GB2066602A - Absolute position encoder - Google Patents

Abstract

An absolute position encoder comprises a scale member carrying a number of digitally-encoded tracks which together define a monostrophic code. A major trace provides two transitions and has a number of reading heads associated with it. The outputs of the reading heads may be decoded by, for example, a read-only memory.

Description

SPECIFICATION Absolute position encoder This invention relates to absolute position encoders, that is to devices for measuring position which are not dependent on measuring and storing changes in position relative to a datum.

The majority of position encoders, or devices for measuring position relative to a datum, function by measuring the movement from the last known position. Hence it is necessary to store the last known position so as to be able to add or subtract, depending from the direction of movement. Such devices, usually referred to as incremental encoders, have two main disadvantages. The first is that if for some reason the stored last known position is lost, then the device ceases to function correctly. The second disadvantage is that the speed at which the position may change is limited by the speed at which the encoder can detect and store the changes. Too fast a movement will result in data being lost.

For the reasons given above the absolute encoder has been developed. This does not rely on any stored data. Whatever its position when switched on, it can determine this position without any ambiguity. Conventional absolute encoders for measuring angular position consist of a disc carrying a number of concentric digitally-encoded tracks. Each track has an associated reading head, and the resolution of the encoder is directly related to the number of tracks. Problems arise in making very high resolution encoders due to the difficulty in actually making discs carrying a large number of tracks.

Similar problems arise with linear encoders, since for a long track length a large number of tracks may be necessary, thus presenting space problems.

It is an object of the invention to provide an absolute position encoder which will provide a given resolution with a smaller number of tracks than has hitherto been possible.

According to the present invention there is provided an absolute position encoder which includes a scale member carrying a plurality of digitally-encoded tracks comprising a major track and a number of minor tracks each extending over a major interval and together defining a monostrophic code, the major track providing only two digital transitions over the major interval; n reading heads associated with the major track and distributed at 1/2n times the major interval so as to divide the major interval into 2n equal minor intervals; a number of reading heads associated with each minor track, the number being given by an integer solution of the expression n.2r where r is an integer less than or equal to 1, the heads being equally spaced at integral multiples of 1/2n times the major interval.

The term "major interval" is used to refer to the extent of the tracks. In the case of rotating encoder this would be one complete 3600 revolution, whilst for a linear encoder the term refers to the full range of movement of the encoder.

The reference to the equal spacing of the reading heads is to be taken to include also the case where a head is displaced from its expected position by half the major interval, a displacement which produces the same results. In addition with a rotating encoder the term "equal spacing" does not include the spacing between the last and the reading heads and the first one.

The invention will now be described with reference to the accompanying drawings, in which: Figure 1 is a sectional side elevation of a rotary position encoder; Figure 2 is a schematic view of the scale member of the encoder of Figure 1; Figure 3 illustrates the production of a suitable code; Figure 4 illustrates a code for use with the encoder described, and; Figure 5 is a schematic view of a linear position encoder.

Referring now to Figure 1, this shows a sectional side elevation of a typical rotary position encoder. A housing 10 carries a bearing 11 which supports a shaft 1 2 for rotation relative to the housing. The shaft carries a disc 1 3 which rotates with the shaft. Located within, and fixed relative to, the housing 10 are three light sources 14, only one of which is visible in Figure 1. Each light source extends along a radius of the disc 13. Also located within the housing 10 are three reading heads 15, again only one being shown in Figure 1.

The reading heads are on the opposite side of the disc to the two light sources 14, and are aligned- one with each light source. Hence light passing from the light source through the disc is detected by the reading head.

Each reading head contains a number of light sensitive devices equal to the number of coded tracks formed on the disc, and a narrow collimating slit is located in front of each lightsensitive device.

The housing 10 may also contain a printed circuit board 1 6 carrying any necessary electronic circuitry necessary to the functioning of the encoder.

Figure 2 shows a plan view of the disc 13, with the positions of three reading heads shown in broken outline. The disc shown carries three tracks, which are not shown in detail, and hence each reading head comprises three detectors. In the particular embodiment to be described the three reading heads are positioned 600 apart around the disc 13.

The disc 13 may be formed from a metal plate which has portions etched away to form the required coded tracks or it may, for example, be a transparent disc with an opaque pattern formed on it.

As already stated, the disc shown in Figures 1 and 2 carries three tracks, each of which is digitally encoded. In order to avoid the difficulties of ensuring that a number of transitions on different tracks occur exactly together, the tracks are arranged to define a monostrophic code, that is a code in which the transitions occur only one at a time. The effect of reading each of two minor tracks in three different positions produces the effect of six separate tracks, that is 6-bit output. In this example, each track has the same number of reading heads, namely three.

Regardless of the coding used, one of the tracks, referred to as the "major track" provides only two transitions. If there are n reading heads associated with the major track, then these heads are distributed at 1/2n times the major interval, which in the case of a rotary encoder is one revolution or 3600. Hence in the present case when n is three, the heads are space at one sixth of a major interval, that is 600 apart. The major interval is thus divided into 6 minor intervals.

Figure 3 illustrates one way in which the disc of Figure 3 may be encoded. Basically this involves writing, or drawing, a six-bit monostrophic code, in this case a Gray code. This conveniently extends over 64 bits, and if only 60 of these are used, each bit may correspoond to 1 O of revolution.

Figure 5 shows the three encoded tracks developed from the code of Figure 4. As already stated, the major track has only two transitions in 3600 One of the two minor tracks is used to carry the three least significant bits of the code, whilst the other minor track carries the three most significant bits. Each successive transition develops a parallel output from the six reading heads which is unique for each 1 0 increment of a revolution of the disc.

The output from a position encoder is required to be in degrees, or some other angular measure, in the case of a rotary shaft encoder. It is therefore necessary to process the outputs from the reading heads in order to provide the appropriate output.

Since each successive output is unique over a major interval, the simplest form of processor is a read-only memory (ROM) which is programmed to recognise and decode each possible parallel output from the reading heads. This form of decoding also means that the code used need not be continuous. For example, the six-bit Graay code referred to above extends over 64 transitions.

However only 60 of these are needed to provide 1 0 increment steps over 600 of movement. Hence four of the transitions may be omitted. Although other forms of decoding may be used, such as complex logic arrangements, problems may arise if such jumps are used.

As already suggested, the encoder need not be of the rotary type. A lineary encoder may be developed using exactly the same principles. In such a case, the major interval is the maximum expected travel of the machine or apparatus to which the encoder is attached. In the case of machine tool, for example, in which two body members move relative to one another the scale member would extend for the full length of one of the body members. The other body member would carry the light sources and reading heads, on opposite sides of the scale member. The spacing between the heads would still be determined by the number of reading heads on the major track, in accordance with the expression given above.

Figure 5 shows such a linear encoder in schematic form. If 1 is the major interval, and there are n reading heads on the major track, then the spacing between the reading heads is 1/2n. Hence, if n equals three, then the reading heads are 1/6 apart.

The tracks will be encoded in exactly the same way as already described. Indeed Figure 4 is a development of the track pattern for a rotary encoder, and may be used directly as a linear encoder.

In both the rotary and the linear encoders, higher resolution may be obtained by using a higher order code. It is then necessary either to increase the number of tracks, or the number of reading heads per track, or both. It is not essential that each track has the same number of reading heads, though the choices are limited. In general, each track may have a number of reading heads given by an integer solution of the expression n x 2 where r is an integer less than or equal to 1. Thus r may be 1,O,-1,-2 and so on. Hence, for example, if the major track has two reading heads, then any minor track may have four, or two, or only one. If the major track has three reading heads, then any minor track may have six or three reading heads.

The maximum obtainable resolution of an encoder, that is the maximum number of increments into which a major interval may be divided is given by the expression 2n x 2', where n is the number of reading heads on the major track, and t is the total number of reading heads used on all the minor tracks. Hence, in the three-head, three-track embodiment of Figure 2, the maximum obtainable resolution is 384. This is not convenient number for a rotary encoder, and hence in practice one would probably use only 360 of the possible increments, as had already been described. The minimum resolution must be a multiple of 2n, giving 360 increments as a possible solution.

The resolution obtained by any particular track arrangement may be reduced by suitable arrangement of the decoder. For example, the track arrangement of Figure 2 may be used to give 100 different outputs over one completion revolution by arranging the ROM so that several successive parallel inputs to the ROM produce the same output from the ROM.

Although it has been stated above that the reading heads must be spaced at 1/2ntimes the major interval, this may be varied. It is possible to move a reading head from its normal position by half the major interval. The net result will be exactly the same, since the major track has only two transitions. This is sometimes useful if it is desired to space the tracks by a distance which is too small to accommodate ali the reading heads.

The above description has referred to light sources and light-sensitive devices for the reading heads, with light either passing from the sources to the sensors or being blocked by the scale member. It is equally possible to use a scale member which either reflects or absorbs light, and hence to located the light sources and the sensors on the same side of the scale members.

Other forms of reading head may be used, relying on the detection of other forms of electromagnetic radiation, on the detection of electric or magnetic fields, or on the detection of electrical potentials.

Claims (9)

1. An absolute position encoder which includes a scale member carrying a plurality of digitallyencoded tracks comprising a major track and a number of minor tracks each extending over a major interval and together defining a monostrophic code, the major track providing only two digital transitions over the major interval; n reading heads associated with the major track and distributed at 1/2n times the major interval so as to divide the major intervals into 2n minor intervals, a number of reading heads associated with each minor track, the number being given by an integer solution of the expression n 2' where r is an integer less than or equal to 1, the heads being equally spaced at integral multiples of 1/2n times the major interval.

2. An encoder as claimed in Claim 1 for the measurement of angular position, in which the scale member comprises a disc carrying a number of concentric tracks.

3. An encoder as claimed in Claim 1 for the measurement of linear position, in which the scale member comprises a strip carrying a number of parallel tracks.

4. An encoder as claimed in any one of Claims 1 to 3 in which the scale member is formed from an opaque material having apertures formed therein to define the tracks.

5. An encoder as claimed in any one of Claims 1 to 3 in which the scale member is formed from a translucent material having areas rendered opaque to define the tracks.

6. An encoder as claimed in any one of Claims 1 to 3 in which the scale member has areas rendered reflective to define the tracks.

7. An encoder as claimed in any one of the preceding claims in which each reading head comprises a light source and an associated light, sensitive device together arranged to detect the pattern on a track of the scale member.

8. An encoder as claimed in any one of Claims 1 to 7 in which the monostrophic code defined by the plurality of tracks is a Gray code.

9. An absolute position encodes substantially as herein described with reference to Figure 1 to 4 or Figure 5 of the accompanying drawings.