WO2003006928A1

WO2003006928A1 - Optical encoder

Info

Publication number: WO2003006928A1
Application number: PCT/GB2002/003079
Authority: WO
Inventors: David Sellars
Original assignee: Ice Technology Limited
Priority date: 2001-07-07
Filing date: 2002-07-05
Publication date: 2003-01-23
Also published as: GB0116646D0

Abstract

An optical encoder for positioning an object is described in the form of: a scale having a pattern formed thereon with a plurality of periods each corresponding to a predetermined position, the pattern having a first portion which is a repeating pattern of constant period and a second portion which identifies each of the plurality of periods. An illumination source is used to illuminate the scale and a detector receives illumination from the scale. A positioning means for causing relative motion between the scale and the object; and A signal analyser analyses the image, by a first step to identifying the period of the pattern in which the object is located by characterising from the image the second pattern uniquely identifying the period, and in a second step reading the first portion of the pattern to determine the position of the object relative to the two extremes of the period, and then actuates positioning means to microposition the object at a desired location within the said period, for example at the centre thereof.

Description

OPTICAL ENCODER

The invention relates to an optical linear or rotary encoder to encode position and/or motion information for an object with high accuracy. In particular, the invention relates to an incremental and/or absolute positional encoder, for example for use in manufacturing systems where objects need to be manipulated and positioned with a high degree of accuracy.

Optical encoders are widely used to measure the absolute and/or relative position of an object and/or to monitor its motion. An optical encoder measures either the angular or linear position of an object by optically detecting marks on a scale relative to which the object moves. The marks take the form of relatively optically transmissive and relatively optically non- transmissive regions for use with an encoder having a light source and a light sensor. Optical encoders are known in which the scale relies on direct transmission of light, in which the markings are relatively transparent and relatively opaque regions, and which rely on reflected light, in which the markings are relatively reflective and relatively non-reflective (e.g. light and dark) regions.

There are two types of prior art encoders. In the simplest type, known as an incremental encoder, the pattern is an alternating repeating coding pattern, and the encoder simply measures translation and/or angular rotation by counting the number of marks that move past the encoders optical detector. Such a device measures only relative position or motion.

For many applications, an absolute positional encoder is to be preferred. In a conventional absolute encoder, each position is given not simply by one mark, but by a unique code pattern of marks which identifies the absolute position of the object. The code pattern comprises a number of optically detectable code bits, which together uniquely identify the position, and a change in position is sensed by detecting a change in the code bits which make up the code pattern. Absolute encoders may lend themselves to both linear and angular encoding of position.

Incremental and absolute encoders may be combined in a practical system, for example to use an incremental code for course determination of, e.g., speed and/or distance travelled, and an absolute encoder for fine position checking. In addition to providing positional information, such a coding system is able to provide information about the motion of the object if processed over time. The present invention relates in particular to absolute positional encoders, where encoding to a high degree of accuracy is desirable, but the principles of the invention could be applied to simple incremental encoders also.

An absolute encoder outputs position information to a detector from a graduation scale. The coder and the detector are assembled to allow them to move relative to one another. The coder is provided with an absolute pattern graduation scale having a plurality of positions each identified by an absolute numerical code. A detector is provided having a sensor or sensors to read the information on the pattern and identify the position from the numerical code. Absolute encoders are of a linear type for movement in a longitudinal direction along a strip scale or of rotary type for angular movement relative to a disk or a cylindrical coder.

In an absolute encoder sensitivity is limited to the size of the smallest code bit which can be recorded. To increase resolution and accuracy of positioning, finer marks are needed. This places extra burden on the printing or other technique used to create the marks, and on the complexity and sensitivity required for the optics of the detector. Both of these factors are likely to increase significantly the expense and complexity of a practical system.

Moreover, since each unique position is identified by unique coding, the higher the resolution of the encoder, the more limited the amount of travel that it can accommodate. Each unique coding pattern is made up of a plurality of bits. For a given number of bits, there is a finite number of unique codes. If the resolution of each pattern is increased, the total distance covered by this finite number of codes will decrease or the number of bits in the pattern required to maintain the travel will need to be increased.

US 5965879 describes an attempt to overcome this problem. In this document, an optical encoder is described having a pattern making up a plurality of periods which includes both an incremental scale in which the pattern is identical for all the periods and an absolute scale which has an identification coding unique to each period. The detector takes the form of a CCD array in a camera head. However, even with this device, the accuracy of positioning remains limited to the resolution of the scale, with the position being measured only to the accuracy of the finest period repeat in the scale.

It is an object of the present invention to achieve an encoder which mitigates some or all of the above disadvantages.

It is a particular object of the present invention to achieve an encoder, and in particular an absolute positional encoder, with higher sensitivity than that of conventional encoders.

It is a particular object of the invention to achieve an encoder, and in particular an absolute encoder, which can achieve high resolution over longer travel distances than conventional encoders. It is a particular object of the present invention to provide an absolute encoder which has a high resolution not dependent on the resolution of the process used to produce the marks on the scale.

Thus, in accordance with a first aspect of the present invention there is provided an optical encoder for positioning an object, comprising: a scale having a pattern formed thereon, the pattern having a plurality of periods each corresponding to a predetermined position, the pattern comprising a first portion which is a repeating pattern of constant period and a second portion which identifies each of the plurality of periods; an illumination source to illuminate the scale; a detector to receive illumination from the scale comprising image forming means having a field of view sufficient to encompass at least one period of the pattern, and having an imaging resolution on a substantially finer scale than the field of view; a positioning means for causing relative motion between the scale and the object; and a signal analyser to analyse the image, wherein the signal analyser is adapted in a first step to identify the period of the pattern in which the object is located by characterising from the image the second pattern uniquely identifying the period, and in a second step to analyse the first portion of the pattern to determine the position of the object relative to the two extremes of the period, and to act on the motion means to microposition the object at a desired location within the said period, for example at the centre thereof.

The invention thus uses the absolute encoded scale to position the object in conventional manner, but then uses the short fixed period repetitive scale to microposition the object for example at the centre of the desired period. The encoding system has both digital and analogue or pseudo-analogue features. The encoding system is first digital in the conventional manner in positioning absolutely generally at one of the uniquely identified periods in the first coarse positioning step.

In the second micropositioning step, an analogue or pseudo-analogue process more accurately positions the object within the period based on distance measurements from the position of the device to the extremities of the period. The object may then be micropositioned, for example to the centre of the period or any other desired position, by any suitable iterative or other process.

It is a crucial feature of an encoder in accordance with the invention that accuracy is not limited by printing resolution but by imaging resolution. Only the first, relatively coarse positioning step is carried out using the unique absolute encoding conventional in the prior art. The second, micropositioning step is carried out by accurate positioning of the device relative to a single mark on the scale, and in particular relative to a desired point along that single mark, such as centrally.

The invention thus overcomes problems in the prior art concerning accuracy of printing verifying scales, and concerning the limited travel for a pattern having a given number of bits when the size of the scale is reduced. For any given scale having a given number of periods used in an encoder in accordance with the invention a significantly higher number of accurate positions may be determined, dependent not merely on the unique identifier for each such period but also on relative positioning within the period, and therefore dependent not merely on the printing resolution to which the uniquely identifying marks are created on the scale but also on the resolution of the imaging device which is adapted to operate on a significantly finer scale. The invention lies primarily in the way that the image is analysed, and hence a number of conventional prior art scales will suggest themselves to the skilled person. In particular for example, the pattern on the scale comprises alternating areas of relatively high and relatively low light transmissivity. For example, the scale is a transparent scale, and the areas consist of relatively high transparency and relatively high opacity. Alternatively, the scale is a reflecting scale and the areas in the pattern consist of relatively reflective and relatively non-reflective, such as relatively light and relatively dark, areas.

The pattern on the scale consists of a first part which is a constant period repeating pattern marking each period on the scale, and a second part which uniquely codes the scale. These two parts may be distinct, for example arranged in parallel in the case of linear scales or concentrically in the case of angular scales. The first pattern may for example be a linear or angular repeating track, for example of alternate relatively light transmissive and relatively light non-transmissive regions of constant lateral or angular pitch. The second part may comprise a plurality of markings which together provide a unique binary encoding identifying each of the plurality of periods.

Conveniently, the two parts of the pattern are combined in a scale which is a multi-track type absolute pattern graduation scale or gray code scale. In this type of scale, an absolute pattern is formed by a plurality of parallel or concentric tracks each having an incremental pattern of different pitch. The different tracks are so arranged that for a given period, reading across the main direction of travel of the tracks, a reading of the scale provides a unique binary encoding characteristic of the particular period. The number of tracks determines the number of bits which can be encoded. The illumination detector must have sufficient field of view to measure the full width across the tracks at any given period. Resolution is limited to the shortest period printing on the pattern, which is conventionally an incremental repeating pattern.

If such a gray code scale is used, the first and second patterns are effectively combined in a conventional gray code scale. The incremental repeating scale which makes up the first part of the pattern comprises the shortest period repeat which is conventionally provided as the first track on such a gray code scale.

In this embodiment the encoded system is then arranged and the analyser is adapted such that in a first image analysis step the absolute position of the object is determined by making a reading across the whole width of the multi- track scale and analysing the detected image to determine the unique identifier code and hence the absolute position in conventional manner. Positioning resolution of this first conventional step is limited to the resolution of the finest pitch on the gray code scale.

However, in accordance with the present invention, this relatively coarse positioning step is augmented in that the analyser is adapted in a second step to determine the precise position of the object along the particular period at which it is positioned by taking measurements relative to the extremities on the shortest period track marking. For these purposes, the image is analysed from this track only, and the longitudinal distance from either end of the mark (as represented for example by the transition to and from an optically transmissive and an optically non transmissive region at the edges thereof) is determined. This information is used to effect a micropositioning of the device to a predetermined position longitudinally along the shortest period marking, for example in the centre thereof. To position in the centre is a simple process. As explained, the object is first coarsely positioned in the usual manner relative to the full gray scale. In a second step, the object is centred to a desired position such as the centre of the finest repeating scale. The image is analysed to determine the distance to the edge of that element of the pattern in one longitudinal direction (a), and the distance to the edge of the pattern in the other longitudinal direction (b). This

gives a relative position from a first edge as . This is optimised to

(a + b) microposition the object at the desired position along the scale. For example,

the object is fully centred when = ^lA, or (a) = (b).

(a + b)

It will be appreciated that when an object is positioned in accordance with the invention the accuracy of positioning is not limited to the resolution imposed by the periodicity of the pattern itself, but is rather limited by the accuracy to which (a) and (b) can be measured, which is a function of the resolution of the optical detector itself. Potentially, much more accurate positioning is possible for a given scale than is possible with prior art devices employing gray code scales.

Any conventional gray code scale can be used in conjunction with a detector of suitable resolution to confer the necessary accuracy. However, whereas in conventional gray code scale patterns the alternating repeating track of shortest period is conventionally the first track, in a scale in accordance with the invention the alternating repeating track of shortest period conveniently comprises a centrally located track. This is preferred because the resolution of most imaging devices is likely to be greater at the centre. Maximum resolution to operate a device in accordance with the invention is required at the centre for accurate determination of the value of (a) and (b) in relation to the edges of the marking on the track of shortest period. More limited resolution is required for the other tracks, since these are merely read in digital manner as is conventional for gray code scales.

In accordance with the invention, by virtue of the second stage of the process, the positioning resolution is determined by resolution of the light detector itself. The encoder is able to position the object with high accuracy to be centred (or otherwise positioned relative to the edges) on the smallest marking. Accuracy of positioning is thus much greater than the accuracy to which the marking can itself be printed. The object is coarsely positioned in the usual manner relative to a full digital scale, but is then micropositioned accurately within the period of the digital scale by an analogue or pseudo-analogue process.

In contrast, the prior devices still rely on essentially digital processing and rely for their accuracy on the accuracy and increased resolution of the printed scale, with all the inherent drawbacks.

It is an important feature of the present invention that the resolution of the imaging detector is maximised. Accordingly, the imaging detector preferably comprises an array of photo detector elements, and in particular an array of charge-coupled devices. The CCD array is conveniently contained in a suitable camera head, which may further include eg illumination source and/or detector.

This simple array alone will give a reasonable degree of resolution for many applications. If higher imaging resolution is required, the imaging detector preferably further comprises a microscopic optical system comprising a suitable lens or lens array to enhance resolution of the image. In this case, the imaging detector preferably comprises a camera head containing the detecting means and further comprising means for focusing emissions received from the scale on to the detecting means to enhance the resolution of the image. In particular, the means for focusing comprises a microscope optical system consisting of one or more lenses.

The image of the pattern formed on the CCD array is analysed by the image analyser to determine first the unique pattern and hence the identifying coding characteristic of the given period at which the object is located, and second the relative position of the object within the period, to allow the second, micropositioning stage to be effected.

The resolution of the imaging detector determines the positioning resolution of the overall device. The invention relies on the ability of the detector to operate at a resolution substantially below the narrowest pitch of the period of the shortest period repeating pattern. Whilst an appreciable improvement in the accuracy of the positioning can be achieved if the resolution of the imaging detector is less than about 0.1 of a period, enhanced accuracy is achieved if the resolution of the imaging detector is less than about 0.01 of a period, and in particular if it is substantially less than 0.01 of a period. In this way, the second positioning step is based on a reading which approximates to an analogue rather than a digital reading.

The encoder may be further adapted to effect an initial very coarse positioning and/or provide a speed determining function by using the first part of the pattern as an incremental position encoder. For example, in the preferred embodiment, where a multi-track gray code scale is used, the device may use the highest frequency regular repeating track as an incremental scale for such a purpose. Means are provided for relative movement of object and scale during the coarse positioning step, and means are also provided for such relative movement during the fine micropositioning step. These may be the same or may be separate.

The invention is equally applicable to a linear or to a rotary encoder. In accordance with the first alternative, the absolute position to be determined is an absolute linear position, and the means for causing relative motion comprises means for causing relative linear movement of the object and the scale. In the second alternative, the absolute position to be determined is an absolute angular position and the means for causes in relative motion comprises means for causing relative rotational motion between the object and the scale about an axis which passes through a centre of a rotary scale.

The pattern may be produced by any suitable patterning technique. In particular, the pattern may be produced by microprinting areas of relatively low light transmissivity on a surface of relatively high transmissivity. For example, areas of relative opacity may be printed on a transparent film. Alternatively, relatively dark and non-reflective areas may be printed on a generally reflective surface. Suitable microprinting techniques include microlithographic processes of conventional type such as are used to produce conventional fine scale optical encoding scales.

The invention allows accurate micropositioning of an object relative to the scale. In a preferred embodiment, the invention comprises a micropositioning device comprising an optical encoder and positioner as hereinbefore described operatively coupled to and controlling a device positioning arm for the positioning and/or transfer of devices in a production system. The invention is particularly suited to the handling in a production system of small objects which require accurate and precise manipulation, such as computer chips and like devices being manipulated in a fabrication line such as a chip coding line.

In accordance with a further aspect of the invention, there is provided a method of determining the position of an object and/or of positioning an object, the method comprising:

providing a scale having a pattern formed thereon, the pattern having a plurality of periods each corresponding to a predetermined position and comprising a first portion which is a repeating pattern of constant period and a second portion which identifies each of the plurality of periods;

providing an illumination source to illuminate the scale and a detector to receive illumination from the scale comprising image forming means having a field of view sufficient to encompass at least one period of the pattern, and having a resolution on a substantially finer scale than the field of view;

causing relative motion between the scale and the object;

analysing the resultant image in a first step to identify the period of the pattern in which the object is located by characterising from the image the second pattern uniquely identifying the period, and in a second step to determine the position of the object relative to the two extremes of the period by analysing the first portion of the pattern;

causing further relative motion between the scale and the object to microposition the object at a desired location within the said period, for example at the centre thereof. Further features of the method will be understood by analogy with the foregoing.

The invention will now be described by way of example only with reference to Figures 1 and 2 of the accompanying drawings, in which:

Figure 1 is a schematic representation of an apparatus in accordance with the invention;

Figure 2 is an example scale for use in accordance with the present invention.

Referring first to Figure 1, an object handling arm (1) is shown for accurate manipulation of an object (3) under control of a motor drive (5). The drive (5) moves the object handling arm (1) and hence the object (3) relative to a gray code scale (4) of the type illustrated in Figure 2.

A camera head (7) and light source (8) are mounted in fixed relationship with the head of the object handling arm (1). The light source (8) illuminates the scale (4), which is made up of areas of opaque printing on a generally reflective surface (for example black on white printing). Reflected illumination is collected by the camera head (7) through optics (9) by a charge coupled device array (11), the image information is passed to an image processor (13).

The image processor (13) processes the image in two stages. In a first stage (a) information about the whole image consisting of the whole gray code scale across the full width of the multiple tracks is compared with stored data in an image memory (14) to determine against a stored database of unique binary identifiers the period in which the object is generally positioned. In the second stage (b) image information about the precise position of the object relative to the two edges of the period, determined by measurement of the distance to the edges of the highest frequency repeat marking along the track in the centre of the scale (4), is analysed, and the result is used to control the drive (5) to microposition the object accurately in the centre of the period. The positioning of the object is now known to a high degree of accuracy, and the object can be manipulated with greater confidence.

This specific example given is of a linear encoder in which the optical encoder measures the linear position of an object by optically detecting marks on a scale relative to which the object moves in linear manner. Nonetheless, the skilled person will be familiar with equivalent devices which serve as angular encoders and which act by optically detecting marks on a angular scale relative to which the object moves angularly. The skilled person would have no difficulty in adapting the technology of the present invention to such an embodiment.

Two examples of suitable scales (4) are shown in greater detail in Figure 2. Two scales are shown at Figures 2a and 2b. In each case, five tracks are represented. In each case, the representations are not to scale, but are compressed in the Y direction for ease of view, in the example by a factor of four.

In essence, both scales are conventional gray coded multi-track scales. Figure 2a is entirely conventional. In the case of Figure 2b, the most notable difference from most such conventional scales is that the highest frequency repeating track (tj) is located centrally to exploit the higher resolution which can be expected of the CCD array (11) towards the centre of its field of view. In use, in the first stage of the process, the object is coarsely positioned relative to a desired period (p) by using the unique binary coding to give the absolute position represented by taking a reading across the full width of the tracks (W). In the second stage, the object is micropositioned to be centred within the period (p) by measuring its precise position relative to the two edges (po and p ) and making such adjustments to the position as are necessary to ensure that the object is located equidistantly therebetween (or in such other predetermined position as may be desired).

Claims

1. An optical encoder for positioning an object, comprising: a scale having a pattern formed thereon, the pattern having a plurality of periods each coπesponding to a predetermined position, the pattern comprising a first portion which is a repeating pattern of constant period and a second portion which identifies each of the plurality of periods; an illumination source to illuminate the scale; a detector to receive illumination from the scale comprising image forming means having a field of view sufficient to encompass at least one period of the pattern, and having an imaging resolution on a substantially finer scale than the field of view; a positioning means for causing relative motion between the scale and the object; and a signal analyser to analyse the image, wherein the signal analyser is adapted in a first step to identify the period of the pattern in which the object is located by characterising from the image the second pattern uniquely identifying the period, and in a second step to analyse the first portion of the pattern to determine the position of the object relative to the two extremes of the period, and to act on the positioning means to microposition the object at a desired location within the said period, for example at the centre thereof.

2. An encoder in accordance with Claim 1 wherein the pattern on the scale comprises alternating areas of relatively high and relatively low light transmissivity.

3. An encoder in accordance with Claim 2 wherein the scale is a transparent scale and the areas consist of relatively high transparency and relatively high opacity.

4. An encoder in accordance with Claim 2 wherein the scale is a reflecting scale and the areas in the pattern consist of relatively reflective and relatively non-reflective areas.

5. An encoder in accordance with Claim 4 wherein the areas consist of relatively light and relatively dark areas.

6. An encoder in accordance with any preceding claim wherein the two portions of the pattern are distinct, for example arranged in parallel or concentrically.

7. An encoder in accordance with any preceding claim wherein the first portion of the pattern is a linear or angular repeating track.

8. An encoder in accordance with any preceding claim wherein the second portion of the pattern comprises a plurality of markings which together provide a unique binary encoding identifying each of the plurality of periods.

9. An encoder in accordance with any preceding claim wherein the two parts of the pattern are combined in a scale which is a multi-track type absolute pattern graduation scale or gray code scale.

10. An encoder in accordance with any preceding claim wherein the imaging detector comprises an aπay of photo detector elements.

11. An encoder in accordance with Claim 10 wherein the imaging detector comprises an array or charge-coupled devices.

12. An encoder in accordance with any preceding claim wherein the imaging detector comprises a microscopic optical system comprising a suitable lens or lens array to enhance resolution of the image.

13. An encoder in accordance with any preceding claim wherein the detector is adapted to operate at a resolution of less than about 0.1 of a period of the shortest period repeating pattern in the scale.

14. An encoder in accordance with Claim 13 wherein the detector operates at a resolution substantially below 0.01 of a period of the shortest period repeating pattern.

15. An encoder in accordance with any preceding claim adapted to function as a linear encoder, in that the absolute position to be determined is an absolute linear position, and the means for causing relative motion comprises means for causing relative linear movement of the object and the scale.

16. An encoder in accordance with one of Claims 1 to 14 is adapted to function as a rotary encoder, in that the absolute position to be determined is an absolute angular position and the means for causes in relative motion comprises means for causing relative rotational motion between the object and the scale about an axis which passes through a centre of a rotary scale.

17. An encoder in accordance with any preceding claim wherein the pattern is produced by micro printing areas of relatively low transmissivity on an area of relatively high transmissivity.

8. A method of determining the position of an object and/or of positioning an object, the method comprising: providing a scale having a pattern formed thereon, the pattern having a plurality of periods each corresponding to a predetermined position and comprising a first portion which is a repeating pattern of constant period and a second portion which identifies each of the plurality of periods; providing an illumination source to illuminate the scale and a detector to receive illumination from the scale comprising image forming means having a field of view sufficient to encompass at least one period of the pattern, and having a resolution on a substantially finer scale than the field of view; causing relative motion between the scale and the object; analysing the resultant image in a first step to identify the period of the pattern in which the object is located by characterising from the image the second pattern uniquely identifying the period, and in a second step to determine the position of the object relative to the two extremes of the period by analysing the first portion of the pattern; causing further relative motion between the scale and the object to microposition the object at a desired location within the said period, for example at the centre thereof.