulse rsp are NO
The signal is ORed by the R circuit 25, becomes the load signal SP, and becomes an input as a set signal of the counter 33. With this signal, the output of the absolute value setting means 27 is set in the counter 33. The setting of the absolute value by rsp is provisional, and the setting by sp is the true absolute position.

An RS flip-flop (hereinafter, RS flip-flop)
−FF) 51 is reset by a reset signal,
Set by the sampling pulse sp, the output of the RS-FF becomes one input of the ANDs 53a and 53b. Also,
Another input of the AND 53a is an up pulse signal of the rotation amount detecting means 31, and another input of the AND 53b is d of the rotation amount detecting means 31.
Connected to own pulse signal. This is because after the power is turned on, after the true absolute value is set in the counter 33 by the sampling pulse sp signal (reference position signal) obtained first, counting is performed using the incremental signals IA and IB. is there. As described above, when the power is turned on, the output of the counter 33 outputs a provisional absolute position by the reset sampling pulse rsp, and then outputs the true absolute position by the sampling pulse sp. Thereafter, the incremental signal is counted. While constantly outputting the true absolute position.

If the sp signals such as sp-2 and sp + 1 are used, a re-preset signal of an accurate absolute position is input while the counter 33 is counting, and the corresponding absolute value is preset again. Even if there is a miscount due to noise or the like during counting, it is corrected to an accurate value. According to this configuration, an encoder that is resistant to noise can be obtained.

Here, the period of the sampling clock clock needs to be shorter than the signal change period of the encoder, and is usually 1/8 or less of the shortest period so that 2 Clock or more is inserted between the shortest phase differences of the IA and IB signals. It is good to Here, an embodiment in which the incremental signals IA and IB are multiplied by 2 and multiplied by 4 will be described. In this case, in the embodiment of FIG. 4, a circuit 34 is added, and in the case of double magnification, the output data from the absolute value setting means 27 to the rotation amount counting means 33 is increased by 1 bit from D0 to D11 to 12 bits. Just do it. Further, when multiplying by 4, it is sufficient to set D0 to D12 and 13 bits. As described above, in order to cope with multiplying signals,
This can be handled by increasing the number of bits.

In FIG. 4, if the rotational speed detection signal RA is sampled by two stages of D-FFs and the outputs of the D-FFs are the inputs A6 and A7 of the absolute value setting means 27, at the change point of the signal RA This has the effect of ensuring the absolute position setting process.

Further, in FIG. 4, D-FF 55 and A
The case where ND56 is used will be described. As a result, of the sampling pulses sp due to the respective edges of the CS signal U, V, and W phases, only the only sp signal generated after reset (sp + 1 for normal rotation, sp-1 for reverse rotation) is valid. And is used as the load signal SP. FIG. 3 shows the state of the signal according to this example.

FIG. 5 shows the D-FF 51 and the AND of FIG.
A circuit example when 53a and 53b are deleted is shown. According to this circuit example, since the up pulse and the down pulse by the incremental signals IA and IB are always input, the value of the counter is always updated as shown in FIG. Also in this case, the true absolute position is preset in the rotation amount counting means 33 (counter) by the sampling pulse sp by each edge of the CS signal U, V, and W phases.
Counting the pulses keeps indicating the true absolute position.

The case of a 4-pole encoder has been described so far, but an 8-pole encoder for generating signals shown in FIG. 4 and an 8-pole motor will be described. In this case,
One cycle of the CS signals U, V, and W phases has a rotation angle of 90 ° of the motor, and four cycles are included during one rotation of the motor. Therefore, by using the rotation number detection signals RA and RB and the CS signals U, V and W, one rotation of the motor can be divided into 24 blocks by 15 °.

As an example of a circuit in the case of an 8-pole encoder, as shown in FIG. 4, each of the rotation speed detection signals RA and RB is sampled by a D-FF to specify one of the 24 blocks, and the absolute position is detected. It is necessary to change the output with respect to the input of the absolute position setting means 27 as the inputs A6 and A7 of the setting means 27. Other circuits and their functions can be handled in the same manner as in the case of the 4-pole encoder.

Tables 1 and 2 show the relationship between the input and output of the absolute position setting means 27.

[0029]

[Table 1]

[0030]

[Table 2]

Next, an embodiment in which an absolute encoder with a low resolution is used will be described with reference to FIG. 1 and FIGS. 2048pulse / rev as an incremental signal
And a gray code signal, GRAY1
An example in which six signals of Nos. To 6 are obtained will be described.
The number of divisions per rotation by the six gray codes is 2 ⁶ = 64, but considering the use for AC servo, a commit signal (C
(S signal) Conversion to U, V, W phases is assumed, and is a multiple of 6. Therefore, a description will be given on the assumption that a gray code signal that divides one rotation into 48 as shown in FIG. 8 is obtained.

The circuit function can be considered in the same manner as in the case of the encoder using the above-mentioned incremental signal with U, V, W.

First, referring to FIG. 8, the rotary disk 11 of the encoder is formed with a slit pattern for generating the next target signal.

Gray code signals GRAY 1 to 6 for roughly dividing the position per rotation. Two-phase signals G having a 90 ° phase difference to count the number of rotations.
RAY5, 6 (included in the gray code signal) IA, IB having 90 ° phase difference to detect the amount of rotation
Two-phase incremental signal consisting of three phases Here, GRAY6
Edge is at 0 °, 180 °, 360 °, GRAY
The edge of No. 5 is at 90 ° and 270 °, and the rotation speed is counted by the edge change of GRAY6 when GRAY5 is High. The edges of GRAY4 are 60 °, 120 °, 240 ° and 300 °; the edges of GRAY3 are 30 °, 150 °, 210 ° and 330 °; the edges of GRAY2 are 15 °, 45 °, 75 ° and 105 °. °, 135
°, 165 °, 195 °, 225 °, 255 °, 285 °, 315 °, 345 °; The edge of GRAY1 is 7.5 °, 22.5 °, 37.5 °, 52.5
°, 67.5 °, 82.5 °, 97.5 °, 112.5 °, 127.5 °, 142.5 °, 15
7.5 °, 172.5 °, 187.5 °, 202.5 °, 217.5 °, 232.5 °, 247.
5 °, 262.5 °, 277.5 °, 292.5 °, 307.5 °, 322.5 °, 337.5
°, 352.5 °. These six signals (GRAY6-1)
, The motor rotation speed can be divided every 7.5 °, and the motor rotation angle area is divided into 48 (RD1 to RD48)
can do.

On the other hand, the number of incremental pulses of the IA and IB phases indicating the amount of rotation is 2048 pulses per rotation.
Assume that the edge of the phase is at 0 ° and 360 ° of the motor. In the encoder as described above, GRAY
5, the number of rotations is counted as -1, 0, +1 by the signal GRAY6, and the number of rotations is counted from 0 to 2047 (multiple amount of the incremental pulse.times.1) by the IA and IB phases. By combining the result of counting the number of revolutions and the count amount per revolution by the incremental pulse, the function of the absolute value encoder is achieved.

The embodiment of the present invention allows a simple operation of the absolute value encoder by allowing the operation during one division among the 48 divided areas and updating the accurate absolute position at the boundary of the divided areas. It is in.

Next, the configuration of the embodiment of the present invention will be described with reference to FIG. The light of the LED (LD1, LD2) passed through the slit on the rotating disk 11 is received by the photodiode (PD), the waveform is shaped by the comparator (CMP), and the signal GR is obtained.
AY6, GRAY5, GRAY4, GRAY3, GRA
Y2, GRAY1, and IA, IB are obtained. Signal G
RAY6 and GRAY5 are counted by the rotation number counting means 15 to count the rotation number.

At least one of the gray code signals GRAY6-1 to GRAY1 is input to the rotational position detecting means 21. In this example, it is assumed that all six signals are input to the rotational position detecting means 21. The rotational position detecting means 21 outputs a sampling pulse sp every time the edge of the designated gray code signal is detected. Further, the rotational position detecting means 21
Represents a signal AD indicating the rotational position of the divided encoder at the time when the sampling pulse sp is output, the absolute value setting means 2
7 is output.

The gray code signal input to the rotational position detecting means 21 does not have to be all of GRAY 6-1.
What is necessary is just to match the required number of divisions. Further, in this example, an example is shown in which edge detection is performed using only the GRAY1 signal. However, an edge of each gray code signal or a composite signal obtained by EX-OR (Exclusive-OR) of each of the GRAY6 to GRAY2 signals may be used.

Table 3 shows position data per rotation corresponding to the edge of each signal shown in FIG. FIG.
As shown in (1), one rotation can be divided every 7.5 ° by a combination of the signals GRAY1 to GRAY6. Therefore, assuming that the resolution of one rotation is 2048, the change point of each division position is given by 2048 × (7.5 × N / 360) (N = 0, 1, 2,...). At this time, the position data corresponding to each division position
In order to make the data integer, a value less than (7.5 × N) ° is a rounded-down value, and a value greater than (7.5 × N) ° is a rounded-up value. That is, when an example of N = 2 is described, 15
At the position of °, 2048 × (15/360) = 85.3, and at −15 ° on the 0 ° side of the edge 15 °, 85 (16
In advance, 54H) At + 15 ° on the 90 ° side of the edge 15 °, 86 (16
In advance, it is 55H). Similar processing is performed for each divided edge, and Table 3
2048P / R data is obtained. Also, if the resolution per rotation is multiplied by 4 and multiplied to 8192, then 8192 × (7.5 × N / 360) (N = 0, 1, 2,...) Table 1
The data at each division position shown in 8192P / R can be specified.

Next, the function of the present embodiment will be described with reference to FIGS. In the description, it is assumed that an accurate position is preset only at each edge of the GRAY1 signal. Thereby, the signal defining the preset position becomes 1
Therefore, the position accuracy of only the GARY1 signal is improved,
The object can be achieved if the detection speed is improved by differential signal detection or the like, and the position accuracy of the GRAY6 to 2 signals is coarsened, and the signal detection speed can be reduced. Therefore, it is possible to easily design and manufacture the slit and the optical signal processing circuit.

Further, according to this method, since the changing points of the GARY6 to GARY2 signals are shifted from each edge of the GRAY1 signal, a stable signal can be obtained at the edge of the GRAY1 signal. The advantage of can be used effectively.

In this example, the case where the power supply Vcc is turned on at a position 32 ° between the rotation angles 22.5 ° and 37.5 ° will be described.
When the Vcc power is turned on, data (HEX data: AA) indicating a rotation angle of 30.00 ° is loaded into the counter 33 by the reset sampling pulse rsp. This value is the rotation angle 22.5 °
Any value between (HEX data: 80) and 37.5 ° (HEX data: D5) may be used, but here, a value of 30.00 ° considered to be the most average is used.

Next, when the encoder rotates (forward rotation),
A sampling pulse sp + 1 is generated at a position of 37.5 °, a sampling pulse sp + 2 is generated at a position of 52.5 °, and 22.5 in the case of reverse rotation.
When the sampling pulse sp-1 is generated at the position of ° and the sampling pulse sp-2 is generated at the position of 7.5 °), the sampling pulse sp is generated from the rotational position detecting means 21. This pulse is generated by the edge of the GRAY1 signal.
It becomes an accurate position signal of the rotation angle and presets the counter value to an accurate value. Thereafter, the incremental signals IA,
By counting IB, the value of the counter always indicates the correct rotation angle. FIG. 10 shows the rotational position detecting means 21, O
A specific example of the R circuit 25 and the reset signal generating means 23 will be described.

The operation of the rotational position detecting means 21 will be described. As shown in FIG. 10, the GRAY6 to GRAY2 signals are
Sampled by a type flip-flop (hereinafter abbreviated as D-FF) in synchronization with the clocK signal,
The signal is sampled by a two-stage D-FF. The output of each D-FF becomes the input of A6 to A0 which is the address of the corresponding absolute value setting means 27, and becomes the signal for specifying the absolute value. The absolute value setting means 27 counts an accurate absolute value per one rotation of the encoder according to the inputs A0 to A6.
Output as 33 preset data. In addition, the GR
The two-stage D-FF output obtained by sampling the AY1 signal is E
When the two signals (A0 and A1) do not match, a sampling pulse sp having a width of 1 clock is output for each edge of the GRAY1 signal.

Here, in FIG. 10, the portions denoted by the same reference numerals as those in FIG. 4 have the same functions as the portions shown in FIG.

From these, as described above, the counter 33
The output of the reset sampling pulse rs at power-on
The temporary absolute position is output by p, and then the true absolute position is output by the sampling pulse sp. Thereafter, the true absolute position is constantly output while counting the incremental signal.

Here, the incremental signals IA, IB
Is multiplied by four. In this case, FIG.
In the embodiment of the present invention, a circuit 34 is added,
The output data from 27 to the rotation amount counting means 33 is D0 to D12 and 2
All you have to do is increase the bits and make them 13 bits. As described above, in order to cope with the multiplying signal, it is possible to cope by increasing the number of data bits. Table 3 shows an output example of the absolute value setting means 27 of this example.

In this example, each of the GRAY6 to 2 signals is latched by the D-FF. However, a stable signal is obtained at the position where the sp signal is generated. Needless to say, -FF can be deleted.

In FIG. 10, D-FF 55 and AND 56
The case in which is added will be described. By these, G
Of the sampling pulses sp by each edge of the RAY1 signal, the only sp signal generated after reset (in the case of normal rotation,
In the case of sp + 1 and reverse rotation, only sp-1) is valid and used as the load signal SP. FIG. 9 shows the state of the signal according to this example.
Shown in According to the present example, there is an advantage that the speed of the signal processing circuit can be reduced at a low speed because the preset is completed in a state where the rotation is slow.

Here, according to FIGS.
RAY1 edge of the signal is not only GRAY6~2 be exact location from power using is described a method of small-fence the operation angle to be preset.

EX-OR of each of GRAY6 to GRAY2 signals
To obtain the GRAY 6-2 signal shown in FIG. FIG. 11 shows a specific circuit example, and Table 4 shows the input / output relationship of the absolute position setting means 27. Each of the signals GRAY6 to GRAY2 is latched by a D-FF synchronously sampled by the same clock at two stages, and the D-FF output of the first stage is supplied to the absolute position setting means 27.
Input AnB (GRAY6: A6B, GRAY5: A5B, GRAY4: A4B, GRAY3:
A3B, GRAY2: A2B), and the D-FF output of the second stage is the input AnA of the absolute position setting means 27 (GRAY6: A6A, GRAY5: A5A, GRAY4: A4).
A, GRAY3: A3A, GRAY2: A2A).

The synthesized GRAY6-2 signal is
DF sampled synchronously by two identical clocks
F, the first stage D-FF output and the second stage D-FF
The -FF output becomes the same EX-OR input to obtain a signal spb that presets the exact position. This spb signal is
The spa signal generated at the edge of the GRAY1 signal is ORed, and the reset sampling pulse rsp and N
The sampling pulse SP is NORed by the OR circuit 25.
Becomes On the other hand, the absolute position setting means 27 receives the inputs A6B to A2A, A
1, Accurate absolute position data per rotation according to the state of A0
Output to the rotation amount counting means 33. An accurate absolute position is set in the rotation amount counting means 33 by this output and the sampling pulse SP, and thereafter, an accurate absolute position is always output by counting an incremental signal.

The signal spb is, as shown in FIG.
It can also be obtained by inputting the latch signals of the respective GRAY signals to the EX-OR and ORing all of them.

FIG. 13 shows an encoder signal output diagram when all the GRAY6 to GRAY1 signals are used. Said,
Similar to FIG. 9, the description will be made on the assumption that the power is turned on at the position where the rotation angle is 32 °. When the power is turned on, data (HEX data) indicating a rotation angle of 33.75 ° is generated by the reset sampling pulse rsp.
(C0) is loaded to the counter 33. Next, when the rotation is performed based on this data, the original signal of the sampling pulse is generated (the edge signal spa by the GRAY1 signal: in the case of normal rotation, the sampling pulse sp + 1a, 52.
In the case of sampling pulse sp + 2a at 5 ° position, reverse rotation,
Sampling pulse sp-1a at 22.5 ° position, sampling pulse sp-2a at 7.5 ° position; edge signal spb by GRAY6 to 2 signals: In the case of normal rotation, sampling pulse sp + 1b, 60.0 ° at 45.0 ° position In the case of the reverse rotation, the sampling pulse sp-1b at the position of 30.0 ° and the sampling pulse sp-2b at the position of 15.0 °).

In the case of normal rotation by a sampling pulse generated by rotation, the position data 214 (HEX data) is obtained at a position of 37.5 °.
When D6) is reversed, position data 1 at 30.0 °
70 (HEX data: AA) is set, and thereafter, incremental signals are counted.

In this embodiment, the signal SP has been described as being obtained from the logical outputs of the signals spb, spa, and rsp as shown in FIG. This logic circuit, after starting rotation with the position provisionally obtained by the rsp signal and before presetting the exact position by the signal spa, also presets the exact position by the spb signal, but once by the signal spa After the exact location is preset, it is only for enabling the signal spa. This is because the precision of the signal and the responsiveness of the detection circuit are paid attention only to the GRAY1 signal in the production of the encoder, and the production is facilitated.
The signal is used as an auxiliary signal. Otherwise,
It becomes necessary to manufacture all of the GRAY 6 to 1 signals with high accuracy. However, the spb signal may always be used. Needless to say, an accurate position can be obtained with a smaller rotation angle by using the spb signal.

In FIG. 5, a circuit D-FF 51 for replacing the circuit 21 with the circuit shown in FIG. 6 and continuing to hold the position data set by the rsp signal until the sp signal is input is provided. It is assumed that ANDs 53a and 53b (described above, FIG. 10) are deleted. According to this circuit example, since the up pulse and the down pulse by the incremental signals IA and IB are always input, after the temporary position data is set by the rsp signal as shown in FIG. Continue to be updated. Also in this case, the sampling pulse sp by each edge of the GRAY1 signal presets an accurate absolute position in the rotation amount counting means 33 (counter). After that, the counter counts up pulses and down pulses to provide accurate counting. Keep showing absolute position.

Thus, there has been described a method of using a gray code to continuously indicate an accurate absolute position. Here, a case where the encoder of the present invention is used as an encoder with a communication signal used for AC servo (hereinafter, referred to as an encoder with CS) will be described with reference to FIG. C
The encoder with S is for detecting the magnetic pole position signal of the AC servo motor.
Rotation is divided into 12, and in the case of an 8-pole motor, one rotation is divided into two.
What is necessary is just to divide into four. That is, one rotation is divided into multiples of 12, and the encoder described above divides one rotation into 48 using gray code. From this, it is obvious that one rotation can be divided into 12 or 24 by the combination of gray codes.

Here, an example of a case where the present invention is used for a four-pole motor is shown in FIG. 8 as a composite signal [for a four-pole motor]. That is, the CS signal corresponding to the case of the 4-pole motor may be assigned as shown in FIG. The same applies to the case of using an eight-pole motor.

The method of presetting an accurate absolute position by the edge of the GRAY signal has been described above. However, it may be necessary to unify the relationship between each edge of each GRAY signal and the incremental signals IA and IB. Understand. A method for this is shown in FIGS. FIG. 1 shows a signal synchronizing circuit for latching each signal (representative signal U or GRAY signal: GRAY *) using the signals IB and IA as shown in FIG. It is added between the shown CMP output and the rotational position detecting means 21.

In the above description, the case where six signals of GRAY 6-1 are used and the signal is divided into 48 is described. As is well known, the maximum number of divisions of the six signals is 2 ⁶ = 64. As an example of this, 32 signals are generated by five signals
FIG. 17 shows an example of division. Heretofore, it has been assumed that the edges of the GRAY6 signal are at the position of 0 ° of the rotation angle in order to actually count at the edge of the GRAY6 signal using the signals GRAY6 and 5 as the signals for detecting and counting the number of rotations. . However, Japanese Patent Application Laid-Open No. 63-83612
As shown in the above, a magnetic encoder is used as an encoder for detecting the number of rotations, and the rotation angle per rotation uses an optical encoder, and multiple rotations are performed by combining two encoders. An absolute value encoder may be constructed. Also, as described in JP-A-1-305315, the edge of the signal of the slit for detecting the number of rotations may not be obtained at the accurate rotation angle of 0 °, but may occur at the position of 359 ° or 1 °. is there. In each case, the edge of the signal of the slit for detecting the number of rotations is shifted from the position of the accurate rotation angle of 0 °. However, it is avoided to reduce the production accuracy of the encoder and reduce the power consumption. In many cases, this is a matter that cannot be done.

The means for correcting the deviation of the rotational speed detection signal will be described below with reference to FIGS. 20 to 23, taking the GRAY signal as an example. First, for simplicity of explanation, as shown in FIG.
Using an encoder composed of signals as an example, a rotation speed detection signal G
It is assumed that the edges of RAY6 and RAY5 are significantly shifted from the position of the rotation angle of 0 °. GRAY6: 168.75 °, 348.75 ° GRAY5: 78.75 °, 258.75 ° In addition, other signals GRAY4 to
The edge of GRAY4: 33.75 °, 123.75 °, 213.75 °, 303.75 ° GRAY3: 11.25 °, 56.25 °, 101.25 °, 146.25 °, 191.25 °,
236.25 °, 281.25 °, 326.25 ° GRAY2: 0.00 °, 22.50 °, 67.50 °, 90.00 °, 112.50 °,
135.00 °, 157.50 °, 180.00 °, 202.50 °, 225.00 °, 247.5
0 °, 270.00 °, 292.50 °, 315.00 °, 337.50 °, 360.00 ° At this time, the rotation angle is set to 0 ° with reference to one of the edges of the GRAY2 signal. However, this example is an example, and it is not always necessary to use the GRAY2 signal as the reference signal. Here, the case where the edge of the GRAY2 signal is 0 ° as in this example and the case where the edge of the GRAY6 signal is 0 °
Table 5 shows the input / output relationship of the absolute position setting means 27 in the case of (1).

One way to achieve this object is to count the number of rotations after correcting the shifted edge of the GRAY6 signal to a rotation angle of 0 °. More specifically, this can be achieved by generating new signals GCON2 and GCON1 with the logic shown in FIG. 20, and using this GCON1 signal instead of the GRAY6 signal. However, in this method, all gray code signals (GRAY) are used to count the number of rotations.
6) -2), there is a problem that power consumption increases when the battery is driven.

Another method for solving the above problem is to count the number of rotations using the GRAY6,5 signals,
Up and down are performed at the edge of the AY6 signal, and when the power supply Vcc is applied, the count value of the number of revolutions is corrected by a signal based on the GRAY2 signal indicating an accurate position. That is, the value counted by the GRAY6 signal is corrected using the GRAY2 signal having the edge at 0 °. The signal for this correction is SEL, and the signals GRAY6 to GRAY2 are used, and the circuit 89 shown in FIG. 1 (the logic is shown in FIG. 21)
Generated by Specifically, when the power supply Vcc is applied, a new signal GCON3 is generated in addition to the above-mentioned signals GCON2 and GCON1. When GCON3 = 1 (when High), the value counted by the GRAY6 signal is GCON3
= 0 (at the time of Low) means that a value smaller than the value counted by the GRAY6 signal is set as the rotation speed count value. The above specific circuit example will be described with reference to FIG. In the present embodiment, the description will be made by adding circuits 85 and 89 which are omitted in the above description.

The rotation direction detecting means 81 outputs a pulse indicating the determination of the rotation direction and the amount of rotation, an up pulse, and a down pulse as in Japanese Patent Publication No. 60-4924. The up pulse and the down pulse are input to the rotation speed counter 83, and count the rotation speed. The rotation direction detection means 81 and the rotation number counter 83
It is driven by a battery and always counts and stores the number of rotations. The output, which is the counting result of the rotation speed counter 83, is input to the rotation speed correction output unit 85. On the other hand, the rotation number correction signal generating means 89 receives the signals GRAY6 to GRAY2 and outputs the signal SEL by a logic circuit. Rotation speed correction output means 85
Outputs the output itself of the rotation speed counter 83 when SEL = 1, and outputs a value one less than the output of the rotation speed counter 83 when SEL = 0. In the case of this example, the signal is the same as the signal GCON3. As mentioned above,
The rotation speed correction output unit 85 may be any unit having a calculation function based on the signal SEL or a unit having a state change function such as a ROM. The output of the rotation speed correction output unit 85 is used as the value of the rotation speed at the multi-rotation absolute position.

Further, in order to more accurately detect the rotational speed with respect to the absolute position, the incremental signals IA, IB
Means for synchronizing with a signal will be described with reference to FIGS. As an example for this purpose, a case will be described in which a signal (Z-phase signal in a normal encoder) that is generated only once per rotation is used. In this example, this signal is
Based on the Y1 signal, the signal is generated in combination with another signal. As a result, edges of a plurality of GRAY1 signals can be formed, and the signal use efficiency can be improved. As an example, a description will be given assuming that the following signal changes occur. Similarly to the above, it is assumed that the edges of the rotation number detection signals GRAY6 and GRAY5 are greatly shifted from the position of the rotation angle of 0 °. GRAY6: 174.0 °, 352.0 ° GRAY5: 78.0 °, 258.0 °. In addition, other signals GRAY4 to
The edge of No. 1 is as shown in FIG. For example, GRAY4 is 30.0 °, 126.0 °, 210.0 °, 306.0 °, and GRAY3 is 6.0 °, 54.0 °, 102.0 °, 150.0 °, 186.0 °,
………… GRAY2 is 0.0 °, 18.0 °, 42.0 °, 66.0 °, 90.0 °,
………… GRAY1 is 12.0 °, 24.0 °, 36.0 °, 48.0 °, 60.0 °,
………… When the rotation angle of the GRAY1 signal is around 0 °,
Less than 1/2 pulse of absolute position 2047 around 0 ° position, 1 of 0
It is assumed that an edge is formed in less than / 2 pulse. FIG. 21 shows this signal waveform. At this time, it is assumed that the GRAY2,1 signal is obtained as an accurate and stable signal with respect to the position by the differential detection circuit.

As shown in the enlarged portion, it is assumed that the GRAY2 signal slightly deviates from the exact position of 0 ° near 0 ° and is generated within less than 0 ° ± 1/4 pulse as shown by the broken line in the figure. . As described above, the GRAY1 signal is 0 °
Suppose that it was generated within less than ± 1/2 pulse. actually,
It is easy to form a slit so that signal accuracy is obtained within any of the deviation amounts.

As described above, GCON1 to GCON3 are given by the logic shown in FIG. 21, and a GCON4 signal is generated as a new signal by the logic shown in FIG. If this GCON4 is used as a new GRAY2 signal, a signal for accurately detecting the rotation speed synchronized with the incremental signals IA and IB can be obtained.
FIG. 22 shows a circuit in which the signal of this example is used. The GRAY2,1 signal is detected by a differential detection circuit (FIG. 18).
(A)) Since the signal is accurate and stable with respect to the position, the two signals are latched by the two-stage D-FF, the respective outputs are input to the absolute position setting means 27, and the edges of both signals are further Used to generate a sampling pulse sp. This makes it possible to obtain an accurate position at 44 positions and to reduce the initial rotation amount. Table 6 shows the input / output relationship of the absolute position setting means 27 in the case of this example.

The detection circuit that has been described several times will now be described with reference to FIG. FIG. 18A shows a differential detection circuit, and FIG. 18B shows a reference voltage detection circuit. In the differential detection circuit, since an optical signal is captured as a reciprocal signal and is compared, a stable signal with high accuracy can be obtained. On the other hand, the reference voltage type detection circuit distinguishes an optical signal based on a value with respect to a reference voltage, and has a feature that the circuit is simplified.

Here, the relationship between the input and output of the absolute position setting means 27 described above is summarized in Tables 3 to 6.

[0072]

[Table 3]

[0073]

[Table 4]

[0074]

[Table 5]

[0075]

[Table 6]

Up to now, the signal pattern of the encoder signal and its processing have been described. Here, the slit pattern formed on the rotating disk 11 is shown in FIG.
23. In order to ensure the signal accuracy, the rotating disk 11 is generated in the circumferential outer direction in order from the signal requiring the accuracy, and in the case of the incremental encoder with U, V, W phases, the signals IB, IA, W, V, U, RB,
RA order. Also, absolute position encoders with low resolution
As an example (FIG. 23), the signals IB, IA,
Gray code signal GRAY1, GRAY2, GRAY3, GRAY4, GRAY
5, slits corresponding to GRAY6 are formed in order from the outer periphery. The light emitted from the LED passes through the shaded area in this figure,
Alternatively, it is reflected and converted into a re-voltage signal which becomes the input of the photodiode. This voltage signal is waveform-shaped by a comparator, and becomes the High level of each signal described above.
That is, the slit pattern shown in FIG.
The signal for the pole motor is generated, and the slit pattern shown in FIG. 23 generates the signal shown in FIG. It should be noted that the slit pattern for generating the signals shown in the other figures can be configured in the same way.

Although the example using the gray code signal has been described so far, a special signal pattern is used below to improve the accuracy, which makes it possible to reduce the amount of rotation after the power is turned on. The following describes an embodiment in which such a case is used. As shown in FIG. 24, the present embodiment uses two-phase signals RGRAY2, 1 shifted by 90 ° as signals for detecting an accurate position.

First, FIG. 24 will be described. GRAY6 ~
The three signals need only be included between the edges of the RGRAY2 to RGRAY1 signals. Here, for ease of explanation, it is assumed that each edge is shifted by about 3 °. Thus, the edges of the rotation speed detection signals GRAY6, GRAY5 are 0 ° of the rotation angle.
GRAY6: 177.0 °, 357.0 ° GRAY5: 81.0 °, 261.0 °, and the edges of the other signals GRAY4-3 are GRAY4: 33.0 °, 129.0 °, 213.0 °, 309.0 ° GRAY3: 9.0 ° , 57.0 °, 105.0 °, 153.0 °, 189.0 °, 23
It is assumed that they are formed at 7.0 °, 285.0 °, and 333.0 °. The edges of the signals RGRAY2 to RGRAY1 which require the position accuracy for generating the sampling pulse are shown in FIG. For example, RGRAY2 is 6.0 °, 18.0 °,
30.0 °, 42.0 °, ... RGRAY1 also changes from 0.0 ° to 12.0 ° every 0.0 °, 12.0 °,
Edges are formed at 24.0 °, 36.0 °, .... Here, RGRAY2, 1
It is assumed that the signal is detected by a differential detection circuit and a positionally stable signal is obtained.

Next, an example of a circuit for achieving this embodiment is shown in FIG. As shown in FIG. 25, each of the GRAY signal and the RGRAY signal becomes an input signal of the slit signal synchronization circuit 91. The slit signal synchronizing circuit 91 is composed of D-FFs, and the GRAY6 to GRAY3 signals are sampled by a single stage D-FF. Further, the RGRAY2 and RGRAY1 signals are sampled by two-stage D-FFs respectively, and EX-
The edges of both signals are detected by the OR circuit. A signal generated by ANDing the detected edge signal, reset signal, and clock provides a D-FF sampling signal for sampling the GRAY signal. By these operations, the GRAY signal becomes a signal synchronized with the RGRAY signal.

By the above operation, the GRAY signal at the time of the forward rotation is GRAY6F: 0.0 °, 180.0 ° GRAY5F: 84.0 °, 264.0 ° GRAY4F: 36.0 °, 132.0 °, 216.0 °, 312.0 ° GRAY3F: 12.0 °, 60.0 ° , 108.0 °, 156.0 °, 192.0 °, 24
GRAY6R: 174.0 °, 354.0 ° GRAY5R: 78.0 °, 258.0 ° GRAY4R: 30.0 °, 126.0 °, 210.0 °, 306.0 ° GRAY3R : 6.0 °, 54.0 °, 102.0 °, 150.0 °, 186.0 °, 23
It is corrected to 4.0 °, 282.0 °, 330.0 °. This is shown in FIG.

The processing after the corrected signal is the same as the above-described invention using the gray code signal, as can be seen from FIG. That is, if the processing in the state of each signal is performed using the corrected signal obtained in FIG. 24 and the same concept as the above-described invention based on the gray code signal,
The present invention can be achieved.

Tables 7 and 8 show a list of data used for signal processing.

[0083]

[Table 7]

[0084]

[Table 8]

In the table, in order to make it easy to understand the state, combinations of the signals are represented in the form of numerical values. RGR from GRAY6 signal
Up to AY1 signal, weight 32,16,8,4,2,1
The gh level is "1" and the low level is "0", and the state of each signal is represented by numerical values 1 to 63. The power-on state is 14
Double (for example, 9 ° to 12 ° and 30 ° indicated by "* W01")
"* W01" ~ "so that ° ~ 33 ° is the same with the numerical value" 54 "
(* 14 locations of W14 "), but after signal correction,
It is understood that it is canceled. That is, after rotation, there is no influence of double and there is no problem in handling as described above, so that only setting data at power-on need to be devised.

Accordingly, Table 7 shows an example of the data of the absolute position setting means 27 which is output by the rsp signal when the power is turned on. There are two types of data setting, general setting data and integrated setting data. General setting data
The data sets the data as fine as possible
Set the data corresponding to the numerical value. However, when the number of the doubling portions becomes three (for example, 9 ° to 12 ° and 18 ° to 24 °
Since the setting data of 30 ° to 33 ° is the same for “DEC: 120”, both ends are set to unique data (“DEC: 100”). As a result, it is clear that the portion is a double portion, so that a special process is performed at the time of turning on the power (for example, only when this data is output, the initial rotation angle is set to ± 3 °). Can be. On the other hand, the integrated setting data is for making the same setting data in all areas having a double portion.

Table 8 shows an example of input data and output data of the absolute position setting means 27 processed by the sp signal after rotation.

Further, the slit pattern formed on the rotary disk 11 in this embodiment will be described with reference to FIG. As described with reference to FIGS. 19 and 23, in order to secure the signal accuracy, the incremental signals IB, IA, the accurate position detection signals RGRAY1, RGRAY2, the gray- Signal GRAY3, GRAY
4, slits corresponding to GRAY5 and GRAY6 are formed.

It should be noted that the signal having a 90 ° phase difference is detected by the detecting element PD.
It is known that the slit can be made one by devising the arrangement of the slits. If you use this, 2
Since the GRAY6,5 signals, RGRAY2,1 signals, and the incremental signals IA, IB, which are phase signals, are obtained from one slit, respectively, the encoder of the present invention can be composed of a total of 5 slits together with the GRAY4,3 signals. . As a result, the size of the encoder can be further reduced.

Thus, one more slit is added to form six slits, and a 60-divided signal without duplication can be obtained. That is, RGRAY1H: 0.0 °, 24.0 °, 48.0 °, 72.0 °, 96.0 °, 12
0.0 °, 144.0 °, 168.0 °, 180.0 °, 204.0 °, 228.0 °, 252.
Signals having edges at 0 °, 276.0 °, 300.0 °, 324.0 °, and 342.0 ° may be provided.

In the description of this embodiment, several examples are described individually, but it is obvious that they may be used in association with each other. Also, the CS signals U, V, W (GRAY
The data preset in the counter 33 by each edge of the signal) may be different from those shown in Tables 1 to 8 due to a problem in the manufacture of the encoder or a problem with the combination with the servo amplifier. Further, in the present embodiment, the preset signal sp
A slit for generating a Z-phase signal, which is generated from each edge of the S signals U, V, and W (GRAY signal) but is normally provided in the encoder, and a special code for performing a preset operation, is provided. Alternatively, a preset signal may be generated by the slit signal, and the preset data may be preset in the counter 33. The magnetic pole position signals U,
FIG. 1 shows a signal synchronization circuit (only the U phase is shown as an example) as shown in FIGS. 15 and 16 in order to unify the relationship between each edge of the V and W phases and the incremental signals A and B phases. It may be added between the CMP output and the rotational position detecting means 21.

In the foregoing, the description has been given of the case of a slit of a commonly used rotating disk such as a gray code. However, the case of using an M-sequence which has begun to be adopted in recent years will be described. An example of an encoder using an M-sequence is described in Japanese Patent Application Laid-Open No. 2-132324.
This is realized by reading an adjacent k-bit pattern by using an M-sequence random number which is a cyclic pattern of one different code. Since this pattern has only one location on the circumference, the absolute position on the circumference can be determined. In this example, the change point of the absolute position obtained by reading the pattern of k bits is used to specify the position where the sampling pulse sp is generated and the data to be preset in the counter. As the preset data at the time of turning on the power, a provisional value may be set as the preset data using an absolute position obtained by reading a pattern of k bits.

Here, an embodiment will be described with reference to FIGS. 32 and 33. FIG. 32 shows a rotary type (rotary type).
, And one rotation is divided into 16 regions. The hunting portion in the figure corresponds to 1 and the blank portion corresponds to 0, and the state is detected by a detector extending over four divided regions. FIG. 33 shows an example of a linear type (linear motion type) slit pattern, in which the state is detected by a detector extending over four divided regions, as described above. The signal detected by the detector is the same for both the rotary type and the linear type.
The values shown in Table 9 are obtained for the 16 divided regions O1 to O16.

[0094]

[Table 9]

The operation of the reset sampling pulse rsp, sampling pulse sp, etc. will be described with reference to FIG. 33 assuming that the rising edge of the incremental signal IA is at a position corresponding to the change point of each divided area of the M-sequence slit pattern. . The signal detected and output by the detector is the state of the M-sequence slit in which the detection unit R3 is arranged (each divided region (D1 to D1) of the M-sequence slit pattern).
6) With respect to one divided area, the output of the detector changes after all the detection areas of R3 have been moved.)

Here, it is assumed that the power is turned on almost at the center of the area D2. At this position the reset sampling pulse rsp
Occurs, and rough data corresponding to the D2 area is preset in the aforementioned counter. Next, by rotating or moving directly by this data, the change point (D2
To D1: a sampling pulse sp-1 is generated, and D2 to D
3: a sampling pulse sp + 1 is generated), and a sampling pulse is generated. Since this change point indicates a specific absolute position, the specific absolute position is preset in the counter by the sampling pulse as described above. Thereafter, when the incremental signal is counted, M
It is possible to obtain an absolute position having a higher resolution than the absolute position obtained as a series.

Although the above embodiment has described the rotary type slit system, it is needless to say that the present invention can be applied to a linear type slit system. Although the present invention has been described by taking an optical encoder as an example, it goes without saying that other types of encoders such as a magnetic encoder can be used.
In this embodiment, the elements 21 and 27 are composed of a combination of the element FF and a function table such as a ROM.
May be replaced by the processing by.

Here, an embodiment of a servo system using the encoder according to the present embodiment will be described with reference to FIG. The absolute rotation of the encoder is serially transmitted to the servo amplifier via the modem 43 and the line driver. The servo amplifier demodulates the signal with the modem 44 via the line receiver, and outputs the data of the multiple rotation absolute position to the signal processing circuit 60. The signal processing circuit 60 processes a signal from the modem and outputs multi-turn absolute value data to a host device. As a result, the host device knows the multi-rotation absolute position of the motor. On the other hand, the motor rotation position indicating the position per one rotation of the motor in the multi-rotation absolute value data is input to the magnetic pole division circuit 64 as position data for magnetic pole division. The magnetic pole dividing circuit 64 divides the current command value from the host device into magnetic poles based on the motor rotation position signal, and outputs a phase current command value. This phase current command value becomes a motor current via a current control unit (ACR unit) 65, and rotates the motor. At this time, the motor shaft and the encoder shaft are mechanically connected, and the encoder also rotates. On the other hand, the A-phase and B-phase two-phase incremental signals are output to the host device via the servo amplifier.

In the servo system shown in FIG. 28, when the encoder power is turned on, the encoder transmits the data set by the reset sampling pulse rsp by the reset signal, and the servo amplifier divides the magnetic poles by this data and outputs the phase current command. Generate a value and rotate the motor. The motor drive at this time is a rectangular wave drive because the position signal of the encoder does not change. When the motor rotates, the counter in the encoder is set to the correct value by the sampling pulse sp due to the signal edge, and thereafter always shows the correct value. From the time when the correct data is preset, the position data of the motor always changes, and the motor is driven by a sine wave. Once the power is turned on, the host device reads the multi-turn absolute value data once, rotates the motor (until the correct motor rotation position is preset at the edge of the signal), and reads the multi-turn absolute value data again. The data read out again indicates the correct motor rotational position, and if the position is controlled based on this data, complete position control becomes possible. After reading the correct data, the position may be counted using the incremental signals A and B.

It is considered that the encoder of this embodiment is mounted on a servomotor and is often used for position control. Therefore, as shown in FIG. 27, the motor is rotated while counting the incremental signals based on the provisional position at the time of turning on the power until the correct position is found. When the correct position is determined by rotation (sampling pulse sp
If the motor is rotated reversely until the signal reaches the same value as the counted incremental signal, the machine returns to the original position and stops. As a result, the deviation from the power-on position can be reduced to zero, and it is easy to secure the conventional start position.

Finally, an example of an industrial robot using the encoder according to the present embodiment will be described with reference to FIGS. Robots are required to have small wrists in order to reduce interference with the working environment. The outer shape of the wrist is disclosed in Japanese Patent Application Laid-Open No. 1-177987 (a representative diagram is shown in FIG. 29).
In the structure in which the two motors of the swing motor 61 and the rotation motor 71 are arranged on the wrist base 5 as described in the above, the size of the wrist base is greatly restricted by the size of the motor and the encoder. Here, Rotary Encoders and Rotation Sensors General Catalog VOL.02 Samtaku Corporation 1990 1
Consider the encoder described in the month as an example. For the absolute encoder AEM-002-2048 size, diameter 75, L63.5mm, the encoder according to the present invention is assumed to be an incremental encoder LHB
Consider a case where these encoders are incorporated in a robot, considering -001-2500. When a conventional encoder is used, the size of the wrist is restricted by the size of the encoder because the encoder 61E (71E) is larger than the motor 61 (71) as shown in FIG. On the other hand, when an encoder using the present invention is used, the outer shape of the motor 61 (71) and the encoder 61E (71E) can be made close to the same size as shown in FIG. 31, and the wrist size can be reduced. be able to. That is, the size of the encoder using this embodiment is 46 mm in diameter and 42 mm in L, and the outer shape can be reduced by about 30 mm from the conventional encoder. This reduction in size directly contributes to the outer shape of the robot wrist, and is particularly effective for a small robot.

When the encoder according to the present embodiment is used, it is not necessary to reduce the number of slits of the encoder in order to reduce the outer diameter of the encoder. For this reason, the resolution of the absolute value does not decrease, and a highly accurate operation can be performed.

In a machine such as a robot having a complex mechanism with multiple axes, if the encoder of this embodiment is used, the approximate position and orientation of the robot can be determined by the position information set by the rsp pulse at power-on. Therefore, it is possible to move the motor of each axis in a direction to bring the robot closer to the home position. This makes it possible to move the robot until a normal position by the sp pulse is obtained without causing interference with the surrounding jig. Here, it is necessary to move the robot until a normal position by the sp pulse is obtained, but the reduction ratio of the speed reducer used for the robot is usually 1/50 to 1/100. Rotation angle of 15
° (Maximum rotation at which the CS signal changes with a 4-pole motor)
On the other hand, the operation amount of the mechanism is 0.3 to 0.15 deg, and there is no problem.

Further, according to the present invention, there is no need for a procedure for performing servo lock at a specific position of the Z phase and correcting the current value to the true current position as in the case of the conventional robot origin adjustment. There is an advantage that the true current position can be obtained only by rotating the motor.

[0105]

According to the present invention, it is possible to use an absolute position signal and an incremental signal having a low resolution so that it is not necessary to generate a slit indicating the entire resolution of the absolute position on the moving disk.
Since the encoder becomes smaller and the number of parts can be eliminated, the reliability is improved. (2) It is not necessary to generate a slit indicating the absolute position per rotation on the rotating disk, and an encoder disk with U, V, and W phases can be used. Therefore, the size of the encoder can be reduced and the number of parts can be eliminated. Reliability is improved. (3) Basically, since the incremental pulses are counted and converted to the absolute value, the resolution can be increased simply by increasing the resolution of only the inquiring signal, and the correspondence can be simplified. (4) Change of encoder pattern with low resolution, U,
The change of the encoder for 4 poles and 8 poles of the encoder with V and W phases can be dealt with by changing the contents of the absolute value setting means.
This can be handled by simple means such as using an OM or changing the contents of the gate array. (5) As an economical effect, it is possible to supply at about half the price (20,000 to 30,000 yen) of the conventional absolute value encoder.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of an encoder according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram of encoder signals with U, V, and W phases.

FIG. 3 is a diagram illustrating an output of an encoder signal according to an embodiment of the present invention.

FIG. 4 is an explanatory diagram (with U, V, and W phases) of a basic specific circuit.

FIG. 5 is a circuit example (part 1) in a case where incremental signals are continuously counted.

FIG. 6 is an example in which FIG. 5 is displayed in gray code.

FIG. 7 is an explanatory diagram of a functional operation by a preset signal (U,
V, W phase).

FIG. 8 is an explanatory diagram of an encoder signal using a gray code.

FIG. 9 is an explanatory diagram (gray code) of a functional operation based on a preset signal.

FIG. 10 is an explanatory diagram of a specific basic circuit (Grayco-
C)

FIG. 11 is a circuit example when the number of preset signals is increased (part 1)

FIG. 12 is a circuit example when the number of preset signals is increased (part 2)

FIG. 13 is an explanatory diagram of a functional operation when the number of preset signals is increased.

FIG. 14 is a circuit example when a preset signal is selected.

FIG. 15 is an explanatory diagram of one embodiment of a signal synchronization circuit diagram.

FIG. 16 is an explanatory diagram of another embodiment of the signal synchronization circuit diagram.

FIG. 17 is an explanatory diagram of an encoder signal (32 divisions) when the uppermost signal is set to 0 °.

FIG. 18 is an example of an optical signal processing circuit.

FIG. 19 shows an example of an encoder disk slit (part 1) (FIG. 2, corresponding to a 4-pole motor)

FIG. 20 is an explanatory diagram of an encoder signal (at the time of division into 32) when the rotation number detection signal is shifted.

FIG. 21 is an explanatory diagram of an encoder signal (at the time of 60 divisions) when a special signal is generated.

FIG. 22 is an example of a specific circuit corresponding to FIG. 21;

FIG. 23 shows an example of a slit of an encoder disk when a special signal is generated.

FIG. 24 is an explanatory diagram of an encoder signal when a two-phase signal is used to generate a preset signal.

FIG. 25 is an example of a specific circuit corresponding to FIG. 24;

FIG. 26 shows an example of a slit of an encoder disk when a two-phase signal is used to generate a preset signal.

FIG. 27 is a processing flow of an initialization operation.

FIG. 28 is an AC servo system diagram using an encoder according to the present invention.

FIG. 29 is an explanatory diagram of a wrist configuration of an industrial robot.

FIG. 30 is an explanatory diagram of a wrist configuration using a conventional encoder.

FIG. 31 is an explanatory diagram of a wrist configuration using the encoder of the embodiment of the present invention.

FIG. 32 shows an example of a slit of a rotary encoder using an M-sequence.

FIG. 33 is an explanatory diagram of a functional operation in the case of an encoder using an M sequence.

[Explanation of symbols]

1. Encoder 2. Servo amplifier 3. Host controller (host device) 11. Rotating disk 15. Rotation counting means 21. Rotational position detecting means 23. Reset signal generating means 25.・ OR circuit 27 ・ ・ Absolute value setting means 31 ・ ・ Rotation amount detecting means 33 ・ ・ Rotation amount counting means 41 ・ ・ Multi-turn absolute value generating means 43 ・ ・ Modem 51 ・ ・ RS-flip-flop 53a, 53b ・ ・ AND 55 D-flip-flop 56 AND 60 Signal processing circuit 64 Magnetic pole splitting circuit 65 Current control unit (ACR unit) 5 Wrist base 61, 71 Motor 61E, 71E Encoder 81rotation direction detecting means83rotation counting means85rotation correction output means89rotation correction signal generation means91 slit signal synchronization circuit

──────────────────────────────────────────────────続 き Continuation of the front page (51) Int.Cl. ⁷ Identification symbol FI G05D 3/12 305 G05D 3/12 305B (JP, a) JP flat 1-305314 (JP, a) JP flat 2-44410 (JP, a) JP Akira 63-271115 (JP, a) (58 ) investigated the field (Int.Cl. ⁷ , DB name) G01D 5/249

Claims (9)

(57) [Claims] 1. A rotating disk having a plurality of absolute position codes indicating an absolute position and a two-phase code having a higher resolution than the absolute position and rotating in a predetermined direction, and detecting the absolute position code to generate a position reference signal. A position reference signal detecting means for detecting the absolute position code, a position specifying signal detecting means for detecting the absolute position code and generating a position specifying signal, and an absolute value signal indicating an absolute position of the rotating disk based on the position specifying signal. a value setting means, absolute indicated by the absolute position code by detecting the two-phase code
A rotation amount detecting means for generating a pulse period higher resolution than versus position, a counter for counting said pulses, and preset means for presetting said absolute value signal to the counter by the position reference signal, the count value of the counter Encoder having means for outputting a signal. 2. A rotating disk having a plurality of absolute position codes indicating an absolute position and a two-phase code having higher resolution than the absolute position and rotating in a predetermined direction, and detecting the absolute position code to generate a position reference signal. a reference position signal detection means for the position specifying signal detecting means for generating position location signals to detect the absolute position code, absolute to make an absolute value signal indicating the absolute position of the rotating disc on the basis of the localization signal a value setting means, absolute indicated by the absolute position code by detecting the two-phase code
A moving amount detecting means for generating a pulse period higher resolution than versus position, a counter for counting said pulses, and a power supply reset signal generating means for generating a reset sampling pulse at power-on, in response to the absolute value signal preset predetermined absolute value on the counter by the Lise <br/> Tsu preparative sampling pulses, and the absolute value signal before by the position reference signal
And preset means for presetting the serial counter, encoder having means for outputting a counted value of said counter. 3. A rotating disk having a plurality of absolute position codes indicating an absolute position and a two-phase code having higher resolution than the absolute position and rotating in a predetermined direction, and detecting the absolute position code to generate a position reference signal. a reference position signal detection means for the position specifying signal detecting means for generating position location signals to detect the absolute position code, absolute to make an absolute value signal indicating the absolute position of the rotating disc on the basis of the localization signal a value setting means, absolute indicated by the absolute position code by detecting the two-phase code
A rotation amount detecting means for generating a pulse period higher resolution than versus position, a counter for counting the pulses, at least one of the plurality of absolute position code indicating the absolute value
One of the preset means for presetting said absolute value signal to the counter edge portion of the code pattern, the encoder having means for outputting a count value of said count <br/> printer. 4. A rotating disk having a plurality of absolute position codes indicating an absolute position and a two-phase code having a higher resolution than the absolute position and rotating in a predetermined direction, and detecting the absolute position code to generate a position reference signal. a reference position signal detection means for the position specifying signal detecting means for generating position location signals to detect the absolute position code, absolute to make an absolute value signal indicating the absolute position of the rotating disc on the basis of the localization signal a value setting means, absolute indicated by the absolute position code by detecting the two-phase code
A rotation amount detecting means for generating a pulse period higher resolution than versus position, a counter for counting said pulses, and a power supply reset signal generating means for generating a reset sampling pulse at power-on, in response to the absolute value signal preset predetermined absolute value on the counter by the Lise <br/> Tsu preparative sampling pulse, and the said absolute value signal at the edge portion of the at least one code pattern of a plurality of absolute position code indicating the absolute value An encoder comprising preset means for presetting a counter, and means for outputting a count value of the counter. 5. A commutation signal (U-phase, V-phase, W-phase signal) for a servomotor and a two-phase incremental signal (A-phase, B-phase signal) having a higher position resolution than the commutation signal are generated. originating in the encoder, and the reference position detecting means for generating a reference position signal indicative of at least one or more reference position by detecting the Konmi station signals, a localization signal by detecting the Konmi station signals
A position specifying signal detecting means for raw, the absolute value setting means for making the absolute value signal indicating the absolute position of the rotating disc on the basis of the localization signal, a counter for counting the incremental signals, the by the position reference signal An encoder comprising: preset means for presetting an absolute value signal to the counter; and means for outputting a count value of the counter. 6. A servo motor for generating a commutation signal (U-phase, V-phase, W-phase signal) and a two-phase incremental signal (A-phase, B-phase signal) having a higher position resolution than the commutation signal. originating in the encoder, and the reference position detecting means for generating a reference position signal indicative of at least one or more reference position by detecting the Konmi station signals, a localization signal by detecting the Konmi station signals
A position identification signal detecting unit that generates the absolute position of the rotating disk based on the position identification signal.
Absolute value setting means for generating an absolute value signal indicating the position, a counter for counting the incremental signal, power reset signal generating means for generating a reset sampling pulse when power is turned on, and a predetermined value corresponding to the absolute value signal. the absolute value preset in the counter by the Lise <br/> Tsu preparative sampling pulses, and preset means and for presetting said absolute value signal to the counter by the position reference signal, means for outputting a count value of said counter Equipped encoder. 7. A code pattern for a servo motor for a commutation signal (U-phase, V-phase, W-phase signal) and a two-phase incremental signal (A-phase, B-phase) having a higher position resolution than the code signal for a commutation signal . Phase signal)
Encoder signal with a rotary disk having a code pattern for servo motor
A position specifying signal detecting means for generating a specifying signal; and an absolute position of the rotating disk based on the position specifying signal.
Absolute value setting means for generating an absolute value signal indicating the position, a counter for counting the incremental signal, and an absolute value signal corresponding to an edge of the code pattern at the edge of the code signal for the commutation signal to the counter. An encoder comprising: preset means for presetting; and means for outputting a count value of the counter. 8. A code pattern for a servo motor for a commutation signal (U-phase, V-phase, W-phase signal) and a two-phase incremental signal (A-phase, B-phase) having a higher position resolution than the code signal for a commutation signal . Phase signal)
Encoder signal with a rotary disk having a code pattern for servo motor
Position specifying signal detecting means for generating a specifying signal; absolute value setting means for generating an absolute value signal indicating an absolute position of the rotating disk based on the position specifying signal; a counter for counting the incremental signal; and a power supply reset signal generating means for generating a reset sampling pulse, the absolute value absolute value determined in advance in response to the signal preset to the counter by the Lise <br/> Tsu preparative sampling pulses, and for the Konmi station signals An encoder comprising: preset means for presetting an absolute value signal corresponding to an edge of the code pattern at the edge of the code pattern to the counter; and means for outputting a count value of the counter. Wherein said means for counting holding the rotational speed of the rotary disk, absolute for outputting the synthesized and multiple rotation absolute position signal the count value of the counter indicating the current position of the rotational speed of the counting result and the rotating disc 9. The encoder according to claim 1, further comprising a position generating unit.