JP4622725B2 - Magnetic absolute encoder - Google Patents

Description

  The present invention relates to a magnetic absolute encoder that can be used for detecting an absolute position of a linear or rotary moving body and can obtain an absolute value signal stably and accurately.

  A conventional magnetic absolute encoder reads an absolute code recorded on a code plate with a plurality of detection heads as disclosed in Patent Document 1, for example. The length corresponding to one code of the absolute code is λ, and the magnetization direction of the magnetized portion of the absolute code consists of two pairs of NS direction and SN direction per λ. An MR element is used as a magnetic sensor for reading an absolute code. The MR element can measure the absolute value of the magnetic field but cannot determine the polarity. Therefore, since there is a part where the direction of the magnetic field is reversed when the code of one magnetized part is read, there is a part where the output of the magnetic sensor becomes 0, and two peaks are formed as shown in FIG. Output. Also, when the magnetized portions are continuous, a peak twice as many as the number of codes corresponding to the magnetized portions is formed. In order to generate an absolute code, a threshold value is provided to the output of the MR element so that an output above the threshold value is output as “1” and an output below the threshold value is output as “0”. ”Should be output, but a portion where the output is“ 0 ”appears in the middle, so the absolute code cannot be accurately detected.

  As a means for solving this problem, the magnetic sensors S1 and S2 are arranged at positions shifted by λ / 2 and their combined values are taken, thereby eliminating the output of “0” generated in the magnetized portion. . However, if the combined value of the MR element outputs at positions shifted by λ / 2 is taken, the length having the output of “1” corresponding to the magnetized portion is expanded by λ / 2. For example, the magnetized portion In the case of one, the length of “1” is (3/2) λ, and the length of three consecutive magnetized portions is (7/2) λ in total. Since the length of the magnetized portion is increased in this way, a method of shortening the length of the magnetized portion is also shown. Thus, the code of the magnetized portion is “1” and the code of the non-magnetized portion. It has been shown that a certain “0” length approaches the correct one and reads the appropriate absolute code. This method is effective in a position where the magnetization direction of the magnetized portion is strongly reflected, that is, in an area where the distance between the code plate and the magnetic sensor is extremely short. Can read the absolute code.

  However, one of the code plate and the magnetic sensor in the magnetic encoder is a movable part, but is spaced by a certain distance in order to reduce component mounting error margin and incremental track reading error when multiple tracks are configured. May be necessary. If the distance between the code plate and the magnetic sensor is increased, the magnetic field distribution and the magnetic sensor exhibit nonlinear characteristics, so that it is difficult to accurately perform the encoding corresponding to the magnetized portion and the non-magnetized portion. For example, when one magnetized part is measured with an MR element, two peaks are formed. However, as the distance between the code plate and the magnetic sensor increases, the valley between the peaks spreads. become. Further, in the vicinity of the boundary between the magnetized portion and the non-magnetized portion, a region of “1” corresponding to the magnetized portion extends to the non-magnetized portion. As a result, “0” is output at the place where “1” should be output, or the area where “1” is output at the position where “0” should be output increases. Even if the magnetic sensors are arranged at positions shifted from each other by λ / 2 as shown in Patent Document 1 and they are combined, it is difficult to fill the valley portion between the peaks. Therefore, there may be a portion where the correspondence between the magnetized portion and the non-magnetized portion cannot be accurately taken. In such a case, it is difficult to read the absolute code correctly, and an error may occur in the detection of the absolute position.

  Also, the vicinity of the boundary between the magnetized portion and the non-magnetized portion is shown in FIG. As described above, with reference to the outputs of the incremental tracks arranged in parallel, the output having the higher reliability among the elements arranged at positions different from each other by λ / 2 is taken. However, the magnetic field at the sensor position also fluctuates due to the influence of the variation in the distance between the code plate and the magnetic sensor, the temperature change, etc., and it becomes difficult to ensure the reliability of the encoding.

Japanese Patent No. 3454002

  The conventional magnetic absolute encoder has a problem that the reliability of discriminating between “0” and “1” near the boundary between the magnetized portion and the non-magnetized portion is low from the characteristics of the magnetic field distribution. The present invention has been made in order to solve the above-described problems, and can perform absolute position detection with high accuracy by performing stable and accurate encoding of an absolute value code. Another object is to obtain an absolute encoder of the type.

A magnetic absolute encoder according to the present invention is a magnetic absolute encoder that generates an absolute value code by a combination of a magnetized portion and a non-magnetized portion of a magnetic scale and outputs absolute data. The magnetization direction of the portion is characterized by having the same magnetization direction as the longitudinal direction of the magnetic scale when the magnetized portions are continuous.

The magnetic absolute encoder according to the present invention generates a first absolute value code by a combination of a magnetized portion and a non-magnetized portion having the same magnetization direction in the longitudinal direction of the magnetic scale when continuous. , A first magnetic sensor for detecting the magnetic flux of the first track, and an absolute detecting means for the detected magnetic flux.

A magnetic absolute encoder according to the present invention is a magnetic absolute encoder that generates an absolute value code by a combination of a magnetized portion and a non-magnetized portion of a magnetic scale and outputs absolute data. Since the magnetization direction of the part is the same magnetization direction as the longitudinal direction of the magnetic scale when the magnetized part continues, the absolute value code can be encoded stably and accurately, and the absolute position can be detected. Can be performed with high accuracy.

The magnetic absolute encoder according to the present invention generates a first absolute value code by a combination of a magnetized portion and a non-magnetized portion having the same magnetization direction in the longitudinal direction of the magnetic scale when continuous. , A first magnetic sensor for detecting the magnetic flux of the first track, and an absolute detection means for the detected magnetic flux, so that encoding of the absolute value code can be performed stably and accurately. The absolute position can be detected with high accuracy.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing a magnetic absolute encoder according to Embodiment 1 of the present invention, and an absolute position is detected by the following operation. In the figure, for example, in the case of 4 bits, the magnetic scale 11 has an absolute value code pattern in which 0 is added to the head of a 4-bit M series, and is composed of 16 codes in total. The length corresponding to one code of the absolute value code is λ, and 1a, 1b, 1c, 1d, 1e, 1f, 1g, and 1h of the length λ are code “0”.
, 2a, 2b, 2c, 2d, 2e, 2f, 2g, and 2h of the length λ are magnetized portions corresponding to the code “1”. The magnetized portion corresponding to the code “1” is magnetized only at a portion having a length α near the center, and α takes a value such as λ / 2, for example. Hereinafter, a portion corresponding to the reference numeral “1” is referred to as a “magnetized portion”, and a portion of the magnetized portion that is actually magnetized is referred to as a “magnetized portion”. Magnetization direction of magnetized matter is, successive 2a, 2b, 2c, 2d are made in the direction of the SN from the left to the right in the figure, below, 2e is NS direction, 2f and 2g is SN direction, 2 h is NS It has become a direction. The portions where the magnetized portions are continuous have the same magnetization direction, and the directions of the magnetization directions are reversed before and after the non-magnetized portion.

  That is, all the magnetized portions corresponding to the continuous code “1” have the same magnetization direction, and the magnetization direction of the next magnetized portion across the non-magnetized portion corresponding to the code “0” is the previous magnetization. The direction is reversed from the magnetization direction of the portion. The magnetic field formed by these magnetization patterns is measured by the magnetic sensor 12. The magnetic sensor 12 includes, for example, a plurality of MR elements (magnetoresistance effect elements). In order to acquire 4-bit absolute value code data, the four MR elements are arranged at positions 3a, 4a, 5a, 6a. In addition, the distance between each element is separated by a length λ corresponding to one code of the absolute value code, and the element is arranged at a position away from the magnetic scale 11 by a predetermined distance. The output of the magnetic sensor 12 is converted into absolute position data by referring to an absolute value table by a known absolute detection means 13.

  By the way, in order to accurately detect the absolute position, signals corresponding to the codes “1” and “0” recorded on the magnetic scale must be accurately detected by the magnetic sensor 12. That is, it is necessary to reliably detect “1” in the portion corresponding to the code “1” and “0” in the portion corresponding to the code “0” by the magnetic sensor. If there is an error in the encoding, accurate absolute position data cannot be obtained.

  FIG. 2 shows a case where the magnetization direction of one magnetized portion is composed of two pairs of NS direction and SN direction, and is for comparison with the present invention. A pattern, magnetic flux density distribution, MR element output, and encoding output are shown. In FIG. 2, the horizontal direction is the X direction. The magnetic flux density distribution indicates the X-direction component of the magnetic flux density at a position away from the magnetic scale by a certain distance r1 or r2. The solid line is the distance r1 and the broken line is the distance r2. . The magnetic flux density distribution in the X direction is obtained by magnetic field analysis calculation and experiment and coincides with the magnetosensitive direction of the MR element. The MR element output is an output obtained in consideration of the sensitivity characteristics of the MR element, and the encoded output is a constant threshold value for the MR element output, and “1” is set above the threshold value. The value less than the threshold value is set to “0”.

  As can be seen from FIG. 2, the magnetic flux density distribution is obtained with alternating positive and negative polarities, and the MR element output has a shape that repeats a peak and a valley. As a result, the encoded output repeats “1” and “0”, and “1” appears twice and “0” appears once in one magnetized part (λ1). When two magnetized portions are continuous (for λ2), “1” appears four times and “0” appears three times. The “0” region appearing between the peaks of one magnetized portion becomes larger as the distance between the magnetic scale and the magnetic sensor is longer as shown in FIG. In the vicinity of the boundary between the magnetized portion and the non-magnetized portion, the “1” region corresponding to the magnetized portion extends to the non-magnetized portion, and becomes more prominent as the distance between the magnetic scale and the magnetic sensor increases. In such a case, “0” is output where it should be encoded as “1” or “1” is output where it should be encoded as “0”. Accurate encoding cannot be performed, and an error may occur in detection of the absolute position.

  On the other hand, FIG. 3 shows the magnetic flux density distribution formed by the magnetic scale 11 by the magnetization method of the present invention, the MR element output, and the encoded output. The definitions of magnetic flux density distribution, MR element output, and encoded output in FIG. 3 are the same as those in FIG. In the drawings, the same reference numerals and the same notations are the same or equivalent, and this is common throughout the entire specification. Further, the description of the constituent elements appearing in the whole specification is merely an example and is not limited to these descriptions.

  In FIG. 3, the length α of the magnetized portion is shorter than the length λ of the magnetized portion corresponding to the reference numeral “1”, and all the continuous magnetized portions have the same magnetization direction and sandwich the non-magnetized portion. The magnetization direction of the next magnetized portion is opposite to the magnetization direction of the previous magnetized portion. As can be seen from the figure, the magnetic flux density distribution shows the same polarity when the magnetized portions are continuous, and the polarity of the next magnetized portion sandwiching the non-magnetized portions is reversed. Since MR element which outputs the absolute value of the magnetic field irrespective of the positive or negative polarity becomes the same magnetization direction when the magnetized component is continuous, the value of a certain value or more is output in magnetized component, toward zero Is not output. That is, in the encoded output, when the magnetized portion is continuous, “0” is not output at all and all “1” is output. On the other hand, only a value close to 0 is obtained in the non-magnetized portion. Further, in the vicinity of the boundary of the portion corresponding to the part and the code "0" corresponding to the magnetic scale of the code "1", "1" never region extends to the portion of "0", and the magnetic scale code The encoded output obtained from the MR element output corresponds appropriately. Therefore, the encoding of the absolute value code can be performed stably and accurately, and the absolute position can be detected with high accuracy.

  Therefore, since the length of the magnetized portion of the magnetized portion is shorter than the length corresponding to one code of the absolute value code, the absolute value code can be encoded stably and accurately. The position can be detected with high accuracy. At this time, the magnetized portion is arranged at the center of the length corresponding to one code of the absolute value code.

  In the M series, the inversion number of “0” and “1” is an even number, and the absolute value conversion code of the present invention also has only “0” added to the head of the M series, so “0” and “1”. The number of inversions is even. Therefore, as shown in FIG. 4, in the case where the magnetic scale 11 is arranged around the rotating body and the rotation angle is measured, the absolute value code pattern circulates, but "0" and "1" ”Is an even number, the magnetization directions of the magnetized portions on both sides of the non-magnetized portion are opposite to each other at any location. Therefore, the absolute position detection according to the present invention can also be applied to a rotating body.

  In the above example, an absolute value code obtained by adding 0 to a 4-bit M series is shown. However, a higher resolution can be achieved by using a larger number of bits. The encoding in the above example uses the code “1” for the magnetized portion and the code “0” for the non-magnetized portion, but the same effect can be obtained even if the code is reversed. In the following embodiments, the magnetization direction of the magnetized portion of the magnetic scale detected by the absolute detecting means 13 is the same magnetization direction when the magnetized portion continues, and the magnetization direction before and after the non-magnetized portion. It is common to reverse the direction.

  From the above, it is a magnetic absolute encoder that generates an absolute value code by combining a magnetized part and a non-magnetized part of a magnetic scale and outputs absolute data, and the magnetization direction of the magnetized part is Since the magnetization direction is the same when the magnetized portions are continuous, the absolute value code can be encoded stably and accurately, and the absolute position can be detected with high accuracy.

  Further, in the case of continuous, a first track for generating an absolute value code by a combination of a magnetized portion and a non-magnetized portion having the same magnetization direction, and a first magnetic for detecting the magnetic flux of the first track Since the sensor and the absolute detection means for the detected magnetic flux are provided, the absolute value code can be encoded stably and accurately, and the absolute position can be detected with high accuracy.

  Further, since the magnetization direction of the magnetized portion is reversed before and after the non-magnetized portion, the encoded output does not output “0” when the magnetized portion continues. All are output as “1”, “0” is output in the non-magnetized portion, and “1” is surely output in the next magnetized portion. Therefore, the encoding of the absolute value code can be performed stably and accurately, and the absolute position can be detected with high accuracy.

Embodiment 2. FIG.
FIG. 5 shows the magnetic flux density distribution, MR element output, and encoded output formed by the magnetic scale according to the magnetization method of Embodiment 2 of the present invention. Unlike the magnetic scale 11 of the first embodiment, the magnetized part and the magnetized part coincide with each other. That is, the case where the length α of one magnetized portion coincides with the length λ corresponding to one code of the absolute value code, and as a result, n codes “1” corresponding to the magnetized portion are obtained. When continuous, the magnetized region has one polarity and the length is nλ. For this reason, the magnetic flux density distribution shows the same polarity when the magnetized portions are continuous, and the polarity is reversed at the next magnetized portion across the non-magnetized portion. As a result, when the magnetized portion continues, “0” is not output at all and “1” is all output, so that the absolute value code can be encoded accurately and stably. Become.

Embodiment 3 FIG.
FIG. 6 shows the magnetic flux density distribution formed by the magnetic scale, the MR element output, and the encoded output formed by the magnetization method according to the third embodiment of the present invention. Compared to the second embodiment, the magnetized portion of the magnetized portion in contact with the non-magnetized portion is narrowed. In order to accurately obtain an output near the boundary between the portion corresponding to the code “1” and the portion corresponding to the code “0” of the magnetic scale, it can be realized by shortening the length of the magnetized portion.

  The length of the magnetized portion may be determined appropriately depending on the distance between the magnetic scale and the magnetic sensor, the characteristics of the MR element, etc. For example, only the end of the magnetized portion in contact with the non-magnetized portion is shortened by λ / 4. If the number “1” of “1” continues, the total length of the magnetized region (magnetized portion) becomes {nλ− (1/2) λ}. As a result, the output near the boundary between the portion corresponding to the code “1” and the portion corresponding to the code “0” becomes appropriate, and the encoding of the absolute value code can be performed more stably and accurately.

  Therefore, when the magnetized portion is continuous, the magnetized portion is set at a predetermined distance from the end in contact with the non-magnetized portion, so that the encoding of the absolute value code can be performed stably and accurately. The absolute position can be detected with high accuracy.

Embodiment 4 FIG.
FIG. 7 is a block diagram showing a magnetic absolute encoder according to the fourth embodiment of the present invention. The magnetic scale is composed of two tracks, and is magnetized in the NS direction and the SN direction at a predetermined interval over the entire area and the first track by the magnetic scale 11 in which the pattern of the absolute value code is magnetized. It consists of a second track by the magnetic scale 15. In addition, although it demonstrates using the magnetic scale 11 of Embodiment 1, you may use the magnetic scale of Embodiment 2 or 3.

In the case of 4 bits, the magnetic sensor 12 of the first embodiment in which four MR elements are arranged at positions 3a, 4a, 5a, 6a separated by a length λ corresponding to one code of the absolute value code. In addition to the arrangement, four MR elements are arranged at positions 3b, 4b, 5b, and 6b that are separated from the positions 3a, 4a, 5a, and 6a where these four MR elements are arranged by λ / 2. That is, the MR element for reading the absolute value code of the first track is provided with 3b, 4b, 5b and 6b which are separated from each other by λ / 2 in addition to 3a, 4a, 5a and 6a. Yes.
Therefore, two sets of magnetic sensors 12 shifted by a half wavelength (λ / 2) are provided.

  The code “0” or “1” is output by the MR elements 3a and 3b, but there is an MR element in the vicinity of the bit-inverted portion such as “0” to “1” or “1” to “0” Sometimes the output is unstable and unreliable. However, when either one of 3a or 3b is in this unstable portion, the other MR element is located at a distance of λ / 2, so that it is sufficiently far from the unstable portion and its output is stable. . That is, since either one of 3a or 3b can obtain a stable output, it is possible to detect the position stably and accurately by switching the output of which element is adopted as necessary.

  For this switching, a switching signal obtained by measuring the magnetic field of the magnetic scale 14 as the second track by the magnetic sensor 15 and obtained by the incremental detection means 16 is used. The magnetic scale 14 repeats this basic unit with a length 2λ as a basic unit in two pairs, one in the NS direction and one in the SN direction per length λ. For this reason, the distribution of the magnetic field generated by the magnetic scale 14 has a sine wave shape, and the output of the MR element 7a constituting the magnetic sensor 15 also has a sine wave shape. The magnetic scale 14 is an alternating magnetism having a wavelength of 2λ parallel to the magnetic scale 11 and can be synchronized with the magnetic scale 11. Here, a representative switching method between 3a and 3b among MR elements 3a, 3b, 4a, 4b, 5a, 5b, 6a and 6b for reading the absolute value code will be described with reference to FIG.

  In FIG. 8, the MR element a output indicates the output of the MR element 3a, and the encoded a output indicates an appropriate threshold value for the output of the MR element 3a. Less than “0”. The MR element b output and the encoded b output are those of the MR element 3b, and the definitions are the same as those of the MR element 3a, but the position of the element is separated from the MR element 3a by λ / 2. The only difference is that. The switching signal is a signal output from the incremental detection means 16, and is generated based on the sine wave signal detected by the MR element 7a.

  In FIG. 7, since the magnetization pattern of the magnetic scale 11 and the magnetization pattern of the magnetic scale 14 are synchronized with a period λ, referring to the sine wave signal that is the output of the MR element 7a, the MR element 3a or 3b The approximate timing at which the output bit is inverted can be specified. Therefore, the switching signal can be output so as to avoid the timing at which the output of the MR element 3a or 3b is bit-inverted. In FIG. 8, when the switching signal level is H, the encoded a output is read, and when the switching signal level is L, the encoded b output is read to avoid the timing of bit inversion of the MR elements 3a and 3b. A detection signal can be generated. As a result, highly reliable data can be read, and the absolute value code can be encoded more stably and accurately.

  The position where the detection signal in FIG. 8 changes from “0” to “1” or “1” to “0” is shifted by a predetermined distance in the X direction with respect to the magnetization pattern of the magnetic scale. Since this amount is a specific value fixed depending on the system configuration, it can be converted into an absolute value.

  Therefore, the first magnetic sensor is composed of two sets of magnetic sensors whose half the length corresponding to one code of the absolute value code is shifted, and is magnetized in the NS direction and the SN direction at predetermined intervals. Since it is switched by the incremental signal obtained from the two tracks, the encoding of the absolute value code can be performed stably and accurately, and the absolute position can be detected with high accuracy.

Embodiment 5 FIG.
FIG. 9 is a block diagram showing a magnetic absolute encoder according to the fifth embodiment of the present invention. Compared with the fourth embodiment, the second embodiment is different in that the configuration of the magnetic sensor 15 that measures the magnetic field of the magnetic scale 14 that is the second track and the phase data calculation means 17 are further provided. The magnetic scale is composed of two tracks. The magnetic scale 11 as the first track is an absolute value code pattern, and the magnetic scale 14 as the second track is in the NS direction at a predetermined interval over the entire area. And a pattern magnetized in the SN direction.

  In the magnetic sensor 15, four MR elements are arranged at positions 7a, 7b, 8a, and 8b whose phases are different from each other by λ / 4. Thereby, for example, one output (7a, 7b side) can be a sine and the other output (8a, 8b side) can be a cosine. Based on the signals of the magnetic sensor 15 that is a pair of sine and cosine, the incremental detector 16 and the phase data calculator 17 can determine the phase angle of the sine wave formed by the magnetic scale 14. By using this phase angle information, the inside of the length λ can be further divided to obtain the position with high accuracy.

  Therefore, the absolute data obtained from the first track is used as the upper bit, the incremental data obtained from the second track is used as the lower bit, and the upper bit and the lower bit are synthesized, so that the absolute position detection with higher resolution can be performed. Can do.

  In Embodiments 4 and 5, two tracks, the first track and the second track, have been described. However, in order to further increase the resolution, a new track can be added to form a multi-track configuration. is there. The embodiments described so far are not limited to the rotary encoder that detects the rotation angle of the rotating body, but can also be applied to a linear encoder that detects the moving distance of the linear moving body.

1 is a configuration diagram of a magnetic absolute encoder according to a first embodiment of the present invention. It is explanatory drawing for demonstrating compared with Embodiment 1 of this invention. It is explanatory drawing of magnetic flux density distribution, MR element output, and encoding output by Embodiment 1 of this invention. It is a block diagram of the magnetic scale of the rotary magnetic absolute type encoder by Embodiment 1 of this invention. It is explanatory drawing of the magnetic flux density distribution by Embodiment 2 of this invention, MR element output, and an encoding output. It is explanatory drawing of the magnetic flux density distribution by Embodiment 3 of this invention, MR element output, and an encoding output. It is a block diagram of the magnetic absolute type encoder by Embodiment 4 of this invention. It is explanatory drawing regarding the switching method of MR element which reads the absolute value-ized code by Embodiment 4 of this invention. It is a block diagram of the magnetic absolute type encoder by Embodiment 5 of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 Magnetic scale, 12 Magnetic sensor, 13 Absolute detection means, 14 Magnetic scale, 15 Magnetic sensor, 16 Incremental detection means, 17 Phase data calculation means

Claims (7)

A magnetic absolute encoder that generates an absolute value code by combining a magnetized portion and a non-magnetized portion of a magnetic scale and outputs absolute data, wherein the magnetization direction of the magnetized portion is the magnetized portion. Is a magnetic absolute encoder characterized by having the same magnetization direction as the longitudinal direction of the magnetic scale . When continuous, a first track that generates an absolute value code by a combination of a magnetized portion and a non-magnetized portion having the same magnetization direction in the longitudinal direction of the magnetic scale, and the magnetic flux of the first track are detected. A magnetic absolute encoder comprising: a first magnetic sensor for detecting the magnetic flux; and an absolute detecting means for detecting the detected magnetic flux. 3. The magnetic absolute encoder according to claim 1, wherein the magnetization direction of the magnetized portion is reversed in the direction of the magnetization direction in the longitudinal direction of the magnetic scale before and after the non-magnetized portion. 4. The magnetic absolute encoder according to claim 1, wherein the length of the magnetized portion of the magnetized portion is shorter than a length corresponding to one code of the absolute value code. The magnetic absolute type according to any one of claims 1 to 3, wherein when the magnetized portions are continuous, the magnetized portion is set at a predetermined distance from the end in contact with the non-magnetized portion. Encoder. The first magnetic sensor is composed of two sets of magnetic sensors whose half lengths corresponding to one of the absolute value codes are shifted, and are attached in the NS direction and SN direction at predetermined intervals in the longitudinal direction of the magnetic scale. The magnetic absolute encoder according to claim 2, wherein the magnetic absolute encoder is switched by an incremental signal obtained from a magnetized second track. 7. The magnetic field according to claim 6, wherein the upper bits and the lower bits are synthesized with the absolute data obtained from the first track as the upper bits and the incremental data obtained from the second track as the lower bits. Type absolute encoder.

Images