WO1994025830A1

WO1994025830A1 - Opto-electronic scale reading apparatus

Info

Publication number: WO1994025830A1
Application number: PCT/GB1994/000933
Authority: WO
Inventors: William Frank Noel Stephens; Raymond John Chaney; Colin Keith Howley; David Roberts Mcmurtry
Original assignee: Renishaw Plc
Priority date: 1993-05-01
Filing date: 1994-04-29
Publication date: 1994-11-10

Abstract

A reference mark for an opto-electronic incremental scale and readhead apparatus is provided by an elongate cylindrical depression in the surface of the scale, which extends parallel to the lines of the scale. The depression generates an image of a light source (provided in the readhead), which is detected at a pair of reference photodetectors (which are also provided in the readhead). As a consequence, the relative velocity of the image and the photodetectors is twice the relative velocity of the scale and the readhead. A plurality of such depressions are provided along the length of the scale, and a given depression may be selected to provide the reference mark by the placing of a non-reflective selector tab over one end of the depression, which prevents the reference photodetector lying in register with the tab from generating a signal.

Description

OPTO-ELECTRONIC SCALE READING APPARATUS

The present invention relates to an opto-electronic scale reading apparatus including a scale defined by a series of spaced-apart lines and a readhead adapted to measure its displacement relative to a reference point on the scale.

An opto-electronic scale reading apparatus is known from European Patent No. 207121, which discloses an embossed scale together with a readhead which projects light onto the scale and generates interference fringes from the interaction of light reflected off the scale with gratings provided inside the readhead. Relative movement of the scale and the readhead results in a movement of the interference fringes relative to the readhead which may be detected with one or more photodetectors. A combination of a plurality of such photodetectors and electronic processing circuitry incorporating an interpolator, generate a series of digital quadrature pulses which are sent to an incremental counter whose instantaneous total represents relative displacement from the reference point; the order in which the quadrature pulses are received is indicative of whether the total in the counter at that time should be incremented or decremented.

A first aspect of the present invention seeks to provide an absolute reference mark for an interference type opto¬ electronic scale reading apparatus. Such a reference mark will enable the readhead to generate a pulse which resets the counter to zero. The provision of an absolute reference mark enables, for example, error maps associated with the scale to be initialised from the correct position. Such reference marks are known per se in connection with "virtual contact" opto-electronic scale reading apparatus, such as the MINILIDA 150 manufactured by Dr. Johannes Heidenhain GmbH. This system employs a pseudo random pattern of light reflective and non-reflective portions on the scale known as an autocorrelation reference mark. A window in the readhead has a correspondingly configured series of light transmitting and opaque portions. A photodetector positioned behind the window in the readhead outputs a sharply peaked electrical signal when the window in the readhead lies in register with the autocorrelation reference mark. This reference mark system is however only easily applicable to virtual contact opto-electronic scale reading apparatus, where the clearance or "ride height" between the readhead and the scale is of the order of 100 microns. Scale and readhead apparatus which operate on the basis of interference fringes typically have a ride height of between 3 and 10mm.

According to a first aspect of the present invention, an opto-electronic scale reading apparatus includes a reference mark in the form of an imaging element on the scale, which element is configured to re-direct light from a light source in the readhead in such a way that the re-directed light is convergent in the direction of spacing of the scale lines, and forms an image of the light source in the readhead at a predetermined ride height above the scale.

Preferably, the imaging feature is provided by one of: a concave indentation, a Fresnel lens, a Fresnel zone plate or a lens provided on the scale.

In one embodiment of the present invention the scale is a reflective tape scale embossed in accordance with the method described in our European Patent 383901, and the imaging feature on the scale is also embossed. The imaging feature may be either one or two dimensional.

The image generated by the feature may be detected typically by one or more photodetectors or a quad cell which in conjunction with suitable processing electronics generate a digital reference pulse. When this reference pulse coincides with a unique pulse emitted by the interpolator once per cycle of quadrature signals, the value of the counter is reset to zero. Achieving coincidence between the digital reference pulse and the unique pulse from the interpolator is frequently difficult, since the timing of the unique pulse is set at a predetermined point in the quadrature cycle. In the prior art, mechanical yaw adjustment of the readhead relative to the scale is often performed as a means of achieving coincidence. Such a mechanical adjustment will alter the timing between the emission of the reference pulse and the unique pulse from the interpolator, and will thus enable the reference pulse to be centred with respect to the unique pulse. However, such a mechanical adjustment is inconvenient and difficult to perform.

Accordingly, a second aspect of the present invention provides opto-electronic scale reading apparatus having a scale and readhead, the apparatus having an absolute track including a reference mark on the scale and a detector in the readhead for producing a peaked electrical output when the reference mark and readhead are in a predetermined spatial relationship, and an incremental track wherein the readhead reads the scale to generate therefrom a pair of sinusoidally varying signals having quadrature relationship, characterised in that a plurality of pairs of said quadrature signals are generated where each pair of said quadrature signals has a different phase with respect to the reference pulse.

By selecting the appropriate pair of quadrature signals, coincidence may be achieved between the reference pulse and the unique pulse from the interpolator without the need for mechanical adjustment of the readhead relative to the scale. Embodiments of the invention will now be described, by way of example, and with reference to the accompanying drawings, in which:

Fig 1 shows a plan view of a scale according to an embodiment of the present invention;

Fig 2 shows a perspective view of the imaging feature shown in Fig 1;

Fig 3 shows a section through a readhead used in conjunction with the scale of Figs 1 and 2; Fig 4 shows a section on IV-IV in Fig 3;

Figs 5A-D are signal diagrams illustrating the output signals of the photodetectors in the readhead in Figs 3 and 4 and processing thereof;

Fig 6 is a functional diagram illustrating the timing adjustment of absolute and incremental tracks in accordance with a second aspect of the present invention; and

Fig 7 shows an interpolator.

Referring now to Figs 1 and 2, a scale 10 has three tracks: a reference track 12, incremental track 14 and a selector track 16. The incremental track comprises a series of substantially parallel light reflecting lines 18, extending in a direction defined as the X direction, and spaced-apart in a direction defined as the Y direction. In one example of scale the light reflective lines are provided by a series of elongate angled facets embossed upon a substrate by a mangling procedure; such a scale is described in our prior published European Patent Application No. 274492. A readhead mounted in register with the scale and offset therefrom in the Z direction generates from the incremental track 14 an incremental count representing the magnitude and direction of relative movement of the scale and the readhead relative to a predetermined datum or reference point at which the counter reads zero.

According to the present invention the reference or datum position is provided by an imaging feature on the scale which reflects light projected from a light source in the readhead, causing the reflected light to converge in the Y direction and generate an image of the light source at a height above the scale corresponding approximately to the ride height of the readhead. In this illustrated embodiment of the invention the reference mark is provided by an elongate depression 20 extending across the scale in the X direction and having an arcuate profile corresponding to a section of a circle. The image formed by the depression 20 is detected at two distinct sets of detectors in the readhead, one of which lies in register with the reference track 12, and the other of which lies in register with the selector track 16. Reference marks in the form of depressions 20 are provided on the scale at intervals of approximately 20mm, and an individual reference mark is selected by fixing a non-reflective selector tab 22 over the part of the reference mark in register with the selector track 16. The depression 20 may be provided by stamping the scale 10 at appropriate intervals with a cylindrical roller of a suitable diameter. In the case of a mangled scale, this is preferably done after mangling of the scale 10.

Referring now to Figs 3 and 4, the readhead 30 includes a light source provided by an LED 32 which projects a beam of light 34 onto the scale 10. The beam 34 is reflected from the scale 10 and passes through an index grating 36, provided on a glass substrate 38, which interacts with the light reflected from the scale to create a pattern of interference fringes extending in the X direction, and spaced apart in the Y direction at an electrograting detector 40. Such detectors are known per se and are more fully described in our co-pending European Patent Application No. 92309807.3. The LED 32 also produces a further light beam 50 (in fact, light beams 34,50 are all part of a single widely divergent beam from LED 32; the beams are described herein as being separate and distinct for the purposes of illustration only) which passes through an aperture 52. The beam 50 comprises two sub-beams 50A, and 50B which extend from LED 32 in the negative and positive X directions respectively (as illustrated best in Fig 4) .

The LED 32 is positioned at a height above the scale 10 which corresponds approximately to the radius of curvature of the circular section of the depression 20. The reference mark provided by the depression thus generates what may be described as a one dimensional image in respect of each part beam 50A,B at detectors 54,56 respectively (the image being "one dimensional" due to the one dimensional nature of the imaging feature, i.e. it images in only one direction) . Because the LED 32 is positioned approximately at the centre of curvature of the depression 20, the image will be approximately the same size as the

LED light source (although this is not essential) . Because the light source and photodetectors are both provided on the readhead and thus move together, the image formed by the cylindrical depression traverses the photodetector at twice the relative velocity of scale and readhead.

Detector 54 is a split detector whose outputs are shown in Fig 5A. Detector 56 is a simple photodetector the output of which is not illustrated. Processing circuitry (not shown) associated with the readhead processes the outputs of detectors 54,56 to generate a gate pulse when no output is obtained from the detector 56 due to the presence of the non-reflective selector tab 22 (Fig 5B) , and the sum of the outputs from the individual cells of split detector 54 exceeds a predetermined threshold. This threshold is normalised by reference to the instantaneous DC light intensity (as determined e.g. from the total output of the electrograting 40) , by setting the threshold at a predetermined fraction thereof. The processing circuitry also generates a further analogue output, shown in Fig 5C, which is the difference between the outputs from the individual cells of split detector 54. When (at the instant T₀) this output has a zero value and the gate pulse is present, the processing circuitry emits a reference pulse (Fig 5D) . The gate pulse thus acts as an enabling pulse, and prevents the emission of a reference pulse due to isolated signal fluctuations.

The reference pulse is used to select the subsequent incremental pulse to reset the counter to zero.

As mentioned above, the reference mark may be provided by any suitable imaging feature on the scale such as a lens, a Fresnel lens, a Fresnel zone plate or other suitable imaging means. The present embodiment illustrates a one- dimensional imaging feature. Two-dimensional imaging features (in the illustrated embodiment, this would be a spherical depression, for example) may also be used. Imaging elements are also advantageous due to the relatively high dirt immunity, i.e. contamination of part of the imaging element will not affect the position at which the image is generated.

A further embodiment of the present invention relates to alignment of the scale and readhead in order to achieve an optimum timing for the generation of a reference pulse from a reference mark.

Referring now to Fig 6, a scale member 110 has an incremental track 112 provided by a series of spaced apart lines 114. A readhead 116 is movable relative to the scale member 110 in the direction of spacing of the lines 114 and is adapted to project light onto the scale 114 and interact with the light reflected from the scale 114 to generate a periodic light pattern at the surface of an analyser. Relative movement between the readhead 116 and the scale member 110 results in movement of the periodic light pattern across the surface of the analyser, and, as a result, photodetectors associated with the analyser emit a series of phase-shifted cyclically varying signals corresponding to the variation of light intensity incident thereon. The optical mechanism by which the periodic light pattern is generated at the surface of the analyser, and the phase-shifted signals are produced at the photodetectors, is not germane to the present invention. Any suitable mechanism may be used such as the mechanism disclosed in GB 1541691 or WO86/03833. The outputs of the photodetectors are sent to signal conditioning circuitry which generates a pair of sinusoidally varying signals having a quadrature relationship (i.e. separated by a 90° phase angle) . The wavelength of each of these sinusoidally varying signals corresponds to one graduation (multiple or fraction thereof) of the scale (this is known from e.g. WO87/07943, incorporated by reference). The signal conditioning circuitry also generates (e.g. with inverting op amps) two further outputs, each of which is the inverse of one of the quadrature outputs. In an alternative embodiment, the conditioning circuitry may generate four sinusoidally varying outputs at phases of 0° ,90° ,180° ,270° directly from the photodetector inputs, by virtue of a suitable combination scheme for said inputs.

The four outputs of the signal conditioning circuitry are sent to an interpolator which, from a given pair of the four possible permutations (cos,sin;sin,cos;cos,sin;sin,cos)

of pairs of quadrature signals, generates digital quadrature signals (not shown) whose wavelengths correspond to the smallest unit of resolution of the apparatus (e.g. 1 micron) . The interpolator also generates, once per quadrature cycle, a unique pulse. The phase of the cycle at which this pulse is generated is determined by the physical construction of the interpolator. In the present example the unique pulse is centred at a phase of 45° of the quadrature cycle of whichever pair of quadrature signals is selected from the 4 possible pairs. In an alternative embodiment the signal conditioning circuitry only generates sin and cos signals, and the sin and cos signals are generated in the interpolator.

Also provided on the scale member 10 is a reference mark 118 such as the reference mark described with reference to Figs 1 to 5. A photodetector associated with the reference mark outputs a peaked electrical output when the readhead 116 lies in register with the reference mark 118. The output from the reference photodetector is converted to a square wave reference pulse by signal conditioning circuitry. The width of the reference pulse is typically about the width of a pitch of the scale; this provides a large gating range while minimising the risk of "jumping" to the next pitch due to drift.

The unique pulses from the interpolator and the reference pulse are "gated" together to enable the reference digital pulse to select a given unique pulse from the interpolator (and hence a given pitch of the scale) .

It is important that the reference pulse be centred with respect to the particular unique pulse with which it coincides, in order that e.g. thermal drift of the reference mark relative to the incremental track of the scale does not cause the reference pulse to "jump" and select a different (adjacent) unique pulse from the interpolator. Referring again to Fig 6, unique pulses generated from the interpolator in respect of the cos,sin pair of quadrature signals at a phase angle of 45° are labelled P→ to P₄. The relative positions at which each of the other three possible pairs of quadrature signals would generate a unique pulse at the same phase angle are shown with dotted lines. It can be seen from Fig 6 that the reference pulse P_pef from the reference mark is coincident with the unique pulse P₃. However, the pulse P_pef is not centred with respect to P₃ and so a small drift of pulse P_ref in the direction of pulse P₂ will result in the selection of pulse P₂. It is therefore more appropriate in this case to select the cos,sin pairs of quadrature signals, since the

unique pulse which would be generated by the interpolator (at the phase angle of 45°), as illustrated by the dashed line L, is substantially centred with respect to the pulse ^Pref

Selection of a quadrature signal pair may be performed by making or breaking a number of conducting links in the readhead or to the interpolator as appropriate.

An example of an interpolator will now be described with reference to Fig 7. In the example of interpolator illustrated, a plurality of sinusoidally varying inputs are provided, only three of which are illustrated, for the purposes of simplicity; in this example three of the plural inputs shown are sin, cos and cos. Each pair of sinusoidally varying inputs is linked by a resistor chain R1,R2;R3,R4. Outputs taken from nodes N1,N2 thus correspond to sinusoidally varying signals having a phase which is shifted to a phase which lies between the phase of the linked pairs of sinusoidally varying inputs (equal resistor values will thus shift the phase by 45°). The outputs from nodes N1,N2 are each sent to a zero crossing comparator compl,comp2 which generate squarewave pulses Q1,Q2 respectively. The number of pairs of linked inputs required will depend on the degree of interpolation required e.g. for an interpolation of four times per cycle, four pairs of sinusoidally varying inputs linked by pairs of resistors giving four evenly distributed phase-shifts will be required. The plurality of squarewave pulses Ql,Q2,....Qn are sent to a logic circuit 200, which generates a train of digital quadrature pulses whose frequency is 1/r of a pitch of the scale, where r is the degree of interpolation, or number of resistor-linked pairs of inputs. Additionally, a further logic circuit 300 is adapted to receive the squarewave inputs Ql,Q2,...Qn and generate a unique pulse once per input quadrature cycle. The timing of the unique pulse with respect to the input quadrature signals depends upon the setting within the logic circuit 300. For example, the settings may be made so that the pulse is emitted at the instant the values of the two input quadrature signals each correspond to a 45° angle (i.e. sin = cos = 1Λ/2) .

Claims

1. Opto-electronic scale reading apparatus comprising a scale defined by a series of spaced-apart lines, and a readhead, the scale and readhead being relatively movable in the direction of spacing of the lines, and the readhead generating, upon said relative movement, at least one cyclically varying electrical signal having a frequency corresponding to the pitch of the scale, wherein: the scale comprises at least one reference mark in the form of an optical imaging element on the scale, said element being configured to re-direct light from a light source in the readhead in such a way that the re-directed light is convergent in the direction of spacing of the scale lines, and forms an image of said light source at a predetermined height above the scale.

2. Apparatus according to claim 1 wherein said scale is reflective, and said element is provided by a cylindrical indentation in the surface of the scale extending parallel to said scale lines.

3. Apparatus according to claim 1 wherein said scale is reflective, and said element reflects said light in such a way that the reflected light is additionally convergent in a direction parallel to the lines.

4. Apparatus according to claim 3 wherein said element is provided by a spherical indentation in the surface of the scale.

5. Apparatus according to claim 1 wherein said element is provided by one of a Fresnel lens, a Fresnel zone plate, and a lens.

6. Apparatus according to claim 2 wherein said light source projects light in a first direction, parallel to said scale lines, and in a second direction, opposite to said first direction.

7. Apparatus according to claim 6 wherein said readhead includes a first photodetector lying in register with one end of said indentation, and a second photodetector, lying in register with the other end of said indentation.

8. Apparatus according to claim 6 wherein a plurality of said indentations are provided along the length of the scale.

9. Apparatus according to claim 8, wherein a first path swept by the first photodetector upon said relative movement defines a reference track, and a second path swept by said second photodetector defines a selector track, and wherein a given indentation of said plurality is selected as a reference mark by the placing or removal of a non- reflective tab over the part of the indentation lying on said selector track.