JPS63245018A - Detecting method for absolute position - Google Patents

Abstract

PURPOSE:To decrease the quantity of information at the side to be detected by devising the pattern of a track of the side to be detected in a way that the pattern is deviated one by bit each in a prescribed direction and each value of a bit string comprising a prescribed number of bits in a predetermined relation of position is made different. CONSTITUTION:A pattern (a) is formed to a track of the side to be detected in a way that the pattern is deviated one by one bit each in a prescribed direction and each value of a bit string comprising the prescribed number of bits in a predetermined relation of position is made different. On the other hand, the absolute position of the track to each value of the bit string is stored in advance and every time the value of the bit string is detected by a detection section (b), the absolute position corresponding to the value of the bit string detected from the absolute position of the track stored in advance is obtained. Thus, the absolute position is detected by having only to provide one track to the side to be detected independently of the bit number to detect the absolute position. Thus, the quantity of information at the side to be detected is decreased.

Description

[Detailed description of the invention] (Industrial application field) The present invention relates to a method for detecting the absolute position of a rotating body or a moving body.

[Conventional technology]

In conventional absolute position detection rotary encoders and linear encoders, the absolute position is detected by providing the same number of tracks on the detected side as the number of bits for detecting the absolute position. As a specific example, when detecting the absolute positions of 16 locations on a rotating body,
In the 4-track detection side pattern a shown in FIG. 9 (1), the filled areas are set as "1" and the unfilled areas are set as 0", and the 4-bit detection part b
When the signals of four tracks are detected with l, the inner track is the upper digit and the outer track is the lower digit, and the 16 positions are B CD (Binary Coded De
cimal) detected at O-F. Figure 9 (2) shows the case where a delay code is used in the same way as Figure 9 (1) (
This is an example of pattern a2 on the detected side and detection part bt). Figure 9 (3) is a pattern a3 in which binary numbers are arranged on one track.
This is a method of detecting 16 positions using the detection unit b3. This is done by providing one slit for each absolute position to determine the detected position on the inner track of the track that detects the absolute position, and when the slit matches the slit on the fixed slit side, the slit group is used to detect the absolute position. This is to read pattern a3. For example, if only the rightmost slit of the thrift group is filled in on the absolute position detection track, the absolute position is 0001"=1
becomes. This method requires that the accuracy of the detection pattern be higher than the accuracy of the detection position.

(Problems to be Solved by the Invention) The conventional absolute position detection methods described above all detect the absolute position from the signals of multiple tracks, which requires a large amount of information on the detected side, and requires miniaturization of the device. The disadvantage is that it is difficult to

[Means for solving problems]

The absolute position detection method of the present invention forms a pattern on a track on the detected side so that six values of a bit string consisting of a predetermined number of bits in a certain fixed positional relationship are shifted one bit at a time in a predetermined direction. On the other hand, the absolute position of the track with respect to the six values of the bit string is stored in advance, and each time the detection unit detects the value of the bit string, the value of the bit string detected from among the pre-stored absolute positions of the track is stored. This is to find the corresponding absolute position.

(Function) Therefore, regardless of the number of bits used to detect the absolute position, the absolute position can be detected by simply providing one track on the detection side, simplifying the pattern on the detection side, and achieving accuracy of the corresponding track. is the same as the accuracy of the detected absolute position. Furthermore, in the case of an absolute value encoder, the number of tracks is only one, so the disk can be made smaller and a compact absolute value encoder can be realized.

〔Example〕

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 shows an embodiment of the absolute position detection method of the present invention, showing a detected side pattern and a detection unit, FIG. 2 is an expanded view of the detected side pattern of FIG. 1, and FIG. FIG. 2 is a diagram showing values when the detected side pattern of FIG. 1 is read from the national origin 4 bits at a time.

In this embodiment, the absolute positions of 16 locations on a rotating body are detected. In pattern a on the detected side of one track shown in Fig. 1, the filled area is 1'', the unfilled area is ``O'', and the 4-bit detection section has already set the left bit to the upper digit. When detecting the 4-bit signal of pattern a with the right bit as the lower digit, 16
The position of the spot can be detected from 0 to F. pattern a as 0 and 1
If you replace it with the sequence of numbers, it will become as shown in Figure 2, and if you read it 4 bits at a time while shifting 1 bit from the national yuan in this sequence, you will get 0137FEC925B6DA48 as shown in Figure 3, which is the number shown in Figure 1. The numbers are shown around pattern a. Since the numbers 0 to F are read once each time, the absolute positions of 16 locations on the rotating body can be distinguished. Absolute positions corresponding to these values of 0 to F can be stored in a ROM or the like and converted into signals such as binary and delay codes.

4 to 8 are diagrams showing other embodiments of the present invention.
In each numerical sequence in the figure, the underlined part indicates a sequence of detected bits, and the number indicated by a leader line below the underline is the value of that detected bit sequence. Also, the numbers in parentheses are repeated parts.

Figure 4 (1), (2), (4), (5)
, (6), and (7) each have 21
, 22 , 23 , 28 , and 211 bits, and the detected bit strings are consecutive 1, 2, 3, and 211 bits, respectively.
The case of 5.4 bits is shown.

Figure 5 (1), (2), (3), (4)
, (5), (6),..., (2B) are 3, 5, 6, 7, 9, 10, respectively. ..., in the case of a number sequence consisting of 31 bits, the detected bit string is 1 bit for (1) in the figure, 3 consecutive bits for (2) to (4) in the figure, and 3 consecutive bits for (5) to (12) in the figure. ) for consecutive 4
For bits (13) to (26) in the same figure, consecutive 5
This is the case for bits.

Figure 6 (1), (2),...-, (8) High deviation is also a number sequence consisting of 2s bits, and when the detected bit string is 5 consecutive bits, the order of the numbers in the number sequence is different. There is.

Figure 7 (1), (2), (3), (su, (
5) is a case where both number sequences are made up of 24 bits, and the number of bits in the detected bit strings is the same but not consecutive. In other words, in the case of (1) in the same figure, the detected bit string is! Every bit, (2) in the same figure is every 2 bits, (3) in the same figure is every 4 bits.
Every bit (4) in the same figure is every 1, 3, or 5 bits, and (5) in the same figure is every 1, 6, or 1 bit. In these cases as well, different numbers can be obtained by shifting one bit at a time.

FIGS. 8(1) and (2) show the case of a number sequence consisting of two 24-bit sequences. These represent the absolute position of 24 bits by 2
This is an example of detection using detection bits in rows and two columns.

For a sequence of 2n-1 bits, a linear $i-distant shift register with maximum sequence length (reference, NIK
KEI ELECRTONIC5BOO can be used as is, S "Electronic Circuit Design Idea Collection", JP-A-58-105619), and by adding one O to a pattern of n-1 consecutive 0s, it is easy to create a random pattern of 2n bits. You can get a sequence of numbers. In this maximum sequence length shift register with linear feedback, all possible permutations other than all 0 words are accommodated. In other words, in the number sequence produced by an n-bit shift register, there is always one place where n-1 consecutive 0s occur, so if one O is added there, a word of all O's is created and a 2n number sequence is obtained. 24t-1 as a shift register with linear feedback of maximum sequence length requiring only 2 feedback taps
2,199,023,255,62
0=241 absolute position detections are also possible with this method.

If we dare to define a random number sequence here, we can say ``m O
If you line up n-m 1's and circulate them, and sequentially extract l Iogt (n-1) l + 1 numbers in a certain fixed positional relationship, you will get a number sequence in which each type of number appears exactly once. However, n and m are natural numbers, 02m.'' Also, sequences with bit numbers other than 2'' can be found by analogy with the 2n-bit sequence as shown in Figure 5. They can also be found using a computer. As a method of ``analogy,'' we obtain shorter sequences by removing numbers from longer sequences. For example, if the number of detected bits is k
In the thing of.

In a number sequence, the same part of the pattern of -1 numbers is 2
Find the spot and remove numbers other than the pattern so that the pattern appears only once. For example, to find the sequence in Figure 5 (24) from the sequence in Figure 5 (26), in the sequence in Figure 5 (26), ...0110101001... Bu ...01
101001 .

When calculating the sequence using a calculator, the number of detection bits (
For n series, l log2 (n-1) l + 1
), add 0 or 1 sequentially from the number sequence, and if the same number occurs twice, return to the length where 0 or 1 can be replaced, and continue adding 0 or 1, until finally n pieces are added. Obtain a cyclic sequence of numbers.

In addition, [an example of a number sequence consisting of 2s bits] shown in Fig. 6
I don't know the rules by which it appears. This is simply a result of arranging numbers. Regarding 2 n,
Since there is the following rule for 20 to 27, it seems to apply even when n is 8 or more.

・Law regarding sequences of length 2n The number of consecutive O or 1 is determined, and the following table is obtained.

For example, in Figure 8 (1), becomes.

Note that the present invention can be used not only for detecting the position of a rotating body but also for detecting a position on a straight line. In addition, it is also possible to use this as an address when processing data other than detecting the absolute position of an object, to make one bit of this number sequence correspond to one piece of data, and to extract desired data using this number sequence. This allows maximum information to be stored within limited storage capacity.

〔Effect of the invention〕

As explained above, the present invention shifts the track pattern on the detected side in a predetermined direction one bit at a time so that six values of a bit string consisting of a predetermined number of bits in a certain fixed positional relationship are different. By doing so, it is possible to significantly reduce the amount of information on the detected side compared to the conventional method, and 2"
For the position detection of , there is an effect that the position can be detected with the conventional information amount of 17n.

[Brief explanation of drawings]

FIG. 1 shows an embodiment of the absolute position detection method of the present invention, showing a detected side pattern and a detection unit, FIG. 2 is an expanded view of the detected side pattern of FIG. 1, and FIG. A diagram showing the values when the detected side pattern in Figure 1 is read 4 bits at a time from 1 left. Figure 4 (1) to (7) are 21 to 2, respectively.
Diagram showing the case of a sequence consisting of 7 bits, Figure 5 <1
) to (26) are diagrams each showing the case of a number sequence consisting of 23 to 231 bits, and Figure 6 (1) to (8) are all 2
A diagram showing the case of a number sequence consisting of 5 bits, Figure 7 (4
) to (5) are diagrams showing the case of a number sequence consisting of 24 bits, Figure 8 (1) and (2) are diagrams showing the case of a number sequence consisting of two columns, and Figure 9 (ri to ( 3) is a diagram showing a conventional method. a--pattern of detected part, b- detecting part.

Claims (1)

[Claims] forming a pattern on a track on the detection side so that each value of a bit string consisting of a predetermined number of bits in a predetermined positional relationship shifted by one bit in a predetermined direction is different;
On the other hand, the absolute position of the track for each value of the bit string is stored in advance, and each time the detection unit detects the value of the bit string, the value of the bit string corresponding to the value of the bit string detected from among the absolute positions of the track stored in advance is detected. Absolute position detection method for determining absolute position.