JP5146963B2 - Encoder error correction method - Google Patents

Description

  The present invention relates to a method for correcting an output pulse output from a pulse generation circuit based on the absolute value of a displacement of a movable object to be detected.

  2. Description of the Related Art Conventionally, a magnetic encoder is known as a device that detects the amount of displacement of a movable object to be detected and the absolute value of the displacement. For example, as one aspect of the magnetic encoder, one pole pair of magnets is rotated, the magnetic field change is detected by an MR element, and the obtained Sin signal and Cos signal are AD-converted into a microcomputer, There is one that detects an absolute value of a rotational position of a magnet by calculating an arc tangent signal and using the arc tangent signal (for example, Patent Document 1). In such a magnetic encoder, when a pulse is output through the pulse generation circuit based on the absolute value, the difference value obtained by subtracting the previous absolute value (one sample previous) from the current absolute value is generated in the microcomputer. Input to the circuit.

JP 2000-65596 A

  However, when converting (correcting) the output resolution to the absolute value that is set to increase the resolution of the pulse output inside the microcomputer, the difference value obtained by subtracting the previous absolute value from the current absolute value The resolution is reduced by shift operation or the like. The number of pulses output with low resolution is shifted from the absolute value inside the microcomputer due to the fractional part that was discarded during low-resolution calculations, so the accuracy of the output pulse decreases due to resolution conversion. . Specifically, this will be described with reference to FIG. FIG. 7 is a block diagram for explaining an output pulse correction method in a conventional encoder.

As shown in FIG. 7, the ABS_POS set to increase the resolution of the pulse output inside the microcomputer is subtracted from the ABS_POS of the previous time (one sample before) by the function of the delay unit 101. Then, the difference value δ is reduced in resolution by the arithmetic unit 102 that divides by 2 ^× . Finally, the integer part of the output value of the arithmetic unit 102 is input to the pulse generation circuit 103. When the output pulse is corrected by such a series of processing, since the decimal part of the input of the pulse generation circuit 103 is rounded down, an error is accumulated in the output pulse with respect to ABS_POS. As a result, the accuracy of the output pulse is reduced.

  The present invention has been made in view of such points, and its purpose is to correct an output pulse that can prevent a decrease in accuracy of the output pulse even when processing such as resolution conversion is performed. It is to provide a method.

  In order to solve the above problems, the present invention provides the following.

  (1) A sine wave A phase signal output from the A phase sensor in response to the displacement of the movable object to be detected, and a sine wave B to be output from the B phase sensor in response to the displacement of the movable object to be detected. In the method for correcting the output pulse output from the pulse generation circuit based on the absolute value, the absolute value of the displacement of the movable detection object is detected by analyzing the phase signal, and the input value of the pulse generation circuit A first step of accumulatively adding integer parts; a second step of multiplying an output value of the first step by a resolution conversion value comprising a predetermined number of integer powers; and an output value of the second step from the absolute value. A third step of subtracting, a fourth step of dividing the output value of the third step by the resolution conversion value, a fifth step of inputting an integer part of the output value of the fourth step to the pulse generating circuit, The Method of correcting the output pulse characterized by Mukoto.

  According to the present invention, first, the absolute value of the displacement of the movable object to be detected is detected using the sinusoidal A phase signal and B phase signal output from the A phase sensor and the B phase sensor. Then, when correcting (in this case, resolution conversion) the output pulse output from the pulse generation circuit based on the absolute value, the integer part of the input value of the pulse generation circuit is cumulatively added, and then a predetermined number (for example, 2) integers Multiply the resolution conversion value consisting of power. The value obtained as a result is subtracted from the above absolute value and then divided by the resolution conversion value (here, the resolution is lowered). Finally, the integer part of the resulting value is input to the pulse generation circuit. By such a series of processes, it is possible to prevent a reduction in accuracy of the output pulse.

  That is, when converting the resolution as one of the corrections, instead of taking the difference between the current absolute value and the previous absolute value as in the prior art, the estimated output position (predetermined number of A value obtained by multiplying the resolution conversion value consisting of an integer power) is created, and the difference between this value and the current absolute value is calculated (feedback of the estimated output position including the error, and this absolute value) The difference is taken between). Therefore, it is possible to prevent the positional deviation between the absolute value inside the microcomputer and the output position from exceeding a certain range, and thus to prevent the accuracy of the output pulse from being lowered.

  The first step of “cumulatively adding” the integer part of the input value of the pulse generation circuit is, for example, the first step when the integer part of the input value of the pulse generation circuit is 1, 1, 1, 2, 2. Are 1, 2, 3, 5, and 7. As described above, the first step means cumulative addition.

  (2) A sine wave A phase signal output from the A phase sensor in response to the displacement of the movable object to be detected, and a sine wave B to be output from the B phase sensor in response to the displacement of the movable object to be detected. In the method of correcting the output pulse output from the pulse generation circuit based on the absolute value, the input value of the pulse generation circuit is obtained by detecting the absolute value of the displacement of the movable detection object by analyzing the phase signal. A first step for cumulative addition; a second step for subtracting the output value of the first step from the absolute value; and a limit for limiting the maximum frequency of the pulse generating circuit to the output value of the second step. A method for correcting an output pulse, comprising: a third step of inputting to a circuit; and a fourth step of inputting an output of the limit circuit to the pulse generating circuit.

  According to the present invention, as described above, first, the absolute value of the displacement of the movable object to be detected is detected using the sine wave A phase signal and B phase signal output from the A phase sensor and the B phase sensor. Then, in correcting the output pulse output from the pulse generation circuit based on the absolute value (here, limit processing), after accumulating the input values of the pulse generation circuit, the resulting value is used as the absolute value described above. Subtract from the value. Then, the value obtained as a result is input to a limit circuit for limiting the maximum frequency of the pulse generation circuit, and finally, the output of the limit circuit is input to the pulse generation circuit. By such a series of processes, it is possible to prevent a reduction in accuracy of the output pulse.

  In other words, when the limit process is performed as one of the corrections, an error occurs in the difference value if the limit process is performed even once. In the present invention, as described above, the estimated output position including the error is fed back. As a result, it is possible to prevent the positional deviation between the absolute value in the microcomputer and the output position from exceeding a certain range, and thus to prevent a reduction in the accuracy of the output pulse.

  (3) The output pulse correction method, wherein the movable object to be detected is a permanent magnet in which a pair of S and N poles are magnetized.

  According to the present invention, since the above-mentioned movable object to be detected is a permanent magnet with a pair of S and N poles magnetized, the absolute value of the displacement of the pair of magnetized permanent magnets is detected. Even in this case, it is possible to prevent a reduction in accuracy of the output pulse. In particular, a single pole pair magnetic encoder has a simple mechanical structure compared to a multipole magnetic encoder. Therefore, it is easy to cope with downsizing while reducing the mechanical cost. On the other hand, circuit cost tends to increase. Even in such a case, according to the output pulse correction method of the present invention, it is possible to prevent a decrease in accuracy of the output pulse in terms of software without increasing the circuit cost.

  According to the present invention, the difference value including the calculation error is integrated and the estimated output position is fed back. Therefore, even when resolution conversion or limit processing is performed, the absolute value and output in the microcomputer are output. It is possible to prevent the deviation of the position from becoming larger than a certain range, and consequently to prevent the accuracy of the output pulse from being lowered.

  The best mode for carrying out the present invention will be described below with reference to the drawings. In the present specification, low resolution means that the high resolution output function set inside the microcomputer is reduced to a necessary resolution in order to meet the output resolution requirement of each customer. Also, if the “output resolution required by the customer (output resolution)” and the “resolution set in the microcomputer” are matched, processing such as shift calculation is not required, but the customer requires several types of output resolution. In this case, a microcomputer program corresponding to each request will be lined up, and it will be necessary to develop multiple models. On the other hand, if the inside of the microcomputer can support high-resolution output as in the model of this application, the output resolution matched to customer requirements can be handled by shift calculation processing, and one model can support multiple models. The complexity of the model can be avoided.

[Hardware configuration]
FIG. 1 is a block diagram showing a hardware configuration of an encoder 1 in which an output pulse correction method according to an embodiment of the present invention is performed. In particular, FIG. 1A shows an outline of the hardware configuration, and FIG. 1B shows an image diagram of the hardware configuration.

  As shown in FIGS. 1A and 1B, the encoder 1 includes an MR element (MR Element) 10 and a microcomputer 20. The Cos signal and Sin are sent from the MR element 10 to the microcomputer 20. A signal is input. More specifically, the MR element 10 outputs an A-phase sensor that outputs a sinusoidal Sin signal corresponding to the displacement of the movable object and a sinusoidal Cos signal corresponding to the displacement of the movable object. And a B-phase sensor for output.

  Here, in the present embodiment, a permanent magnet 50 in which a pair of S and N poles is magnetized is used as the “movable object” (see FIG. 1B). In FIG. 1A, only the MR element 10 that is one component of the A-phase sensor and the B-phase sensor is shown. However, for example, a rectifier circuit, a low-pass filter, a differential amplification amplifier, and an MR element 10 are used. A phase sensor and a B phase sensor are constituted by various electric elements such as a driver for supplying an exciting current to the A phase sensor.

  The microcomputer 20 has an electrical element such as an Interpolator or an RS422 driver (may be an Open Collector), but may have any other element. Further, as shown in FIG. 1A, an A / D conversion circuit (ADC 21) is provided, and the analog signal is digitized by the ADC 21. Although not particularly shown in FIG. 1, the microcomputer 20 has various electric elements such as a CPU, a ROM, and a RAM, and of course has a function of detecting the absolute value of the displacement of the movable detection object, as well as an output pulse. It also has a function of correcting the above. That is, various programs for executing an output pulse correction method are stored in a storage medium such as a ROM, and the CPU reads the program as appropriate and executes correction processing while using the RAM as a working area. Thus, the microcomputer 20 detects the absolute value of the displacement of the permanent magnet 50 by analyzing the A phase signal (Sin signal) and the B phase signal (Sin signal and a Cos signal having a phase difference of π / 2). And a function of performing correction processing such as resolution conversion and limit processing. Although not shown in FIG. 1, the microcomputer 20 includes a pulse generation circuit 14 (see FIG. 2) and also has a function of outputting a pulse.

[Output pulse correction method]
FIG. 2 is a block diagram for explaining an output pulse correction method according to the first embodiment of the present invention. FIG. 3 is an explanatory diagram for explaining the state transition of the block diagram shown in FIG.

2, the microcomputer 20 includes an arithmetic unit 11 and dividing by ^{2 x,} an arithmetic unit 12 for multiplying the ^{2 x,} has a (1 sample) delay unit 13, a pulse generating circuit 14, a. Note that the computing unit 11, the computing unit 12, the delay unit 13, and the like may be separate independent computing units, or may be realized by a single chip in terms of software. ABS_POS in the figure indicates an absolute value set to increase the resolution of the pulse output inside the microcomputer, and δ in the figure indicates a value obtained by subtracting a feedback value (estimated output position) from ABS_POS. Show.

First, the feedback loop will be described. Value input to the pulse generating circuit 14 performs output value S ₄ containing fraction output by the processing of the fourth step, the process for the rounded value to an integer closer to 0 in the fifth step ("Processing for obtaining a value rounded to an integer close to 0" is indicated by "Int" in the dotted frame in the fifth step in FIG. 2). This process results in a value rounded to the nearest integer from 0 to a value that includes a decimal (the state of the output value defined by S _5), i.e., the integer part cumulative addition (dotted first step in FIG. 2 Frame). The state of this output value is defined by S _6.

Then, multiplying the resolution conversion value 2 ^x for this output value S ₆ of two of x square (dotted frame in the second step in FIG. 2). This is for returning the low resolution value to be input to the pulse generation circuit 14 to the original high resolution value in the feedback loop, and for adjusting the dimension of the arithmetic processing in the next step. , and the define the state of the output value S _2. Then, from ABS_POS, it subtracts the value in the state of _{S 2} (dotted frame in the third step in FIG. 2). Then, the difference value δ is obtained. The state of ABS_POS defined _{S 1,} the state of the difference value δ is defined by _{S 3.}

Next, the difference value δ is divided by the resolution conversion value ½ ^x (dotted line frame in the fourth step in FIG. 2). The output value is comprise a fraction, to define the state of the output value S _4. Then it performs processing to round the S ₄ to an integer close to zero of the. (Dotted line frame in the fourth step in FIG. 2). Finally, a process int for the output value S ₄ including the decimal to the value rounded to an integer closer to 0, inputs the integer part to the pulse generation circuit 14 (broken line in the fifth step in FIG. 2 frame).

  Even if the arithmetic unit 11 performs resolution conversion on the absolute value ABS_POS set so as to increase the resolution of the pulse output inside the microcomputer, such a series of processes is set inside the microcomputer 20. Thus, it is possible to prevent the positional deviation between the high-resolution absolute value and the position output with a reduced resolution from exceeding a certain range, and thus to prevent the accuracy of the output pulse from being lowered.

This will be described more specifically with reference to FIG. Here, the resolution conversion value is set to 1/2 (that is, x = 1). For example, as shown in the second row, when the first state S ₁ of ABS_POS is 2.2, there is no feedback at first, so S ₂ = 0, while S ₃ is 2.2 (= S ₁ -S _{2), S 4} is _{1.1 (= S 3/2)} , S 5 is 1 (= integer portion of the _{S 4),} the initial value _{of S 6} is 1 (delay unit 13 is set to 0 ).

Next, as shown in the third row, when the ABS_POS state S ₁ is 4.4, the arithmetic unit 12 feeds back the value S _{2 that} is twice the S ₆ state (= 1) described above. Therefore, the state S ₃ of the difference value δ becomes 2.4 (= S ₁ −S ₂ = 4.4-2). Then, _{S 4} is _{1.2 (= S 3/2)} , S 5 is 1 (= integer portion of the _{S 4), S 6} is 2 (= _{S 5} + delayer previous sample by 13 _S 6 = 1 + 1 )

Next, as shown in the fourth stage, when the state _{S 1} of ABS_POS was 6.6, the arithmetic unit 12, twice the value _{S 2} states _{S 6} described above (= 2) are fed back Therefore, the state S ₃ of the difference value δ becomes 2.6 (= S ₁ −S ₂ = 6.6-4). Then, _{S 4} is _{1.3 (= S 3/2)} , S 5 is 1 (= integer portion of the _{S 4), S 6} is 3 (= _{S 5} + delayer previous sample by 13 _S 6 = 1 + 2 )

Thereafter, the calculation proceeds in the same manner. When reaching the 10 stage from the top, when the state _{S 1} of ABS_POS was 19.8, the arithmetic unit 12, twice the value _{S 2} of one sample previous state of _{S 6} (= 8) is Since the feedback is provided, the state S ₃ of the difference value δ becomes 3.8 (= S ₁ −S ₂ = 19.8−16). Then, _{S 4} is _{1.9 (= S 3/2)} , S 5 is 1 (= integer portion of the _{S 4), S 6} is 9 (= _{S 5} + delayer previous sample by 13 _S 6 = 1 + 8 )

Here, the cumulative error is calculated. Since the state S ₁ of ABS_POS is 19.8 originally value before multiplying the resolution conversion values of a predetermined integer number of power for the estimated output position, i.e. the value before multiplying the 2 in the present embodiment (S ₆ state) must be 19.8 ÷ 2 = 9.9. However, since the actual state of S ₆ is 9, it can be seen that the cumulative error is 0.9.

However, as shown in 11 paragraph from the top, when the state _{S 1} of ABS_POS was 22.0, calculator 12 by, 1 2 times the value of the previous sample state of _{S 6} (= 9) _{S 2} Is fed back, the state S ₃ of the difference value δ becomes 4.0 (= S ₁ −S ₂ = 22−18). Then, _{S 4} is _{2 (= S 3/2)} , S 5 is 2 (= integer portion of the _{S 4), S 6} and 11 (= _{S 5} + delayer previous sample by 13 _S 6 = 2 + 9) Become.

Here, the cumulative error is calculated again. Since the state S ₁ of ABS_POS is 22.0, the value before multiplying the resolution conversion values of a predetermined number of integral power with respect to the estimated output position, i.e. the value before multiplying the 2 in this example (the state S ₆₎ Where 22 ÷ 2 = 11, the actual state of S ₆ is 11, and it can be seen that the cumulative error is 0. Thus, at the timing when the integer part of the state of the S ₅ increases 1, it can be seen that the accumulated error is canceled.

[Main effects of the embodiment]
As described above, according to the method of correcting the output pulse according to the present embodiment, since the estimated output position by multiplying the difference value comprising a calculation error are to be fed back (the state in FIG. 2 S ₆₎ Even when resolution conversion is performed by the computing unit 11 (in FIG. 2, the resolution is reduced to 1/2), the deviation between the absolute value in the microcomputer and the output position is within a certain range (0.9 in FIG. 2). It can be prevented from becoming larger than that, and as a result, a reduction in accuracy of the output pulse can be prevented. Further, in the present embodiment, a magnetic encoder of a single pole pair (using the permanent magnet 50) is assumed, and a reduction in the accuracy of the output pulse can be prevented while suppressing the circuit cost.

  Specifically, according to the output pulse correction method according to the present embodiment, the output pulse can be corrected stepwise as shown in FIG. FIG. 8A is a graph of the result of the conventional correction method, and FIG. 8B is a graph of the result of the correction method according to the present embodiment.

  8A and 8B, the line indicated by a broken line represents a change in the cumulative number of output pulses when there is no cumulative error. Originally, the cumulative number is preferably changed. However, as shown in FIG. 8A, it can be seen that in the conventional correction method, the error gradually increases with each sampling step, that is, with time (as the permanent magnet 50 rotates).

  On the other hand, as shown in FIG. 8B, the correction method according to the present embodiment corrects in a staircase pattern, so that it follows that the line indicated by the broken line follows. In other words, the error is not accumulated, and even if a certain error occurs, the error is corrected to a range without the original accumulated error, so that it is possible to prevent a reduction in accuracy of the output pulse.

[Modification]
4 to 6 are block diagrams for explaining the output pulse correction method according to the second embodiment of the present invention. FIG. 4 shows a conventional block diagram when the limit process is not performed, FIG. 5 shows a conventional block diagram when the limit process is performed, and FIG. 6 shows a block of the present embodiment when the limit process is performed. The figure is shown.

  As shown in FIG. 4, conventionally, resolution conversion 21 (lower resolution) is performed on the absolute value ABS_POS, instead of performing resolution conversion on the difference value. The pulse generator 23 receives a value δ obtained by subtracting the POS ′ after resolution conversion of one sample before from the POS ′ after resolution conversion by the delay unit 22.

  In this case, no particular error occurs. However, as shown in FIG. 5, a problem arises when a limit value is required for the value of δ due to the maximum frequency limitation of the pulse generation circuit 23. Specifically, POS ′ whose resolution has been reduced by the resolution conversion 31 becomes δ by the delay unit 32, then becomes δlim through the limit circuit 33, and then is input to the pulse generation circuit 34. Then, since δ and δlim are different values, an error occurs if the limit is reached even once.

  Therefore, as shown in FIG. 6, according to the output pulse correction method according to the second embodiment of the present invention, this error can be reduced. Specifically, the estimated output obtained by integrating and feeding back the difference value δlim including the calculation error by the delay unit 44 is subtracted from POS ′ whose resolution is reduced by the resolution conversion 41. The obtained δ is input to the limit circuit 42.

  That is, the input value δlim of the pulse generation circuit 43 is cumulatively added (dotted line frame in the first step in FIG. 6), and the output value in the first step is subtracted from the absolute value POS ′ to obtain δ (in FIG. 6). ) Is input to the limit circuit 42 for limiting the maximum frequency of the pulse generation circuit 43 (the third step dotted line frame in FIG. 6), and finally, the limit circuit 42 Is input to the pulse generation circuit 43 (the dotted frame in the fourth step in FIG. 6). As a result, even when limit processing is performed, it is possible to prevent the deviation between the absolute value inside the microcomputer and the output position from exceeding a certain range, as in the first embodiment described above. As a result, it is possible to prevent a reduction in accuracy of the output pulse.

  The output pulse correction method according to the present invention is useful as a method capable of preventing a decrease in accuracy even when corrections such as resolution conversion and limit processing are performed.

It is a block diagram which shows the hardware constitutions of the encoder 1 with which the correction method of the output pulse which concerns on embodiment of this invention is performed. It is a block diagram for demonstrating the correction method of the output pulse which concerns on the 1st Embodiment of this invention. It is explanatory drawing for demonstrating the state transition of the block diagram shown in FIG. It is the conventional block diagram in the case of not performing a limit process. It is the conventional block diagram in the case of performing a limit process. It is a block diagram of this embodiment in the case of performing a limit process. To explain a method of correcting an output pulse in a conventional encoder It is a graph for demonstrating the comparison with the former about the correction method of the output pulse which concerns on the Example of this invention. Is a block diagram of

Explanation of symbols

11, 12 arithmetic unit 13 delay unit 14 pulse generation circuit

Claims (3)

A sinusoidal A-phase signal output from the A-phase sensor corresponding to the displacement of the movable detection object, and a sine-wave B-phase signal output from the B-phase sensor corresponding to the displacement of the movable detection object In the method of correcting the output pulse output from the pulse generation circuit based on the absolute value, the absolute value of the displacement of the movable detection object is detected by analyzing
A first step of cumulatively adding an integer part of an input value of the pulse generation circuit;
A second step of multiplying the output value of the first step by a resolution conversion value consisting of a predetermined number of integer powers;
A third step of subtracting the output value of the second step from the absolute value;
A fourth step of dividing the output value of the third step by the resolution conversion value;
And a fifth step of inputting an integer part of the output value of the fourth step to the pulse generation circuit. A sinusoidal A-phase signal output from the A-phase sensor corresponding to the displacement of the movable detection object, and a sine-wave B-phase signal output from the B-phase sensor corresponding to the displacement of the movable detection object In the method of correcting the output pulse output from the pulse generation circuit based on the absolute value, the absolute value of the displacement of the movable detection object is detected by analyzing
A first step of cumulatively adding the input values of the pulse generation circuit;
A second step of subtracting the output value of the first step from the absolute value;
A third step of inputting the output value of the second step to a limit circuit for limiting the maximum frequency of the pulse generation circuit;
And a fourth step of inputting the output of the limit circuit to the pulse generation circuit.   3. The output pulse correction method according to claim 1, wherein the movable object to be detected is a permanent magnet in which a pair of S and N poles are magnetized.

Images