JP2005526951A - Reference point Talbot encoder - Google Patents

Abstract

The disclosed optical encoder includes a scale and a sensor head. The scale includes optical elements. The sensor head includes a light source, a detector array, and an index detector, all of which are disposed on the substrate. The scale is disposed to face the sensor head and is disposed to move relative to the sensor head. The distance between the scale and the sensor head is selected to position the detector array near the Talbot imaging plane. The light source emits a divergent light beam that is directed to the scale. Light from the divergent light beam is diffracted by the grating and travels to the detector array. The light from the divergent light beam is diffracted by the optical element and travels to the index detector. The detector array provides a relative measurement of the position of the sensor head relative to the scale. The index detector provides a reference measurement of the position of the sensor head relative to the scale.

Description

[Reference to related applications]
This application is assigned to the assignee of the present invention and filed at the same time as the present application, co-pending US patent application Ser. No. 60 / 316,121, entitled “Harmonic Suppressed Photodetector Array”. Agent reference number MCE-018 (111390-140)]. This application is hereby incorporated by reference in its entirety.

  The present invention relates to an optical encoder. More specifically, the present invention relates to an improved reference point optical encoder.

  Diffractive optical encoders are well known in the field of position displacement detection systems. Such devices are commercially available from the assignee of the present invention and from several other vendors. US Pat. Nos. 5,559,600 and 5,646,730 describe examples of known optical encoders.

  A recent trend is to develop a diffractive encoder of small size. U.S. Pat. Nos. 5,559,229, 5,671,052 and 5,991,249 disclose examples of such small size encoders. In general, what characterizes such a small size encoder is a quasi-monochromatic (or nearly monochromatic) solid-state illumination source, a binary grating, one or more detection elements, and a reduced number of additional optical components. Is use.

  One problem with known small size encoders is that size reduction generally has a negative impact on their accuracy. Accordingly, there is a need for a small size diffractive optical encoder characterized by improved accuracy.

[Disclosure of the Invention]
These and other objectives are provided by an improved diffractive optical encoder. The encoder includes an index detector to provide a reference position measurement. This index detector can be implemented by using a three cell configuration. The present invention also provides an algorithm for processing the signal generated by the index detector. The present invention further provides other features for improving the accuracy of diffractive optical encoders.

  Still other objects and advantages of the present invention will be readily apparent to those skilled in the art from the following detailed description. Here, several embodiments are shown and described to merely illustrate the best mode of the invention. As will be realized hereinafter, the invention is capable of other and different embodiments, and its several details are capable of modifications in various respects, all without departing from the invention. Accordingly, the drawings and description are to be regarded as naturally presented, rather than as restrictive or restrictive, in scope of application as indicated in the claims.

  For a better understanding of the nature and objects of the present invention, reference should be made to the following detailed description taken in conjunction with the accompanying drawings. In the drawings, the same reference numerals are used to indicate the same or similar parts.

Detailed Description of Preferred Embodiments
FIG. 1 is a perspective view of a diffractive optical encoder 100 constructed in accordance with the present invention. As shown, the encoder 100 includes a photoelectric assembly or sensor head 110, a scale 160, and a signal processor 190 as three basic components.

  FIG. 2A is a side view of the encoder 100. FIG. 2B is a diagram of the sensor head 110 viewed in the direction of line 2B-2B shown in FIG. 2A. FIG. 2C is a view of the scale 160 as viewed in the direction of the line 2C-2C shown in FIG. 2A. 2D is an end view of encoder 100 as viewed in the direction of line 2D-2D shown in FIG. 2A. For convenience of explanation, the signal processor 190 is not shown in FIGS.

  Referring to FIGS. 1 and 2A-2D, the sensor head 110 includes a light source 112, a primary detector array 120, and an index (or reference point) detector 140. As shown, the light source 112 and the detectors 120 and 140 are all mounted on a common substrate 111. The primary detector array 120 and the index detector 140 are preferably mounted on a single piece of silicon. The scale 160 includes a substrate 161, on which a diffraction grating 162 and two diffractive optical elements (DOE) are mounted. Scale 160 and sensor head 110 are disposed generally opposite each other at a fixed distance d (as shown in FIG. 2D). As a result, the scale 160 and the sensor head 110 can move relative to each other in the direction of the arrow AA shown in FIG. 2A. In operation, the encoder 100 monitors the relative movement of the scale 160 (in the direction of arrow AA) relative to the sensor head 110 and generates a signal representing the relative position of the scale 160 relative to the sensor head 110.

  In operation, the light source 112 emits an expansible or divergent light cone 102. The light source 112 is preferably a quasi-monochromatic (or nearly monochromatic) light source that can be implemented using a vertical cavity surface emitting laser (VCSEL). As shown in FIG. 1, since the sensor head 110 and the scale 160 are preferably arranged, the light cone 102 is incident sufficiently wide on a part of the grating 162 and one of the DOEs 166 when reaching the scale 160. To do. Some of the light cone 102 propagates through the scale 160 and is diffracted by the scale 160. It is preferable that this light does not return to the sensor head 110. Some of the light in the cone 102 is also reflected, diffracted and returned to the sensor head 110. Since the sensor head 110 and the scale 160 are preferably configured, (1) light that is diffracted by the grating 162 and returned to the sensor head 110 is incident on the detector array 120 primarily, and (2) is diffracted by the DOE 166. The light returned to the sensor head 110 is incident on the index detector 140 temporarily. As described in more detail below, the light incident on the detector array 120 allows the encoder 100 to provide a relative measurement of the position of the sensor head 110 relative to the scale 160. On the other hand, the encoder 100 can provide an index point measurement value (that is, a reference point measurement value) of the position of the sensor head 110 with respect to the scale 160 by the light incident on the index detector 140.

  3A, 3B and 3C show the scale 160 in more detail. Specifically, FIGS. 3B and 3C are enlarged views of region 310 shown in FIG. 3A. The scale 160 is preferably formed on a glass-like substrate 161. As shown in FIG. 3B, the grating 162 includes light reflective stripes 164 and light transparent stripes 163 alternately. The light reflective stripe 164 is preferably formed by covering a region of the substrate 161 with a highly reflective material. In this embodiment, the light transmissive stripe 163 is formed simply by not covering the substrate 161. Instead of the transmissive stripe, a light-absorbing stripe can also be used. As shown in FIG. 3C, in another embodiment, the stripes can all be reflective and the stripes can be alternately arranged at different depths. A grating 162 of the type shown in FIG. 3B is known as an “amplitude grating”. A grating 162 of the type shown in FIG. 3C is known as a “phase grating”.

  Regardless of whether the grating 162 is implemented as shown in FIG. 3B or 3C, each stripe has its short side parallel to the displacement direction of the scale (ie, parallel to the arrow AA in FIG. 1). A thin, oriented rectangle is preferred. The center spacing of the stripes (ie, from the left edge to the left edge of the stripe spacing as shown in FIGS. 3B and 3C) defines the period P of the grating 162. The stripes are preferably equally spaced. In addition, the short side of each stripe is preferably approximately equal to ½ of the period P of the grating 162. Depending on the desired system performance, the period P is typically in the range of 5-40 microns, with 20 microns being a preferred value. Ideally, the scale has an anti-reflective coating on the exposed glass areas on both sides.

  Returning to FIG. 1, the grating 162 diffracts the light from the cone 102 into a plurality of light cones directed toward the sensor head 110. FIG. 4, which shows the encoder 100 in the same orientation as that shown in FIG. 2A, shows some of the cones of light 103 that are diffracted by the grating 162 toward the sensor head. The diffracted light cones 103 optically interfere with each other to produce a composite fringe-like pattern in the space between the scale 160 and the sensor head 110.

FIG. 5 shows the intensity of the fringe pattern formed by interference between the diffracted light cones 103 at different distances from the grating 162. As shown, the optical fringe pattern formed by interference between the light cones 103 at distances d ₂ and d ₄ from the grating 162 is a relatively high contrast periodic pattern. Conversely, the optical fringe pattern formed by interference between the light cones 103 at distances d ₁ and d ₃ from the grating 162 has a relatively low contrast. The planes at distances d ₂ and d ₄ from the grating 162 are called self-imaging planes, or “Talbot imaging planes”. In these Talbot imaging planes, the divergent cones of diffracted light are combined with the same relative phase that they have in the grating, essentially forming an image of the grating 162 itself. As generally discussed in US Pat. No. 5,991,249, these high contrast imaging planes occur regularly, and the distance between the grating and any imaging plane is given by ).

In equation (1), z ₀ is equal to the distance between the light source 112 and the grating 116, z ₁ is equal to the distance between the grating 162 and the Talbot self-image plane, N is an integer, and P is Is the period, and λ is the wavelength of the light emitted by the light source 112.

As shown in FIG. 5, the first Talbot plane (in the distance d ₂ from the scale) is deviated (at a distance d ₄ from the scale) 180 degrees out of phase with respect to the second Talbot plane. In general, adjacent Talbot planes are 180 degrees out of phase with each other. The reason for this 180 degree phase shift between adjacent Talbot planes is that in an even plane (ie a Talbot plane where N is an even number) all orders of diffracted light are combined with the same relative phase they have in the grating. In contrast, in the odd plane (ie, the Talbot plane where N is odd), the 0th order is 180 degrees out of phase and all other orders are combined with the same relative phase they have in the grating.

  It should be noted here that the pattern shown in FIG. 5 has a zero-order beam when the zero-order beam contributes to the pattern (for example, the fringe pattern is similar to other higher-order diffracted beams in the zero-order, +1 Next, when it is formed by interference between the -1st order), it is a feature of the stripe pattern. When the zero order beam is erased, the fringe pattern appears to be significantly different from that shown in FIG. In the case of a phase grating with a ½ wavelength delay, the low contrast plane is the Talbot imaging plane and the high contrast plane is between the Talbot imaging planes. In the high contrast region, the fringe pattern does not appear as an image of the original grating as in the amplitude grating. Rather, the fringe pattern of the phase grating is a complex combination of harmonic components that are typically dominated by components having a period that is typically one-half the period shown in the Talbot plane of FIG. As with the amplitude grating, the period of the fringe pattern from the phase grating increases in proportion to the distance from the scale. In general, it is difficult to predict a plane in which the fringe pattern from the phase grating exhibits minimum harmonic distortion and / or noise.

  Therefore, the cancellation of the zero order beam is considered to cause degradation of the periodic signal monitored by the encoder. However, it is still advantageous to construct an encoder in which the zero order beam is canceled for at least the following reasons. That is, as a practical matter, in encoders of the present design, high diffraction orders are quickly filtered by propagation, and the resulting fringe pattern is often close to a pure sinusoidal shape.

  Despite the advantages of the phase grating described above, the preferred grating for the present invention is an amplitude grating. The amplitude grating (shown in FIG. 3B) is more widely commercially available than the phase grating (shown in FIG. 3C). Therefore, the design of an encoder using an amplitude grating has the advantage that it is not very expensive and that it is easy to capture the scale. However, the use of an amplitude grating means the presence of a zero order beam. The design of an encoder in which a zero order beam is present is discussed below.

In the encoder constructed according to the present invention, since the sensor head 110 and the scale 160 are preferably arranged, the detector array 120 is located in one of the Talbot imaging planes (ie, calculated by the above equation (1)). The distance between the scale and the sensing surface of the detector array will be equal to z ₁ ). As is apparent from FIGS. 2A and 2D, in an encoder constructed in accordance with the present invention, the upper light emitting surface of light source 112 is preferably substantially coplanar with the upper (ie, sensing) surface of detector array 120. Therefore, in an encoder constructed in accordance with the present invention, distance z ₀ is approximately equal to distance z ₁ . When z ₀ is equal to z ₁ , the above equation (1) is simplified to the following equation (2).

Therefore, in order to ensure that the detector array 120 is disposed in one of the Talbot imaging planes, the encoder constructed in accordance with the present invention has a sensor head 110 and a scale (as shown in FIG. 2D). The distance d between 160 and 160 is preferably adjusted so that the separation distance between the scale 160 and the detector array 120 is approximately equal to z ₀ as calculated for integer values of N by equation (2). However, since it is almost impossible to ensure that the actual distance between the scale 160 and the detector array 120 is equal to z ₀ , this distance is preferably chosen so that the sensing surface of the detector array 120 is the Talbot plane. It should be located in a region close to one of the above. The desired size of this region is discussed below.

As shown in FIG. 5, the distance between the scale and the first Talbot plane is d _2. The distance between the scale and the nth Talbot plane is nd ₂ (that is, n times d ₂ ). If it is desired to place the detector array in the nth Talbot plane, the distance between the scale and the detector array is preferably equal to nd ₂ + or −0.5d ₂ . Thus, for example, if it is desired to place the detector array to a third Talbot plane, the detector array, between location and the scale 2.5d ₂ away from the scale of 3.5D ₂ away Should be placed in the area. Continuing with this example, when the distance between the scale and the detector array is equal to 3.0d _2, the detector array is located to ensure that the third Talbot plane. If this distance is slightly larger or smaller than 3.0d ₂ , the fringe pattern contrast will be slightly less than optimal and the encoder accuracy will be correspondingly slightly reduced. As the detector array is moved away from the desired position 3.0d ₂ , the fringe pattern contrast continues to decrease and eventually reaches a minimum at distances 2.5d ₂ or 3.5d ₂ (ie, the contrast is These values have a minimum because the Talbot planes are separated by equally spaced planes characterized by minimum contrast). Since the Talbot planes are separated by an equidistant plane with minimum contrast, nd ₂ + or −0.5d ₂ will indicate the maximum size of the region in which the detector array is to be placed. If the detector array is placed nd ₂ + or −0.5d ₂ away from the scale, the performance of the encoder increases. Encoder performance is further increased when the detector array is placed at a position nd ₂ + or −0.1d ₂ away from the scale. More generally, the detector array 120 is in an area limited by two planes: a first plane nd ₂ + xd ₂ away from the scale and a second plane nd ₂ −xd ₂ away from the scale. Preferably it is located. Here, x is 1/2 or less. A preferable value of x is 0.2, and a more preferable value of x is 0.1.

  As described above, when the zero-order beam is erased, a high-contrast fringe pattern enters the detector array regardless of the distance between the detector array and the scale. Therefore, it is advantageous to use the scale 160 with a phase grating (shown in FIG. 3C) that substantially cancels the zeroth order beam to relax the limitations discussed above for the distance between the detector array and the scale. . In such an embodiment, the distance between the upper stripe and the lower stripe (or the depth of the lower stripe) is preferably approximately equal to N times the quarter wavelength of the light generated by the light source 112. Here, N is an odd number. Another advantage of using such a phase grating is that it reduces the period of the optical fringe pattern by a factor of 2, thereby potentially increasing the resolution of the encoder by a factor of 2. Alternatively, if it is desired to manufacture the encoder using a phase grating in which a zeroth order beam is present, the distance between the upper and lower stripes is (N + x) × (light generated by the light source 112 Is preferably equal to ¼) of the wavelength. Here, N is an odd number, and x is a small number less than 1/2.

  As shown in FIG. 5, the interference fringes are periodic and are characterized by a period T. Since the grating 162 is illuminated by an expanding cone of light, the fringe period T is generally a function of the distance from the grating, as shown by equation (3) below.

In equation (3), z ₀ is the optical path length between the light source 112 and the scale 160, z ₁ is the optical path length between the scale and the detector array 120, P is the period of the grating, and e Is the offset between light source 112 and detector array 120 (ie, the difference between z ₀ and z ₁ ), and K is the scale factor.

As can be seen from equation (3), the special case (ie, e = 0) where the distance between the light source and the grating (z ₀ ) is equal to the distance between the detector array and the grating (z ₁ ). , The scale factor K is 2, so that the fringe period T is always equal to a constant value that is twice as large as the grating period P (ie T = 2P). As described above, it is preferable that the upper light emitting surface of the light source 112 is substantially coplanar with the detector array 120, so in the encoder constructed according to the present invention, the distance (z ₀ ) between the light source and the grating is It is approximately equal to the distance (z ₁ ) between the grating and the detector array. Therefore, in the encoder 100, the period T of the fringes incident on the detector array 120 is always substantially equal to the constant 2P.

  In operation, when the scale 160 moves relative to the sensor head 110 in the direction of arrow AA shown in FIG. 2A, the fringe pattern incident on the detector array 120 causes the detector array 120 to move in the direction of arrow AA. Move across. Movement of the incident fringe pattern across the detector array is equivalent to a change in phase angle between the incident fringe pattern and the detector array. The detector array 120 and its associated signal processor 190 monitor this phase angle, thereby monitoring the relative position of the sensor head 110 with respect to the scale 160.

  The detector array is preferably constructed as an array of a plurality of photodetectors configured to facilitate measurement of the phase angle between the detector array and the fringe pattern incident on the detector array. Co-pending US Patent Application No. 60 / 316,121 [Attorney Docket No. MCE-018 (111390-140)] entitled "Harmonic Suppressed Photodetector Array", incorporated by reference, is a detector array. Several detector arrays are disclosed that can be used in 120 implementations. However, any detector array that allows measurement of the phase angle between the array and the incident fringe pattern can be used to implement the detector array 120. The output signal generated by the detector array 120 is input to the signal processor 190. The signal processor 190 preferably generates an output signal representative of the phase angle between the detector array 120 and the fringe pattern incident on the array 120.

  FIG. 6 is a view showing the top surface of the sensor head 110 in the same manner as the view shown in FIG. 2B, but FIG. 6 shows additional details. As shown, the detector array 120 includes a major axis that extends in the line LL direction (ie, along the length of the photodetector) and a line WW direction (ie, photodetection). A plurality of rectangular photodetectors having a minor axis extending along the width of the device. The detector array 120 is preferably configured for use with a 4-bin algorithm. Therefore, the photodetectors in the array are preferably electrically connected to the four bonding pads 121. The processing circuit 190 (not shown) is electrically connected to the bonding pad 121. This is to allow the array 120 to be monitored. The light source 112 is preferably electrically connected to the two bonding pads 113 and controlled by an electrical signal applied thereto. The opening 114 of the VCSEL 112 is also shown in FIG. All light emitted by the VCSEL passes through this aperture.

  As also shown in FIG. 6, the index detector 140 includes a central photodetector 142 and two end photodetectors 144 disposed on opposite sides of the central photodetector 142. It is preferably implemented in a three cell configuration. The central photodetector 142 is electrically connected to the bonding pad 143. Each of the end photo detectors 144 is electrically connected to the bonding pad 145. Processing circuit 190 (not shown in FIG. 6) is electrically connected to bonding pads 143 and 145. This is to enable the index detector 140 to be monitored. Since the central photodetector 142 is preferably aligned with the light source 112, a line extending parallel to the line LL from the opening 114 bisects the central photodetector 142.

  Referring to FIG. 1, a diverging light cone 102 emitted from a light source 112 is shown illuminating DOE 166. It should be appreciated that DOE 166 enters and exits light cone 102 when one of scale 160 and sensor head 110 is moved in the direction of arrow AA shown in FIG. is there. When DOE 166 is illuminated by light cone 102, DOE 166 diffracts light from cone 102 toward index detector 140. DOE 166 is preferably implemented using an anamorphic zone plate lens. The DOE 166 preferably generates a “line image” of the light source 112 when illuminated by the light cone 102. That is, the DOE 166 preferably diffracts the “light line” and returns it to the index detector 140. The line image generated by the DOE 166 and incident on the sensor head 110 is preferably substantially parallel to the line LL shown in FIG.

  For clarity, only one DOE 166 is shown on the scale 160 of FIG. However, as shown in FIGS. 2C and 3A, the scale 160 can include two DOEs 166 disposed on opposite sides of the grid 162. The cone of light 102 reaching the scale 160 is preferably large enough to illuminate a portion of the grating 162 and one of the DOEs 166. However, when two DOEs 166 are included in the scale 160, the scale 160 and the sensor head 110 are assembled regardless of the orientation when the encoder 100 is formed. That is, if the scale 160 includes two DOEs 166, one of the DOEs 166 is illuminated by the light cone 102 regardless of whether the scale 160 is installed correctly or upside down. The scale 160 may also include two DOEs 166 that are not symmetrically disposed (eg, one DOE is disposed near the center of the scale and the other DOE is disposed near the end of the scale. ing).

In operation, one of scale 160 and sensor head 110 is moved relative to the other (in the direction of arrows AA shown in FIG. 2A) so that the line image generated by DOE 166 crosses index detector 140. Sweep. As the scale 160 moves relative to the sensor head 110 in the direction of arrow AA by a distance D, the line image generated by the DOE 166 moves across the sensor head 110 by a distance equal to KD. Where K is the scale factor from equation (3). Therefore, when e is 0 (ie, when Z ₀ is equal to Z ₁ as described in equation (3)), scale 160 is displaced relative to sensor head 110. The line image generated by the DOE 166 moves across the sensor head 110 at twice the speed of the scale movement. The line image generated by the DOE 166 is centered on the central photodetector 142 of the index detector 140 only when the DOE 166 directly exceeds the light source 112 (ie, when the encoder is configured as shown in FIG. 1). . The processing circuit 190 generates an output signal representing the light incident on the index detector 140. This output signal is sometimes called an index signal. This index signal is preferably characterized by a pulse each time the line image generated by DOE 166 sweeps across index detector 140. It should be noted that such a pulse provides a measure of the index (or reference point) of the relative orientation of scale 160 and sensor head 110. The distance or displacement measurement generated by the detector array 120 between the scale 160 and the sensor head 110 is a relative measurement. This is because the fringe pattern incident on the array 120 is a periodic signal. However, the line image generated by DOE 166 is incident on index detector 140 only when light source 120, DOE 166, and index detector 140 are all in one particular orientation. Thus, the index signal provides a reference measurement.

  The processing circuit 190 generates an index signal using various algorithms. The output signal generated by the index detector 140 may change due to changes in the intensity of the light source, stray light, or mismatch between the sensor head 110 and the scale 160, but the processing circuit 190 does not react to such changes. It is preferable to use an algorithm. This index signal is characterized by a pulse each time the line image diffracted by the DOE 166 sweeps across the index detector 140. The width of this pulse is preferably approximately equal to the period P of the grating 162. With such a pulse width, this pulse can uniquely identify or correspond to a single stripe of the pattern generated by the grating 162. In one preferred embodiment, the width of the center photodetector 142 is approximately equal to twice the period P of the grating 162 (as measured in the direction of the line WW shown in FIG. 6). In this embodiment, the index signal is preferably high whenever the center of the line image generated by the DOE 166 is incident on the central photodetector 142 and is preferably low at all other times.

FIG. 7 shows the general shape of the output signal generated by the index detector 140 as the line image 700 moves across the array from left to right as indicated by arrow 702. Curve A shows the shape of the output signal generated by the left edge photodetector 144 as the line image 700 moves over the photodetector. Curve B shows the shape of the output signal generated by the central photodetector 142 as the line image 700 moves over the photodetector. Finally, curve C shows the shape of the output signal generated by the right edge photodetector 144 as the line image 700 moves over the photodetector. One preferred method for generating the index signal from the unprocessed output signals A, B, C is that the processing circuit 190 generates signals S ₁ and S ₂ according to equation (4) below.

FIG. 7 also shows the signals S ₁ and S ₂ generated according to equation (4) from the raw signals A, B, C shown in FIG. The obvious point from equation (4) is that both signals S ₁ and S ₂ are independent of stray light. This is because the light incident on all three photodetectors of the index detector 140 is subtracted or does not contribute to S ₁ and S ₂ .

As shown, the signal S ₁ generally has a positive peak when the center of the line image 700 is incident on the central photodetector 142. In addition, the signal S ₁ has a number of side lobes or ringings that can trace the intrinsic diffraction effects in the line images. Similarly, signal S ₂ generally has a negative peak when the center of line image 700 is incident on central photodetector 142 and a number of side lobes from diffraction effects in the line image. One preferred method for generating an index signal from signals S ₁ and S ₂ is shown in equation (5) below.

In the formula (5), O is a constant offset, which is preferably preferably greater than the predicted sidelobe peaks in S ₁ and S _2, also smaller than the minimum predicted maximum value of S _1.

  FIG. 7 also shows the index signal (INDEX SIGNAL) generated by equation (5). As shown, this index signal is 1 or higher when the center of the line image generated by the DOE 166 is incident on the central photodetector 142 and 0 or lower at all other times. Has the desired characteristics. Such an index signal is characterized by a pulse every time the line image generated by DOE 166 sweeps across index detector 140.

Although the use of equation (5) is the preferred method of generating the index signal, it should be noted that other methods can be used as well. For example, whenever the signal S ₁ is greater than a certain selected value, it is possible to set the index signal simply high.

The width of the end photo detector 144 is preferably equal to the width of the central photo detector 142. Thus, stray light is not reliably contribute to signals S ₁ and S _2. However, it should be appreciated that in other embodiments, the width of the edge photodetector 144 can be different from the width of the center photodetector 142. One advantage of using an end photodetector 144 of different width than the central photodetector 142 is that such a configuration averages the diffraction effects of the line images, thereby allowing side lobes of the signals S ₁ and S ₂ . Can be reduced. Further, by adjusting the width and interval of the detector, it is possible to cancel the ringing of the signal from the edge photodetector with the ringing of the signal from the central photodetector. If such a method is used, the weighting of the raw signal in equation (4) is preferably changed so that the signals S ₁ and S ₂ are still unresponsive to stray light. In yet another embodiment, the index detector 140 can be configured by using only the central photodetector 142 and omitting the end photodetector 144. However, such a method is not preferable. This is because the resulting index signal is too sensitive to noise and mismatch.

FIG. 8 shows another embodiment of the index detector 140. In this embodiment, detector 140 includes two two-cell detectors 140A and 140B. The two-cell 140A has a center detector 142 and a left end detector 144. The two cells 140B have a center detector 142 and a left end detector 144. Since the two two cells are preferably positioned, a line extending from the light source 120 in the direction of line LL bisects the central detector 142 of both two cells 140A and 140B, as shown in FIG. It should be noted here that the signals S ₁ and S ₂ are simply generated according to the above equation (4) using the two cells 140A, 140B. For example, signal S ₁ is simply generated by adding the output signals generated by two center detectors 142 and subtracting the output signal generated by two end detectors 144 from the sum. .

  FIG. 9 is an end view of a preferred embodiment of a diffractive optical encoder 100 constructed in accordance with the present invention. FIG. 9 shows a view of encoder 100 as seen in the direction of line 2D-2D shown in FIG. 2A. The main difference between FIG. 2D and FIG. 9 is that in FIG. 9, the sensor head 110 is shown tilted with respect to the grid 160 (instead of being substantially parallel). More specifically, the sensor head 110 is tilted about an axis that is substantially parallel to the moving direction of the scale 160 (ie, parallel to the line 2A-2A shown in FIG. 2A). A preferred embodiment of the encoder 100 includes the slope shown in FIG. Inclining the sensor head 110 relative to the scale 160 as shown in FIG. 9 provides at least two advantages. First, the tilt reduces the amount of light reflected from the scale 160 and returning to the light source 120. Second, the slope increases and balances the amount of light reaching the detector array 120 and index detector 140.

  In general, it is not desirable for light reflected from the scale to enter the light source 112. First, even the preferred VCSEL light source is adversely affected by reflected light that reenters the lasing medium. Secondly, since the laser emitting surface is somewhat reflective, light reaching this surface is reflected towards the scale 160. This compositely reflected and / or diffracted stray light can produce extraneous components in the detection signal if not properly controlled. In the present invention, the intentional tilt between the photoelectric plane and the scale is selected to keep these extraneous beams away from the detector. As shown in FIG. 9, when the sensor head 110 is tilted with respect to the scale, (1) the light reflected from the scale is effectively prevented from re-entering the light source 112, or the amount of such light is significantly increased. And (2) ensure that light reflected from the light source does not reach the detector or significantly reduce the amount of such light.

  The second function of introducing a tilt between the sensor head 110 and the scale 160 is to increase and balance the level of light reaching the detector. Preferably, the sensor head 110 is tilted so that the peak intensity of the specularly reflected light cone is located approximately halfway between the detector array 120 and the index detector 140. This maximizes the amount of light incident on the two detector regions 120, 140 while minimizing the drop in light intensity on both regions.

  As described above, it is advantageous to configure encoder 100 such that the optical path length between light source 112 and scale 160 is approximately equal to the optical path length between scale 160 and detector array 120. By doing so, the period of the fringe pattern incident on the detector array 120 is reliably independent of the distance between the sensor head 110 and the scale 160. When the light source 112 is implemented as a VCSEL that emits light in a direction orthogonal to the plane of the sensor head 110 (as shown in FIG. 1), the upper surface of the detector array 120 is coplanar with the emission surface of the light source 112. By doing so, it is possible to make these optical path lengths equal. However, it is difficult to actually make these surfaces coplanar because the light source and photodetector are each typically characterized by a particular thickness.

  FIG. 10A shows one technique for making these surfaces coplanar. As shown in FIG. 10A, a trench 900 is formed in the substrate 111 of the sensor head 110 by etching. Either the photodetector of the detector array 120 or the light source 112 is disposed in the trench 900 as indicated by box 910. It should be appreciated that the use of such a trench can compensate for the difference in thickness between the detector array 120 and the light source 112. A trench, such as trench 900, is provided by machining the substrate 111 or using photolithography techniques.

  FIG. 10B shows another technique for making these surfaces coplanar. As shown in FIG. 10B, the spacer 912 is disposed on the upper surface of the substrate 111 of the sensor head 112. As indicated by box 910, either the photodetector of the detector array 120 or the light source 112 is disposed on such a spacer. A spacer such as a spacer 112 having a desired thickness is formed on the substrate 111 by, for example, a material volume, or by adhering a preformed spacer to the upper surface of the substrate 111.

  FIGS. 10C and 10D show how the light source 112 can be implemented using an edge emitting laser diode instead of a VCSEL. They also illustrate other strategies for making the optical path length between the light source 112 and the scale 160 equal to the optical path length between the scale 160 and the detector array 120. In FIG. 10C, the light source 112 is implemented using an edge emitting laser diode that emits light in a direction essentially parallel to the top surface of the sensor head 110. In this embodiment, the sensor head 110 also includes a reflective mirror 920 disposed in the optical path of the light source 112. The mirror 920 reflects the cone of light emitted by the light source 112 upward toward a scale (not shown). Also in FIG. 10D, the light source 112 is implemented using an edge emitting laser diode. The light source of this embodiment is disposed in a trench 900 provided in the substrate 111 of the sensor head 110. One edge 903 of the trench 900 is formed to be reflective. For this reason, the edge 903 reflects the cone of light emitted by the light source 112 upward toward a scale (not shown). It should be noted that the mirror 920 and the reflective edge 903 can be implemented using a reflective prism or etched folding mirror as disclosed in US Pat. No. 6,188,062. That is. The arrangements shown in FIGS. 10C and 10D each affect the optical path length between the light source 112 and the scale. It should be appreciated that such an arrangement can be used to make the optical path length between the light source 112 and the scale equal to the optical path length between the scale and the detector array 120.

Alternatively, the optical path between the light source 112 and the scale 160 to the trench and spacer suggested in Figure 10A~10D for equalizing the optical path _{(Z 1)} between the scale 160 and the detector array 120 _{(Z 0)} To avoid the cost of use, the concept of equal optical path lengths and fringe periods independent of the distance between the scale and the detector array can be abandoned. In such a case, the fringe period T incident on the detector array 120 is proportional to the grating period P and is given by equation (3) above. When designing such an encoder, it is desirable to calibrate the scale factor between the scale and the detector array and to optimize the encoder.

In the ideal case, i.e., z ₀ and z ₁ are equal and there are no other mismatches, the encoder scale factor is approximately equal to 2 (ie, the period T of the fringes incident on the detector array is equal to that of the grating). Because it is equal to twice the period P). In practice, however, the actual scale factor associated with an optical encoder constructed in accordance with the present invention is not necessarily equal but tends to approach two. In general, the reason that the scale factor does not always equal 2 is that it is difficult to measure the components sufficiently accurately and to make the spacer / trench precisely enough to ensure that z ₀ equals z _1. . Other factors such as inconsistencies also contribute to changing the scale factor from the ideal value of 2. Finally, the preferred scale factor for an optical encoder is one that gives the highest accuracy performance without being directly related to the actual value of the fringes or detector period.

  By giving this criterion (best accuracy), a preferred method for determining the scale factor of an optical encoder constructed according to the invention and calibrating the encoder in terms of the measured scale factor is discussed. . A calibration sensor head and a calibration scale are preferably manufactured. However, the calibration scale is not characterized by a substantially uniform period, but rather has a calibration grating similar to the grating 162 (as preferably the grating 162 is). This calibration grid has several different sections, each section being characterized by a unique period. One section is manufactured with a design period P (eg, P is equal to 20 microns). The other segment is characterized by a period slightly offset from P. The various sections of the calibration grid preferably extend over a period range around P with incremental steps corresponding to about 0.5% of P. That is, the various segments have periods that are approximately P, 0.995P, 1.005P, 0.990P, and the like. The inventor has observed that a period range of +/− 3% typically includes an optimal period. Of course, as will be apparent to those skilled in the art, if the best performance is observed at the end of the range, a new calibration grid with a wider range should be manufactured. The various sections of the calibration should be spatially distributed on the common substrate and separated for easy identification and selection. To facilitate use and alignment, the axes of the various sections should be parallel. The calibration sensor head is used to measure the phase angle of a fringe pattern incident on an array with a period T approximately equal to the design point 2P (eg, US patent application Ser. 316, 121 [Attorney Docket No. MCE-018 (111390-140)], including a preferably configured calibration detector array. The calibration sensor head and the calibration scale are configured to form a calibration encoder (eg, as shown in FIGS. 2A-2D).

  When the encoder scale factor of the calibration encoder is exactly equal to 2 (and there is no other variation effect), the calibration encoder is most useful when the calibration detector array is used in the section of the calibration grid characterized by period P. Give precise results. However, the most accurate results are usually given when the calibration detector array is used in other sections of the calibration grid. The calibration encoder is preferably tested using each of the calibration grid sections. This is to determine which section of the calibration grid gives the most accurate results. Typically, the accuracy of each test is determined by the rms (root mean square) difference between the displacement truth sensor that makes simultaneous measurements of grid motion and the encoder output. Laser interferometers are successfully used as truth sensors.

  Since the calibration encoder is designed to work with a grating of period P, the most accurate results are generally obtained from a calibration grid section of period FP. It can therefore be assumed that the measured calibration factor F should be used during the manufacture of the operational encoder. Specifically, the operating encoder should use a grating with period FP or the detector array period T should be modified to T / F.

  In this regard, a number of encoders according to the present invention can be manufactured by using the scale 160 instead of the calibration scale and by using the sensor head 110 instead of the calibration sensor head. One method of constructing an encoder according to the present invention measures (1) a scale having a grating 162 characterized by a period FP and (2) a phase angle of an incident fringe pattern characterized by a period T approximately equal to 2P. And a sensor head having a detector array 120 configured for this purpose. One problem with this method is that the period FP of the grating 162 is unlikely to be an integer in standard length units (ie, microns or mils). Thus, for example, the grating period of such a grating may be 20.2 microns instead of the more typical 20 microns. Accordingly, a preferred method for constructing an encoder according to the present invention includes (1) a grating 162 characterized by a period P and (2) an incident fringe pattern having a period approximately equal to 2P divided by the scale factor F. Using a sensor head having a detector array 120 configured to measure the phase angle. The latter method is more preferred because it allows any generation of sensor heads constructed according to the present invention to be used interchangeably with industrial standard scales.

  It should be appreciated that when index detector 140 is included in the encoder, it is also desirable to adjust the width of the index detector element by a calibration scale factor. For example, it may be advantageous to make the width of the central photodetector of the index detector 140 approximately equal to the period P of the grating 162 divided by the scale factor F.

  FIGS. 11A and 11B illustrate additional features incorporated into an encoder constructed in accordance with the present invention. 11A and 11B each show a side view of the diffractive optical encoder 100 viewed from the same perspective as FIG. 2A. FIG. 11A shows the diverging light cone 102 going upward from the sensor head 110 to the scale 160. FIG. 11A also shows three light beams that are diffracted by the grating 162 of the scale 160 and directed downward to the detector array 120. Specifically, FIG. 11A shows a zero order beam whose left and right boundaries are indicated by reference numeral 1000, a −1st order beam whose left and right boundaries are indicated by reference numeral 1001, and a left and right boundary indicated by reference numeral 1003. The third-order beam is shown. As shown, the zero order beam, the minus first order beam, and the minus third order beam all enter the detector array 120. It should be noted that other beams (eg, positive first and third order, and positive and negative fifth order beams) are also incident on the detector array 120, but for convenience of explanation these beams are It is not shown in FIG. 10A. One problem with the encoder shown in FIG. 11A is that many diffracted beams are all incident on the detector array 120, and the presence of these beams degrades the quality of the resulting interference fringes that are incident on the detector array 120. It is.

  The encoder 100 shown in FIG. 11B is similar to that shown in FIG. 11A, but the encoder of FIG. 11B additionally includes a mask 1010. As illustrated, the mask 1010 is disposed between the sensor head 110 and the scale 160 in proximity to the scale 160. Mask 1010 also defines a central opening 1012. Mask 1010 prevents most of the light in cone 102 from reaching scale 160. That is, only the light passing through the opening 1012 reaches the scale 160. The mask 1010 is preferably made of an absorbent material. This is because light incident on the mask 1010 is simply absorbed and is not reflected toward the sensor head 110. The mask 1010 has the advantage of limiting the angular range of the beam that is diffracted by the scale 160 and returns to the scale head 110. In the encoder shown in FIG. 11B, the 0th order beam and the −1st order beam are incident on the detector array 120, but the −3rd order beam is not incident on the detector array 120. It should be noted that if the third order beam is not incident on the detector array 120, then all higher order beams are also not incident on the detector array (ie, the higher order beam is more than the third order beam shown). Is also displaced to the right or left of the detector array 120). Accordingly, the mask 1010 advantageously improves the quality of the interference fringes incident on the detector array 120 by removing unwanted higher order beams. In operation, the mask 1010 and the sensor head 110 are preferably in a fixed relationship with each other. The scale 160 is moved relative to the sensor head 110 (left and right in the configuration shown in FIG. 11B).

  As discussed in US Patent Application No. 60 / 316,121 [Attorney Docket No. MCE-018 (111390-140)] entitled “Harmonic Suppressed Photodetector Array”, a preferred detector array is 3 Does not respond to second harmonics. Also, using a grating characterized by a 50-50 duty cycle can prevent all even order beams from reaching the detector array 120. Thus, the aperture 1012 need not be small to ensure that the third or fourth order beam does not reach the detector array. The aperture 1012 is rectangular and the width of the aperture is preferably small enough to prevent the fifth order diffracted beam from reaching the detector array 120. The height of the aperture 1012 is preferably selected so that light from the cone 102 illuminates both the grating 162 and the DOE 166.

  In one preferred embodiment of an encoder constructed according to the present invention, the distance d between the sensor head 110 and the scale 160 is approximately equal to 4.7 mm, the light source 112 is implemented using a VCSEL, and its cone angle is Equal to approximately 17 degrees, the wavelength of light emitted by the VCSEL is approximately equal to 850 nm, the tilt angle between the sensor head 110 and the scale 160 is approximately equal to 8 degrees, and the period P of the grating 162 is approximately equal to 20 microns. The detector array 120 is configured to monitor an incident fringe pattern having a period approximately equal to 40 microns. In another preferred embodiment, the rectangular opening 1012 is characterized by a width approximately equal to 0.4 millimeters and a height approximately equal to 1.2 millimeters. A mask 1010 that defines the rectangular opening 1012 is disposed between the sensor head 110 and the scale 160. Also, the mask 1010 is separated from the scale 160 by a distance approximately equal to 250 microns.

  Several methods of constructing an improved diffractive optical encoder have been disclosed. It should be appreciated that several encoders are constructed in accordance with the present invention by incorporating one or more of these methods. For example, one encoder constructed in accordance with the present invention includes an index detector but does not include a mask (eg, as shown in FIGS. 11A and 11B). Similarly, one encoder constructed in accordance with the present invention includes a mask but does not include an index detector. One encoder constructed according to the present invention includes both a mask and an index detector.

  Certain modifications may be made to the apparatus without departing from the scope of the invention encompassed herein, so that all matters contained in the above description or shown in the accompanying drawings are restrictive. It should be construed as illustrative rather than mere.

1 is a perspective view of a diffractive optical encoder configured according to the present invention. 1 is a side view of a diffractive optical encoder configured according to the present invention. FIG. 2B is a top view of the sensor head viewed in the direction of line 2B-2B shown in FIG. 2A. It is the figure of the scale seen in the direction of line 2C-2C shown by FIG. 2A. FIG. 2B is an end view of the encoder viewed in the direction of line 2D-2D shown in FIG. 2A. It is a figure of the scale used with the diffractive optical encoder comprised by this invention. 3B is an enlarged view of a portion of the scale shown in FIG. 3A, showing two different approaches for assembling the scale used in a diffractive optical encoder constructed in accordance with the present invention. It is a side view of a diffractive optical encoder, Comprising: A part of beam which is diffracted by a scale and goes to a sensor head is shown. Fig. 5 shows an interference fringe pattern at different distances from the scale. It is a more detailed top view of a sensor head constructed in accordance with the present invention. 2 shows a graph of raw signals generated by an index detector of an encoder constructed in accordance with the present invention and a graph of signals generated by the present invention in response to these raw signals. 6 shows an alternative embodiment of an index detector constructed in accordance with the present invention. 1 is an end view of a diffractive optical encoder configured according to the present invention, in which a sensor head is inclined with respect to a scale. Fig. 4 shows different strategies for equalizing the optical path length between the light source and the scale and the optical path length between the scale and the detector array according to the invention. 2 shows a part of a beam diffracted from a scale and directed to a sensor head in an optical encoder constructed according to the present invention. Figure 2 shows a diffractive optical encoder constructed in accordance with the present invention having a mask to prevent higher order beams from reaching the detector array.

Claims (13)

A. Scale and
B. An optical encoder comprising a sensor head,
The scale includes an optical grating and an optical element;
The sensor head includes a light source, a detector array, and an index detector, all of which are disposed on a substrate,
The scale is disposed to face the sensor head, and is disposed to move relative to the sensor head.
The distance between the scale and the Talbot imaging plane closest to the scale is equal to d;
The sensor head is disposed in a region limited by the first plane and the second plane;
The first plane is separated from the scale by a distance approximately equal to n × d + d × x;
The second plane is separated from the scale by a distance approximately equal to n × d−d × x, n is an integer, and x is ½ or less,
The light source emits a divergent light beam that is directed to the scale;
Light from the divergent light beam is diffracted by the grating and travels toward the detector array,
The light from the divergent light beam is diffracted by the optical element and directed to the index detector;
The detector array provides a relative measurement of the position of the sensor head relative to the scale;
The optical encoder, wherein the index detector provides a reference measurement value of the position of the sensor head with respect to the scale.   The encoder of claim 1, wherein the substrate defines a trench, and at least one of the light source and the detector array is disposed in the trench.   The encoder according to claim 1, further comprising a spacer disposed on the substrate, wherein at least one of the light source and the detector array is disposed on the spacer.   The encoder according to claim 1, wherein the index detector includes a central photodetector, a left photodetector, and a right photodetector.   The encoder according to claim 1, wherein x is 0.2 or less.   The encoder according to claim 1, wherein x is 0.1 or less. A. Scale and
B. A sensor head;
C. An optical encoder comprising a mask disposed between the scale and the sensor head,
The scale includes an optical grating and an optical element;
The sensor head includes a light source and a detector array, both of which are disposed on a substrate,
The scale is disposed to face the sensor head, and is disposed to move relative to the sensor head.
The distance between the scale and the Talbot imaging plane closest to the scale is equal to d;
The sensor head is disposed in a region limited by the first plane and the second plane;
The first plane is separated from the scale by a distance approximately equal to n × d + d × x;
The second plane is separated from the scale by a distance approximately equal to n × d−d × x, n is an integer, and x is ½ or less,
The light source emits a divergent light beam that is directed to the scale;
Light from the divergent light beam is diffracted by the grating and travels toward the detector array,
The mask defines an opening;
The mask maintains a substantially fixed state with respect to the sensor head,
The optical encoder is characterized in that the aperture is sized and positioned so as to substantially prevent a fifth-order beam diffracted from the grating from reaching the detector array. A. Scale and
B. An optical encoder comprising a sensor head,
The scale includes an optical grating and an optical element;
The sensor head includes a light source, a detector array, and an index detector, all of which are disposed on a substrate,
The scale is disposed to face the sensor head, and is disposed to move relative to the sensor head.
The light source emits a divergent light beam that is directed to the scale;
Light from the divergent light beam is diffracted by the grating and travels toward the detector array,
The light from the divergent light beam is diffracted by the optical element and directed to the index detector;
The detector array provides a relative measurement of the position of the sensor head relative to the scale;
The index detector provides a reference measurement of the position of the sensor head relative to the scale;
The index detector includes three optical detectors.   9. The three photodetectors of the index detector are a central photodetector, a left photodetector, and a right photodetector, and each of the three photodetectors generates an output signal. Encoder described in.   And further comprising a processing circuit for generating a first signal, wherein the first signal is an output signal generated by the left and right photodetectors from twice the output signal generated by the central photodetector. An encoder according to claim 9 representing the difference minus the sum of.   The encoder according to claim 10, wherein the processing circuit also generates a second signal, the second signal representing -1 times the first signal.   The processing circuit also generates an index signal, which is equal to the first value when the first signal is greater than the second signal plus an offset value, and not otherwise. 12. The encoder according to claim 11, wherein the first signal is equal to a second value.   The encoder of claim 9, wherein the central photodetector comprises more than one photodetector.

Images