GB2236847A - Incremental position transducer - Google Patents

Abstract

A transducer is provided for determining the extent and direction of movement of a first member (10) relative to a second member (11). The first member carries a scale grating (12) having a pattern of measuring marks extending along the measuring direction. The second member carries a reading head (13) having a window (14) positioned close to the scale grating (12). A linear array (15) of at least three radiation-sensitive detectors arranged in a regular pattern is mounted behind the window (14) and radiation passing through the window is incident upon the detector array (15). Circuit means are povided responsive to the outputs of the detector array (15) to define a waveform matching the detector outputs and to analyse the phase of said waveform so as to indicate changes in the relative position of the first and second members. The window (14) may be replaced by an index grating (20, Figure 2). The transducer may be used to measure the position of a machine tool slide. <IMAGE>

Description

INCREMENTAL POSITION TRANSDUCER Incremental position transducers are used to measure the linear or angular movement of a body from an arbitrary start position. Many techniques are known, several using optical methods involving the use of gratings with optical reading heads. In the case of a linear transducer of this type for use on, say, a machine tool slide, it is common practice to attach a long grating scale to one part of the machine, say the bed. A short length of grating, known as the index, is attached to the other part of the machine and arranged to move'along the scale close to it. A light source and reading head are provided to detect the Moire fringes or similar pattern formed by the two gratings using light passing through the two gratings from source to reading head or reflected from one grating.The detector usually comprises four light-sensitive detectors arranged to be spaced apart by one-quarter of the pattern wavelength and connected in alternate pairs to provide two signals for processing. Typically these signals are 900 apart so that the distance and direction of movement may be determined. The distance is determined by counting complete cycles. from the arbitrary start point and interpolating within the grating pitch.

The problems associated with the accurate interpolation of the grating signal and reliable cycle counting are many, usually resulting from an asymmetric reduction in the intensity of light reaching the detectors. Dirt of any sort on or between the gratings, scratches or other blemishes on the gratings, mechanical misalignment and distortion are only a few of the potential sources of error. Techniques have been developed to overcome or compensate for many of these errors, but these techniques have their own problems. Accurate systems tend to be complex and expensive and systems which are immune to one error tend to be sensitive to another.

It an an object of the present invention to provide an incremental position transducer which is insensitive to the normal causes of interpolation and cycle counting errors.

According to the present invention there is provided an incremental position transducer for determining the extent and direction of the movement of a first member relative to a second member moving in close proximity thereto in a measuring direction, which transducer includes a scale grating carried by one of said two members for movement therewith and carrying a pattern of measuring marks extending along the measuring direction, a linear array of radiation-sensitive detectors carried by the other of said two members for movement therewith in close proximity to the scale grating and including at least three detectors arranged in a regular pattern so as to interact with the pattern - formed by the measuring marks upon the sensitive surface of the detector array, and circuit means responsive to outputs of the detector array to define a waveform matching said outputs and to analyse the phase of said waveform so as to indicate changes in the relative positions of the first and second members.

The invention will now be described with reference to the accompanying drawings, in which: Figure 1 is a schematic diagram showing the relatively movable members and reading head according to one embodiment of the invention; Figure 2 is a similar diagram of the relatively movable members and reading head according to a second embodiment; Figure 3 is an expanded isometric view showing detail of the reading head of Figure 2; Figure 4 is a schematic block diagram of one form of the circuit means; and Figure 5 is a schematic block diagram of part of the circuit means of Figure 4; Referring now to Figure 1, this shows a side view of the mechanical parts of a transducer as used on a machine having two members 10 and 11 capable of relative motion parallel to and close to one another.Member 10 may, for example, be the bed of a machine tool whilst the member 11 is part of a slide or carriage moving on guides or bearings supported by the bed 10.

The bed carries a linear scale grating 12 which may conveniently be a metal strip carrying a continuous pattern of measuring marks and extending in the direction of movement and over the full range of movement of the carriage 11. The measuring marks may, in one form, be a series of close-spaced parallel lines substantially perpendicular to the direction of movement of the carriage 11.

The carriage 11 carries a reading head comprising a housing 13 which has a window 14 in the face adjacent to the scale grating 12 and is preferably sealed to prevent the ingress of unwanted ambient light. Light passing through the window 14 is incident on a linear array of light-sensitive detectors 15 arranged in a regular pattern. The arrangement shown is a Vernier system.

The scale grating 12 needs to be illuminated in the region adjacent to the reading head. This may be the result of ambient illumination or of light passing through a transparent scale grating 12 from a suitable source (not shown).

In the arrangement of Figure 1 light from the pattern of measuring marks carried by the scale grating 12 is incident upon the sensitive surface of the detector array 16. Since th elements of the array themselves form a regular pattern there is interaction between the two patterns. This results in differing electrical outputs from the different elements of the array. This arrangement becomes more sensitive as the number of elements in the array is increased.

A more common arrangement, having greater similarity to previously-used systems, is shown in Figure 2. This differs mainly in the inclusion of a transparent index grating 20 taking the place of window 14 in the first embodiment. The index grating 20 is a short length of grating similar to scale grating 12 and carrying a similar pattern of measuring marks. The marks may be at a slightly different pitch from those on the scale grating 12 so as to generate a Vernier effect. The optical efficiency of the arrangement may be futher increased by including a cylindrical lens 21 arranged to focus light passing through the index grating 20 on to the detector array. The interaction between the measuring marks on the scale grating and those on the index grating produces a further pattern which itself interacts with the pattern of detector elements.

Figure 3 shows the optical components of Figure 2 in an exploded view. The illumination of the scale grating 12 is denoted by the arrows 30 and the drawing also shows how the light passing through index grating 20 is focussed to a line focus on the sensitive area of the array 15 by the cylindrical lens 21.

Using the form of grating described above a nominally sinusoidal waveform of light intensity is produced along the length of the index grating 20 and hence the intensity of light received by the separate detector elements of the array 15 will vary substantially periodically along the length of the array.

Figure 4 is a block diagram of one form of circuit suitable for processing the output of the detector array, shown here at 40. Such an array may comprise a large number of separate detectors and the outputs from those detectors pass, in digital form, to a head processing circuit 41 which will be described in detail later. The stream of bits from the processing circuit 41 are stored in a RAM 42 acting as a buffer. From the buffer the digital data is passed to a waveform analyser 43 in which the phase of the relevant waveform is determined. Analysers of this type are well known, the simplest being a processor using Fourier analysis software, though other more sophisticated forms exist. The phase data from the analyser 43 is passed to a processor 44 using an algorithm to determine the position of the reading head, and hence of the carriage, relative to a datum position. A simple form of the algorithm, could be, for example, d,QXP 360 where d is the distance moved between successive readings of the array, 0 is the change in phase between successive readings, in degrees, and p is the grating pitch.

The output from the processor 44 is a series of numbers each representing an incremental change of relative position and an indication of the direction of that increment. Passing these to a counter 45 therefore gives an indication of change of position relative to the arbitrary datum.

The head processing circuit 41 is shown in more detail in Figure 5. This is a digital circuit which, for the sake of explanation, is for use with a 100 element linear array. In practice arrays are more likely to have 256 elements.

The detector array 50 receives the optical radiation via the cylindrical lens and has two clock inputs. The 5MHz clock input steps around the elements of the detector array 50, causing it to provide as an output a sequence of analogue signals representing the cumulative radiation received by each detector element in the period since the last cycle. In this example the period is 20s and corresponds with the 50KHz clock pulse train which synchronises with the loading of the first detector array element. The analogue output signals from the array 50 are applied to an 8-bit analogue-to-digital converter 51, which produces an output as an eight-bit parallel binary signal at 5MHz. This is applied to an 8-bit parallel in/serial out converter 52. The converter output, at 40MHz, is applied to the buffer RAM 42 of Figure 4.The remaining elements shown in Figure 5 are the 40 MHz clock pulse generator 54 and other circuit elements necessary to derive other clock pulse rates.

As already indicated, the main sources of error in this type of position transducer are due to dirt or scratches on the scale grating. The conventional four-detector Vernier or Moire fringe transducer, for example, was placed in considerable difficulty in such a situation because this resulted in a false output from one of the two pairs of detectors with consequent interpolation errors or even a cycle count error.

Conventional reading heads have four detectors coupled in differential back-to-back pairs. This gives immunity to several sources of error, such as changes in absolute light level, but the interpolation process is dependent upon a single parameter of tan~1 (90 /O). Therefore any defect which is not cancelled by the differential coupling, or by the ratio inherent in the tan 1 factor, remains uncorrected and causes interpolation errors. Examples of such error sources range from a non-sinusoidal waveform to unbalanced signals.

Imbalance can easily cause interpolation errors as large as 20b of the scale grating pitch, and in extreme cases may prevent correct cycle counting completely by appearing to oscillate by +/- 1/4 cycle or less rather than progress when the reading head is moving continuously.

The invention replaces back-to-back differential pair with equivalent but better signal processing techniques. For example, by evaluating the raw signals from all detectors, redundant information becomes available. The more detectors are used the better the information and the minimum number of detectors is three. By analysing the detector signals to determine the phase of the periodic waveform many errors associated with conventional reading heads are eliminated. For example, the waveform does not need to be a perfect sinusoidal waveform since the analysis produces a waveform which best fits all the detector outputs. Non-uniformity in light levels has no effect because the analysing algorithms may compensate for the mean pattern.The waveform analysis processor attempts to define a relevant waveform which best fits all of the separate detector outputs, thus avoiding the need for the output to relate to a perfect periodic waveform. False readings from a small proportion of the detectors will not affect the analysis and hence no accuracy is lost. By way of example only, consider a 256-element detector array viewing 2-3 cycles of a periodic waveform along a 10mm length of index grating. If a blemish lmm wide exists on the grating then 26 out of 256 detectors will provide incorrect signals whilst the remainder give correct signals. The waveform analysis circuit is able to filter out the incorrect signals with ease. Signal changes due to variations in light levels are also easily corrected.

The invention does not necessarily require the linear array of detectors to be provided with focussing means in the form of the cylindrical lens described above, though if a large number of detectors is required in a small space then this arrangement is to be preferred. However, discrete detectors each with a separate lens to focus a point of the pattern on to it may be used.

Similarly, as already stated, Moire fringe or Vernier techniques do not need to be used and the periodic waveform produced by the interaction of the scale and index gratings need not be a sinusoidal waveform. Known waveform analysis arrangements will work with any periodic waveform.

The scale grating may be mounted in a position where a transparent grating may be used, in which case grating illumination may be provided behind the scale grating.

In the above description it has been considered that the relative movement between the first and second members is linear. However, as with known forms of transducer using gratings the invention is equally applicable to the measurement of angular movement between two members which remain at a constant distance from one another.

Claims (8)

Claims:

1. An incremental position transducer for determining the extent and direction of the movement of a first member relative to a second member moving in close proximity thereto in a measuring direction, which transducer includes a scale grating carried by one of said two members for movement therewith and carrying a pattern of measuring marks extending along the measuring direction, a linear array of radiation-sensitive detectors carried by the other of said two members for movement therewith in close proximity to the scale grating and including at least three detectors arranged in a regular pattern so as to interact with the pattern formed by the measuring marks upon the sensitive surface of the detector array, and circuit means responsive to outputs of the detector array to define a waveform matching said outputs and to analyse the phase of said waveform so as to indicate changes in the relative positions of the first and second members.

2. A transducer as claimed in Claim 1 which includes an index grating carried by the other of said two members between the scale grating and the detector array and carrying a pattern of measuring marks similar to that carried by the scale grating, the interaction between the pattern carried by the scale grating and that carried by the index grating producing a pattern upon the sensitive surface of the detector array.

3. A transducer as claimed in either of Claims 1 or 2 which includes focussing means for focussing radiation from selected regions of the periodic waveform on to different ones of the detectors.

4. A transducer as claimed in Claim 3 in which the focussing means comprise a cylindrical lens located between the index grating and the detector array to form a line image of selected regions of the periodic waveform on the detector array.

5. A transducer as claimed in any one of Claims 1 to 4 which includes a source of radiation operable to illuminate that area of the scale grating in close proximity to the index grating.

6. A transducer as claimed in any one of the preceding claims in which the circuit means is operable to determine the instantaneous phase of the periodic waveform relative to a reference.

7. A transducer as claimed in any one of Claims 1 to 6 which includes counter means responsive to the output of the circuit means to determine the cumulative distance moved by the second member relative to the first member.

8. An incremental position transducer substantially as herein described with reference to the accompanying drawings.