JPH04130221A - Rotary encoder and apparatus using rotary encoder - Google Patents

Abstract

PURPOSE:To obtain a simple, low-cost rotary encoder by using a diffraction grating which can be manufactured by plastic molding as the configuration of a cylindrical grating and has the composite function of two functions for an amplitude-type diffraction grating and for wavefront splitting. CONSTITUTION:A light emitting means is composed of a semiconductor laser 1 and a collimator lens system 2 which converts the divergent luminous flux from the laser into the approximately parallel luminous flux. A rotary optical scale 3 having a cylindrical grating part is rotated in either direction shown by arrows. The scale 3 comprises a light transmitting material, and at least the grating part has the light transmitting property. Many V grooves are aligned along the entire surface in the circumferential direction at an equal interval, and the grating part is formed. Photodetectors 4a, 4b, and 4c which are the light receiving means are arranged at the positions facing the light emitting means with the scale 3 in-between. The outputs of the photodetectos 4a - 4c are connected to a signal processing circuit 6. The signal processing circuit 6 has a pulse counting circuit, a rotating-direction judging circuit, a signal interpolating circuit and the like.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a rotary encoder, and more particularly to an optical rotary encoder using a cylindrical optical scale.

[Prior Art] As an example of an encoder for measuring the amount of rotation of a cylindrical optical scale, there is a rotary encoder proposed by the applicant of the present invention in Japanese Patent Application Laid-open No. 1212/1983. This rotary encoder uses a 0-rotation grating, which is an excellent method that can measure the amount of rotation of a cylindrical optical scale with a slit-like grating as shown in Figure 23, with a simple configuration and relatively high resolution. By making it cylindrical, conventional - numerical 2
In addition to eliminating the need for relative positioning of the two gratings (rotating grating and fixed grating), it also has the effect of canceling detection errors caused by eccentricity of the rotating shaft, achieving high accuracy and ease of installation. This effect is achieved by installing an imaging optical system inside the scale (hollow part), and using this imaging optical system, the image of the grating in the first area on the side of the scale is placed on the opposite side of the scale from the first area with respect to the axis of rotation of the scale. This is achieved by projecting onto the grid of the second area on the side.

On the other hand, as another type of encoder using a similar cylindrical optical scale, the present applicant has filed Japanese Patent Application No. 1-33922.
In No. 1, we proposed a rotary encoder that applied the principle of so-called Talbot interference, which combines the Talbot effect of a grating and Moiré technology, instead of the above-mentioned imaging optical system. According to this, in addition to the effects of the prior art example described above, it is possible to further improve the simplification, miniaturization, and low inertia of the entire device configuration. The configuration and measurement principle of this encoder will be explained below using FIGS. 24 and 25.

In FIG. This parallel light beam is diffracted by the grating in the first region 31, and diffracted lights such as 0th order, ±1st order, ±2nd order are generated from the grating in the first region 31,
The Fourier image of the grating in the area 31 is projected onto the grating in the second area 32 of the scale 3 due to interference between two or three beams of the 0th-order light and the ±1st-order diffracted light. The pitch of brightness and darkness of this Fourier image is equal to the pitch P of the grating in the first region 31. Further, as described above, this Fourier image is curved, but this curvature occurs along the curved surface of the second region 32, and does not have a large effect on measurement accuracy.

Here, as shown in FIG. 25, scale 3 is indicated by arrow 100.
If the Fourier image is rotated in the direction (counterclockwise), the Fourier image moves in the direction of arrow 110 (clockwise).

At this time, the grid of the area 32 on which the Fourier image is projected is moving in the direction of the arrow 100. Therefore, when the scale 3 is rotated by an angle θ, the relative angle change between the Fourier image and the grating in the region 32 is 2θ, and the rotation angle can be measured with a resolution twice the grating pitch.

The 21 gratings in the second region 32 are illuminated by the Fourier image of the grating in the region 31, and the superimposition of the two produces moiré fringes, and the bright and dark light that has passed through the grating in the region 32 enters the light receiving surface 40 of the photodetector "4". Photodetector 4
converts the received light into an electrical signal, and the rotation angle of the scale 3 is measured based on this signal. In this rotary encoder, when the scale 3 rotates by an angle θ as described above, the Fourier image of the grating in the area 31 and the grating in the area 32 rotate by an angle of 200 relative to each other, so the total number of slits 3 in the scale 3 is If n, per revolution of scale 3,
2n sine wave pulses are output from the photoelectric conversion element 4. The rotation angle is measured by sequentially counting these sinusoidal pulses.

[Lesson 1 that the invention is trying to solve! ]however,
In the conventional rotary encoder mentioned above, cylindrical gratings such as amplitude type diffraction gratings and phase type diffraction gratings are used.
It is difficult to obtain these with high precision and good productivity.

Generally, the signal output of a rotary encoder is a two-phase output signal having a 90' phase difference in order to discriminate the rotational direction and perform interpolation for high resolution. Due to their structure, it is not possible to simply obtain a multi-phase output signal with a cylindrical diffraction grating or a phase-type diffraction grating.Therefore, some kind of ingenuity is required to obtain a multi-phase output signal with a cylindrical diffraction grating. .

[Object of the Invention] An object of the present invention is to provide a simpler and cheaper rotary encoder and a device using the same.

[Means for Solving the Problems] According to the present invention, as a form of a cylindrical grating used in an encoder, a novel diffraction grating having multiple functions, which is different from the conventional amplitude type diffraction grating or phase type diffraction grating, is used. The above problem is solved by using a grid. The complex function here mainly means the following two functions. (1
) function as an amplitude type diffraction grating, (2) wavefront splitting function (
Phase difference signal generation function) More specifically, the grating part of the cylindrical scale has translucency, and the grating part has an inclined surface with respect to the incident light along the circumferential direction of the inner surface of the scale. By using an optical scale formed by arranging concavities and convexities at equal intervals,
achieve one function.

[Example] Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 1(A) is a block diagram of an embodiment of the present invention. In the figure, 1 is a semiconductor laser with a wavelength λ (=780
generates a coherent light beam of 10 nm). 2 is semiconductor laser 1
The semiconductor laser 1 and the collimator lens system 2 constitute a light irradiation means. 3 is a rotating optical scale having a cylindrical grating section, which rotates in either direction shown by the arrow. ′! Figure J2 is a perspective view of this scale 3. The scale 3 is made of a translucent optical material and has several V-grooves lined up at equal intervals to form a lattice portion. Returning to FIG. 1(A), at positions facing the light irradiation means with the scale 3 in between, there are photodetectors 4a, 4b, 4c, which are light receiving means.
is located. The output of each photodetector is connected to a signal processing circuit 6. The signal processing circuit includes a pulse counting circuit, a rotation direction/direction discrimination circuit, a signal interpolation processing circuit, and the like.

It is used as an optical scale to detect the amount of rotation of the drive shaft. Figures ``t%4'' to ``f%8'' show several variations of how to attach the scale 3. In each case, the scale 3 is attached by direct fitting to the rotary drive shaft 5, and has a grating that serves as a scale for rotation detection. The coaxiality of the surface and the fitting part and the precision of the surface metal can be maintained at high precision. Especially's
In the form shown in Figure 8, the inside of the cylinder in which the lattice is formed is sealed.
Therefore, there is no adhesion of dust, oil, etc. to the grating section, and it has the advantage of very high dustproof ability.

Figure 3 is a detailed view of the lattice part of scale 3, and shows n V-grooves arranged at equal intervals on the inner surface of the cylinder, in which grooves and flat parts are alternately arranged to form a lattice. Pitch P (r
ad), and the V groove width is 34.
P (rad), and ■ The two planes forming the groove are each I
It has a width of P (rad), and each inclined surface is inclined at an angle of at least a critical angle, which in this example is θ; 45@, with respect to a straight line connecting the bottom and center of the V-groove.

The distance d (diameter inside the scale) between the grating in the first area 31 and the grating in the second area 32 of the scale 3 along the optical axis is d=N in this example, where the grating pitch is P and the wavelength is λ. -P"/λ (N=3) P=πd/n (n is the total number of slits). By setting the diameter d of the scale 3 in this way, the hollow part of the scale 3 The first region 3 on the side surface of the scale 3 is
The image of the first grating can be directly projected onto the grating of the second area 32.

The grating image projected here is called a Fourier image, and is generated by the self-imaging effect of the grating due to the optical diffraction phenomenon.Since the scale 3 in this embodiment has a cylindrical shape,
Although the Fourier image tends to be somewhat curved and the contrast decreases, the light irradiation means (
1, 2) and scale 3, there will be no practical problem.

(No.N) P2/λ<d<(N tens) P2/λ (N is a natural number) P=πd/n (n is the total number of slits) In this example, the material of scale 3 is plastic, It is suitable for mass production because it is produced by a manufacturing method such as injection molding or compression molding.

That is, it can be provided at an extremely low cost compared to a processing method using a conventional photolithography process.

In addition, in the encoder having the configuration of this embodiment, when there is a change in the external environmental temperature, the scale diameter d1, the grating pitch P1, the wavelength λ of the semiconductor laser change slightly, and the image formation position of the Fourier image and the grating plane change accordingly. There is a risk that relative positional deviation may occur and cause a decrease in the S/N of the detection signal6.
For example, when the temperature increases, the diameter d of the scale increases, the value of the pitch P of the grating increases accordingly, and the wavelength λ shifts to the longer wavelength side. At this time, the position of the Fourier image is L
From the equation -N-P2/λ, it changes at the rate of p2/λ. Therefore, by selecting the material of the scale and the characteristics of the semiconductor laser so that the amount of change in the diameter d of the scale (Δd) due to temperature change and the amount of movement of the Fourier image (ΔL) are as close as possible, it is possible to change the position of the lattice plane and the Fourier image. Relative positional deviation of the imaging position can be reduced, and even if external temperature changes occur, the S/N ratio of the detection signal is less likely to deteriorate. The semiconductor laser with a wavelength of 780 nm used in this example has a wavelength fluctuation of about 10 nm in response to a temperature change of 50°C, but it is preferable to use a material with a relatively large coefficient of thermal expansion as the material for the scale. In the example, the material of scale 3 is plastic (n-1, 49 acrylic resin), which has a higher coefficient of thermal expansion than glass etc., so the S/H of the output signal decreases less due to temperature fluctuations. Considering this advantage together with the advantage that it can be provided at low cost, it is a very suitable material for the scale of the encoder of this embodiment.

Next, the measurement principle of this embodiment will be explained using FIGS. 1(A), 1(B), and 1(C).

The light beam from the semiconductor laser 1 is converted into a converging destination by adjusting the position of the collimator lens system 2, and this convergent light beam is made incident on the first region 31 of the scale 3. The reason for using convergent light here is that the side surface of the scale 3 has a refractive power equivalent to that of a concave lens due to the difference in curvature between the outer and inner surfaces, and the light that enters the scale 3 becomes almost parallel light due to the concave lens action. Become.

This convergent light beam is transmitted to the grating portion of the first region as shown in Fig. 1 (B
), the light beam reaching the grating part 30a is 30a
Since the inclined surface is set at a critical angle or more, the light beam that passes through the surface and enters the cylinder, and reaches the grating section 30b-1 surface, is totally reflected as shown in the figure and reaches the grating section 30b-2. Since the light beam is directed toward the surface and is also totally reflected on the surface 30b-2, the light beam that reaches the surface 30b-1 is returned almost to the direction of incidence without entering the interior of the rotating body. Similarly, the light beam reaching the surface 30b-2 undergoes total internal reflection and is returned. Therefore, the light beam that reaches the range of the two inclined surfaces 30b-1 and 30b-2 forming the V groove in the first region 31 is reflected without entering the cylinder, and only the light beam that reaches the portion 30a is inside the cylinder. will proceed to. That is, in the first region 31, the V-groove diffraction grating has the same effect as a transmission type amplitude grating.

The light beam is diffracted by the grating portion of the first region 31, and 0th-order, ±1st-order, ±2nd-order, etc. diffracted light is generated due to the action of the grating, and two or As a result of the interference between the three light beams, a Fourier image of the grating in the first region 31 is formed inside the scale 3 . Fourier images are repeatedly formed at positions that are integral multiples of the distance behind the lattice plane.

In this embodiment, the light source wavelength λ
, the grating pitch P1 and the position of the collimator lens system 2 are set. The brightness pitch of this Fourier image is the first region 31
and the grating pitch P of the second region 32.

The light beam incident on the surface 30a in the second region 32 is
As shown in Figure (C), since the light rays are almost perpendicularly incident, they are transmitted in a straight line and reach the photodetector 4c. Furthermore, since the light rays that reach the two inclined surfaces 3°b- and 30b-2 forming the V-groove surface are incident on each surface at an incident angle of approximately 45°, they are largely refracted in different directions, and each It reaches detectors 4a and 4b. In this way, in the second region, the luminous flux is divided into three types by a total of three types of surfaces with different inclination directions: two inclined surfaces inclined in different directions with respect to the incident light beam, and a plane between the V grooves. 4a, 4b, which are separated in different directions and set up at positions corresponding to each surface.
It will reach each photodetector of 4C. That is, in the second region 32, the V-groove grating functions as a lightwave wavefront splitting element.

As described above, the scale of this embodiment has two functions: the first region 31 functions as an amplitude diffraction grating, and the second region functions as a wavefront splitting element for two-phase detection. It is characterized by having the following.

Here, changes in the amount of light detected by each of the photodetectors 4a, 4b, and 4c when the scale 3 rotates will be described below. Here, it is assumed that the scale 3 rotates counterclockwise.

′! FIG. J9 is a diagram showing how bright and dark grating images with a period equal to the grating pitch P are superimposed on the second region 32. In this case, the bright portion of the bright and dark grating image overlaps with the portion 30a, and the light beam passes through 30a and is concentrated on the photodetector 4C. 's10 diagram shows the state when the diffraction grating is rotated by 1/8P in the counterclockwise direction (100 direction) from the state of the previous 's9 diagram. In this case,
The bright and dark grid image moves in 110 directions. At this time, a portion of the light beam passes through the portion 30a, and the remaining light beam reaches 30b-2. Therefore, 1/2 of the luminous flux incident on the second region 32 will be incident on the photodetector 4c, and the remaining 1/2 of the luminous flux will be incident on the photodetector 4a.

As mentioned above, if the balance of the amount of light incident on each photodetector changes according to the relative displacement between the position of the grating and the position of the Fourier image, and as a result, the scale 3 rotates counterclockwise, as shown in Fig. 11 ( As shown in A), a change in the amount of light due to the rotation of the grating is obtained. Here, the horizontal axis is the amount of rotation of the cylindrical grating, and the vertical axis is the amount of received light. Signals a, b, and c correspond to photodetectors 4a, 4b, and 4c, respectively.

Conversely, if scale 3 rotates clockwise, a
is the output of 4b, b is the output of 4a, and c is the output of 4C. This difference allows the direction of rotation to be determined. In addition, the 11th
Figure (A) shows the theoretical light intensity change when the contrast of the Fourier image is very high and close to ideal. ), the amount of light changes approximately sinusoidally.

Next, electrical processing within the signal processing circuit 6 based on these signals will be specifically explained.

FIG. 13 shows an example of a signal processing circuit. The lattice is P(ra
d) When rotated, the output waveform is a two-cycle sine wave,
Particularly in this case, since the phase relationship between a and b is 90°, only the output signals a and b are used, and they are passed through a comparator circuit and converted into a rectangular wave as shown in FIG. 12 (A).
Further, by obtaining pulse signals at the rising and falling portions of each rectangular wave, it is possible to obtain 8 pulses at a rotation angle of P (rad) as shown in FIG. 12(B). Therefore, if the number of gratings in one rotation is n, rotation angle signals of 8 n P / R can be detected. In this case, since the signal C is not necessary, the device configuration may be the same as that shown in FIG. 1A except that the photodetector 4c is removed.

Now, in the above method, only signals a and b with a phase difference of 90'' are used, and signal C is basically unnecessary, but the accuracy can be further improved by using signal C. Below, The method will be explained below.

■Ideally, the groove width is exactly HP, but in reality it may deviate slightly from the ideal value due to processing accuracy, etc. Then, the output signal is not the ideal value, and the signals a and b
The phase difference will not be exactly 90'' and the accuracy of the finally obtained pulse will deteriorate. Figure 's14 is a diagram explaining the output waveform in this case.

Now, suppose that the width of the V-groove in the cylindrical lattice used is machined to be wider than Hp over the entire 360° circumference, then signal a,
The phase difference between B and B is slightly larger than 90'', resulting in a relationship as shown in FIG. 14(A).

Therefore, a circuit configuration as shown in FIG. 15 is adopted, and the output signal of the photodetector 4c is used to obtain the differential output of the signal a, the C1 signal, and the C1 signal, respectively, and a new C1 signal as shown in FIG. 14(B) is obtained. , C2. By appropriately adjusting the amplitude gains of the c, b, and a signals in this circuit, the phase difference between the composite signals C1 and C2 can be set to exactly 90#. These two signals C1 and C2 can be used to determine the rotational direction and perform interpolation processing. By utilizing the output of the photodetector 4c in this manner, it is possible to compensate for the manufacturing precision of the grating and perform highly accurate rotation detection.

Next, we will show another embodiment that can achieve even higher precision. In this embodiment, the grating has a groove-shaped V shape as shown in FIG.
The groove width is set to 2/3P of the grating pitch P. Figure 17 (A)
are the theoretical waveforms of the output signals a, b, and c of the photodetectors 4a, 4b, and 4c at this time, and FIG. 17(B) shows the actually obtained waveforms.In this case, In this embodiment, three-phase output signals having a 120° phase difference with uniform amplitude levels are obtained.High resolution is achieved using these three-phase output signals.

Specifically, using the signal processing circuit shown in Figure 19, the three-phase signals are converted into rectangular waves as shown in Figure 18 (A) through comparators, and then pulsed as shown in Figure 18 (B). . Thus, when the cylindrical grating is rotated by P (rad), 12 pulses are obtained. Therefore, if the number of gratings is n, a rotation angle signal of 12nP/R can be obtained. first two
In contrast to the 4-fold pulse in the form using phase signals, the present form using 3-phase signals provides 6-fold pulses with even higher precision.
Double pulsing becomes possible.

In the embodiments described so far, the V-grooves are arranged at regular intervals to form an uneven lattice part, but as a modification, the second
As shown in FIG. 0 (A), the unevenness may be formed as an alternating array of V-shaped chevrons and flat parts. Or Figure 20 (B)
It is also possible to have an uneven shape as shown in FIG. Furthermore, each uneven grid may be formed by a curved surface. Moreover, the shape of the unevenness does not necessarily have to be symmetrical, and may be asymmetrical. In any case, any scale can be used as long as it has a grating section formed by arranging concavities and convexities having sloped surfaces with respect to the incident light beam at equal intervals, and has the function of an amplitude type diffraction grating and the function of wavefront division.

In addition, in the previous embodiments, multi-phase signals were obtained using a plurality of photodetectors, but if a single phase is sufficient, the first
In Figure (A), it is sufficient to provide only one photodetector 4c.

Furthermore, although the above embodiments are examples of rotary encoders that use the principle of so-called Talbot interference, the present invention is not limited to this. The above optical scale may be used in an encoder having the following configuration.

In this case as well, the effects of lower cost, smaller size, and higher precision can be obtained.

Further, the light source that is not used in the present invention is not limited to a semiconductor laser, but may be a point light source LED, for example.

By using LEDs, which are cheaper than semiconductor lasers, it is possible to further reduce costs.

Now, FIG. 21 shows an example of the use of the above-mentioned encoder, and is a block diagram of a motor encoder in which the encoder is attached to the rotational output section of the motor and integrated. 21(A) is a top view, FIG. 21(B) is a side view showing the state of assembly, and FIG. 21(C) is a side view of the completed motor encoder. First, the scale 3 is attached to the rotating shaft 5 on the rear end side of the motor 300 in one of the forms shown in FIGS. 4 to 8 above. On the other hand, semiconductor laser 1, collimator lens 2
, the lens holder 2', and the photodetector 4 are covered with a detection head unit 120 that integrates each member, and the fitting part Z of this detection head unit 120 and the fitting part Z' of the rear end of the case of the motor 300 are fitted. Figure 21 (C
) to integrate.

As described above, since the rotation scale is directly attached to the rotation output part of the motor and the detection head unit for reading it is integrated with the motor case, a motor encoder with a simple configuration and easy alignment can be achieved. Furthermore, by making the rotary scale and the detection head unit separate structures, it is possible to simplify assembly.

Further, FIG. 22 shows an example of a system using the above encoder, and is a system configuration diagram of a drive system having a rotary encoder.

The above-described encoder is connected to the rotation output section of the drive means having a drive source such as a motor, an actuator, or an internal combustion engine, and detects the drive state such as the amount of rotation and the rotation speed. The detection output of this encoder is fed back to the control means, and the control means transmits a drive signal to the drive means so as to achieve the state set by the setting means. By configuring such a feedback system, the rotational state set by the setting means can be obtained. Such a drive system can be widely applied to various machine tools, manufacturing machines, measuring instruments, robots, cameras, audio equipment, information equipment, and not only these, but also to all devices having drive means.

[Effects of the Invention] As explained above, the cylindrical optical scale used in the present invention can be manufactured, for example, by plastic molding, and can be provided at a very low cost. By using a shaped scale, it is possible to achieve miniaturization and low inertia, and furthermore, it becomes easier to assemble.

[Brief explanation of the drawing]

Figure 1J1 is a configuration diagram of an embodiment of the present invention, Figure 2 is a diagram of an optical scale according to an embodiment, Figure 3 is a detailed diagram of the grating portion of the scale, and Figures 4 to 8 show how to attach a rotating grating. Diagram showing some variations,! Figures 9 and 10 are diagrams explaining the principle of the embodiment, Figure 11 is a waveform diagram of the output signal of the encoder of the embodiment, Figure 12 is a diagram explaining signal processing, and Figure 413 shows an example of the circuit configuration. Figures, Figures 14 to Y
Figure S19 is a diagram explaining a modification of signal processing, Figure 20 is a diagram of a modification of the grid shape, Figure 21 is a configuration diagram of a motor encoder, Figure 22 is a system configuration diagram of a drive system using an encoder, Figure 23 is a diagram of an optical scale used in a conventional encoder, Figure 24 and Figure 25 are explanatory diagrams of a conventional encoder, and the main symbols in the figure are: 1... Semiconductor laser; 2...Collimator lens system, 3...Optical scale, 4a, 4b, 4c...Photodetector, 5...
・Rotation axis, 31...first area, 2...second area

Claims (9)

[Claims] (1) A light irradiating means, a cylindrical optical scale, and a light receiving means are provided, and a grating is formed on the side surface of the cylindrical optical scale, and the light from the light irradiating means is irradiated onto a first area on the side surface of the scale. In a rotary encoder that detects the rotational state of a scale by directing light that has passed through a grating in a first area to a second area different from the first area, and receiving the light that has passed through the grating in the second area with a light receiving means. The cylindrical scale has a translucent grating portion, and the grating portion is formed by arranging concave and convex portions at equal intervals along the circumferential direction of the inner surface of the scale, each having an inclined surface with respect to the incident light beam. A rotary encoder characterized by: (2) The rotary encoder according to claim 1, wherein the light receiving means has at least two light receiving elements, each of which obtains a signal having a different phase. (3) The rotary encoder according to claim (2), further comprising means for determining the rotation direction based on the signals having different phases and/or means for performing signal interpolation processing. (4) comprising a motor, a cylindrical optical scale attached to the rotational output part of the motor and forming a grating along the rotational direction, a light irradiation means and a light reception means, and configured to receive light from the light irradiation means. By irradiating a first area on the side surface of the scale, directing the light through a grating in the first area to a second area different from the first area, and receiving the light through the grating in the second area with a light receiving means, a detection head that detects the rotational state of the motor; the cylindrical scale has a grating portion that is translucent; A motor encoder characterized in that unevenness having an inclined surface is formed by arranging them at equal intervals. (5) The motor encoder according to claim (4), wherein the rotation output section of the motor and the optical scale are connected by fitting. (6) The motor encoder according to claim (4), wherein the motor and the detection head are coupled by fitting. (7) A driving means having a driving source, a setting means for setting the driving state of the driving source, and a cylindrical optical scale attached to the drive output part of the driving means and forming a grating along the rotation direction. and a light irradiation means and a light reception means, the light irradiation means irradiates the first region on the side surface of the scale, and the light passing through the grating of the first region is directed to a second region different from the first region. , a detection means for detecting the driving state by receiving light through the grating in the second area with the light receiving means, and comparing the output of the detection means with the setting state set by the setting means, and determining the set state by comparing the output of the detection means with the setting state set by the setting means. a control means for controlling the driving source so as to be in a driving state, the cylindrical scale has a lattice part that is translucent, and the lattice part is arranged along the circumferential direction of the inner surface of the scale. What is claimed is: 1. A drive system having a rotary encoder, characterized in that concavities and convexities each having an inclined surface relative to an incident light beam are arranged at equal intervals. (8) The lattice part of the cylindrical scale has translucency, and the lattice part is formed by arranging concavities and convexities at equal intervals along the circumferential direction of the inner surface of the scale, each having an inclined surface with respect to the incident light beam. An optical scale for rotary encoders that is characterized by: (9) The optical scale for a rotary encoder according to claim (8), wherein the material of the optical scale is plastic.