US 4,736,187 AGrant

Encoder for Measuring Both Incremental and Absolute Positions of Moving Elements

Issue Date:1988-04-05

•69 Claims

•16 Drawing Sheets

Abstract

An optical encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member. For example, three stationary light sources may be supported by the stationary member for emitting three narrow light beams on respective paths. An encoding band may be supported by the movable member in the paths of the light beams. The band has a timing and incremental encoding track comprising at least one series of alternating light-blocking stripes and light-passing stripes, and a data track comprising a series of absolute position tags alternating with spacers, each tag being framed by start bits. A read head may be supported by the stationary member and has three optical sensors, one for each light beam, two preferably in quadrature with each other for the timing and incremental encoding track, and one for the data track. Each optical sensor includes light sensors for generating an analog signal voltage. A decoding system actuated by these analog signal voltages determines the incremental and absolute position of the movable member relative to the stationary member.

Metadata

Assignee

The Regents of The University of California

Inventors

Robert I. Kibrick
Calvin R. Delaney

Application Information

Application Number:US 9381643

Filing Date:1986-12-04

Priority Date:1986-12-04

Art Unit:217

Classifications

IPC:

H03M 122H03M 130

Field of Search:

340250347 P231 SE;237 G

Patent Drawings (16 sheets)

Description

This invention relates to an encoder for measuring both the incremental and absolute positions of moving elements, including rotating members.

The encoder of this invention is especially suitable for measuring the azimuthal position of telescope domes. It provides an inexpensive and highly reliable dome-position encoder which has few moving parts and which eliminates the need for a mechanical engagement to the telescope dome. It may also be used to measure accurately the angular or linear position of not only comparably sized moving or rotating objects, such as hangar doors, conveyor belts, and elevators, but also may be used as part of a standard rotary or linear encoder.

Since telescope domes provide an excellent example of how the invention may be used, the application of the invention to such apparatus will be used as the prime demonstrative example in this application, but without limiting the invention to the prime example.

BACKGROUND OF THE INVENTION

A number of techniques have been used to measure the position of large rotating objects such as telescope domes. The most common approach has been a mechanical coupling between a commercial rotary position encoder and the moving surface. Thus, telescope dome position has traditionally been encoded by mechanically coupling a rotary encoder to the edge of the dome.

For example, a pinch roller has been pressed against an inner circumference of the dome, to rotate as the dome revolved. However, if this pinch roller slipped, large errors in position measurement resulted. An alternative approach to the problem has been to attach a chain to the inner circumference of the dome, the chain then driving a sprocket. In either case, the rotating axle of the pinch roller or the sprocket has been used to turn the shaft of a rotary position encoder.

Both incremental and absolute position encoders have been used. When an absolute position encoder was used, a custom-made gear reduction was usually required, so that one rotation of the dome would produce one rotation of the absolute position encoder. Such a custom gear reduction often proved to be more expensive than the absolute position encoder itself. It also tended to introduce errors in position measurement, due to mechanical backlash, and to inexact gearing. When the gear reduction was not exact, the encoder incurred a small incremental error on each rotation; then when the dome was rotated repeatedly in the same direction, this error would accumulate and grow unacceptably large.

While these mechanical couplings sometimes worked well for many years, they could become and eventually became unreliable, especially after decades of mechanical wear. Older domes that shook, wobbled, and nutated as they revolved caused these parts to wear out even faster. For example, the 100-year-old dome on the 1-meter Nickel Telescope and the 30-year-old dome on the 3-meter Shane Telescope, both at Lick Observatory, have suffered from unreliable dome pointing, caused by worn mechanical couplings between the position encoder and the dome.

If an incremental position encoder is used, then some other mechanism is required to establish an initial absolute position. This has often been accomplished by a switch which was tripped whenever the dome was rotated to an index or reference position. However, for a large dome, it can take a considerable period of time to rotate the dome to its initializing position.

In either case, while one can easily obtain commercial position encoders (either absolute or incremental) that are by themselves relatively accurate, reliable, and inexpensive, the accuracy, reliability, and economy of the position measurement that is ultimately achieved is often compromised by the inaccuracy, unreliability, and expense of the mechanical coupling between the encoder and the dome.

Even if an inexpensive and reliable mechanical coupling could be found, currently available rotary encoders do not provide redundancy and are not self-diagnosing. They can and do malfunction, and there is often considerable delay in determining that a malfunction has occurred. In the meantime, valuable observing time is lost when the dome does not point in the proper direction.

Other techniques have been employed, such as having a series of separate switches spaced at regular intervals around the circumference of the stationary part of the dome building; the separate switches were then tripped by a single detent on the moving part of the dome. This method required a tremendous number of switches and wires, and so it was quite costly and complex to maintain. To provide angular resolution comparable to that provided by the present invention, even on an average size dome, would require several thousand separate switches. Clearly, this technique does not offer the requisite economy, accuracy, and reliability.

One object of the invention is to solve these problems and prevent their recurrence.

Because of these problems, various schemes of optical encoding have been tried or considered. All of these schemes involved placing one or more bands of stripes or codes around an inner surface of the rotating portion of the dome. These codes would be sensed by one or more optical sensors attached to the fixed portion of the dome. One such scheme, that of Calvin Delaney in 1979, was a single-track incremental encoder. A similar scheme was used in 1971 for an incremental encoder on the dome of the 40-cm photometric telescope of the Observatoire de Haute Provence, France. Other optical-based schemes that were considered would have used a wider coded band to make a multi-track Gray-code absolute encoder. However, none of these went into general use because of their excessive electronic and mechanical complexity and their susceptibility to skewing errors.

One proposed idea for solving these problems was to use inexpensive commercial fixed-beam bar-code readers to encode the absolute dome position. The idea was to place bar-code symbols around the inside of the rotating part of the dome, with the absolute position at each point encoded in the corresponding bar-code symbol. As the dome rotated, the bar-code symbols would be drawn past a fixed-beam bar-code reader, which would decode them and provide the absolute position of the dome.

Unfortunately, standard bar-code symbols cannot be scanned this way. Although such symbols can be scanned at many different speeds and from either direction, the speed and direction during any single scan must remain reasonably constant. Since the dome speed or direction can change at any time, the normal motions of the dome cannot be used reliably to scan standard bar-code symbols.

Another object of the present invention is to overcome the problems that are inherent in prior-art optical encoders.

A further object is to provide accurate position encoding for both incremental and absolute indications.

Another problem is presented by the fact that many domes are out-of-round. For example, each of the domes at both the 1-meter and 3-meter Lick Observatory telescopes is out-of-round by more than one inch, and this eccentricity exceeds the depth of focus of most inexpensive optical sensors. Further, besides being out-of-round, the encoder track surface at the dome of the 1-meter telescope at Lick Observatory exhibits considerable vertical warping.

An object of this invention is to overcome the problems presented by out-of-round domes and by vertical warping.

A further object is to provide an encoder system having self-diagnosis capabilities, and a related object is to provide such a system with automatic detection of errors.

Other objects and advantages of the invention will appear from the following description.

SUMMARY OF THE INVENTION

The present invention provides both incremental and absolute position encoding. As related to telescope domes it uses the dome itself as a major component of the encoder. In the preferred form of the invention, the encoding is done optically. A band (or series of bands) of reflective material imprinted with a coded pattern of stripes is attached to an inner surface of the rotating portion of the dome. This pattern is made up of two separate tracks: a timing track, which is used in encoding incremental positions, and a data track, which is used to encode the absolute position. The timing track, adjacent to the data track makes the code on the data track self-clocking. Similar effects can be obtained by using electrically conductive bands and brush contacts.

As the dome rotates, these two tracks move past a read head attached either to a stationary part of the telescope housed inside the building or to a stationary part of the building itself. The read head contains three optical sensors that detect changes in reflectivity as the stripes move past the sensors. Two of the three optical sensors scan the timing track, and the horizontal spacing between them is adjusted so that the signals they produce are in quadrature. By comparing the phase relationship between the signals from these two sensors, the direction of dome motion is determined. The amount of motion is given by counting the number of signal transitions, as in most commercial incremental encoders.

As an alternative, the two tracks may be on the stationary member and the read head supported on the movable member.

Each sensor preferably pairs a light-emitting diode with a phototransistor. The light-emitting diode (LED) emits a small infrared beam that, in one form of the invention, is either reflected or absorbed by the light or dark marks on the coded band. The phototransistor measures the reflected light and generates an analog voltage proportional to the intensity of the reflected beam. In another form of the invention, the light is passed through transparent portions of the band or is barred by opaque portions. Thus, in general the optical encoder of this invention uses bands with light-blocking stripes or light-passing stripes.

The resulting analog signals may be used directly with some types of computers but generally are converted to digital TTL levels by a separate operational amplifier and a Schmitt trigger for each phototransistor output. These three TTL levels are fed to separate input lines on an input/output port of a microcomputer, which then computes the dome position from these inputs.

Although it is preferable (for economy and flexibility) to use a microcomputer to process the sensor-derived TTL levels, these functions can also be performed by using conventional digital logic circuits. To simplify the description of these functions, they are described in terms of digital logic circuits, although, preferably, these functions are actually performed by the microcomputer.

In this instance, the TTL levels generated from the two timing track sensors are fed into the computer at a portion programmed to function as a phase comparator; the phase comparator outputs a direction level signal and a clock pulse corresponding to each incremental tick of motion. The computer is so programmed that these outputs control an "up/down counter" function, which measures the incremental position. The computer is also programmed so that the direction level signal and clock pulse control two "shift register" functions that process the signals from the data track. The "up/down counter", "shift registers", and some other registers may be located in the microcomputer's RAM. The logic for the phase comparator and for the clocking and gating of the various registers is implemented in software.

The third optical sensor scans the data track, which consists of discrete absolute-position tags separated by spacers, which may appear as areas of blank space. Each tag is encoded as a serial bit stream, and framed by start bits on either side. Each tag is decoded by the software into a corresponding absolute position. Preferably, the tags may be used as indices to access a look-up table of absolute positions, which is in the microcomputer's memory.

As stated before, the invention is applicable generally to both rotary and linear encoders. The light beam may be reflected, as described above, or the light source and detector may be on the opposite sides of the timing and data tracks, and the reflective portions are then replaced by transparent portions. There are still light-blocking portions (dark or opaque or non-reflective) and light-passing portions (transparent or reflective). The tracks, instead of being on a cylindrical surface, may be on a flat disk for a rotary encoder. For a linear encoder, they will be on a flat surface, whether reflective or transparent. Instead of a single timing track with the light sources and sensors spaced to be in quadrature, it may be better in some circumstances to use two timing tracks themselves spaced or coordinated with the light system to be in quadrature.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a view in elevation of the exterior of an observatory telescope dome building, showing the window or gap in the dome.

FIG. 2 is a view in elevation and in vertical section taken along the line 2--2 in FIG. 1.

FIG. 3 is an enlarged fragmentary view in perspective of a portion of the dome building, showing portions of the data and timing tracks and a portion of the read head.

FIG. 4 is a view in side elevation, partly in section, of the read head and its mounting and a portion of the timing and data tracks.

FIG. 5 is a view in front elevation of the read head and its mounting.

FIG. 6 is a top plan view of the read head and its mounting and a portion of the timing and data tracks.

FIG. 7 is a block diagram of the encoder system.

FIG. 8 is a diagrammatic view of a portion of the timing and data tracks showing how quadrature of the timing track may be achieved by the sensors and its relation to the data track and its sensor.

FIG. 9 is a block diagram of the functional operation of the microcomputer of FIG. 7.

FIG. 9A is a continuation of FIG. 9, showing the fault detection system therefor.

FIG. 10 is a detailed diagram of the framing gating logic in the microcomputer of FIG. 9.

FIG. 11 is a simplified view in elevation and in section of an observatory dome having an alternative form of encoding system embodying the principles of the present invention.

FIG. 12 is an enlargement of a portion of the dome of FIG. 11 and pertinent encoding system.

FIG. 13 is a view similar to FIG. 11 of a dome with a modified form of encoding system according to this invention.

FIG. 14 is an enlargement of a portion of FIG. 13.

FIG. 15 is a view similar to FIGS. 11 and 13 of a dome with yet another form of encoding system embodying the invention.

FIG. 16 is an enlargement of a portion of FIG. 15.

FIG. 17 is a plan view of a glass disk having a pattern for an incremental-absolute optical rotary encoder embodying the principles of the invention.

FIG. 18 is a plan view like FIG. 17 of a glass disk with a modified pattern for the rotary encoder.

FIG. 19 is a diagrammatic view in elevation of the glass disk of FIG. 18 in combination with a light source and a sensor.

FIG. 20 is a view like FIG. 18, showing another modified pattern.

FIG. 21 is a view like FIG. 20 of a disk like that of FIG. 20 except that the data track is different in order to accommodate operation of its phase comparator at 2.times. the frequency used for the disk of FIG. 20.

FIG. 22 is another similar view with a data track for use with a phase comparator operated at 4.times. the frequency used for the disk of FIG. 20 .

DESCRIPTION OF THE PREFERRED EMBODIMENT

The dome 22 (FIGS. 1 and 2)

To exemplify the invention, FIG. 1 shows a telescope dome building 20 comprising a stationary cylindrical structure 21 with a rotatable dome 22 having a window, slit, or gap 23 with which a telescope 24 is aligned, the remainder of the dome 22 excluding unwanted stray light and affording shelter. Inside the building 20, as shown in FIG. 2, is the telescope 24 rotatably and pivotally mounted on an upright member 25 that is stationary and is affixed to a floor or foundation 26.

Inside the building 20 is the telescope 24, which moves independently of the rotatable dome. It is very important that they both can be properly aligned so that the telescope 24 can look through the dome slit 23. The basic purpose of the encoder of this invention is to know the absolute position of the slit or window 23 at all times, so that the independent motion of the dome 22 can be coordinated with that of the telescope 24. Since the dome 22 and the telescope 24 may move within different coordinate reference frames, coordinated motion usually requires that they be driven at different rates of speed.

As shown in FIG. 2, the dome 22 has a lower circular edge 27 mounted on rollers 28 that ride on a circular track 29 at the upper edge of the stationary cylindrical structure 21.

The encoder tracks 32 and 33 (FIG. 3)

As shown in FIGS. 2 and 3, the dome 22 itself is a major component of the encoder of this invention. A band 30 imprinted with a coded pattern of stripes is attached to a generally cylindrical inner surface portion 31 of the rotating dome 22. This pattern is made up of two separate tracks: a timing track 32, which is used for encoding incremental positions, and a data track 33, which is used to encode the absolute position.

The timing track 32 may be a continuous, dome-encircling series of identical light-reflecting rectangles 34 alternating with identical, non-reflecting dark rectangles 35, preferably identical in size to the rectangles 34. The rectangles 34 and 35 are narrow stripes that may be numbered adjacent to one edge of each dark stripe 35. They may be about 5 mm. wide and 25 mm high. The stripes 34 may be made of highly reflective metal foil.

The data track 33 is adjacent to the timing track 32. The light-reflecting rectangles 34 and non-reflecting dark rectangles 35 on the timing track 32 define a corresponding series of bit positions on the data track 33. The width of each bit on the data track 33 equals the spatial width of each incremental tick on the timing track, which is in turn determined by the operation of a phase comparator 71 shown in FIG. 9 and discussed in a later section of this description.

A reflective (or light-passing) data bit on the data track 33 is treated as a "one", while a non-reflective (or light-blocking) data bit is treated as a "zero". This convention of light-passing meaning "one" and light-blocking meaning "zero" is arbitrary, and can be reversed should that prove convenient.

The data track 33 is divided into a series of distinct absolute-position tags 36. Each tag is composed of a fixed number of consecutive data bits. The tags 36 are separated by spacers 37, each composed of a minimum number of consecutive data bits. The number of bits in the tags 36 need not be the same as the number of bits in the spacers 37, although they may be the same. These numbers are a function of the size and resolution of a given encoder.

Each tag 36 is composed of a unique bit sequence that distinguishes it from all other tags 36 and spacers 37. Each tag 36 is also surrounded on either side by start bits 38 or 39 which are used to insure a transition (i.e., "one" to "zero" or "zero" to "one") at the boundary between the tag 36 and its neighboring spacers 37.

Each spacer 37 is composed of an identical bit sequence so that all the spacers look alike. The specific bit sequence used for the spacers 37 must not be contained within (i.e., embedded in) the bit sequence of any of the tags 36. Preferably, the spacer bit sequence may be specified as a sequence of consecutive "one" bits so that each spacer will appear as a solid area of blank (i.e., reflective or light-passing) space in between each tag 36 on the data track 33. In this case, any bit sequence containing at least as many consecutive "one" bits as contained in the spacer bit sequence cannot be used as a tag 36. For example, if the spacer bit sequence consists of the sequence "111111", then any bit sequence containing six or more consecutive "one" bits (such as the sequence "01011111110010"), cannot be used as a tag 36.

This arrangement of tags 36, start bits 38 and 39, and spacers 37 simplifies detection of the tags 36 from either direction and insures that it is always possible to recognize when a spacer 37 has become properly aligned (or framed) within the shift registers described below. Once a spacer 37 is properly framed, the adjacent tag 36 must be framed as well, and this allows the data on the data track 33 to be self-synchronizing.

The absolute position tags 36 are not a direct binary encoding of absolute position of the encoder band 30, but are instead codes which are translated or decoded into corresponding absolute positions. This may be accomplished in a number of ways, but preferably the tags 36 may be used as indices to access a look-up table (or equivalent computer memory) of absolute positions. This coding scheme is used because certain bit sequences (i.e., those which have the spacer 37 bit pattern embedded within them) cannot be used as tags 36, and because it reduces the number of data bits needed to represent each position.

The band 30 may be comprised of clear mylar strips upon which the patterns for the timing track 32 and data track 33 are printed, as by a device such as a laser printer, and to which a layer of aluminized mylar tape is applied, against the printed side of the strips, using a clear adhesive. The taped side of the band 30 may be secured to the surface 31 by a suitable adhesive. The printed patterns, which may consist of laser printer toner particles, of the two tracks 32 and 33 are then visible through the clear, non-printed side of the mylar strips and are physically protected since they are sandwiched between these strips and the aluminized mylar tape, which provides a reflective underlayer.

In this case, the light-reflecting rectangles 34, the spacers 37, and the "one" bits within each tag 36, appear as reflective areas of aluminized mylar tape viewed through the clear mylar strip. The non-reflecting rectangles 35, the start bits 38 and 39, and the "zero" bits within each tag 36 appear as non-reflecting areas of opaque laser printer toner particles (which hide the reflective aluminized mylar tape underlayer) viewed through the clear mylar strip.

The read head 40 (FIGS. 3-6)

As the dome 22 rotates, these two tracks 32 and 33 move past a read head 40 at one end of a stationary column 41 mounted either on the stationary telescope support 25 (see FIGS. 4 and 6) or (as shown in FIGS. 11 to 16) on a stationary part of the dome 107. The read head 40 contains three optical sensors 42, 43, and 44 (FIG. 5) that detect changes in reflectivity as the stripes 34 and 35 and tags 36 move past the sensors. If desired, the read head 40 may be mounted on the rotatable dome 22, and the tracks 32 and 33 may be supported by a stationary portion of the building 20.

Theoretically, the read head 40 could be mounted rigidly on the column 41, and if the dome 22 is perfectly round it may be. In fact, the read head 40 can be mounted rigidly with respect to radial motion so long as the out-of-roundness of the dome surface 31 is significantly less than the depth of focus of the optical sensors 42, 43, and 44 that are used. For example, if sensors 42, 43, and 44 with a depth of focus of 0.75 inch were to be used (these are available but at much greater expense than the sensors currently recommended) then the dome's cylindrical surface 31 could be out-of-round by as much as 0.5 inch, (i.e., .+-.0.25 inch).

However, affordable off-the-shelf sensors do not have sufficient depth of focus (i.e., greater than 1 inch) to enable the use of badly out-of-round domes. Also, while a rigid mount would simplify installation, it might prove less forgiving of earthquakes, which are quite common at some observatories. Unfortunately, many domes 22 are out-of-round. For example, each of the domes at both the 1-meter and 3-meter Lick Observatory telescopes is out-of-round by more than one inch, and this eccentricity exceeds the depth of focus of the optical sensors 42, 43, and 44. As a result, the read head 40 is mounted to move radially, so as to keep the optical sensors 42, 43, and 44 focused on the encoder timing and data tracks 32 and 33.

This is accomplished by a spring-loaded parallelogram linkage 45 between the read head 40 and the column 41. This linkage 45 introduces a small (less than 1/8 incremental tick of motion) uncertainty in the azimuthal position of the read head 40. This, however, is tolerable. A piston (keyed to prevent rotation of the concentric cylinders) may be used instead, but would require more complicated machining.

The proper spacing of the parallelogram linkage 45 relative to the encoder track surface 31 may be provided by two wheels 48 and 49 which are preferably connected to a block 47 in the middle of the parallelogram linkage 45 and which roll along the encoder track surface 31. Neoprene O-rings may be used as tires 50 and 51 on these wheels 48 and 49 because of neoprene's exceptional resistance to ozone. (At altitudes above 4,000 feet, the ozone concentration is sufficient to degrade normal buna-N O-rings.) The traction provided by these tires 50 and 51 is unimportant, for even if they slip, the functioning of the encoder is not affected.

Besides being out-of-round, the encoder track surface 31 at the dome 22 of the 1-meter telescope at Lick Observatory exhibits considerable vertical warping. To correct for such a problem, the optical sensors 42, 43, and 44 are preferably mounted on the block 47 in the center of the parallelogram linkage 45, and the block 47 pivots vertically. The two wheels 48 and 49 which roll along the encoder track surface 31 are preferably connected to this block 47. The wheels 48 and 49 and the tires 50 and 51, which provide for the proper spacing of the parallelogram linkage 45, also provide for the proper alignment of the vertically pivoting block 47. Were the encoder track surface 31 not warped, the block 47 would not need to pivot.

The wheels 48 and 49, their tires 50 and 51, the pivoting block 47, and the parallelogram linkage 45 are the only moving parts in the read head assembly 40, and none of these require machining to any precise tolerance. Even if the more traditional approach of coupling a rotary encoder to the dome were used, a similar assembly would be required to keep the mechanical pick-off for that encoder in contact with an out-of-round or vertically warped dome. Thus, the cost of this assembly should not be viewed as an added cost of using the encoder of this invention.

The Optical Sensors 42, 43, and 44 and Their Generated Analog Voltages (FIGS. 7 and 8)

The three optical sensors 42, 43, and 44 of the encoder read head 40 are used to scan the timing and data tracks 32 and 33. The sensors 42 and 43 are accurately spaced to develop their signals in quadrature although they respond to a single timing track 32. Each sensor 42, 43, and 44 pairs a light-emitting diode 52 (as a light source) with a phototransistor 53. The light-emitting diodes 52 for the sensors 42 and 43 may each emit a narrow, low-intensity infrared beam that is either reflected by a light reflecting stripe 34 or absorbed by a non-reflecting stripe 35 on the timing track 32 of the coded band 30, and reflection or nonreflection is sensed by its phototransistor 53. Similarly, the light-emitting diode 52 of the sensor 44 sends a beam to the data track 33 and its sensor 53 detects whether that beam is reflected or absorbed.

Two optical sensors 42 and 43 scan the timing track 32, and the signals they produce are in quadrature, as shown in FIG. 8. By comparing the phase relationship between the signals from these two sensors 42 and 43, the direction of dome motion is determined. The amount of motion is given by counting the number of signal transitions.

Each phototransistor 53 measures the reflected light and generates an analog voltage proportional to the intensity of the reflected beam. Thus, the sensor 42 generates an analog voltage 54; the sensor 43 generates an analog voltage 55, and the sensor 44 generates an analog voltage 56.

Conversion of the Analog Voltages (FIG. 7)

As shown by FIG. 7 the resulting three analog signals 54, 55 and 56 are converted to three digital TTL levels 57, 58 and 59, respectively, by a separate operational amplifier 60, 61 or 62 and a Schmitt trigger 63, 64, and 65 for each phototransistor output. These three TTL levels 57, 58, and 59 may be fed to separate input lines of an input/output port 69 of an 8-bit microcomputer 70, which then computes the dome position from these inputs. As shown in FIG. 8, the signals 57 and 58 are in quadrature.

The Microcomputer 70 (FIGS. 9 and 10)

The microcomputer 70 is programmed to give a series of functions and the following discussion treats these functions as though they were separate circuits. In fact, all the circuits now to be described are part of the microcomputer 70. A printout of a suitable program is given in the Appendix to this specification.

As shown in FIG. 9, the TTL levels 57 and 58 generated from the two sensors 42 and 43 are fed to a portion of the microcomputer 70 programmed to act as a phase comparator 71; the phase comparator 71 outputs a direction level signal 72 and a clock pulse 73 corresponding to each incremental tick of motion.

A 1.times. phase comparator 71 that produces the clock pulses shown in FIG. 8 operates as follows:

Any positive-going transitions of the signal 57 that occur when the signal 58 is in the low state cause the direction level signal 72 to be set to the high state and a clock pulse 73 to be generated. Any negative-going transitions of signal 57 that occur when the signal 58 is in the low state cause the direction level signal 72 to be set to the low state and a clock pulse 73 to be generated. Transitions of the signal 57 that occur when the signal 58 is in the high state are ignored.

2.times. and 4.times. phase comparators operate in a similar fashion, but are more complicated since they use a greater number of the possible combinations of transitions of the two signals 57 and 58.

The phase comparator 71 can be programmed to respond either to all possible combinations of transitions of the signals 57 and 58, or to only some subset of these combinations. If programmed to respond to all combinations, the phase comparator 71 outputs the direction level signal 72 and clock pulse 73 at four times (4.times.) the spatial frequency of the stripe-pairs (34 and 35) on the timing track 32. If programmed to respond only to some combinations, it can generate these outputs at either once (1.times.) or twice (2.times.) the spatial frequency of the stripe-pairs.

The spatial width of the incremental tick generated by the encoder is a function of the frequency at which the phase comparator 71 is operated. When the phase comparator is operated at four times (4.times.) the spatial frequency of the stripe-pairs, the spatial width of each incremental tick corresponds to one half the width of a single stripe (34 or 35) on the timing track 32. When operated at 2.times., the spatial width of each tick equals the width of a single stripe (34 or 35). When operated at 1.times., the tick width equals the width of a stripe-pair (34 and 35).

The choice of operating frequency (1.times., 2.times., or 4.times.) for the phase comparator 71 involves trade-offs between resolution and robustness, with 1.times. giving the greatest robustness (i.e., immunity to skewing or quadrature misadjustments) and 4.times. giving the greatest resolution. However, the choice of operating frequency does not change the logic of any of the other portions of the encoder. For the purposes of simplifying the remaining discussion and figures, the operating frequency of the phase comparator 71 will be taken as 1.times., and thus the spatial width of an incremental tick will equal the width of a stripe-pair (light-blocking stripe 35 plus light-passing stripe 34) on the timing track 32.

The computer is programmed so that the direction level signal 72 and clock pulse 73, (which are generated by the phase comparator 71 for each incremental tick of motion), control an "up/down counter" function 74, which measures the incremental position. The computer is also programmed so that the direction level signal 72 and clock pulse 73 control two "shift register"functions 75 and 76 that process the signal 59 from the data track 33. The up/down counter 74, shift registers 75 and 76, and some other registers are located in the RAM of the microcomputer 70. The logic for the phase comparator 71 and for the clocking and gating of the various registers is implemented in appropriate software.

When the direction level signal 72 is in the high state, the clock pulse signal 73 causes the counter function 74 to be incremented, and the "shift register" functions 75 and 76 to be shifted right. When the direction level signal 72 is in the low state, the clock pulse signal 73 causes the counter function 74 to be decremented, and the "shift register" functions 75 and 76 to be shifted left.

When the dome 22 rotates, the direction level signal 72 and the clock pulse 73 that the phase comparator 71 derives from the timing track 32 are used to shift the bits obtained from the data track 33 into one of the two multi-bit bi-directional serial shift registers 75 and 76. (See FIGS. 9 and 10.) For example, the shift registers 75 and 76 may be 24 bits wide. The number of bits is not absolute, and may be larger or smaller depending on how many bits are being used to encode the absolute position tags 36. This in turn depends on the number of tags 36 and the spacing between them, and this in turn depends on the size and resolution of the encoder. Conceptually, these two registers 75 and 76 can be imagined as existing side-by-side and, when using 24-bit shift registers they form a 48-bit-wide window through which the microcomputer 70 "views" the data track 33. (For N-bit shift registers, the window is 2N bits wide.) After proper initialization, a minimum of N consecutive bits (e.g., 24) in this window always corresponds to an identical sequence of N bits (e.g. 24) on the section of the encoder data track 33 that is currently positioned in front of and immediately adjacent to the read head 40. The bits in the window are always aligned such that the bit nearest the "center" of the window corresponds to the data bit that has just passed under the optical sensor 44 for the data track 33.

The direction of dome rotation determines in which direction the registers 75 and 76 are shifted, and the distance moved determines the number of shifts that occur. For each incremental tick of motion, the registers 75 and 76 are shifted by one bit. The two registers 75 and 76 can be shifted in either direction, but both registers 75 and 76 perform the identical shift sequence. The registers 75 and 76 are not connected in cascade; so the contents of one register 75 will never shift into the other register 76 and vice versa. In fact, none of the four serial shift outputs is connected to anything; so the bits which shift out of the registers 75 and 76 are shifted into oblivion. The two registers 75 and 76 share a common serial input, namely, the signal 59 from the optical sensor 44 that scans the data track 33.

The actual operation of these shift registers 75 and 76 proceeds as follows. On power up, the registers 75 and 76 are set to zero. The dome 22 must be moved a small amount (at least N incremental ticks, e.g., 24 incremental ticks) in either direction to initialize either the register 75 or the register 76. When the dome 22 rotates to the right by one incremental tick, both registers 75 and 76 shifted right one bit (see FIG. 9). The bit from the data track 33 is shifted into the "dome right motion shift register 75", while a zero is shifted into the "dome lift motion shift register 76." When the dome 22 moves to the left, the converse is true.

When an absolute position tag 36 shifts into either register 75 or 76 and becomes properly framed (i.e., the tag 36 is centered in either register 75 or 76 and surrounded on both sides by the appropriate framing pattern made up of a start bit 38 or 39 and the spacer bit sequence 37), then a framing gating 78 or 79 generates a "tag-framed" strobe 80 or 81. This signals a tag decoder 82 or 83 (which may be implemented as a look-up table) to check the tag 36 for parity and adjacency, i.e., is this tag located physically adjacent to the last tag decoded. If the tag 36 fails either test, an error signal 84 is generated, and the tag 36 is ignored. If the tag 36 is valid, it is decoded by the tag decoder 82 or 83 into an absolute position. The tag decoder 82 or 83 also generates a "tag valid" strobe 85 or 86, which causes the decoded absolute position to be passed through an absolute position multiplexer 87 to an absolute position register 90. The multiplexer 87 generates an "absolute position valid" strobe 88, which causes the decoded absolute position that it is passing to be latched into the absolute position register 90 (FIG. 9) and which also signals a position comparator 91 to compare the absolute position in the register 90 to the incremental position in the incremental position up/down counter 74. If they do not match, an error signal 92 is generated, and the incremental position is reset to the absolute position.

A given tag 36 will decode to one of two different absolute positions, depending on whether it is framed in the "dome right motion" shift register 75 or the "dome left motion" shift register 76. The positions correspond, respectively, to the tag 36 being positioned either just to the right or just to the left of the read head 40.

Some Alternative Dome and Track Structures

First alternative structure (FIGS. 11 and 12):

FIG. 11 shows a dome 100 mounted by wheels 101 on a rail 102 supported by a stationary cylindrical concrete wall 103. There is an exterior shroud 104 to shield the interior of the domed structure from the weather. The dome 100 has a walkway 105 with a railing 106, and below the inner edge of the walkway 105 there may be a sheetrock wall 107. Slip rings 108 transfer electricity onto the moving dome 100 from an electrical panel 109.

A reflective encoder-band 110 is attached, preferably with glue, to a vertical edge 111 of the rotating walkway 105. A read head 112, like the read head 40, is rigidly mounted by a fixed portion 113 to the sheetrock wall 107, with a parallelogram linkage portion 114 joining the movable part of the read head 112 to the fixed portion 113. There is an opening 115 cut at the top of the sheetrock wall 107 to make the encoding band visible to the read head 112. The wheels and pivoting block of the read head 112 have been omitted from FIGS. 11 and 12, for the sake of clarity.

This system functions almost exactly like that described previously.

Second alternative structure (FIGS. 13 and 14):

In the structure shown in FIGS. 13 and 14, an annular flat encoder band 120 is placed on an horizontal surface 121 (see FIG. 14) on the bottom of the walkway 105 (or other rotating part of the dome 100, provided that such a surface is available and accessible), rather than on the inside vertical surface as in FIGS. 11 and 12. In this configuration, a read head 122 is mounted underneath the walkway 105, and its optical sensors look up at the encoder band 120 above. As a result, radial run-out in the dome 100 does not change the spacing between the up-looking read head 122 and the encoder band 120, and thus does not cause the optical sensors 42, 43, and 44 to go in and out of focus. Rather, radial run-out of the dome 100 causes a motion of the encoder band 120 in the plane of dome rotation and perpendicular to the direction of rotation. This motion can be accommodated by making each of the two (timing and data) tracks wider in the direction corresponding to this motion (i.e., the direction perpendicular to the direction of rotation.) Thus, in this configuration, the read head 122 can be rigidly mounted if the dome 100 in question exhibits only radial run-out and is not subject to other mechanical anomalies in its rotation.

Another advantage of this location is that it may provide an easier way for coping with domes which exhibit both axial and radial run-out. For such domes, if the encoder band 110 is mounted vertically, as shown in FIG. 12, the side-looking read head assembly 112 requires two separate parallelogram linkages 45 (one to compensate for the radial run-out and the other to compensate for the axial run-out) operating at right angles. (On the read head used for the encoder on the dome of the Lick 1-meter telescope, only a single parallelogram linkage 45 is needed since this dome has negligible axial motion but does have significant radial run-out.) If, instead, the encoder band 120 is mounted horizontally, as shown in FIG. 14, the up-looking read head assembly 122 requires only a single parallelogram linkage 123 to compensate for axial run-out. The radial run-out is handled as described in the previous paragraph. Note that in FIGS. 13 and 14, the wheels and pivoting block of the read head 122 have been omitted from the drawings, in the interest of clarity.

Locating the encoder band under the rotating part of a dome 100 does have some disadvantages. A major disadvantage of this configuration is that the encoder band 120 and read head 122 are located in a dirtier environment where they may be exposed to oil and lubricants from the dome "trucks" and dome drive motors. Another disadvantage is that the segments of the encoder band 120 become sectors of an annular ring, rather than segments of a cylinder. Thus, apparatus such as mylar strips used to form the band 120 may need to be printed as slightly curved pieces, the degree of curvature depending on the radius of the encoder band. This is easily accomplished, since the patterns for the mylar strips may be computer generated. However, it might be more difficult to apply the layer of aluminized mylar tape to the curved segments of mylar since curved segments of aluminized tape are not presently readily available. This can be compensated for by using extra tape to cover the rectangular area which bounds the curved segment.

Fortunately, the curvature problem becomes less pronounced as the radius of the encoder band 120 increases, and for large domes the curvature of each segment may be sufficiently small so that the band can be assembled from straight segments. In this case, instead of forming a perfect circle, the encoder band 120 forms an "N" sided polygon. For a dome comparable in size to the one on the Lick 3-meter telescope, "N" would be about 350 if a 10-inch segment size is used. A 350-sided polygon looks very much like a circle.

A flat band type of encoding track (FIG. 17):

For use in a structure like that of FIGS. 13 and 14, a band pattern 120 like that shown in FIG. 17 may be used. There is an outer circle providing a timing track 124 and an inner circle providing a data track 125. The timing track 124 alternates reflective stripes 126 with non-reflective stripes 127, and the data track alternates tags 128 with spacer 129.

This pattern 120 may comprise an annulus. Also, for somewhat smaller encoders, it may be placed on a glass disk.

Third alternative structure--FIGS. 15 and 16:

In the third alternative implementation, illustrated in FIGS. 15 and 16, a transmissive band 130 is used in place of a reflective one. The transmissive band 130 may be hung like a skirt from the bottom of the walkway 105 or other rotating part of the dome 100. This may be accomplished on a dome such as that of the Shane 3-meter telescope by gluing the 1/2-inch upper margin of 3-inch high by 10-inch long mylar strips, which are used to form the band 130, onto the bottom 1/2-inch of the inner vertical surface of the walkway on the rotating part of the dome. Each of the three optical sensors 134 (only one of which is shown in FIGS. 15 and 16) in the read head assembly 132 is composed of two physically separate parts. Each sensor's light source 133 may be on the radially inner side of the band 130 and aimed outwardly, while the optical detector 134 may then be on the other, radially outer, side of the band 130 and aimed inwardly. By using a reasonably fast, well-focused beam, the depth of focus can be made very large. Thus, the image of the band 130 on the optical sensor 134 remains in focus, regardless of radial or axial run-out, or vertical warping. Accordingly, both parts of the optical sensor 133, 134 can be rigidly mounted, so that the only moving part of the encoder is the dome 100. Further, there is no physical contact between the fixed read head 132 and the rotating dome 100 and its encoder band 130. The only "contact" is the optical sensor's light beam passing through the transmissive band 130. Finally, using a transmissive band 130 rather than a reflective one eliminates the need for the layer of aluminized mylar tape, thus removing the cost of this material and the labor to apply it to the mylar strips.

The main disadvantage of this scheme is that the mylar strips, which are far from rigid, are literally dangling in the air, and are thus vulnerable to damage. This may be solved by gluing the mylar strips onto a clear but rigid material, such as Plexiglas, which is in turn rigidly mounted to the dome 100 itself. Alternatively, if computer-controlled milling equipment is available, the encoder band 130 may be made from strips of inexpensive aluminum stock, which may then be rigidly mounted to the dome 100. The patterns, instead of being printed on mylar would instead be cut as holes in the aluminum, and the logical sense of the signals complemented.

Another disadvantage is that the optical sensors 134 for the read head 132 which scans the transmissive band 130 may not be as readily available nor as inexpensive as those used in a read head 40 or 112 for scanning a reflective band 110 or 120.

Because the "dangling transmissive band" scheme eliminates any physical contact between the read head 130 and the surface being encoded, it may also be useful in encoder applications where one must tolerate radial and/or axial run out of the encoder band surface but cannot tolerate the mechanical/frictional drag created by an encoder pick-off or read head assembly which does make contact. While this is not a concern for massive objects like telescope domes, it may be a concern when measuring the position of low-mass, low-inertia rotating devices that spin on high-quality bearings.

Some Advantages of the Invention

This system of the present invention has several advantages over the more traditional solution of mechanically coupling a rotary encoder to the dome. To start with, mechanical complexity is reduced by eliminating most moving parts, such as pinch rollers, chains, sprockets, and reducing gears. If the dome 22 does not suffer from warping or run-out (either radial or axial) that causes the motion of the encoder track surface 31, 111, or 121 to exceed the depth of focus of the optical sensors 42, 43, and 44, then the read head 40 or 112 or 122 can be completely stationary. In this case, the parallelogram linkage 45 or 114 or 123 can be eliminated, and the block 47 need not pivot. The dome 22 itself will be the only moving part in the system, and all mechanical connections between the read head 40 or 112 or 122 and the dome 22 will be eliminated.

Otherwise, out-of-round domes are readily taken care of by the parallelogram linkage 45, in conjunction with the wheels 48 and 49, together with their tires 50 and 51. Similarly, warped tracks are taken care of by mounting the block 47, which holds the sensors 42, 43, and 44, so that it pivots vertically to compensate for the vertical warping.

Another advantage of the encoder of this invention is that a gear reduction is not required, as is the case of an absolute rotary encoder coupled to the dome. A gear reduction is expensive and is also a significant source of error. If the gearing does not produce the exact reduction required, a rotary encoder will incur a small incremental error on each rotation. Then, as said above, if the dome is rotated repeatedly in the same direction, this incremental error accumulates and may grow unacceptably large. Additional errors result from mechanical backlash in the gears. These errors are eliminated by the present invention.

Automatic detection of errors:

The encoder of this invention provides redundancy, and this enables the hybrid encoder to provide automatic detection of errors, because, unlike most encoders, there is redundant information which is cross-checked. Without some type of redundancy, it is very difficult to detect, diagnose, or correct errors.

In conventional optical incremental encoders, there has been no way to detect whether spurious counts have occurred as a result of dust or scratches on the glass disk of the encoder, dimming of the encoder's light source, or electrical noise. Incorrect position information has thus tended to go undetected. These problems have occurred many times during use of such encoders.

Similar problems can occur with absolute encoders as well, although these could be reduced by the addition of some redundancy in the form of a parity track. However, the addition of such a track would cut the available resolution of the encoder in half.

Gray-code absolute encoders provide some error detecting capability, because on any incremental step only a single bit should make a transition. Thus, if more than one bit changes at once (as might happen if there were dust, scratches, or electrical noise), then a Gray-code encoder can signal an error. On the other hand, if a given bit or group of bits get stuck in one state (as would happen if a light source dimmed or burned out, or a light sensor failed), those bits will make no transitions. Thus for some motions of the encoder, no motion will be reported and no error will be detected.

In the hybrid (absolute and incremental) encoder of this invention, the comparison of redundant sources of information greatly increases the likelihood that most typical malfunctions which will cause an error will be reported. Scratches, dust, or electrical noise affecting the signal from the data track 33 will result in the corruption of an absolute position tag 36. This will cause either a tag parity error or adjacency error to be reported, and the tag 36 to be rejected. The addition of the parity bit to the tag 36 of the hybrid encoder does not cut the encoder's resolution in half, (unlike the case of the absolute encoder) it does not require the addition of a separate track. Rather, it requires only the addition of one bit to a tag 36 that is already at least seven bits long and which has a typical effective width of twenty bits. Accordingly, the addition of the parity bit only reduces the absolute position resolution by 5%, and does not reduce the incremental position resolution at all.

Similarly, scratches, dust, or electrical noise affecting the signals from the timing track 32 will result in corruption of the incremental position and sometimes in the corruption of a nearby absolute position tag 36, since errors on the timing track may cause incorrect shifting of the tag 36 into the shift registers 75 and 76. In either case, this will cause an error to be reported. If a tag 36 has been corrupted, either a tag parity error, adjacency error, or missed-tag error will be reported. (A missed-tag error occurs when a tag is missed, i.e., not detected after a reasonable amount of incremental motion has been measured.) Ultimately, when the next valid tag 36 becomes "properly framed", a mismatch will be detected between the corrupted incremental position and the absolute position corresponding to the "properly framed" tag 36. This will cause a position mismatch error to be reported. The corrupted incremental position will then be corrected by resetting it to match the absolute position corresponding to the "properly framed" valid tag 36.

Diagnosis of errors (FIG. 9A):

So far, it has been demonstrated how the hybrid encoder will detect most typical errors, but it has not indicated how it can diagnose or isolate the cause of an error. For the types of errors discussed thus far (noise, scratches, dust), a precise diagnosis is not possible, although the different classes of error (tag parity error, adjacency error, missed tag error, position mismatch error) help to isolate the probable causes.

For example, if some tags 36 repeatably read correctly while others repeatably cause errors, the problem is probably (although not necessarily) with the encoder track surface 31 and not with the read head sensors, electronics, or signal processing logic. Knowing which tags cause errors identifies where on the encoder track surface 31 scratches or dust might be. For non-repeatable errors, more likely causes are electrical noise or an incorrect mechanical adjustment of the read head assembly 40, which causes the optical sensors to drift in and out of focus.

There are, however, certain failure modes in which a more definite diagonsis can be made. In particular, failures of either the light source, light sensor, or associated electronics for one of the three signal channels (two channels "A" and "B" corresponding to the signals 57 and 58 from the timing track 32 and one channel "D" for signal 59 from the data track 33 as shown in FIG. 9A) produce a recognizable signature which will point to the failed channel. The failed channel is located by comparing in pairs the signals from each of the three channels. If analysis of the signals from a given pair of channels indicates that the encoder is making significant motion (i.e., the incremental position is changing by more than a minimum amount, usually one count), and the signal from the third channel shows no transitions, then the signal from the third channel has failed.

In the case of the conventional incremental encoder, where there are only the two channels from the timing track, there is only a single channel available to use to cross-check the opposing channel. (Note that many conventional incremental encoders have a third track which provides a single index pulse per revolution. However, since this pulse is generated so infrequently, it cannot be relied on for making a rapid diagnosis of the failure of the other channels.) A single channel cannot provide sufficient information to determine unambiguously whether or not the encoder is making a significant motion.

It is tempting to suggest that a fault could be detected simply by counting the signal transitions from each of the two channels, channel signal 57 and channel signal 58. If no transitions are seen on the channel signal 58 after some given number of transitions are seen on the channel signal 57, the conclusion could (erroneously) be made that the signal 58 has failed. Unfortunately, although this scheme will detect real failures, it will also erroneously detect failures where none has occurred. If the encoder is making a fine (i.e., the incremental position change is only one count or less) rocking motion back and forth across a single stripe on the timing track 32, then the signal 57 could produce many transitions without any transitions being generated from the signal 58, even though the signal 58 had not failed. Such rocking motions are indeed possible, particularly for domes and telescopes subjected to gentle wind buffeting.

In the case of the encoder of this invention, there is always sufficient information from any pair of channels to diagnose unambiguously a single-channel failure, (i.e., loss of the signal) from the third channel. Such failures are detected as soon as some minimal amount of motion actually occurs.

If the signal 59 from the single channel of the data track 33 fails, this can be detected quickly by using a second, smaller up/down counter 93. This counter 93 is clocked by the same direction level 72 and clock pulse 73 used by the incremental position up/down counter 74. This second up/down counter 93 is zeroed (cleared) whenever any transition is detected on the signal 59. There should always be some transition detected from the signal 59 whenever the encoder moves an amount equal to the width of a tag 36 or a spacer 37, whichever is greater. If this second up/down counter 93 counts (in either direction) a motion exceeding this amount and does not reset to zero, then the signal 59 from the data track (channel D) has failed.

The signals 57 and 58 from the two channels (A and B) of the timing track 32 are equivalent with one exception. The transitions on the signal 59 are 225 degrees out-of-phase with those of the signal 57 and 135 degrees out-of-phase with those of the signal 58, as shown in FIG. 8. Clearly, the signal 57 is 90 degrees out-of-phase with the signal 58. Thus, phase information can be obtained by comparing any combination of two channels, and failures of the signals 57 and 58 (A or B) can be treated much the same.

If there is a failure of the signal 58, phase information can be recovered instead from the signal 59, although not as often, since the signal 59 makes transitions less often than the signal 58. Because of this phase relationship, any transitions from either of the signals 58 or 59 will occur between opposing transitions of the signal 57. If during the interval between two opposing transitions of the signal 57, a transition is detected on the signal 59 but is not detected on the signal 56, then the signal 58 has failed.

Failures of the signal 57 can be dealt with in much the same way, since there is sufficient phase information from the other two channels to determine if significant motion has occurred. Any transitions of the signals 57 or 59 will occur between opposing transitions of the signal 58. If during the interval between two opposing transitions of the signal 58, a transition is detected on the signal 59 but is not detected on the signal 57, then the signal 57 has failed.

In summary, the hybrid encoder provides robust error detection and diagnosis through the use of redundant information, and does so, without significantly compromising positional resolution, and without requiring additional optical or analog electronic components inside a standard rotary encoder package.

Most encoders currently available (either incremental or absolute) do not provide this type of needed redundancy, and cannot detect whether they are malfunctioning.

The novel combination of incremental and absolute encoding:

A distinctly novel feature of this invention is that the encoder combines the features of incremental and absolute encoders into a single system, using only three optical sensors 42, 43, and 44. This number is far fewer than the number of sensors normally used in absolute encoders alone. It is also the same number of sensors found in encoders which provide only incremental position plus a single index pulse. In fact existing incremental encoders can be redesigned to use this technique with only minimal changes to the encoder optics and geometry.

General use rotary encoder (FIGS. 18-20):

The encoder of this invention may also be fabricated in a conventional rotary encoder package for general use, not just in telescope domes. This may be accomplished by converting the optical track which is normally used to provide a single index position into data for providing absolute position.

Conventional optical incremental encoders normally use two separate optical tracks as timing tracks, plus a third track if an index pulse is provided. Each of these tracks is scanned by its own optical sensor. The azimuthal alignment of the two timing tracks is arranged so that they are in spatial quadrature, i.e., the resulting signals are 90-degrees out-of-phase. The reason for using two separate tracks is that it allows the two optical sensors scanning these tracks to be mechanically locked together, thus eliminating the need for extremely precise mechanical adjustment of the spacing between the sensors and insuring that the phase relationship remains stable despite thermal variation or mechanical vibration.

In the specific implementation of the hybrid encoder heretofore described, only one optical track is used for the timing track 32. However, this single track is scanned by two optical sensors 42 and 43, and quadrature is achieved by adjusting the spacing between these two sensors 42 and 42. This can be done because the physical scale of the dome is so large, and consequently the spacing between stripes on the timing track is so coarse that the adjustments in the spacing of these two sensors required to achieve proper quadrature are correspondingly coarse. The pattern of FIG. 17 may be used on glass disks in appropriate cases.

In implementations in which finer incremental line spacings are required, as would be the case where a hybrid encoder is fabricated in a standard rotary encoder package, the adjustment of the spacing between these two optical sensors requires greater precision and stability than is usually possible. Therefore, if one were to make a hybrid encoder for a standard rotary encoder package, there should be two separate timing tracks with the azimuthal alignment of these two tracks providing the quadrature, as in a conventional incremental encoder and as shown in FIGS. 18 and 20.

FIG. 18 shows a transparent glass disk 200 having two timing tracks 201 and 202 and a data track 203. The timing tracks 201 and 202 are offset to provide quadrature. As FIG. 19 shows, light sources 204 and 205, respectively, are aligned radially with the tracks 201 and 202 and, as shown in FIG. 19, are on opposite sides of the disk 200 from their similarly aligned phototransistors 206 and 207. A light source 208 is used for the data track 203 and addresses a transistor 209. All of the light sources may be LED's or something else. FIG. 18 shows 180 incremental lines per revolution, giving an incremental resolution of two degrees. The respective dark stripes 210 and 211 of the tracks 201 and 202 are opaque and the respective clear strips 212 and 213 are transparent. The data track 203 has nine absolute tags 214, giving an absolute resolution of forty degrees.

FIG. 20 shows a similar glass disk 220 with two timing tracks 221 and 222 and a data track 223. Here there are 360 incremental lines per revolution, for an incremental resolution of one degree. There are eighteen absolute tags 224, for an absolute resolution of twenty degrees. Typically, such glass disks may have thousands of incremental lines per revolution.

As described earlier, incremental resolution can also be increased by operating the phase comparator 71 at either twice (2.times.) or four times (4.times.) the spatial frequency of the incremental lines on the timing track. These higher operating frequencies also yield a tighter packing of the tags 36 and spacers 37 on the data track 33. Thus, both incremental and absolute resolution can be increased together. This is illustrated in FIGS. 21 and 22, which are identical to FIG. 20, except that they show, respectively, glass disks 225 and 226 designed for use with a phase comparator 71 operated at twice (2.times.) or four times (4.times.) the spatial frequency of the lines. Thus, while the timing tracks 221 and 222 are identical in FIGS. 20, 21, and 22, the data tracks 223 (FIG. 20), 227 (FIG. 21), and 228 (FIG. 22) are different, in accordance with the different frequencies involved.

If the optical sensor for the data track were to be locked in the same mounting with the sensors for the timing track, an azimuthal offset between the data and timing tracks would be needed to provide the phase difference required for detection of single channel failures.

In either case, whether one or two timing tracks are used, two sensors (two light sources 204 and 205 and two phototransistors 206 and 207) are always required for the incremental portion of the hybrid encoder, while a third sensor (light source 208 and phototransistor 209) is required for scanning the data track. Thus, while the complete hybrid encoder always requires exactly three optical sensors, the number of separate optical tracks may be either two or three depending on the specific implementation and the fineness of the incremental line spacing. Note that the number of optical sensors is the same number required for a conventional incremental encoder that provides only incremental position and a separate index pulse. Also note that unlike conventional absolute encoders, the number of optical sensors (three) and optical tracks (two or three) never increases beyond three, regardless of changes in the size or resolution of the encoder.

The only part that would need to be changed inside the rotary encoder package itself would be to install the glass disk 200 or 220 in place of the conventional disk; the sensors and electronics for what was the index track should be usable for what would become the data track. All of the signal processing logic described above (the shift registers, up/down counters, etc.) could and probably should be located outside this package in a separate box.

Further Advantages of the Invention

(1) All forms of the invention provide both incremental and absolute position.

(2) The phase information from the timing track provides the clocking of the incremental position.

(3) the phase information from the timing track also controls the clocking of the data track and hence the absolute position.

(4) The incremental and absolute positions can be clocked by a phase comparator operating, e.g., at either once (1.times.), twice (2.times.), or four times (4.times.) the spatial frequency of the stripes on the timing track.

(5) Only three optical sensors are ever needed, regardless of encoder size or resolution.

(6) At most, three optical tracks are needed, and in some implementations, two tracks will suffice.

(7) Single channel failures can be readily detected and isolated.

(8) The encoding of absolute positions into absolute position tags that are surrounded by start bits and separated by spacers enables the data track to be self-synchronizing.

These eight features distinguish the encoder of this invention from all other existing encoders that we are aware of.

To those skilled in the art to which this invention relates, many changes in construction and widely differing embodiments and applications of the invention will suggest themselves without departing from the spirit and scope of the invention. The disclosures and the descriptions herein are purely illustrative and are not intended to be in any sense limiting. ##SPC1##

Claims

What is claimed is:

1. An encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member, including in combination: an encoding band supported by one said member and having a timing and incremental encoding track comprising at least one series of alternating activating stripes and non-activating stripes, and a single data track comprising a series of absolute position tags disposed at regular intervals along said data track and alternating with spacers that are identical to each other, each said tag being framed by start bits and being distinctly different from every other said tag and each comprising a serial bit stream for encoding the absolute position corresponding to the location of its tag along said data track, reading means supported by the other said member and having three non-sweeping sensing means, two in quadrature with each other for said timing and incremental encoding track, and one for said data track, each said sensing means having a single energy-source-energy-sensor pair including means for generating an analog signal voltage, said members being capable of both discrete and continuous motion in both directions, and decoding means actuated by said analog signal voltages for determining the incremental and absolute positions of said movable member relative to said stationary member, whereby said absolute and incremental positions can be determined solely from said system without dependence or any additional information external to said system.

2. The system of claim 1 wherein said band is cylindrical and is supported perpendicular to the plane of motion of said movable member.

3. The system of claim 1 wherein said band is flat and annular and is supported parallel to the plane of motion of said movable member.

4. An encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member, including in combination: an encoding band supported by one said member and having a timing and incremental encoding track comprising at least one series of alternating activating stripes and non-activating stripes, and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, and reading means supported by the other said member and having three sensing means, two in quadrature with each other for said timing and incremental encoding track, and one for said data track, each said sensing means including means for generating an analog signal voltage, and decoding means actuated by said analog signal voltages for determining the incremental and absolute positions of said movable member relative to said stationary member, said decoding means comprising (1) timing track phase comparator means connected to said two in-quadrature sensing means for comparing the phase relationship of their two signal voltages and producing a direction level signal and a clock pulse signal, (2) up/down counter means connected to said direction level and clock pulse signals and having readout means for indicating the current position of said movable member, (3) left-motion serial shift register means connected to said direction level and clock pulse signals and also to said third sensing means, (4) right-motion serial shift register means connected to said direction level and clock pulse signal and also to said third sensing means, (5) two framing gating means, each connected to said clock pulse signal and to one of said shift registers, (6) absolute position register means connected to both said framing gating means, and (7) position comparator means connected to said absolute position register means and to said up/down counter means.

5. The system of claim 4 wherein said sensing means send their voltage signals through operational amplifier means and Schmitt trigger means en route from said two in-quadrature sensing means, to said phase comparator means and, from said third sensing means en route to said shift register means.

6. The system of claim 4 whereby said signals from said framing gating means pass, through respective decoder and error checking means and from there to absolute position multiplexer means, while en route to said absolute position register means.

7. The system of claim 4 having three fault-detection means, to which said three sensing means and said timing track phase comparator means also send their respective signals, for outputting a fault signal when a signal derived from any of the respective sensing means has by itself failed.

8. A decoding apparatus actuated by three analog signal voltages from an encoder for determining the incremental and absolute positions of a movable member relative to a stationary member, said encoder having three sensing means for producing said three analog voltages from a timing track and a data track, comprising: (1) timing track phase comparator means connected to two timing sensing means for comparing the phase relationship of their two analog signal voltages and producing a direction level signal and a clock pulse signal, (2) up/down counter means connected to said direction level and clock pulse signals and having readout means for indicating the current position of said movable member, (3) left-motion serial shift register means connected to said direction level and clock pulse signals and also to said third analog signal voltage, (4) right-motion serial shift register means connected to said direction level and clock pulse signal and also to said third analog signal voltage, (5) two framing gating means, each connected to said clock pulse signal and to one of said shift register means, (6) absolute position register means connected to both said framing gating means, and (7) position comparator means connected to said absolute position register means and to said up/down counter means.

9. The decoding apparatus of claim 8 wherein said sensing means send their voltage signals through operational amplifier means and Schmitt trigger means en route from said two timing sensing means to said phase comparator means and, from said third sensing means en route to said shift register means.

10. The decoding apparatus of claim 8 whereby said signals from said framing gating means pass through respective decoder and error checking means and from there to absolute position multiplexer means, while en route to said absolute position register means.

11. The decoding apparatus of claim 8 having three fault detector means, connected to said three sensing means and said timing track phase comparator means so that they also send their respective signals to said three fault detector means, said fault detector means outputting a fault signal when the signal derived from the respective sensing means has by itself failed.

12. An optical encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member, including in combination: three light sources supported by one said member for emitting three narrow light beams on respective paths, an encoding band supported by the other said member in the paths of said light beams, said band having at least one timing and incremental encoding track comprising at least one series of alternating light-blocking strips and light-passing stripes, and a single data track comprising a series of absolute position tags disposed at regular intervals along said data track and alternating with spacers that are identical to each other, each said tag being framed by start bits and distinctly different from every other said tag and each comprising a serial bit stream for encoding the absolute position corresponding to the location of its tag along said data track, a read head supported by said one said member and having three non-sweeping optical sensors, one for each said light beam, two in quadrature with each other for said timing and incremental encoding track, and one for said data track, each said optical sensor including light sensing means for generating an analog signal voltage, said members being capable of both discrete and continuous motion in both directions, decoding means actuated by said analog signal voltages for determining the incremental and absolute positions of said movable member relative to said stationary member, and whereby said absolute and incremental positions can be determined solely from said system without dependence or any additional information external to said system.

13. The system of claim 12 wherein said light-passing stripes comprise reflecting stripes, reflecting light from said beams to said optical sensors.

14. The system of claim 12 wherein said light-passing stripes are transparent stripes through which said light beams pass en route to said optical sensors.

15. The system of claim 12 wherein there is a single timing track, and its two associated light sources and light sensing means are spaced apart at a predetermined distance to provide serial quadrature.

16. The system of claim 12 wherein there are two timing tracks, staggered with respect to each other to provide said quadrature, each having one said light beam and one said optical sensor associated therewith.

17. The system of claim 12 wherein said band is cylindrical and is supported perpendicular to the plane of motion of said movable member.

18. The system of claim 12 wherein said band is flat and annular and is supported parallel to the plane of motion of said movable member.

19. An optical encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member, including in combination: three light sources supported by one said member for emitting three narrow light beams on respective paths, an encoding band supported by the other said member in the paths of said light beams, said band having a timing and incremental encoding track comprising at least one series of alternating light-blocking stripes and light-passing stripes, and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, and a read head supported by said one said member and having three optical sensors, one for each said light beam, two in quadrature with each other for said timing and incremental encoding track, and one for said data track, each said optical sensor including light sensing means for generating an analog signal voltage, and decoding means actuated by said analog signal voltages for determining the incremental and absolute positions of said movable member relative to said stationary member, said decoding means comprising a computer programmed to function as (1) a timing track phase comparator connected to said two timing optical sensors for comparing the phase relationship of their two signal voltages and producing a direction level signal and a clock pulse signal, (2) an up/down counter connected to said direction level and clock pulse signals and having readout means for indicating the current position of said movable member, (3) a left-motion serial shift register connected to said direction level and clock pulse signals and also to said third optical sensor, (4) a right-motion serial shift register connected to said direction level and clock pulse signal and also to said third optical sensor. (5) two framing gatings, each connected to said clock pulse signal and to one of said shift registers, (6) an absolute position register connected to both said framing gatings, and (7) a position comparator connected to said absolute position register and to said up/down counter.

20. The system of claim 19 wherein said optical sensors send their voltage signals through an operational amplifier and a Schmitt trigger en route from said two optical sensors to said phase comparator and, from said third optical sensor en route to said shift registers.

21. The system of claim 19 whereby said signals from said framing gatings pass, in said computer, through elements functioning as respective decoders and error checking means and from there to elements functioning as an absolute position multiplexer, while en route to said absolute position register.

22. An optical encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member, including in combination: three light sources supported by one said member for emitting three narrow light beams on respective paths, an encoding band supported by the other said member in the paths of said light beams, said band having a timing and incremental encoding track comprising at least one series of alternating light-blocking stripes and light-passing stripes, and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, and a read head supported by said one said member and having three optical sensors, one for each said light beam, two in quadrature with each other for said timing and incremental encoding track, and one for said data track, each said optical sensor including light sensing means for generating an analog signal voltage, decoding means actuated by said analog signal voltages for determining the incremental and absolute positions of said movable member relative to said stationary member, and three fault detector means, each connected to some subset of the signals from said three optical sensors and said timing track phase comparator, for producing a fault signal if the signal derived from its respective sensor has by itself failed.

23. An encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member, including in combination: three light sources supported by one said member for sending three narrow light beams on respective paths, a continuous circular encoding band supported by the other said member in the paths of said light beams, said band having a timing and incremental encoding track comprising at least one series of alternating light-blocking stripes and light-passing stripes, and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, a read head supported by said one said member and having three optical sensors, one for each said light beam, two in quadrature with each other for said timing and incremental encoding track, and one for said data track, each said optical sensor including light sensing means for generating an analog signal voltage, comparator means connected to said two timing optical sensors for comparing the phase relationship of their two signal voltages and producing a direction level signal and a clock pulse signal, up/down counter means connected to said direction level and clock pulse signals and having readout means for indicating the current position of said movable member, left-motion serial shift register means connected to said direction level and clock pulse signals and also to said third optical sensor, right-motion serial shift register means connected to said direction level and clock pulse signal and also to said third optical sensor, two framing gating means, each connected to said clock pulse signal and to one of said shift register means, absolute position register means connected to both said framing gating means, and position comparator means connected to said absolute position register means and to said up/down counter means.

24. The system of claim 23 wherein said optical sensors send their voltage signals through an operational amplifier and a Schmitt trigger en route from each of said two optical sensors to said phase comparator means and, from said third optical sensor en route to said shift register means.

25. The system of claim 23 whereby said signals from said framing gating means pass through respective decoding means and error checking means and from there to absolute position multiplexer means, while en route to said absolute position register means.

26. The system of claim 23 having first fault detector means connected to the signals from the three optical sensors for producing a fault signal if the signal derived from the first of said two in quadrature sensors has by itself failed, second fault detector means connected to the signals from the three optical sensors for producing a fault signal if the signal derived from the second of said two in quadrature sensors has by itself failed, and third fault detector means connected to said direction level signal, to said clock pulse signal, and to the signal from said optical sensor for said data track, for providing a fault signal if the signal derived from said data track sensor has by itself failed.

27. The system of claim 23 wherein there is a single timing track and its two associated light sources and light sensing means are spaced apart at a predetermined distance to provide said quadrature.

28. The system of claim 23 wherein there are two timing tracks, staggered with respect to each other to provide said quadrature, each having one said light beam and one said optical sensor associated therewith.

29. The system of claim 23 wherein said light-passing stripes comprise reflecting stripes, reflecting light from said beams to said optical sensor.

30. The system of claim 29 wherein said band is supported by said movable member perpendicular to the plane of said movable member and is cylindrical.

31. The system of claim 29 wherein said band is a flat annulus supported by said movable member parallel to the plane of said movable member.

32. The system of claim 23 wherein said light passing stripes are transparent stripes through which said light beams pass en route to said optical sensor.

33. The system of claim 32 wherein said band is cylindrical and is supported by said movable member.

34. The system of claim 32 wherein said band is a flat annulus supported by said movable member.

35. An encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member, including in combination: an encoding band supported by one said member, said band having a single timing track for timing and incremental encoding, comprising a series of alternating dark stripes and reflective stipes, and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, a read head supported by the other said member and having three optical sensors, two for said timing track, being spaced apart to provide quadrature, and one for said data track, each said optical sensor including a light source coupled with light sensing means for generating an analog signal voltage, timing track phase comparator means connected to said two timing-track optical sensors for comparing the phase relationship of their two signal voltages and producing a direction level signal and a clock pulse signal, up/down counter means connected to said direction level and clock pulse signals and having readout means for indicating the current position of said movable member, left-motion serial shift register means connected to said direction level and clock pulse signals and also to said third optical sensor, right-motion serial shift register means connected to said direction level and clock pulse signal and also to said third optical sensor, two framing gating means, each connected to said clock pulse signal and to one of said shift register means, absolute position register means connected to both said framing gatings, and position comparator means connected to said absolute position register means and to said up/down counter means.

36. The system of claim 35 wherein said optical sensors send their voltage signals through an operational amplifier and a Schmitt trigger en route from said two optical sensors to said phase comparator means and, from said third optical sensor en route to said shift register means.

37. The system of claim 35 whereby said signals from said framing gating means pass through respective decoding means and error checking means and from there to absolute position multiplexer means, while en route to said absolute position register means.

38. The encoding system of claim 35 having a first fault detector circuit connected to the signals from said three optical sensors and producing a fault signal if the signal derived from the first of said two optical sensors for said timing track has by itself failed, a second fault detector connected to the signals from said three optical sensors and producing a fault signal if the signal derived from the second of said two optical sensors for said timing track has by itself failed, and a third fault detector connected to said direction level signal, to said clock pulse signal, and to the signal from said optical sensor for said data track and providing a fault signal if the signal derived from said data track sensor has by itself failed.

39. The system of claim 35 wherein said band is cylindrical and is supported perpendicular to the plane of said movable member by its said member.

40. The system of claim 35 wherein said band is a flat annulus and is supported parallel to the plane of said movable member by its said member.

41. An encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member, including in combination: an encoding band supported by one said member, said band having a single timing track for timing and incremental encoding comprising a series of alternating opaque stripes and transparent stripes, and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, three light source means supported by the other said member and directing respective beams toward said band, a read head supported by said other member and having three optical sensors on the opposite side of said band from said light source means, two for said timing track, being spaced apart to provide quadrature, and one for said data track, each said optical sensor including light sensing means for generating an analog signal voltage, timing track phase comparator means connected to said two timing-track optical sensors for comparing the phase relationship of their two signal voltages and producing a direction level signal and a clock pulse signal, up/down counter means connected to said direction level and clock pulse signals and having readout means for indicating the current position of said movable member, left-motion serial shift register means connected to said direction level and clock pulse signals and also to said third optical sensor, right-motion serial shift register means connected to said direction level and clock pulse signal and also to said third optical sensor, two framing gating means, each connected to said clock pulse signal and to one of said shift register means, absolute position register means connected to both said framing gating means, and position comparator means connected to said absolute position register means and to said up/down counter means.

42. The system of claim 41 wherein said optical sensors send their voltage signals through an operational amplifier and a Schmitt trigger en route from said two sensors to said phase comparator means and, from said third optical sensor en route to said shift register means.

43. The system of claim 42 whereby said signals from said framing gating means pass through respective decoding means and error checking means and from there to absolute position multiplexer means, while en route to said absolute position register means.

44. In a system for determining the absolute and the incremental positions of a movable telescope dome of a dome building having a stationary portion, the combination of: a circular encoding band supported by said movable dome, said band having a timing track for timing and incremental encoding comprising at least one series of narrow stripes all around said band, alternating light-absorbing and light-passing stripes, and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, and a read head supported by said stationary portion of said dome building and having three optical sensors aimed at said band, two for said timing track, being spaced apart a predetermined distance for quadrature, and one for said data track, each said optical sensor including a light-emitting diode coupled with a phototransistor for generating an analog signal voltage from the transmission of the light-emitting diode light from said tracks, said read head being mounted on wheels in rolling contact with said band and mounted to said stationary portion through a parallelogram linkage and spring-urged toward a set distance from said band, so that out-of-roundness of said dome is compensated.

45. The system of claim 44 wherein said read head is pivotally mounted relative to said parallelogram linkage.

46. The system of claim 44 wherein there is a single timing track and its two associated light sources and light sensing means are spaced apart at a predetermined distance to provide said quadrature.

47. The system of claim 44 wherein there are two timing tracks, staggered with respect to each other to provide said quadrature, each having one said light beam and one said optical sensor associated therewith.

48. An optical encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member, including in combination: three light source supported by one said member for emitting three narrow light beams on respective paths, an encoding band supported by the other said member in the paths of said light beams, said band having a single timing and incremental encoding track comprising at least one series of alternating light-blocking stripes and light-passing stripes, and a single data track comprising a series of absolute position tags disposed at regular intervals along said data track and alternating with spacers that are identical to each other, each said tag being framed by start bits and being distinctly different from every other said tag and each comprising a serial bit stream for encoding the absolute position corresponding to the location of its tag along said data track, and a read head supported by said one said member and having three non-sweeping optical sensors, one for each said light beam, two in quadrature with each other for said timing and incremental encoding track, and one for said data track, each said optical sensor including light sensing means for generating an analog signal voltage, said members being capable of both discrete and continuous motion in both directions around a circular path, whereby said absolute and incremental positions can be determined solely from said system without dependence or any additional information external to said system.

49. An encoding system for determining the absolute and the incremental positions of a movable telescope dome of a dome building having a stationary portion, including in combination: an encoding band supported by a generally cylindrical surface of said movable dome around the complete circle thereof, said band having a timing track for timing and incremental encoding comprising a series of narrow stripes all around said band, alternating dark and reflective stripes, and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, and a read head supported by said stationary portion of said dome building and having three optical sensors aimed at said band, two for said timing track, being spaced apart a predetermined distance, and one for said data track, each said optical sensor including a light-emitting diode coupled with a phototransistor for generating an analog signal voltage from the reflection of the light-emitting diode light from said tracks, said read head being mounted to its rigid support through a parallelogram linkage, and spring means urging it toward a set distance from said band, so that out-of-roundness of said dome is compensated.

50. An encoding system for determining the absolute and the incremental positions of a movable telescope dome of a dome building having a stationary portion, including in combination: an encoding band supported by a generally cylindrical surface of said movable dome around the complete circle thereof, said band having a timing track for timing and incremental encoding comprising a series of narrow stripes all around said band, alternating dark and reflective stripes, and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, and a read head supported by said stationary portion of said dome building and having three optical sensors aimed at said band, two for said timing track, being spaced apart a predetermined distance, and one for said data track, each said optical sensor including a light-emitting diode coupled with a phototransistor for generating an analog signal voltage from the reflection of the light-emitting diode light from said tracks, said read head being mounted on wheels in rolling contact with said cylindrical surface.

51. An encoding system for determining the absolute and the incremental positions of a movable telescope dome of a dome building having a stationary portion, including in combination: an encoding band supported by a generally cylindrical surface of said movable dome around the complete circle thereof, said band having a timing track for timing and incremental encoding comprising a series of narrow stripes all around said band, alternating dark and reflective stripes, and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, a read head supported by said stationary portion of said dome building and having three optical sensors aimed at said band, two for said timing track, being spaced apart a predetermined distance, and one for said data track, each said optical sensor including a light-emitting diode coupled with a phototransistor for generating an analog signal voltage from the reflection of the light-emitting diode light from said tracks, timing track phase comparator means connected to said two timing-track optical sensors for comparing the phase relationship of their two signal voltages and producing a direction level signal and a clock pulse signal, up/down counter means connected to said direction level and clock pulse signals and having readout means for indicating the current dome position, dome left-motion serial shift register means connected to said direction level and clock pulse signals and also to said third optical sensor, dome right-motion serial shift register means connected to said direction level and clock pulse signal and also to said third optical sensor, two framing gating means, each connected to said clock pulse signal and to one of said shift register means, absolute position register means connected to both said framing gating means, and position comparator means connected to said absolute position register means and to said up/down counter means.

52. The system of claim 51 wherein said optical sensors send their voltage signals through an operational amplifier and a Schmitt trigger en route from said two optical sensors to said phase comparator means and, from said third optical sensor en route to said shift register means.

53. The system of claim 51 whereby said signals from said framing gating means pass through respective decoding means and error checking means and from there to absolute position multiplexer means, while en route to said absolute position register means.

54. The system of claim 51 having first fault detector means connected to the signals from said three optical sensors for producing a fault signal if the signal derived from the first of said two timing track optical sensors has by itself failed, second fault detector means connected to the signals from said three optical sensors for producing a fault signal if the signal derived from the second of said two timing track optical sensors has by itself failed, and third fault detector means connected to said direction level signal, to said clock pulse signal, and to the signal from said optical sensor for said data track, for providing a fault signal if the signal derived from said data track sensor has by itself failed.

55. An encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member, including in combination: a circular encoding band on said movable member, said band having a single timing track means for timing and incremental encoding comprising a series of alternating opaque stripes and transparent stripes, and a single data track comprising a series of absolute position tags disposed at regular intervals along said data track and alternating with spacers that are identical to each other, each said tag being framed by start bits and distinctly different from every other said tag and each comprising a serial bit stream for encoding the absolute position corresponding to the location of its tag along said data track, three light source means mounted on said stationary member and directing respective beams toward said band, and a read head secured to said stationary member and having three non-sweeping optical sensors on the opposite side of said band from said light source means, two for said timing track, being spaced apart, and one for said data track, each said optical sensor including light sensing means for generating an analog signal voltage, said members being capable of both discrete and continuous motion in both directions around a circular path, whereby said absolute and incremental positions can be determined solely from said system without dependence or any additional information external to said system.

56. An encoding system for determining the absolute and the incremental positions of a movable member relative to a stationary member, including in combination: a circular encoding band on said movable member, said band having timing track means for timing and incremental encoding comprising a series of alternating opaque stripes and transparent stripes, and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, three light source means mounted on said stationary member and directing respective beams toward said band, a read head secured to said stationary member and having three optical sensors on the opposite side of said band from said light source means, two for said timing track, being spaced apart, and one for said data track, each said optical sensor including light sensing means for generating an analog signal voltage, timing track phase comparator means connected to said two timing-track-means optical sensors for comparing the phase relationship of their two signal voltages and producing a direction level signal and a clock pulse signal, up/down counter means connected to said direction level and clock pulse signals and having readout means for indicating the current position of said movable member, left-motion serial shift register means connected to said direction level and clock pulse signals and also to said third optical sensor, right-motion serial shift register means connected to said direction level and clock pulse signal and also to said third optical sensor, two framing gating means, each connected to said clock pulse signal and to one of said shift register means, absolute position register means connected to both said framing gating means, and position comparator means connected to said absolute position register means and to said up/down counter means.

57. The system of claim 56 wherein said optical sensors send their voltage signals through an operational amplifier and a Schmitt trigger en route from said two optical sensors to said phase comparator means and, from said third optical sensor en route to said shift register means.

58. The system of claim 56 whereby said signals from said framing gating means pass through respective decoding means and error checking means and from there to absolute position multiplexer means, while en route to said absolute position register means.

59. A method for determining the absolute and the incremental positions of a movable member relative to a stationary member, comprising the steps of: sending three separate beams of light toward one said member, intercepting said beams by an encoding band secured to the other said member, said band having a timing track for timing and incremental encoding portion comprising a series of narrow alternating light-beam blocking and light-beam passing stripes and a data track comprising a series of absolute position tags alternating with spacers, each said tag being framed by start bits, two of said light beams being directed at spaced-apart loci along said timing track and one directed to a locus along said data track, generating an analog signal voltage from the light from each beam after the interception, two such voltages from said timing track and one from said data track, comparing the phase relationship of the two analog voltages from said timing track and producing direction level signals and clock pulse signals therefrom, counting the clock pulse signals, thereby indicating the current incremental position of said movable member, sending said direction level and clock pulse signals and a logic level derived from the third analog voltage to left-motion serial shift register means, and also to right-motion serial shift register means, framing the output from each said shift register means and sending the resultant output of each to absolute position register means, comparing the absolute position register values with those from said counting step, and resetting said incremental position to match the value contained in the absolute position register means, should they disagree.

60. The method of claim 59 wherein said sending step comprises additionally sending said direction level and clock pulse signals and the signals derived from the step of intercepting said three light beams to a fault detection circuit and there comparing, two at a time, the signals derived from the step of intercepting said three light beams and determining whether the signal derived from any signal from said light beam interception has by itself failed.

61. The method of claim 59 wherein said timing track comprises a single track.

62. The method of claim 59 wherein said timing track comprises two tracks.

63. The method of claim 59 wherein said intercepting step comprises impinging said separate beams of light on said encoding band, said band's having a timing and incremental encoding portion comprising a series of narrow alternating dark and reflective stripes and a data track comprising a series of absolute position tags alternating with spacers.

64. The method of claim 59 wherein said intercepting step comprises sending the three light beams through said band via a timing and incremental encoding portion comprising a series of narrow alternating opaque and transparent stripes and a data track comprising a series of absolute position tags alternating with spacers.

65. A method for encoding the absolute and the incremental positions of a movable member relative to a stationary member, comprising the steps of: sending three separate beams of light toward a stationary position, intercepting said beams by an encoding band secured to said movable member, said band having a timing track for timing and incremental encoding portion comprising a series of narrow alternating light-beam blocking and light-beam passing stripes and a single data track comprising a series of absolute position tags disposed at regular intervals along said data track and alternating with spacers that are identical to each other, each said tag being framed by start bits and being distinctly different from every other said tag and each comprising a serial bit stream for encoding the absolute position corresponding to the location of its tag along said data track, two of said light beams being directed at spaced-apart loci along said timing track and one directed to a locus along said data track, generating an analog signal voltage from the light from each beam after the interception, two such voltages in quadrature from said timing track and one from said data track, said members being capable of both discrete and continuous motion in both directions around a circular path, and whereby said absolute and incremental positions can be determined solely from said system without dependence or any additional information external to said system.

66. The method of claim 65 wherein said timing track comprises a single track.

67. The method of claim 65 wherein said timing track comprises two tracks.

68. The method of claim 65 wherein said intercepting step comprises impinging said separate beams of light on said encoding band, said band's having a timing and incremental encoding portion comprising a series of narrow alternating dark and reflective stripes and a data track comprising a series of absolute position tags alternating with spacers.

69. The method of claim 65 wherein said intercepting step comprises sending the three light beams through said band via a timing and incremental encoding portion comprising a series of narrow alternating opaque and transparent stripes and a data track comprising a series of absolute position tags alternating with spacers.

Patent Citations (5)

Patent	Date	Inventor
US3531798	1970-09-01	Dureau
US3702471	1972-11-01	Kennedy
US3982106	1976-09-01	Stutz
US4628298	1986-12-01	Hafle
US4633224	1986-12-01	Gipp

Abstract

Description