JP2001255173A - Processing circuit for signal for counting and apparatus for generating angular position signal of sine wave/ cosine wave signal - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize a processing circuit, which processes so that a digital signal for counting having a quantization error removed is outputted, when an analog signal is converted into a digital signal, and an angular position signal-generating apparatus utilizing the same. SOLUTION: This processing circuit for processing digital signals for counting to be outputted synchronously with clocks has shift registers 51-57 for delaying digital signals by each cycle of clocks and holding delayed signals, change- detecting circuits 58, 59, 61, 62, 64 and 65 for detecting from the digital signals held in the shift registers that the digital signals are the same logic value for a period of a prescribed number of cycles of clocks, and output circuit 60, 63 and 66 for changing the outputs to the logic value, when the digital signals are the same logic value for the period of the prescribed number of cycles of clocks and holding the outputs in other cases, on the basis of the detected result of the change-detecting circuits.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a circuit for processing a digital signal for counting when a signal output from a moire scale, a magnescale or the like is digitized and the amount of movement is detected by counting the digitized signal. From a wave signal and a cosine wave signal, a device that detects a finer angular position of the signal and outputs a counting digital signal indicating the angular position,
In particular, the present invention relates to a counting digital signal processing circuit and an angular position signal generation device that reduce the influence of fluttering of the counting digital signal.

[0002]

2. Description of the Related Art A pulse signal is output from a moire scale or a magnescale (registered trademark) as a signal indicating a moving amount. One pulse corresponds to a predetermined unit movement amount, and the movement amount is obtained by counting the number of pulses. FIG.
Is a diagram showing an example of such a signal. The analog signal output from the detector changes according to the movement,
The change of the cycle corresponds to one scale (unit movement amount) of the scale. Therefore, the analog signal changes in a small cycle when the moving speed is high, and changes in a large cycle when the moving speed is low. If the moving speed is constant, the analog signal becomes a sine wave or a cosine wave. Such an analog signal is binarized by a comparator or the like, converted into a digital signal, and the change is counted to determine the movement amount.

However, in this case, the unit movement amount is the resolution of detection, and the resolution is determined by the grating pitch or the like. Therefore, to perform detection with higher resolution, the grating itself must be miniaturized. However, since it is not easy to miniaturize the grating, a high resolution is performed by detecting a signal change within one pulse. For example, if a signal having a duty ratio of 50% is output as the above pulse, the resolution is 2
It becomes possible to double. Currently, a method used to detect changes in one pulse more finely is to output a sine wave signal and a cosine wave signal instead of a pulse signal, and determine the angle of the signal within one cycle from the phase relationship. Position detection methods are widely used.

FIG. 2 is a diagram showing a sine wave signal, a cosine wave signal, and a binarized signal thereof, and also shows a double frequency signal obtained by calculating an exclusive OR (EXOR) of the binarized signal. is there. Since the sine wave signal and the cosine wave signal show changes as shown in FIG. 2, when they are input to a comparator having a threshold of zero level, a binarized signal as shown is obtained.
This corresponds to a pulse signal having a duty ratio of 50%. As is clear from the figure, the sine wave binarized signal and the cosine wave binarized signal are shifted in phase by 1/4 cycle, and if they are EXORed, a double frequency signal as shown is obtained. As a result, the angular position of the signal within one cycle can be detected with a resolution of 1/4 cycle.

If the sine wave signal and the cosine wave signal are compared more finely, the angular position of the signal can be detected more finely. If the amplitude of the sine (sin) signal and the amplitude of the cosine (cos) signal are the same, a coordinate point having the two signals as the Y coordinate and the X coordinate draws a circular locus when the angle changes, and the angular position of the signal is The straight line connecting the coordinate point and the origin is represented by an angle θ formed with the X axis. Therefore, from the ratio of the sin signal and the cos signal, the tangent (ta
n) If the signal is obtained, the angular position of the signal can be detected in detail. Therefore, one cycle is divided into a plurality of parts with the required angular resolution, the above θ is calculated, which part is determined, and a signal corresponding to the part is output to detect the angular position of the signal. Is being done. The angular position can be calculated by converting a sine wave signal and a cosine wave signal into a digital signal with an A / D converter and calculating the value with a computer. Angular positions indicated by the coordinate values are stored in a memory in association with each other, and corresponding angular positions are read out using the amplitude values of the sine wave signal and the cosine wave signal as coordinate values. FIG. 3 is a diagram for explaining the detection principle.

In the XY orthogonal coordinates, a square portion centered on the origin is divided into a grid as shown in the figure. Each divided part is defined by an X coordinate and a Y coordinate and has a corresponding angular position. For example, the portion P in the figure represents coordinates x,
y. Therefore, if each angular position is stored in correspondence with the coordinate value of each part, the angular position of the part defined by the coordinate value can be obtained without calculation by inputting the coordinate value. If the amplitude value of the cos signal corresponds to the X coordinate and the amplitude value of the sin signal corresponds to the Y coordinate, the angular position of the portion defined by the coordinate value is the angular position of the signal. Therefore, the angular position of the signal can be immediately detected.

The angular position can be set to various resolutions as required. For example, in FIG. 3, a signal corresponding to an angle range to which one cycle belongs is divided into 16 equal parts.
It may be equal or more. In FIG. 3, if the number of bits of the output data is 4, the angle range shown can always be expressed, and the angular position can be known without counting the output. Similarly, if the number of bits of the output data is increased, the number of divisions of the angle range that can be expressed increases. Note that the total movement amount means that when the angular position changes from the fourth quadrant to the first quadrant, it has moved by one unit, and thus the number of times is calculated and obtained. In addition, since the angular position is obtained, the direction of rotation can be identified, and the moving direction can be known. In this case, the number of times counted when the angular position changes from the first quadrant to the fourth quadrant is reduced.

In FIG. 3, the X direction and the Y direction are each divided into 16 equal parts, and the whole is divided into 256 parts. The number of divisions is defined by the resolution of the A / D converter, and as the number of divisions increases, the storage capacity must be increased accordingly. As is clear from the figure, near the origin of the coordinates,
Although the number of divisions becomes insufficient as compared with the change in the angle range, there is no problem since the amplitude of the signal does not actually become too small. Even if the amplitude of the signal changes, an accurate angle can be detected within a predetermined range. A ROM is used as storage means for storing the angle position, and a value obtained by digitally converting a sin signal and a cos signal is input as an address signal. In general, a configuration in which a value is stored in association with an address is called a look-up table, so this term is used here.

[0009]

As described above,
The signal output from the moire scale or magnescale is an analog signal, which is converted into a binary digital signal, or converted into a multi-bit digital signal by an A / D converter, and then a digital signal indicating an angular position. Is digitized. When converting an analog signal into a digital signal, there is a problem of quantization error that the value of the digital signal varies ± 1 in a minimum unit due to noise or the like. For example, A / D
Since the value of the digital signal converted by the converter fluctuates by ± 1, it may occur that the divided angle range indicates an adjacent range. This is the same in the case of binarization. When the signal strength is near the threshold of the comparator,
Even if the signal strength is the same, it may be H or L.

The problem of such a quantization error is not a problem when the detection signal is changing at a certain speed or more. For example, the detection signal is set to 10
If the change is monitored by sampling at more than twice the clock, even if the intensity of the detection signal is near the threshold value and an error occurs in binarization at the time of sampling, the signal changes in one direction and the next sampling Since the sampling is sometimes performed normally, there is no problem in the count value itself. The same applies to the case where the angle position is obtained from the sine wave and the cosine wave, and the angle position does not reverse.

However, when the change of the detection signal is delayed, the angular position signal may indicate the forward and the backward movement even though it changes in one direction. The same applies to the binarized signal. When the detection signal slowly changes near the threshold, the output becomes H.
And L may change in a short cycle. The binarized signal is sent to the counter and counted, and the count value indicates the amount of movement, but if the binarized signal changes between H and L in a short cycle even though it hardly moves as described above. The change is counted and an erroneous moving amount is output.

When the moving amount is controlled by the angular position signal, the control is changed by the angular position signal when the moving amount is controlled in one direction as described above, but the angular position signal indicates forward and backward. This is not preferable. Further, there is an application in which a displacement amount synchronization pulse having no direction is generated and used from the minimum bit of the angle position signal. In such a case, an erroneous pulse is generated.

An object of the present invention is to solve such a problem and to provide a processing circuit for processing so as to output a counting digital signal from which a quantization error has been removed when an analog signal is converted into a digital signal. An object of the present invention is to realize an angle position signal generation device that outputs an angle position signal from which a quantization error has been removed.

[0014]

In order to achieve the above object, a counting signal processing circuit according to the present invention counts when a counting digital signal changes and the changed state is maintained for a predetermined clock cycle. The quantization error is removed by processing to change the digital signal for use. That is, the counting signal processing circuit of the present invention is a counting signal processing circuit for processing a counting digital signal output in synchronization with a clock, and the digital signal is delayed by one cycle of the clock and the delayed signal. , A change detection circuit that detects from the digital signal held in the shift register that the digital signal has the same logical value for a predetermined number of clock cycles, and a digital signal based on the detection result of the change detection circuit. An output circuit for changing an output to the logical value when the signal has the same logical value for a predetermined number of clock cycles, and maintaining the output at other times.

When the digital signal for counting changes slowly, it is desirable to increase the number of cycles of the clock in which the digital signal for determining whether to change the output has the same logical state, and reducing the number of cycles has no effect. On the other hand, when the counting digital signal changes rapidly, if the number of clock cycles in which the digital signal is in the same logical state is increased, the signal delay becomes a problem and the counting digital signal changes. However, there is a problem that the output does not change and an erroneous signal is output. Therefore, the change detection circuit includes a plurality of sub-change detection circuits that detect that the digital signal has the same logical value during different numbers of clock cycles, and the output circuit has a clock corresponding to the plurality of sub-change detection circuits. A plurality of sub-output circuits for changing the output to a logical value when the logical value is the same for the corresponding predetermined number of cycles, and maintaining the output at other times, the signal processing circuit for counting includes a plurality of sub-output circuits. A selection circuit for selecting one of the outputs of the circuit is provided, and the output from the selection circuit can be selected according to the change period of the digital signal for counting.

Further, the angle position signal generating device of the present invention comprises:
A first A / D converter for converting a sine wave signal to a digital signal, and a second A / D converter for converting a cosine wave signal to a digital signal
A D converter and an angular position indicated by the coordinate values are stored in correspondence with the coordinate values of the rectangular coordinates, and a sine wave signal and a cosine signal are output according to the outputs of the first and second A / D converters input as address signals. A memory for outputting a digital signal indicating an angular position indicated by the wave signal, wherein the counting signal processing circuit is provided as a circuit for processing the digital signal output from the memory.

The change detection circuit has a plurality of sub change detection circuits, the output circuit has a plurality of sub output circuits, and the counting signal processing circuit selects one of the outputs of the plurality of sub output circuits. When a selection circuit is provided and an output from the selection circuit is made selectable in accordance with a change cycle of the counting digital signal, a speed detection circuit for detecting a change speed of the sine wave signal or the cosine wave signal is provided. The selection in the selection circuit is controlled in accordance with the changed speed.

[0018]

FIG. 4 is a view showing the structure of a uniaxial moving mechanism according to an embodiment of the present invention. As shown in the drawing, the moving mechanism of the embodiment is provided with support plates 5 and 6 on both sides of a base 1, a ball screw 3 and a guide rod 4 provided therebetween, and rotating the ball screw 3. As a result, the moving table 2 moves along the guide bar 4. The ball screw 3 is rotated by a pulse motor 7. The laser beam output from the head 8 of the laser length measuring device is reflected by a mirror 9 provided on the moving table 2 and returns to the head 8 to generate an interference fringe according to the optical path length therebetween. When the moving table 2 moves, the optical path length changes and the interference fringes change. Therefore, by detecting the change, the moving amount of the moving table 2 can be detected. In the head 8, interference fringes having different phases by 90 ° are captured by the optical fiber 21 and sent to the counter unit 10. The counter unit 10 converts the interference light signal sent through the optical fiber 21 into an electric signal, detects the number of changed interference fringes and the phase thereof, and sends them to the control device 11. The control device 11 controls the pulse motor 7 based on the number of changed interference fringes detected by the counter unit 10 and the phase thereof so that the movable table 2 moves to a position specified from the outside.

FIG. 5 is a diagram showing the configuration of the counter unit 10. As shown in FIG. As shown in the figure, the interference light signals having different phases by 90 ° sent by the optical fiber 21 are respectively converted into electric signals by light receiving elements 22A to 22D such as photodiodes. The electrical signals of the 0 ° and 180 ° interference light signals are input to the differential amplifiers 22A and 22B, and the offset is removed to generate a sine wave signal. Also, 90 ° and 27
The electrical signal of the 0 ° interference light signal is supplied to the differential amplifiers 22C and 22C.
D and the offset is removed to generate a cosine signal. Each of the sine wave signal and the cosine wave signal has 9 bits, and is input as an address of the angle position table ROM 25. On the other hand, the output of the light receiving element 22A is
The speed is input to the speed detection circuit 26, and the change speed of the interference light signal is detected. Specifically, the output of the light receiving element 22A is a signal that swings positively and negatively with respect to the zero level, and counts the number of clocks while the output crosses zero.

The angular position table ROM 25 stores 64 times for one round for 9-bit sine wave data and cosine wave data.
Divided angle position data is stored. That is,
The angular position data is represented by 6 bits, and the resolution is 5.
625 °. The 6-bit angle position data output from the angle position table ROM 25 is stored in the filter circuit 27.
And filtered. This filtering is a feature of the present invention. The 6-bit angular position data that has passed through the filter circuit 27 is input to the origin memory 28 and the cycle counter 29, and is also input to the adding circuit 30. The cycle counter 29 increases the count value by 1 when the 6-bit angular position data changes from the fourth quadrant to the first quadrant, and decreases the count value by 1 when the 6-bit angular position data changes from the first quadrant to the fourth quadrant. The number of interference fringes is counted. When the reset signal is input, the cycle counter 29 resets its value to zero. The count value of the cycle counter 29 is 27 bits, and is added to the 6-bit angle position data by the adding circuit 30.

When the reset signal is inputted, the angular position data shows a certain value, but this cannot be set to zero. Therefore, the value at the time of reset is stored in the origin memory 28, and the value is added by the subtraction circuit 31. Correction is performed by subtracting from the output of the circuit 30. As described above, displacement data indicating the number and angular position of interference fringes changed from the reset state is generated. The pulse generation circuit 32 generates a non-directional synchronization pulse using some bits of the displacement data.

Although the counter unit 10 of the embodiment has the above-described configuration, the difference from the conventional example is the filter circuit 27, which will be described in detail. FIG. 6 is a diagram showing a configuration of the filter circuit 27.
As shown, each bit of the angular position data is processed by a 0-bit filter circuit 41-0 to a 5-bit filter circuit 41-5. Each bit filter circuit has the same configuration and outputs four data signals a, b, c, and d. The set of four data signals a, b, c, and d is input to the selector 42, and one is selected and output according to the stage number switching signal from the speed detection circuit 26.

FIG. 7 is a diagram showing a configuration of each bit filter circuit. As shown, seven D-type flip-flops (FF) 51 to 57 are connected in series. The first FF 51 is a circuit that latches each bit of the angle position data output from the angle position table ROM 25 according to a clock. The FFs 52 to 57 are shift registers that delay data output from the FF 51 in the order of six clock cycles. Therefore, the angular position data delayed by the number of clock cycles corresponding to the number of stages up to that time is output from the FF of each stage. The output of the first FF 51 is the output a.

AND gates 58 and 59 are connected to FFs 51 and 5
The logical product of the output of 2 and the inverted output is calculated,
The output is sent to the RS flip-flop (FF) 60 S
Output as input and R input. The RS-type FF maintains its output when the S input and the R input are both zero (L),
When input is 1 (H) and R input is zero (L), output is 1
(H), and when the S input is 0 (L) and the R input is 1 (H), the output is 0 (L). S input and R input are both 1
In the case of (H), the output is undefined, but here, such a case does not occur. Therefore, the output of the RS-type FF 60 is
For each bit of the angular position data, the value changes and changes when two clock cycles continue, and at other times the original value is maintained. Therefore, when the value returns to the original value in one clock cycle, the output is not changed.
The output of the RS FF 60 is b.

Similarly, AND gates 61 and 62 and RS
The type FF 63 changes when the value of each bit of the angular position data changes and the value changes for four consecutive clock cycles, and maintains the original value at other times. For this reason,
If the value returns to the original value within three clock cycles or less, the output is not changed. The output of the RS FF 63 is c.
The AND gates 64 and 65 and the RS type FF 66 are
With respect to each bit of the angular position data, the value changes and changes when seven clock cycles continue, and at other times the original value is maintained. Therefore, when the value returns to the original value within six clock cycles or less, the output is not changed. The output of the RS FF 66 is d. The above output a,
b, c, and d are input to the selector 42.

FIG. 8 is a time chart showing changes in the outputs b, c and d of the bit filter circuit with respect to the angular position data a. As described above, b changes when a has the same value for two clock cycles, c changes when a has the same value for four clock cycles, and d changes when a has the same value for seven clock cycles. The selector 42 selects one of the angular position data a, b, c, and d according to the change speed of the interference light signal detected by the speed detection circuit 26.

In the counter unit of this embodiment, the output is changed only when the angular position data changes and the changed state is maintained for a predetermined clock cycle, so that the influence of the quantization error due to noise or the like can be eliminated. Moreover, the rate of change of the original signal is detected, and when the signal changes slowly, the number of clock cycles that must be maintained before the output changes is increased, so that the effects of quantization errors can be sufficiently eliminated. When the signal changes rapidly, the number of clock cycles that need to be maintained before the output changes is reduced, so that the signal delay is small,
There is no problem that an erroneous signal is output. In any case, the rate of change of the signal is detected, and the number of clock cycles that need to be maintained before the output is changed is changed according to the detection result. Errors can be eliminated.

[0028]

According to the present invention, a quantization error is removed from a counting digital signal generated by converting an analog signal into a digital signal, thereby enabling accurate counting and generation of an angular position signal. Thereby, the accuracy of position control of the moving device or the like can be improved.

[Brief description of the drawings]

FIG. 1 shows an analog signal output from a detector and the analog signal 2
FIG. 3 is a diagram illustrating an example of a digitized counting digital signal.

FIG. 2 is a diagram illustrating an example of a sine wave signal and a cosine wave signal output from a detector and a signal obtained by binarizing the sine wave signal and the cosine wave signal.

FIG. 3 is a diagram illustrating a principle of detecting an angular position based on a sine wave signal and a cosine wave signal.

FIG. 4 is a diagram illustrating an overall configuration of a mobile device according to an embodiment of the present invention.

FIG. 5 is a diagram illustrating a configuration of a counter unit according to the embodiment.

FIG. 6 is a diagram illustrating a configuration of a filter circuit according to an embodiment.

FIG. 7 is a circuit diagram showing a configuration of an individual filter circuit of the embodiment.

FIG. 8 is a time chart showing the operation of the individual filter circuits of the embodiment.

[Explanation of symbols]

 24A, 24B ... A / D converter 25 ... angular position table ROM 26 ... speed detection circuit 27 ... filter circuits 41-0 to 41-5 ... individual bit filter circuits 42 ... selector

 ──────────────────────────────────────────────────続 き Continuing on the front page F term (reference) 2F065 AA02 AA06 AA09 DD04 FF55 GG04 GG12 GG22 HH13 JJ01 JJ05 JJ18 KK01 LL02 LL12 PP12 PP22 QQ04 QQ12 QQ30 QQ51 UU05 2F077 AA21 CC07 PP19 TT49 TT49 Q44 TT49 TT49 TT46

Claims (4)

[Claims] 1. A counting signal processing circuit for processing a counting digital signal output in synchronization with a clock, the shift register delaying the digital signal by one cycle of the clock and holding the delayed signal. And from the digital signal held in the shift register,
A change detection circuit for detecting that the digital signal has the same logical value for a predetermined number of cycles of the clock; and, based on a detection result of the change detection circuit, the digital signal has the same logic for the predetermined number of cycles of the clock. An output circuit for changing an output to the logical value when the value is a value, and maintaining the output at other times. 2. The counting signal processing circuit according to claim 1, wherein the change detection circuit detects that the digital signal has the same logical value during different numbers of cycles of the clock. A change detection circuit, wherein the output circuit changes the output to the logic value when the logic value is the same for a predetermined number of cycles of the clock corresponding to the plurality of sub-change detection circuits, Sometimes comprising a plurality of sub-output circuits for maintaining the output, comprising a selection circuit for selecting one of the outputs of the plurality of sub-output circuits, and in response to a change period of the digital signal for counting, Counting signal processing circuit whose output can be selected. 3. An apparatus for generating a signal indicating an angular position from a sine wave signal and a cosine wave signal input from a detector, wherein the first A / D converts the sine wave signal into a digital signal.
A second A / D for converting the cosine signal into a digital signal;
A converter that stores an angular position indicated by the coordinate value in association with the coordinate value of the rectangular coordinate, and that outputs the sine wave signal and the 2. A memory for outputting a digital signal indicating an angular position indicated by the cosine wave signal, and a counting signal processing circuit according to claim 1 provided as a circuit for processing the digital signal output from the memory. A sine-wave cosine-wave angular position signal generation device. 4. The apparatus according to claim 3, wherein the change detection circuit of the counting signal processing circuit detects that the digital signal has the same logical value during different cycles of the clock. A plurality of sub-change detecting circuits, wherein the output circuit of the counting signal processing circuit outputs an output when the logical value is the same for a predetermined number of corresponding cycles of the clock corresponding to the plurality of sub-change detecting circuits. A plurality of sub-output circuits for changing to the logical value and maintaining the output at other times, a selection circuit for selecting one of the outputs of the plurality of sub-output circuits, and the sine wave signal or the cosine wave A speed detection circuit for detecting a change speed of a signal; and a sine / cosine angular position signal generation device for controlling selection in the selection circuit according to the detected change speed.