JPS636415A - Method for determining slit number of absolute encoder - Google Patents

Abstract

PURPOSE:To obtain the best number of slits which make an error permitted for phase detection larger when obtaining target accuracy by setting the number of slits to an integral value close to N<1/2>/2 for the target accuracy + or -1/N. CONSTITUTION:The light and dark of light corresponding to the positions of the slits 21 and 22 are converted into phase signals Va and Vb by scanning a photodiode array 4. At this time, the phase signal Va changes in phase between '0' and 2pi (n) times in every turn of a code plate 2 and the phase difference phiab between the phase signals Va and Vb changes once. For the purpose, the section of the phase signal Va is found from the phase difference phiab and the section is interpolated with the phase signal Va to find an angle theta of rotation. In this case, an error in the measurement of the phase of the signal Va needs to be <2pi/4n so as to find the section accurately and <2pin/N so as to obtain the target accuracy + or -1/N, so the best number of slits is N<1/2>/2.

Description

Detailed Description of the Invention [Field of Industrial Application] The present invention provides two sets of slit rows with different numbers in the circumferential direction of a code plate, and detects the movement of the slit rows using a detection means such as a photodiode array. In an absolute encoder that detects the absolute rotation angle from the phase difference between two output signals and the phase of either output signal, the slit determines the optimal number of slits to obtain high detection accuracy. It is related to the number determination method.

[Conventional technology]

Generally, when designing an encoder, the number of slits provided in the code plate is a major factor that affects the rotation angle detection accuracy (resolution). For example, in an incremental encoder, the greater the number of slits, the higher the resolution and the higher the detection accuracy. .

[Problem that the invention seeks to solve]

However, in the above-mentioned absolute type encoder, the number of slits must be determined not only in consideration of the target accuracy but also in consideration of the accuracy in reading the phase of the output signal. That is, if the number of slits is increased, the final resolution will be increased, but the phase difference signal formed from the two output signals will be required to be read with high precision, making signal processing difficult.

For this reason, conventionally, no clear method for determining the number of slits has been established, and the number of slits considered to be appropriate has been sought through repeated trial production.

The present invention eliminates the drawbacks of the conventional methods as described above, and provides a method for determining the number of slits that easily determines the optimal number of slits that can increase the error allowed for phase detection in order to obtain the target accuracy. It is intended to provide.

[Means for solving problems]

The method for determining the number of slits in an absolute encoder of the present invention is to provide two sets of slit rows with different numbers in the circumferential direction of a code plate, detect the movement of the slit rows with a detection means such as a photodiode array, and generate two output signals. In an absolute encoder that calculates the absolute rotation angle from the phase difference of The number of slits is set to an integer value close to the optimal number of slits, n.

[For production]

In this way, by expressing the optimum number of slits using a mathematical formula, the number of slits can be easily determined according to the target accuracy. In addition, the number of slits determined by this formula is the number of slits that can most alleviate the tolerance of the detection means, so it is easy to implement.

〔Example〕

A method for determining the number of slits in an absolute encoder according to the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of an absolute encoder to which the slit number determining method of the present invention is applied. In the figure, 1 is a rotating shaft, 2 is a code plate fixed to the rotating shaft 1, 2. , 2. are two slit rows provided on the code plate 2, 3 is a light source that irradiates the code i72, and 4 is the slit row 2. , 2. This is a photodiode array that receives the light (slit image) that has passed through the slit. Here, slit row 2. , 2. are 1 round nm and (TI
-1) slits are provided.

Figure 2 shows slit row 2. , 2. and an enlarged view of 4 portions of the photodiode array.

In the absolute encoder configured in this way, the brightness and darkness of the light depending on the position of the slit are converted into alternating current phase signals Va and Wb, respectively, by electrically scanning the photodiode array 4 at an angular frequency ω. Ru. Further, the phase of the output signals Va, l/b of the photodiode array 4 changes by 2π when the code plate 2 rotates by one pitch of the slit. Therefore, when the code plate 2 rotates once, the output signals Va, Vb
The phases ma and cbb of (z fx yl ) are respectively
and changes by (2wX (Tl-1)). Therefore,
The absolute value θ of the rotation angle can be determined from the two phase differences (d) a −dab ). Expressing the above relationship in a formula, Va= 5in(ωt + d)a)
(1) Vb=sin(ωt + d
rb) (2) Also, dr
awnθ, mb −(n-1)θ, so Va,
The phase difference jab of Wb is d+abs+wd+a −d
sb mθ (3).

FIG. 3 shows a method of calculating the rotation angle θ using the output signals Va and Wb. 1 of the code plate 2, as shown in the figure.
Per rotation, the phase 6a of the output signal Va changes n times between 0 and 2t, and the phase difference jab between the output signals Va and Wb
changes once. Therefore, the section i of the output signal Va is determined by the phase difference crab, and the section i is defined as the output signal Va.
If measured by interpolation using (small a), θ= 2 v i
/n+(6a/TI(4)), and the rotation angle θ can be measured with the same precision as d+a.

Now, in the above, in order to find the interval j correctly,
The error tab that occurs when measuring dumb must satisfy the following conditions.

tabmaa+tb-2ea<2f/2n (5);
, s a < 2 v / 4 n
(6) Here, 6a is the measurement error of d+a (-gb
) That is, since d+ab is the phase difference between d)a and 6b, the error zab is the sum of their respective errors, and this must not cause a half interval of 6a.

Also, since the error of the measuring mirror θ in equation (4) is aa/r+, in order to achieve the target accuracy ±1/N, 6a
The error εa must satisfy the following equation.

5 a/n < 2 da/N ■,
′, εa <2qη/ N (8
) In this way, when the target accuracy is ±1/N, the number of slits of 1 also affects the tolerance 6a of the detection means, so the number of slits of 1 realizes the tolerance 5a required for the detection means. It must be determined to be a possible value.

FIG. 4 shows the tolerance 6a0) region shown in equations (6) and (8), and the shaded area in the figure is the region where both equations are satisfied. As is clear from the figure, the intersection point (n,) of the two graphs.

εas) is the point at which the allowable error 6a is maximized, and selecting the number of slits at this point is a preferable choice from the standpoint of device design.

Therefore, if we calculate the number of slits n from FIG. 4, we get 2vn, /N-2f/4·n.

, “, n, = J”W/2
(9) becomes. Note that since n, which is determined by this formula, is generally not an integer, an integer value around n is adopted as the number of slits. Further, in this case, the value of the tolerance 5a becomes smaller as the number of slits is changed, but there is no problem as long as the tolerance εa after the change is a realizable value.

Here, an actual design example is shown. For example, set the target accuracy to ±1
/4000, from equation (9), n is n, = 3
1.622... Therefore, 32 may be used as the number of slits, but in this case, considering the ease of manufacturing the code plate, it is preferable to use n-36, in which the pitch of the slit rows is 10 degrees. Further, the tolerance εa required for the detection means at this time is ±5°, which is a fully achievable value.

In this way, when the number n of slits is determined using equation (9), the goal of determining the number of slits is clear, and the number of slits can be easily determined according to the desired resolution. Further, the number n of slits determined by this equation is the number of slits that can most alleviate the tolerance 6a of the detection means, so that the detection means etc. can be easily realized.

In addition, in the above explanation, the case where the optimal number of slits that can maximize the tolerance of the detection means from the target resolution is determined, but the approximate accuracy [tolerance] of the detection means is given in advance. If you are
Based on this, the number of slits that can obtain the highest resolution can also be determined.

〔Effect of the invention〕

As explained above, in the method for determining the number of slits in an absolute encoder of the present invention, two sets of slit rows with different numbers are provided in the circumferential direction of the code plate, and the movement of the slit rows is detected by a detection means such as a photodiode array. In an absolute encoder that detects and calculates the absolute rotation angle from the phase difference between two output signals and the phase of one of the output signals, if the target accuracy is ±1/N, then n, = J The optimal number of slits, n, is determined from the formula Fr/2, and the number of slits is set to an integer value close to this optimal number of slits, n, so the goal of determining the number of slits is clear and the target accuracy can be achieved. It is possible to provide a method for determining the number of slits that easily determines the optimum number of slits that can increase the tolerance required for the detection means.

[Brief explanation of drawings]

1 to 3 are block diagrams showing one embodiment of an absolute encoder to which the method for determining the number of slits in an absolute encoder of the present invention is applied, and diagrams for explaining its operation;
The figure is an explanatory diagram showing the concept of a method for determining the number of slits in an absolute encoder according to the present invention. 1... Rotating shaft, 2... Code plate, 2. , 2.・・・
- Slit row, 3... light source, 4... photodiode play (detection means). Figure 1 Figure 2

Claims (1)

[Claims] Two sets of slit rows with different numbers are provided in the circumferential direction of the code plate, and the movement of the slit rows is detected by a detection means such as a photodiode array, and the phase difference between the two output signals is detected. In an absolute encoder that calculates the absolute rotation angle from the phase of one output signal, if the target accuracy is ±1/N, then the optimal number of slits n_■ can be calculated from the formula n_■=√N/2. A method for determining the number of slits in an absolute encoder, in which the number of slits is determined to be an integer value close to the optimum number of slits n_■.