WO1997040345A1

WO1997040345A1 - A quasi-absolute encoder

Info

Publication number: WO1997040345A1
Application number: PCT/SE1997/000680
Authority: WO
Inventors: Lennart Stridsberg
Original assignee: Stridsberg Innovation Ab
Priority date: 1996-04-22
Filing date: 1997-04-22
Publication date: 1997-10-30
Also published as: SE9601551D0; AU2657197A

Abstract

An encoder disc (104) for a position detector has a conventional incremental position track (103), an absolute position track (102) and a commutation track (104), the latter used for providing signals for commutating a brushless electric motor. The absolute position and commutation tracks have fields differing in a physical characteristic such as the area or height of fields. When these fields are detected they provide different signal levels. In the absolute position track there are thus sync fields (115) having a maximum height and other lower fields (110 - 114, 117, 118) providing information on the absolute position of the encoder disc (104). Thus, for only a small rotation of the disc, the absolute position thereof is determined. In the commutation track (101) the fields have heights varying as a sine function. These fields are detected by using mask windows (107, 108) having a suitable offset from each other in the direction of the commutation row for providing also information on the signal amplitudes when detecting the fields of this row.

Description

A QUASI-ABSOLUTE ENCODER TECHNICAL FIELD

The invention is concerned with transducers for sensing position and recording position information. BACKGROUND

Position encoders for sensing the position of a part that moves in relation to another part are well known in the art. In a common rotary case the parts comprise a disc such as an encoder disc having a shaft and rotating in relation to a stationary detection device carrying light sources, mask windows and light sensors. Position encoders can generally be based on arranging, on one of the parts, e.g. an encoder disc, fields having varying measurable, physical characteristics such as fields having different light transmission or fields having different light reflectivity. Fields are here localized, definite areas or regions on a coding member. Such transducers are in the rotary case well known under the name of optical encoders. A first one of the two large groups of such encoders comprises absolute encoders that permit the detection of a position of a rotating part within one full turn, and a second group comprises incremental encoders that permit the detection of movements in relation to an initial position of the two parts in relation to each other. Incremental encoders are often combined with a zero reference. To find the absolute position within one turn, the encoder disc must be rotated until a signal is received from the zero reference device. This provides a reference position and thereafter the absolute position can be calculated by adding the incremental displacement according to the incremental encoder to the reference position given by the signal from the zero reference device.

Absolute encoders have the advantage that the position of the parts in relation to each other is known immediately after electrical power is applied to the encoder without moving or rotating the encoder disc shaft but they have the disadvantage that they require a large number of sensor channel tracks, basically one sensor channel track per bit of position in¬ formation. In addition, absolute encoders are sometimes provided with mechanical gear¬ boxes and supplementary encoder disc or discs in order to permit recording the angular position within several turns directly after power up. Incremental encoders have the disadvantages that after power up, the position is not known until the encoder disc shaft has been rotated to set the disc in the reference position and that they require a low noise data transfer in order to avoid that false position increment signals are detected, but they have the advantage that they require only one sensor channels track or possibly two in the case where a zero reference track is included. It is also possible to make an absolute encoder having only one incremental track and one absolute position track, see U.S. patent 5,068,529. The absolute position track here carries binary codes that are arranged in a special pattern, so that when a sufficient number of adjacent fields in the absolute position track are detected simultaneously, by using a plurality of sensors for this track, the absolute position is directly found from the detected 0 binary code. However, such a solution requires several light sensors and associated

CONFIRMATION COPY transmission lines, what obviously increases the complexity and cost of the encoder system . and increases its sensitivity to noise.

There is also known in the prior art a group of encoders that can be called quasi- absolute encoders and are related to the encoder system of the cited U.S. patent 5,068,529. Such encoders have at least one conventional incremental encoder track and another track that allows a determination of the absolute position of the encoder disc after only a small displacement such as a small rotation for determining a sufficiently long binary code. In U.S. patent US-A 5,231,596 the absolute position track has binary codes, each corresponding to a quarter of the pitch of the incremental code fields. The binary codes are made according to a special mathematical pattern. U.S. patent US-A 4,914,437 discloses an encoder strip having an absolute position track using chain codes, each bit corresponding to one field of the incremental tracks. SUMMARY

It is an object of the invention to provide an optical encoder that permits the acquirement of absolute position information after only a minor angular turn of an encoder shaft in rotary case or after only a minimal displacement of a linear encoder slide in the linear case and to achieve this without arranging a multitude of sensor channel tracks or a multitude of sensors.

It is another object of the invention to provide a device that permits, in the rotating case, sensing the angular position of part movable in relation to another part within several turns or sensing, in the linear case, the linear position of a part almost immediately after power up without having to rely on mechanical gearboxes and supplementary encoder disc or discs or other mechanical devices.

These objects are achieved by the invention, the characteristics of which appear from the appended claims.

Thus generally, the position detector is arranged to detect or determine or measure any of a plurality of positions of a first part movable in relation to a second part, the first part for example being a rotating disc attached to a motor shaft and the second part being a stationary detecting device. Generally a coding member is provided in the conventional way, the coding member comprising some suitably configured fields, areas or regions on e.g. the moving part. The coding member comprises a first row of incremental codes and a second row of absolute codes, these rows for the rotating case being circular rows having the same rotary axis as the rotating disc. The rows thus have extensions or longitudinal directions which are parallel to a direction of movement of the first part in relation to the 5 second part. One row, that can be taken as the first row, comprises in the conventional way regularly spaced, identically shaped incremental coding fields and the other row, the second row, is a row of absolute codes and comprises fields representing coded positional data, these data then informing directly on the position of the first part in relation to the second part, not requiring any direct position information from the fields of the first row. Between 0 neighbouring fields in such a row there are can be areas where the physical characteristic does not exist or has a zero level or value, these zero level areas for the first row then being regularly spaced and having all identical shapes. First detection means are provided for detecting, recording or measuring the physical characteristic of the fields of the first row of incremental codes. Second detection means detect or measure the physical characteristic of the fields of the second row of absolute codes.

The second detection means are in the conventional way arranged to provide a signal dependent on a physical characteristic of a field of second row detected by the second detection means. This physical characteristic is typically the areas of the fields or some dimension thereof such as the lengths for equally wide fields. Also, the fields can have a varying reflectivity or a variable transmission of light. The fields of the second row of absolute codes comprises fields having at least two different, non-zero levels of the physical characteristic, so that the fields of the second row then will be detected by the second detection means which then provide two different, non-zero signals. Together with the absence of a field, for example for areas located between the fields or for other fields, the position of which is provided by some other means such as the detection of the fields of the incremental row, this can give a non-binary detection situation, where the detection means then can provide at least three different signal levels when the coding member moves past the detection means. This will save space on the coding member but make the evaluation circuits a little more complicated than in the conventional case where a field is or is not detected, the detecting device basically providing only two signals, one for a field and an¬ other for the absence of a field. This latter detecting method is used for the first row of incremental codes.

In the typical cases the fields of the second row comprise fields having at least two different, non-zero areas or having at least two different, non-zero heights or lengths, these height or length dimensions being taken perpendicularly to the direction of movement or to the longitudinal direction of the rows.

Further, the positions of the fields of the two rows can be so arranged in relation to each other, that phase or position information obtained from the first detection means when reading the fields of the first row of incremental codes can be used to determine positions where the fields of the second row of absolute position codes are located, for example when using zero level fields as already indicated. This means that there can exist a fixed relationship between the locations of the fields of the two rows where this relationship can be used for by the second detection means for determining that at calculated time there should be a relevant signal detected. The fields in the two rows have preferably the same pitch, meaning that the fields are located in pairs, a field of a pair in a row being opposite to or adjacent to the other field of the pair in the other row, these two fields having the same position in the direction of movement.

The second row of absolute codes can also use fields having a height or length equal to zero, these fields being detected by the second detection means as providing a zero signal. Such zero fields can be used, since the positions of the zero fields are given by the positions of the fields of the first incremental code row.

Advantageously, the centres of the fields of the second row are located on a line ex¬ tending in the direction of movement or in the longitudinal direction of the row. When using the position detector for an electric motor, the coding member can have a third row of commutation codes, for providing commutation signals to the motor. The third row is then also formed in the coding member and extends in the direction of movement of the first part in relation to the second part. Also, this third row can comprise fields having different levels of a physical characteristic such as different areas, or heights or lengths as taken in a direction perpendicular to the direction of movement. Third detection means must then be arranged for detecting the fields of the third row of commutation codes.

Like the fields of the first and second rows, the fields of the third row also has the same pitch, so that for example the fields in the first and third rows are located opposite to or in parallel with each other or with the same spacing in the direction of movement, so that for a field in one row there is an adjacent field in the other row, where these two fields have the same position in the direction of movement.

Like the fields of the second row, the centres of the fields of the third row can be located on a line extending in the direction of movement.

The levels of the physical characteristic such as areas, or heights or lengths of the fields of the third row are preferably substantially a sine function of the displacement amount of one part in relation to the other part in the direction of movement. The third detection means can comprise a first, single detector for detecting simultaneously a multi¬ tude of adjacent fields in the third row, in particular two or three adjacent fields, for pro¬ viding a signal being a measure of the total area of the detected fields. The third detection means advantageously then comprise a second, single detector for detecting simultaneously a multitude of adjacent fields in the third row, in particular two or three adjacent fields, for providing a signal being a measure of the total area of the detected fields. The second detector has such a position in relation to the first detector, that the distance between the first and second detectors corresponds to a quarter of the period of the sine function. The commutation information can instead be contained in a continuous device, such at least one contour or edge line of the coding member, where this contour or edge line is generally wave-shaped or has the shape of a sine function as above. Then a point on the contour or edge line will generally have a varying position in a direction perpendicular to the direction of movement when the point moves in the direction of movement. The contour 5 or edge line can be the outer edge of one of the two parts which move in relation to each other. In another case, two contours or edge lines are provided which are two opposite edges of a transparent track having for instance its centre line in parallel with the centres of the fields of the incremental and absolute position rows. The continuous case comprises in its most general form that the commutation information is formed in the coding member and 0 has a longitudinal direction parallel with the direction of movement of the first part in relation to the second part. A region of the coding member that forms the information has a physical characteristic that varies substantially continuously in the direction of movement. The third detection means are then arranged for detecting the physical characteristic of the commutation information at an area having only a small extension in the direction of move- ment and thus in the longitudinal direction of the commutation information, so that the third detection means will output a substantially continuously varying signal, when the two parts move in relation to each other. A small extension means that the width of the detected area is small compared to the period of a function to be detected, such as a the period of a sine function like above. The coding member can also have a third periodic information track which can be used for commutation and comprises some physical quantity or characteristic that varies along the longitudinal direction of the track so that the level of the quantity or characteristic follows a curve substantially composed of a sequence of straight linear segments. In particular, the characteristic can be the transverse position of a contour that is partially linear, i.e. is substantially constituted of straight, connected segments. At least two geometrically separate, identical detection means such as mask windows having identical shapes are provided and associated light sources and light detectors. In order to obtain some information on the average level and the amplitude of the detected signals the detection means have such a location in relation to each other that one of the detectors provides a signal that is either close to or substantially at the maximum signal level or close to or sub¬ stantially at the minimum signal level, meaning that these detector means then are placed about 180° apart from each other, in regard of the period of the information track.

Advantageously at least three such geometrically separate, identical detection means are provided. Then two of these detection means have such location in relation to each other, that they provide output signals which are 180° apart in regard of the period of the information track. Summation means are provided for summing the output signals of said two detection means in order to obtain a signal representing an average level of the maximum and minimum signal levels output from the detection means. Thereby a reference level is obtained permitting signal processing means to adjust for drift of the detection systems due to e.g. contamination, ageing or temperature variations.

For the third information track, signal processing means can use a reference level, that is the sum of the output signals of suitable ones of the at least two detection means, to estimate the average signal level. Further, they can use this average level and that signal level in the signals output from the detection means which deviates most from the reference level to estimate the amplitudes of the signals output from detection means to adjust for drift of the detection systems due to e.g. contamination, ageing or temperature variations. BRIEF DESCRIPTION OF THE DRAWINGS

The invention will now be described by way of non-limiting embodiments with reference to the accompanying drawings in which: Figure 1 is a fragmentary axial view of an encoder disc and an associated mask disc, where a part of the mask disc is removed,

Figure 2 is an axial view of an encoder disc comprising a binary coding of the absolute position,

Figure 3 are waveform diagrams illustrating the signals obtained from the commutation channel of the device of figure 2, and

Figure 4 is schematic sectional view illustrating light detecting devices. DETAILED DESCRIPTION

Figure 1 shows an axial view of a portion of an encoder disc 104 which is partly hidden behind a masking device 105 that covers the left part of the figure. The encoder disc 104 has three sensor channel tracks 101 - 103 which thus are concentric patterns of fields having different measurable amounts of a physical characteristic such as the absorbtion, transmission or reflectivity of light of some suitable wavelength. For example, as assumed in the following, the fields may comprise cut-outs having different sizes and in particular slots having all the same angular width but different heights or radial lengths allowing different amounts of the light to pass therethrough.

The innermost track 103 is a conventional 500 line incremental encoder track and can be treated as is common in incremental encoder designs. It comprises identically shaped, substantially rectangular windows which are elongated in the radial directions and are uniformly distributed in the circumferential direction meaning that the spacing of adjacent windows equals an angle of 360°/500. The material between adjacent slots can be assumed to have the same width in the circumferential direction as the slots, meaning that the slots have an angular width of half the spacing, i.e. l/2 (360°/500). All tracks have the same distribution of slots in the circumferential direction meaning that the angular width of the slots in different tracks is always the same, i.e. l/2 (360°/500). Further, the slots in a track have their centres located on the same circle concentric with the disc 104. Moreover, the slots in different tracks are placed along the same radii, so that e.g. the same radius of the encoder disc 104 extends centrally through one window of each track 101 - 103. Thus, within an angle interval of l/2-(360°/500) there is one slot from each track.

The embodiment shown of the incremental encoder track uses the most common decoding principle comprising four groups 120 - 123 of mask windows. Each group consists of seven identically shaped, rectangular windows, which have half the radial lengths compared to the slots in the track 103 and which have the same circumferential spacing as the slots in the incremental track 103 to be detected. The first two groups 120, 121 are located so that the windows therein cover the innermost half of the track slots and the 5 windows of the second two groups 122, 123 cover the outer half of the track slots. The windows of the two inner groups 120, 121 are further displaced a little in relation to each other from the equal spacing of the track windows, the displacement corresponding to a 90° offset in the detection signal, meaning that the spacing between the windows of the two groups equals an angle of (n + l/4)-(360°/500) where n is an integer, n >_ 1, e.g. n = 1. 0 The same spacing in relation to each other is true for the windows in the two outer groups 122, 123, however the windows of these groups are further spaced in relation the windows of the two inner groups 120, 121 corresponding to an angle offset of 180° in the detected signal meaning that the angle is l/2-(360°/500).

When the disc 104 moves behind the masking windows of an inner groups 120, 121 and of an outer group 122, 123, where these two groups have this latter angular displacement of 1/2- (360^*7500), on the masking device, light detectors, see figure 4, behind the mask groups will detect light intensities that thus are basically 180° offset in phase. The difference between these two signals is processed to give a net signal, the average value of which is fairly independent of ageing phenomena of light sources, etc. In this way two net signals which are 90° offset in phase are obtained from the differences of the intensities sensed through a first pair of an inner mask group 120 and an associated outer mask group 122 and through a second pair comprising an inner mask group 121 and an associated outer mask group 123.

The outermost track 101 is primarily intended for the commutation of a brush-less electric motor, not shown. It contains a set of rectangular, elongated slots or windows having their longitudinal directions located radially, the heights h of the slots obeying a relation h = hø 4- h j -sin(p/2-</>), where φ is the mechanical angle of the encoder disc 104 and p is the number of poles of the motor to be commutated and hø and h_j are constants. The case shown in figure 1 corresponds to a motor having 20 poles. For the outermost track 101 , at power up, the system detects the light intensity as obtained through a first rectangular window 107 and a second rectangular window 108, two opposite straight sides of the windows being located substantially radially and the other two opposite sides located substantially in the same tangential direction of the disc 104. Each of these windows has a width covering three slots in the considered track and a height corresponding to a little more than the largest height of a slot in the track 101. The two detection windows 107, 108 are located so that the signals which are detected for each window 107, 108, for a movement of the disc 104, will be 90° out of phase, meaning that the actual mechanical angle separating the detection windows 107, 108 is (l/4) ^« 360°/(20/2) for the twenty motor pole case. Assuming that the system has a good estimate of the signal amplitude that can be expected when a set 116 of three adjacent slots having maximum heights faces one of the detection windows 107, 108, it can calculate the angular position within one commutation period and thus determine the suitable commutation for a motor, not shown, to which the encoder disc 104 is attached. A good estimate of the signal amplitude that can be expected when a maximum height slot set 116 faces a window detection 107, 108 can often be obtained provided that the temperature of semiconductor light receivers and light trans¬ mitters, see figure 4, used for the detection of light through the slots and windows, are known. In the case where no good estimate of the signal amplitude that can be expected when a maximum height slot set 116 faces a mask window 107, 108 can be obtained, a third window, not shown, can be arranged, which has a shape identical to that of the detection windows 107, 108 and is located so that the signal detected for this window is 180° away one of the detection windows 107, 108 and is 90° away from the other one of the detection windows 107, 108. The average amplitude from two detection windows which are located so that the signals therefrom are 180° apart should then be constant 5 independently of the angular position of the disc 104.

The middle track 102 occupies approximately the same radial space that is normally required for the zero patterns of a conventional incremental encoder. Track 102 does however utilise the available space much more efficiently. Track 102 is read through a rectangular mask window 106 that has a radial height corresponding to a little more than the 0 maximum height of the slots in the track 102 and a width in the circumferential direction equal to a little more than the width of a slot in the track to be read. Thus, detecting the light through the mask window 106 the slots are read one by one. There are a small number of different heights of the slots in track 102, and the heights thereof can be measured by sensing the amount of light that passes the mask window 106.

15 In this track 102 six different slot heights are shown. Slots 115 having one of these heights, the largest of the heights used, are used as sync characters. Slots having other smaller heights are used to represent different digits. In the case shown, the five digits 0 to

4 are represented by the slots 110, 111, 112, 113 and 114 which thus all have different heights. Three such other slots are placed between neighbouring sync slots 115, so that

20500/4 = 125 sync slots 115 are arranged in the track and the sync slots thus being located with an angular spacing of 360°/125. In the case shown thus, this also gives a total of 5-5-5

= 125 codes between neighbouring sync slots 115, these codes then corresponding to 125 different reference positions obtained between this angular interval of 360°/ 125.

When the system is started after power down, a movement of slightly more than 25 1/ 125th of a full turn is required to find the absolute position of a full turn. The servo driver, not shown, controlling the system will at initialisation move the encoder disc 104 in some direction, for example to the left. As the positions of the track 102 slots have the same pitch as the slots in track 103, a track 102 slot will be in front of the 106 window whenever the track 103 is in a certain phase, for example when the difference signal from 30 windows 120 and 122 gives a maximum signal and (consequently) the difference signal from windows 121 and 123 is zero.

When the servo system has moved the disc 104 so much that the signals from track 103 satisfy the condition given above, the system reads the intensity through the window 106, the intensity depending on the height of the track 102 slot in front of window 106. 35 Assume for example that disk slot 113 happens to be slightly to the right of the mask window 106 at power up. If so, the first slot to be measured is slot 113. Assume that the reading is 0.121 units and that a further turning of the disc 104 will give amplitudes of 0.052 from slot 118, 0.046 from slot 117 and 0.500 from slot 115. As each fourth slot of track 102 is a sync slot having the maximum amplitude, the largest signal from four 40 consecutive readings can be assumed to be from a sync slot. In figure 1, the transmission for a "0" slot like 110 is 10% of that of a sync slot; the transmission for a "1 " slot like 111 is 15% of that of a sync slot; the transmission for a "2" slot like 112 is 22% of that of a sync slot; the transmission for a "3" slot like 113 is 34% of that of a sync slot; and that of a "4" slot like 114 is 50% of that of a sync slot. Dividing the obtained values by 0.500 (equal to the largest value read) will give a relative amplitude of 0.242 for slot 113, which will be interpreted as a "2" (22% nominal). The two slots 117 and 118 will in the same way be inteφreted as being "0". These data will identify the 102 track as being at position 2- 1 + 0-5 + 0-25 or position No. 2 of the 125 positions around the disc. As the track 102 is synchronised with the 500 line encoder patterns of track 103, the absolute position of the system is now as well defined as it would have been after having found a conventional, prior art zero pulse.

The small width of the mask window 106 will only permit a small amount of light to pass. This will not however in most applications raise any problem, since most systems will read the absolute encoder track 102 only at initialisation and then the speed will necessarily be very low as the movement is starting from zero speed and the controller only has to rotate the shaft of the disc some 1/125 of a turn. A very high gain, low bandwidth sensor system can therefore be used.

The commutation track 101 is basically only read at power up and at standstill, and the necessary bandwidth is therefore also very low and the sensitivity of the system used for detecting it can be kept very high.

The commutation track 101 can be used together with a low power battery supported electronic system to trace the movements, if any, during power off. The light source can either be pulsed or use a very low constant light intensity. In the application shown, one revolution will give 10 periods of the commutation track signals. Therefore even a very low frequency system can trace the normally rather slow movements expected during power down. This permits space saving electronic devices for replacing the mechanical gearboxes and multiple encoder discs normally required for multiturn absolute encoders.

The embodiment shown concerns a rotary encoder. However, the basic principle of an absolute position code track like 102 having a plurality of different heights is also applicable to linear encoders, thus permitting an absolute position to be obtained after an encoder movement of only a fraction of a mm. Using a 100 μm encoder track period and a four digit position number using 8 different sloth heights, the absolute position along a 1.6 meter linear path would be known after a movement of some 450 μm.

The commutation/power off track 101 shown having discrete fields is designed to be produced using etched metal discs or linear strips. If a plastic or glass disc is used, the track can instead be made as a continuous track having a varying height or radial width along the circle, that is in a direction perpendicular to the direction of movement.

Another embodiment is illustrated in figure 2 where a disc 204 made for example of etched or deposited nickel is shown. Here, the outer periphery of the disc 204 gives a commutation track 201 similar to that (101) of figure 1. The track 202 used for absolute position is in this case made with only 32 segments, and the segment number is binary encoded using a low slot like 205 for a " 1 " and no slot for a "0". The positions of the "no slots" is given by position of the respective field in the incremental position track 203. The sync characters comprise a high slot like 206. The masking device is partially shown at 5210. The windows used for detecting the conventional incremental track 203 are not shown. The window 211 is used to read the absolute encoder track 202, and in the position shown a " 1" slot like 205 is shown behind this window 211. Three other mask windows 212 - 214 are used to read the position of the periphery commutation slot 201. Two of these windows 213 and 214 have such a distance from each other, that the signals obtained, when detecting io through these windows, have a 180° phase offset; in the position illustrated in figure 2, the window 213 is almost quite open whereas the associated window 214 is almost quite closed, covered by the marginal portion of the rotary disc 204. The third mask window 212 has such a distance from the other two windows 213, 214, that the signal obtained from this window is 90° out of phase compared to the signals detected in the other two windows 213, 15 214, when the encoder disc is rotated.

Figure 3 shows output signals 301, 302, 303 as functions of time or displacement. These output signals are obtained from three sensors using windows having positions like the three windows 212 - 214 of figure 2, and thus the signals have phase positions of 0, 90 and 180° in relation to each other. The contour shown in figure 2 of the commutation 20 contour or edge line 201 comprises straight line segments forming a profile somewhat similar to a sine curve, the profile thus having a definite period. The profile contains basically four differently positioned or shaped linear segments 304 - 307, one 307 of which has a constant level corresponding to the minimum value of the signal and one 305 has a constant level equal to the maximum value of the signal. The other two types of straight 25 line segments 304, 306 have the same slopes but with different signs, i.e. these segments are substantially mirrored pictures of each other, and they interconnect the other two types of segments. This makes it possible for the evaluating electronic circuits to identify the signal levels based on the fact that at least one of the channels will be at or close to the maximum or minimum level; this permits the use of temperature dependent sensors having 30 no or a limited accuracy for varying temperatures. In order to obtain an even more simple evaluation criterion, four windows, which are located so that the detected signals therefrom have phase offsets of 90° from each other, can be used. In such a case one of the four sensors using these windows will provide a signal having a level close to the maximum signal level and another sensor will provide a signal having a level that is close to the 35 minimum signal level. Alternatively, the average of the signals from two sensors, which are located so that their output signals are 180° apart, can be used to constitute a reference level that is independent of the disc position.

Two of the output signals from the commutation track like 301 and 302 as obtained from the corresponding mask windows 212 and 214 can be used as a low frequency or low

40 pitch incremental encoder. During power down, low bandwidth electronic circuits powered by a low power battery can track the slow motions expected during system power down.. After power up, a minor movement that permits the system to read a whole position data word will restore the position information without having to rely on mechanical gearboxes and supplementary encoder disc or discs.

5 In figure 4 the arrangement of light sources and light sensors is schematically shown.

Light sources 401 having optical focusing systems such as lenses 402 are arranged at the side of the movable coding disc or coding strip 403. The light from the light sources is stopped by the coding device 403 or passes through slots therein, further through windows in the stationary mask plate 404 to reach the surface of optical sensors or light detectors

10405. The signals from the optical sensors 405 are through signal lines 406 provided to a signal processing unit 407 evaluating the signals received for producing relevant position data.

Claims

1. A position detector for detecting any of a plurality of positions of a first part movable in relation to a second part, comprising a coding member having a first row of incremental codes and a second row of 5 absolute codes, both formed therein in a direction of movement of the first part in relation to the second part, the first row comprising regularly spaced incremental coding fields and the second row comprising fields representing coded positional data, first detection means for detecting the fields of the first row of incremental codes, second detection means for detecting the fields of the second row of absolute codes, o characterized in that the second detection means are arranged to provide a signal dependent on a physical characteristic of a field detected by the means, and that the fields of the second row of absolute codes comprises fields having at least two different, non-zero levels of the physical characteristic, the fields of the second row then ie being detected by the second detection means as providing two different, non-zero signals.

2. A position detector according to claim 1 , characterized in that the physical characteristic is the area of a field, the second detection means thus being arranged to provide a signal dependent on the area of a field detected by the means and the fields of the second row comprising fields having at least two different, non-zero areas.

20 3. A position detector according to claim 1 , characterized in that the physical characteristic is the height or length of a field, as seen perpendicularly to the direction of movement, the second detection means thus being arranged to provide a signal dependent on the height or length of a field detected by the means and the fields of the second row of absolute codes comprising fields having at least two different, non-zero heights or lengths,

25 as seen perpendicularly to the direction of movement.

4. A position detector according to one of claims 1 - 3, characterized in that the positions of the fields of the two rows are so arranged in relation to each other that phase or position information obtained from the first detection means when reading the fields of the first row of incremental codes provide information to be used for selecting or determining

30 positions of the fields of the second row of absolute position codes.

5. A position detector according to one of claims 1 - 4, characterized in that the fields in the two rows have the pitch, so that for a field in one row there is an adjacent field in the other row, these two fields having the same position as taken in the direction of movement.

35 6. A position detector according to claim 5, characterized in that the second row of absolute codes also comprises fields having a height or length equal to zero, these fields being detected by the second detection means providing a zero signal.

7. A position detector according to one of claims 1 - 6, characterized in that the centres of the fields of the second row are located on a line extending in the direction of

4o movement.

8. A position detector according to one of claims 1 - 7, characterized in that the coding member has a third row of commutation codes, for providing commutation signals to an electric motor, the third row also being formed in the coding member in the direction of movement of the first part in relation to the second part, the third row comprising fields having different levels of a physical characteristic as taken in a direction perpendicular to the direction of movement, third detection means being arranged for detecting the fields of the third row of commutation codes.

9. A position detector according to claim 8, characterized in that the physical characteristic is the areas of the fields.

10. A position detector according to claim 8, characterized in that the physical characteristic is the heights or lengths of the fields as taken in a direction perpendicular to the direction of movement,

11. A position detector according to one of claims 8 - 10, characterized in that the fields in the first and third rows have the same pitch or spacing in the direction of movement, so that for field in one row there is an adjacent field in the other row, these two fields having the same position in the direction of movement.

12. A position detector according to one of claims 8 - 11, characterized in that the centres of the fields of the third row are located on a line extending in the direction of movement.

13. A position detector according to one of claims 8 - 12, characterized in that the levels of the physical characteristic of the fields of the third row are substantially a sine function of the displacement amount of one part in relation to the other part in the direction of movement.

14. A position detector according to claim 13, characterized in that the third detection means comprise a first, single detector for detecting simultaneously a multitude of adjacent fields in the third row, in particular two or three adjacent fields, for providing a signal being a measure of the total area of the detected fields.

15. A position detector according to claim 14, characterized in that the third detection means comprise a second, single detector for detecting simultaneously a multitude of adjacent fields in the third row, in particular two or three adjacent fields, for providing a signal being a measure of the total area of the detected fields, the second detector having such a position in relation to the first detector, that the distance between the first and second detectors corresponds to a quarter of the period of the sine function.

16. A position detector according to one of claims 1 - 7, characterized in that the coding member comprises commutation information for providing commutation signals to an electric motor, the commutation information being formed in the coding member and having a longitudinal direction parallel with the direction of movement of the first part in relation to the second part and a region of the coding member having a physical characteristic that varies substantially continuously in the direction of movement, third detection means being arranged for detecting the physical characteristic at an area having a small extension in the direction of movement and in the longitudinal direction of the commutation information, so that the third detection means output a substantially continuously varying signal, when the two parts move in relation to each other.

17. A position detector according to claim 16, characterized in that the commutation information comprises at least one contour or edge line, so that a point on the contour or edge line has a varying position in a direction perpendicular to the direction of movement when the point moves in the direction of movement.

18. A position detector according to claim 17, characterized in that the contour or edge line is the outer edge of one of the parts.

19. A position detector according to claim 17, characterized in that two contours or edge lines are provided constituting two opposite edges of a transparent track.

20. A position detector according to claim 19, characterized in that the track has its centre line extending in parallel with the centres of the fields of the two rows.

21. A position detector according to one of claims 16 - 20, characterized in that the physical characteristic of the region comprising the commutation information varies as substantially a sine function of the displacement amount of one part in relation to the other part in the direction of movement.

22. A position detector according to claim 21, characterized in that the third detection means comprise a first detector for detecting at a first area having a small extension in the direction of movement and a second detector for detecting at a second area having a small extension in the direction of movement, the second detector having such a position in relation to the first detector, that the distance between the first and second detectors corresponds to a quarter of the period of the sine function.

23. A position detector according to one of claims 1 - 7, characterized in that the coding member has a third periodic information track comprising a contour that is substantially partially linear, i.e. is substantially constituted of straight, connected line segments, and that at least two geometrically separate, identical detection means are provided, the detection means having such a location in relation to each other that one of the detectors provides a signal that is either close to or substantially at the maximum signal level or close to or substantially at the minimum signal level.

24. A position detector according to claim 23, characterized in that at least three geometrically separate, identical detection means are provided, two 5 of which having such location in relation to each other, that they provide output signals which are 180° apart in regard of the period of the information track, and summation means being provided for summing the output signals of said two detection means in order to obtain a signal representing an average level of the maximum and minimum signal levels output from the detection means, thereby providing a reference level 0 permitting signal processing means to adjust for drift due to contamination, ageing or temperature variations.

25. A position detector according to one of claims 23 - 24, characterized in that signal processing means are arranged to use a reference level, the reference level being the sum of the output signals of suitable ones of the at least two detection means, to estimate s the average signal level and to use that signal level in the signals output from the detection means which deviates most from the reference level to estimate the amplitudes of the signals output from detection means to adjust for drift due to contamination, ageing or temperature variations.