Laser encoder system

Abstract

A laser encoder system (1) is disclosed which comprises a detector head (4) which is adapted to generate a detection signal from which positional data relating to the position of a target (8) is derivable. The detection signal is an analogue quadrature signal comprising first and second constituent signals (c, s) having a predetermined phase difference between them. First and second electrical pathways (43, 45, 47) respectively carrying the first and second constituent signals (c, s) within the laser encoder system are arranged (M) to have substantially the same length in order to maintain the predetermined phase difference between the first and second constituent signals (c, s), thereby to avoid introducing (or to reduce) a sub-divisional error in the derived positional data.

Description

[0001] Laser Encoder System

[0002] The present invention relates to a laser encoder system. The present invention relates in particular, but not exclusively, to improvements in the setup and operation of such a laser encoder system.

[0003] Figure 1 of the accompanying drawings illustrates a fibre optic laser encoder system 1 which is made and sold by Renishaw pic. The laser encoder system 1 provides position feedback signals suitable for use in precision position feedback applications such as machine calibration and motion control.

[0004] The main components of the laser encoder system 1 are a laser unit 2, a detector head (or detector unit) 4 and a machine interface 6. The detector head 4 is the core of the optical measuring system and will be described in further detail below with reference to Figure 2 of the accompanying drawings. The laser unit 2 comprises a laser source and signal processing electronics, with an electrical cable 3 for receiving signals from and providing power to the detector head 4 and a fibre optic conduit 5 that delivers laser light directly to the detector head 4 through a fibre optic cable (not visible in Figure 1) within the fibre optic conduit 5. The machine interface 6 forms part of a controller 31 and communicates with the laser unit 2 via an electrical cable 7.

[0005] To complete the configuration, a target optic 8 is provided in the path of a laser beam 9 emitted from the detector head 4, such that the laser beam 9 is reflected off the target optic 8 and returned to the detector head 4. In this example, the target optic 8 is in the form of a retroreflector but with a variant of the detector head 4 the target optic 8 could instead be a plane mirror. The beam 9 is a measurement beam, with the return measurement beam interfering with a reference beam which in this example is internal to the detector head 4, with the distance to the target optic 8 (or rather changes in this distance relative to a chosen datum position) being determinable from the interference signal in a known way.

[0006] Figure 2 is a schematic illustration of the main components of the detector head 4 of Figure 1. The fibre optic cable 11 passes through into the body or housing 24 of the detector head 4 and into a collimator 17. The fibre optic cable 11 is terminated within and held in place via a ferrule 21, with laser light being emitted from the end of the fibre optic cable 11 in a diverging cone. The role of the collimator 17 is to collimate this diverging beam, using a lens 23, before it passes further through the detector head 4. As shown by the arrows, the collimated beam passes first to a beam splitter 14, with some of the light being reflected up to a reference retroreflector 18 (as a reference beam) and the remainder of the beam passing out through the laser aperture 16, via a beam steerer 30, and onwards (as the measurement beam 9) to the retroreflector target optic 8. Also shown in Figure 2 is a circuit board 12, which supports various processing, detection, and control electronics (such as a light detector 19), as well as an optical shutter 26 which can be used to shut off the measurement beam 9.

[0007] The return measurement beam from retroreflector target optic 8 re-enters the detector head 4 via the laser aperture 16, and through the beam splitter 14 where it joins (and interferes with) the measurement beam from the reference retroreflector 18 and is incident on a light detector 19. An analogue quadrature interference signal (or detection signal) from the light detector 19 then passes out from the detector head 4 via the electrical cable 3 where it is received at the laser unit 2 shown in Figure 1. Although there may be some processing performed on the interference signal at the laser unit 2, the main processing is typically performed at the interface 6, having received the interference signal from the laser unit 2 via electrical cable 7.

[0008] By digitising, interpolating and processing the interference signal the interface 6 can determine with high accuracy how far the retroreflector target optic 8 has moved by counting fringes, or rather pulses in the digitised / interpolated version of the signal. The positional data from the interface 6 can then be used by the controller 31 for the intended purpose, such as machine calibration or motion control. It should be noted that the laser unit 2 can be set to output a digital rather than analogue quadrature output signal, in which case the digitising and interpolating would be performed at the laser unit 2 rather than at the interface 6.

[0009] The exterior of the detector head 4 of Figure 1 is shown in more detail in Figure 3 of the accompanying drawings. The fibre optic cable 11 can just be seen within the fibre optic conduit 5. The fibre optic conduit 5 is itself coupled to the body of the detector head 4 via a strain relief 15, which is intended to prevent or at least limit forces on the fibre optic conduit 5 and the enclosed fibre optic cable 11 being transferred to any internal optical components to which the fibre optic cable 11 is connected. A laser aperture 16 is also apparent in Figure 3, through which both the outgoing and returning measurement beams 9 will pass.

[0010] The detector head 4 as shown in Figures 1 to 3 is just one type of detector head made and sold by Renishaw pic. Figure 4 of the accompanying drawings shows another type of detector head 4, which differs from that shown in Figure 3 mainly in that the measurement beam 9 is emitted at a ninety-degree angle to main axis of the detector head 4 (rather than zero-degree angle). The detector head 4 of Figure 4 would typically be used as a pair, with one detector head 4 of the pair measuring along an X machine axis and the other measuring along a Y machine axis, and would typically use a target optic 8 in the form of a plane mirror (though a retroreflector target could also be used). It can be seen that the laser unit 2 of Figure 1 has a spare set of connections available for accommodating a second detector head 4 in this way.

[0011] Figure 5 of the accompanying drawings shows yet another type of detector head 4, which differs more substantially from that of Figures 3 and 4. The detector head 4 of Figure 5 is a differential interferometer detector head 4, with a pair of measurement beams 9M emitted from the laser aperture 16 as well as a pair of reference beams 9R. By using an external reference beam 9M, the differential interferometer detector head 4 is able to measure the relative displacement between two plane mirror targets, one of which (the reference target) would typically be in a fixed position, for example on a fixed column of the machine, and the other of which (the measurement target) would be moving, for example on a moving stage of the machine on which a semiconductor wafer or other workpiece is supported. This helps to ensure accurate positioning between process critical components and to eliminate common mode errors.

[0012] For a detector head 4 of a type as shown in Figures 3 and 4 the electrical cable 3 is fixedly connected to the detector head 4, but for a detector head 4 of a type as shown in Figure 5 an electrical connector 13 is provided for releasably connecting the electrical cable 3 to the detector head 4. For all of these detector heads 4 the electrical cable 3 is releasably connectable at the other end to the laser unit 2. On the other hand, for all of these detector heads 4 the fibre optic conduit 5 (with enclosed fibre optic cable 11) is detachable from the detector head 4 (as will be explained in more detail below) but is fixedly coupled into the laser unit 2. In this respect, the fibre optic cable 11 is continuous from the fibre launch within the laser unit 2 all the way to the collimator 17 and therefore cannot easily be disconnected from the laser unit 2. The detector head 4 typically incorporates a feature that prevents the laser beam 9 being emitted if either the fibre optic cable 11 or electrical cable 3 is disconnected.

[0013] The collimator 17, the strain relief 15, fibre optic conduit 5 (with enclosed fibre optic cable 11) and fibre optic connector 34 (see Figure 1) can be considered to form a single optical fibre assembly 32, with the collimator 17 being considered as an integral part of the optical fibre assembly 32 because it terminates the fibre optic cable 11 in a manner required by the detector head 4 (and the laser encoder system 1 as a whole). The components of the optical fibre assembly 32 (and the internal components of the collimator 17) are assembled and aligned precisely in the manufacturing facility, and supplied to the customer as a unit, and as such are considered to be inseparable in normal use. The way in which the detector head 4 is adapted to receive the optical fibre assembly 32 will be more apparent from Figures 6 and 7 of the accompanying drawings.

[0014] When the laser encoder system 1 is being configured for operational use, the supplied optical fibre assembly 32 is connected to the detector head 4 simply by pushing the collimator 17 through a correspondingly sized opening formed in the housing 24 and into the body of the detector head 4, with the strain relief 15 remaining outside the body and in contact with the housing 24. This connection operation is illustrated in Figure 6 for a differential interferometer detector head 4 of a type shown in Figure 5, but the connection would be entirely equivalent for a detector head 4 of a type shown in Figures 3 and 4. The combination of the strain relief 15 and collimator 17 can be referred to as a fibre barrel 10, with the optical fibre assembly 32 being terminated by the fibre barrel 10.

[0015] Figure 7 shows a view of the differential interferometer detector head 4 of Figure 4 with the upper part (or lid) of the housing 24 removed. This illustrates how the collimator 17, having been pushed into the body of the detector head 4, is held in place tightly by a clamp 20, which is in turn tightened via a locking screw 22. It will also be apparent from Figures 6 and 7 that the collimator 17 is rigidly coupled to the strain relief 15, consistent with the above explanation that the collimator 17 and the strain relief 15 form part of a unitary optical fibre assembly 32. The strain relief 15 and collimator 17 are prevented from being pulled away from the detector head 4 by action of the clamp 20 on the collimator 17, but there could be an additional coupling of the strain relief 15 to the housing 24 (e.g. via a screw thread connection).

[0016] For a proper connection, the fibre barrel 10 (and in particular the collimator 17) must be inserted into the detector head 4 in the correct orientation. To achieve this, and as shown in Figure 6, a line L marked on the fibre barrel 10 is aligned visually with an alignment dot D marked on the detector head 4. The fibre barrel 10 is inserted fully and then rotated slightly until a ball bearing inside the clamp 20 can be felt to engage in a recess 38 in an outer surface of the collimator 17 (specifically in a lid of the collimator 17), with the ball bearing being biased by a spring 39 (see Figure 7) into the indent 38 to form a detent feature. There would also be an end stop in the lid of the detector head 4 which sets the position along the insertion axis, so that the indent 38 is in the correct axial position to receive the ball bearing that is biased into it by the spring 39.

[0017] Figure 8 of the accompanying drawings shows the front panel of the laser unit 2 of Figure 1 in more detail. The laser unit 2 of Figures 1 and 8 is adapted to support two detector heads 4 simultaneously, which might for example be arranged to measure along two orthogonal machine axes X and Y as described above with reference to Figure 4. Accordingly, many of the connectors for the laser unit 2 are duplicated between the two axes and are differentiated by AX1 and AX2 marked on the relevant labels visible in Figure 8. With the system setup as shown in Figure 1, only the first of the two available axes is in use, with the detector head 4 being connected via cable 3 to connector DI (of pair DI, D2) and the interface 6 (which can also be referred to as a controller) being connected via cable 7 to connector Cl (of pair Cl, C2). The laser light itself is delivered to the detector head 4 via optical fibre assembly 32 which is connected to connector Al (of pair Al, A2). Also shown in Figure 8 is a pair of status lights LI, L2, a pair of sensor connectors Ml, M2, a set of configuration switches W (normally covered but shown partially revealed in Figure 8), a pair of reference mark connectors Rl, R2, a shutter connector S, a reset connector R, an AUX I / O connector A, a laser status light L, and a 24V power supply input P. These are labelled as follows:

[0018] LI : STATUS AX1 (similarly for L2)

[0019] Cl : SIGNAL OUT AX1 (similarly for C2) Ml : SENSOR AX1

[0020] W: CONFIGURATION SWITCHES (COVERED) S: SHUTTER R: RESET

[0021] R1 : REF. MARK AX1 (similarly for R2) DI : DETECTOR AX1 (similarly for D2) Al : LASER FIBRE OPTIC AX1 (similarly for A2) A: AUX I / O

[0022] L: LASER STATUS

[0023] P: 24 V POWER SUPPLY INPUT

[0024] Further information regarding the purpose and function of each of the above can be found in the literature relating to the RLU10 or RLU20 laser unit products made and sold by Renishaw pic.

[0025] As mentioned above, the interference signal sent from the detector head 4 to the laser unit 2 is in practice in the form of an analogue quadrature signal, which will now be described in more detail with reference to Figures 9 and 10 of the accompanying drawings. Figure 9 illustrates the reference and measurement beams forming a fringe pattern 29 on the light detector 19 in the detector head 4, while Figure 10 shows a simplified circuit diagram of the detection and processing circuitry 41 in the detector head 4. The fringe pattern 29 will move across the light detector 19 as the target 8 moves towards and away from the detector head 4. The light detector 19 is a multi-channel light detector comprising four photodiodes 19a, 19b, 19c and 19d (see Figure 10) providing four corresponding respective detection signals, which are effectively samples at 90° intervals across the fringe pattern 29 (where a single fringe period covers 360°). These provide a cosine and minus cosine signal pair, which are fed to a differential amplifier to give a clean cosine signal (cleaned of noise), and a sine and minus sine signal pair, which are likewise fed to a differential amplifier to give a clean sine signal. The clean cosine and sine signals are each fed through a DC offset adjustment stage and a gain adjustment stage to provide a quadrature output signal which comprises the adjusted cosine and sine signals, which are made available at the quadrature signal output 42. The cosine and sine signals of the quadrature signal can be referred to as constituent signals of the quadrature signal (or component signals or simply components).

[0026] If the sine and cosine signals of the quadrature output signal are plotted on the X and Y axes of a graph, they produce a circular “Lissajous” figure as shown in Figure 10. When the target optic 8 moves, the fringe pattern 29 on the light detector 19 will likewise move, and hence so will the sine and cosine signals from the detection and processing circuitry 41. The moving sine and cosine signals will therefore sweep out a circular path on the graph as the target optic 8 moves, and when the target optic 8 stops, so will the point that is being swept out on the circular path. The distance moved by the target optic 8 is measured by counting revolutions around the circular path (or Lissajous). By way of example, with the present laser encoder systems from Renishaw pic one revolution around the Lissajous corresponds to 316 nm for a single pass interferometer (this would be described as having a “316 nm Lissajous”) and 158 nm for a double pass interferometer (this would be described as having a “158 nm Lissajous”). If the laser unit 2 is set to output digital quadrature signals, then the analogue sine and cosine quadrature signals received from the detector head 4 are digitised and interpolated in the laser unit 2 by a signal conversion unit before being output to the interface 6. Otherwise, an analogue quadrature signal would be output from the laser unit 2 and the digitising and interpolating would be performed elsewhere. The laser unit 2 can be switched between analogue and digital quadrature output using the configuration switches W (the output from the detector head 4 is analogue, with the digitisation being in the laser unit 2).

[0027] Interpolation enables a higher resolution of measurement to be obtained, effectively enabling sub-revolutions around the Lissajous to be counted rather than just complete revolutions. For example, for the double pass interferometer mentioned above that has a 158 nm Lissajous (i.e. where the analogue output signal period is 158 nm), nominal output resolutions for the digital quadrature signal output from the laser unit 2 (when using a signal conversion unit within the laser unit 2 itself) include 10, 20, 39.5 and 79 nm. An even higher resolution of 0.39 nm is achievable using the REE interpolator from Renishaw pic (receiving an analogue quadrature signal from the laser unit 2), and an even higher resolution of 38.6 pm is achievable with the RPI20 and RPI30 parallel interfaces from Renishaw pic (which would be used as the interface 6 described above). The concepts of digitisation and interpolation are also discussed further below.

[0028] The present applicant has appreciated that, due to manufacturing tolerances the optical power of the laser light produced by the laser unit 2 (and delivered to the detector head 4 through the fibre optic cable 11 of the optical fibre assembly 32) will vary from one laser unit 2 to the next. The optical power will even vary between the two laser outputs Al, A2 of the same laser unit 2 in a dual axis system such as illustrated in Figures 1 and 8. Variation in optical power can result from tolerances in the optical power of the laser tube itself, and also of variation in the optical coupling efficiency, which is in turn determined by how accurately the launch end of the fibre is positioned relative to the coupling lens. It can also be affected by tolerances on the focal length of the coupling lens and the mode field diameter of the fibre. The optical power of the laser light delivered to the detector head via laser outputs Al, A2 has a direct impact on the signal strength received back from the detector head 4 via connectors DI, D2, with signal strength being directly proportional to the optical power of the laser light. In simple terms, the brighter the light falling on the light detector 19 within the detector head 4 (see Figure 2), the stronger the output signal from the detector head 4 (via cable 3) will be. Furthermore, there will in practice be variations in the sensitivity of the light detector 19 between different detector heads 4, and manufacturing variations will also be present elsewhere in the optical and electrical pathways.

[0029] Accordingly, each laser unit 2 is currently matched to one or two specific detector heads 4 prior to shipping to the customer as a complete fibre optic laser encoder system 1. The matching process is designed to give the best chance of achieving a nominal 100% signal strength for the signal sent from the laser unit 2 to the interface 6, which is equivalent to 1 Vpp measured on the appropriate signal strength output pin on the AUX I / O connector A using a multimeter M as shown schematically in Figure 11 of the accompanying drawings. This ideal 1 Vpp is also shown in relation to the output signal of Figure 10. This matching process is currently achieved by careful manual adjustment of a potentiometer associated with each of the gain and offset stages within the detector head 4 (see Figure 10) as part of the production process. As illustrated in the table shown in Figure 12 of the accompanying drawings, the system should be set up to achieve a signal strength in a range from 25% to 120%, though for optimum performance this should normally be around 100%. The system will still function with a signal strength in a range from 12.5% to 25% and above 120% but it may not achieve the best possible accuracy. The system is not intended to function with a signal strength below 12.5% and will typically assert an error output.

[0030] Because of the sensitivity of signal strength to laser output power (as well as other optical and electronic components), it is strongly recommended to the customer that the detector head(s) 4 that are supplied with the laser unit 2 are used only with that laser unit 2 to ensure optimal performance. For this reason, a label showing the serial numbers of the supplied detector heads 4 is currently affixed to the rear of the supplied laser unit 2. The consequences of changing the detector head 4 that is connected to a laser unit 2 may include one or more of the following:

[0031] (a) low signal strength (below the acceptable threshold), which will trigger a beam low error such that the relevant axis status light LI, L2 will appear continuously amber.

[0032] (b) high signal strength (exceeding 120%), which will trigger a beam saturation error such that the relevant axis status light LI, L2 will appear continuously amber, and which may also cause interpolation errors (see below).

[0033] (c) low signal strength (but still within the acceptable range), which will make system alignment more difficult, especially at longer distances.

[0034] (d) non-ideal offset and / or amplitude of the sine and cosine components of the quadrature signal, leading to a sub -divisional error or SDE.

[0035] Referring to part (d) above, unequal offset between the sine and cosine signals and unequal sine and cosine signal levels can both cause imperfect Lissajous plots, as represented respectively in Figures 13 and 14 of the accompanying drawings. In turn, this results in a sub -divisional error (also referred to as a non-linearity error or interpolation error), which is typically expressed in nanometres (nm). The interpolation error E is represented as the difference between the same corresponding point on the ideal Lissajous 35 and the actual Lissajous 36. Sub- divisional error is cyclic, occurring within each full signal period based on a difference between where the point is on the actual Lissajous 36 and where it should be on the ideal Lissajous 35, and does not therefore accumulate.

[0036] If a detector head 4 does have to be moved from one system to another, this would currently require the detector head 4 to be matched to the new laser unit 2, by adjusting an internal gain and offset of the detector head 4 to ensure that it is tuned correctly to the laser unit 2. Under normal circumstances there should be no requirement to adjust the gain and offset of the detector head 4. Incorrect adjustments to the detector head 4 can cause system performance errors as described above. Furthermore, the making of such adjustments is a highly skilled task which the normal user would not typically be capable of performing, thereby requiring the system to be returned to the manufacturer for tuning, which is inconvenient for the customer.

[0037] In view of the above, the present applicant has appreciated that it would be beneficial to provide a laser encoder system with an improved interpolation error or sub -divisional error (SDE). According to a first aspect of the present invention there is provided a laser encoder system comprising a detector head (or detector unit) which is adapted to generate a detection signal from which positional data relating to the position of a target is derivable, the detection signal being an analogue quadrature signal comprising first and second constituent signals having a predetermined (or expected) phase difference between them, wherein first and second electrical pathways respectively carrying the constituent signals within the laser encoder system are length matched to maintain the phase difference between the first and second constituent signals within a predetermined acceptable range.

[0038] The laser encoder system may comprise a laser unit that is connected in use to the detector head to provide laser light to the detector head.

[0039] The laser unit may be connected in use to the detector head via an optical fibre assembly to provide the laser light to the detector head via the optical fibre assembly.

[0040] The detection signal may be an interference signal resulting from interference between a reference laser beam and a measurement laser beam, with the measurement laser beam being directed onto and at least partially reflected back from the target.

[0041] The detection signal may be an electrical signal.

[0042] Length matching may be employed in relation to first and second electrical tracks within a printed circuit board carrying the first and second constituent signals respectively within the laser encoder system.

[0043] Length matching may be employed at least in relation to a printed circuit board in the detector head. Length matching may be employed in relation to first and second electrical cores within an electrical cable carrying the first and second constituent signals respectively within the laser encoder system.

[0044] Length matching may be employed at least in relation to an electrical cable connected between the detector head and the laser unit.

[0045] The first and second electrical pathways may be length matched at least until a point at which the analogue quadrature signal is digitised and interpolated by a signal conversion unit.

[0046] Length matching may be employed in relation to electrical pathways in the signal conversion unit, such as electrical tracks formed on a printed circuit board.

[0047] The signal conversion unit may be or may form part of or may be implemented as a functional unit of the laser unit.

[0048] The signal conversion unit may be or may form part of or may be implemented as a functional unit of an interface component.

[0049] The signal conversion unit may be arranged to output the digitised and interpolated signal to a positional data deriving unit.

[0050] The positional data deriving unit may be adapted to derive positional data from the digitised and interpolated signal.

[0051] The positional data deriving may be or may form part of or may be implemented as a functional unit of an interface component or controller.

[0052] The predetermined phase difference may be 90°. The detection signal may relate to a distance between the detector head and a target (or target optic).

[0053] The first and second constituent signals may be electrical signals.

[0054] The first and second constituent signals may be sinusoidal signals (as a function of time).

[0055] The first and second constituent signals may be at least nominally 90 degrees out of phase.

[0056] The first and second constituent signals may be cosine and sine signals respectively (as a function of time).

[0057] The first and second constituent signals may have a frequency of less than 500 MHz.

[0058] The first and second constituent signals may have a frequency of less than 100 MHz.

[0059] The first and second constituent signals may have a frequency of less than 60 MHz.

[0060] The first and second constituent signals may have a frequency of less than 30 MHz.

[0061] The predetermined acceptable range (or the degree of length matching between the first and second electrical pathways) may be dependent on or determined in dependence upon a predetermined acceptable sub -divisional error in the derived positional data which is attributable to the phase difference. The predetermined acceptable range may be dependent on or determined in dependence upon a speed of the target (or target optic).

[0062] According to a second aspect of the present invention there is provided a method of arranging (or designing) electrical pathways for a laser encoder system as claimed in any preceding claim, comprising length matching the first and second electrical pathways carrying the constituent signals within the laser encoder system to maintain the phase difference between the first and second constituent signals within a predetermined acceptable range.

[0063] The method may comprise length matching the first and second electrical pathways at least until a point at which the analogue quadrature signal is digitised and interpolated by the signal conversion unit.

[0064] The method may comprise determining the predetermined acceptable range (or the degree of length matching between the first and second electrical pathways) in dependence upon a predetermined acceptable sub -divisional error in the derived positional data that is attributable to the phase difference.

[0065] The method may comprise determining the predetermined acceptable range (or the degree of length matching between the first and second electrical pathways) in dependence upon a speed of the target (or target optic).

[0066] The predetermined acceptable range may be within 1° of the predetermined (or expected) phase difference.

[0067] The predetermined acceptable range may be within 0.5° of the predetermined (or expected) phase difference.

[0068] The predetermined acceptable range may be within 0.2° of the predetermined (or expected) phase difference. Reference will now be made, by way of example, to the accompanying drawings, in which:

[0069] Figure 1, discussed hereinbefore, illustrates a known fibre optic laser encoder system comprising a laser unit, detector head and machine interface;

[0070] Figure 2, also discussed hereinbefore, is a schematic view of some of the internal components of the detector head of Figure 1;

[0071] Figure 3, also discussed hereinbefore, shows in more detail a detector head of the type shown in Figures 1 and 2;

[0072] Figure 4, also discussed hereinbefore, shows a different type of detector head in which the laser beam is emitted from the body at a different angle;

[0073] Figure 5, also discussed hereinbefore, shows a differential interferometer type of detector head which emits both measurement and reference beams;

[0074] Figure 6, also discussed hereinbefore, shows how the fibre barrel of an optical fibre assembly is connected to a differential interferometer type of detector head;

[0075] Figure 7, also discussed hereinbefore, shows a view of the detector head of Figure 6 with the upper part of the housing removed;

[0076] Figure 8, also discussed hereinbefore, shows the front panel of the laser unit of Figure 1 in more detail;

[0077] Figure 9, also discussed hereinbefore, illustrates a fringe pattern formed on a multi-channel light detector by the reference and measurement beams in the detector head; Figure 10, also discussed hereinbefore, shows a simplified circuit diagram for the detection and processing that is carried out in the detector head, including gain and offset correction, and which also explains the concept of a quadrature signal and a Lissajous plot;

[0078] Figure 11, also discussed hereinbefore, shows the laser unit of Figure 8 connected to a multimeter to measure a signal level of the detection signal from the detector head;

[0079] Figure 12, also discussed hereinbefore, is a table showing different ranges for the signal level and the associated condition which is indicated via a status light on the front panel of the laser unit;

[0080] Figure 13, also discussed hereinbefore, illustrates an interpolation error which results from a DC offset between the sine and cosine parts of the quadrature signal from the detector head;

[0081] Figure 14, also discussed hereinbefore, illustrates an interpolation error which results from unequal sine and cosine levels in the quadrature signal from the detector head;

[0082] Figure 15 is a schematic representation of a circuit board used in the detector head, and is used to explain a problem caused by a difference in the lengths of the tracks used to carry the sine and cosine signals on the printed circuit board;

[0083] Figure 16 is a schematic representation of a generic circuit board to illustrate the type of track layout that would typically result from use of printed circuit board design software;

[0084] Figure 17 is a schematic representation of a circuit board based on that of Figure 15 but modified by length matching of tracks to address the problem in accordance with an embodiment of the present invention;

[0085] Figure 18 illustrates how length matching can be applied along the entire signal pathways carrying the signals from the detector head to the signal conversion unit in the laser unit or interface;

[0086] Figure 19 is based on Figure 18 but shows the key functional components of the signal conversion unit; and

[0087] Figure 20 provides a graphical representation of how the analogue quadrature signal is converted into a digital quadrature signal by the signal conversion unit.

[0088] As described above, the quadrature signal from the laser unit 2 is received and processed by the interface 6 to derive positional data, which involves interpolating the constituent sine and cosine signals of the quadrature signal (before or after digitisation). As also mentioned above, for example with reference to Figures 13 and 14, the present applicant has appreciated that excessive DC offset and unequal sine and cosine levels which will cause non-linearity errors (also known as interpolation error or sub -divisional error or SDE, or more simply the measurement error).

[0089] The present applicant has determined that, in a laser encoder system 1 such as this that relies on interpolation of an analogue quadrature signal, it is critical that the constituent sine signal leads (or lags) the constituent cosine signal by exactly 90°. In the illustration provided in Figure 10 the sine signal is depicted as lagging the cosine signal by 90°. The present applicant has determined that any phase error will also increase the sub -divisional error (SDE).

[0090] The present applicant has determined that, at the high frequencies that a laser interferometer may produce, a difference in the lengths of the physical tracks (or traces) of the sine / cosine signals on the printed circuit board (PCB) could cause a short delay that is large enough to increase the measurement error. This problem is illustrated in Figure 15, which is a very schematic and selective representation of a printed circuit board 40 used in the detector head 4. The printed circuit board

[0091] 40 would comprise, amongst other things, the detection and processing circuitry

[0092] 41 shown in Figure 10, and just the quadrature signal output 42 from the circuitry 41 of Figure 10 are represented in the printed circuit board 40 of Figure 15. At the quadrature signal output 42, the sine signal is shown as lagging the cosine signal by exactly 90° as intended by the circuitry 41.

[0093] From the quadrature signal output 42, a pair of tracks 43 carry the quadrature signal along a first layer of the printed circuit board 40 to a pair of vias 44, which connects to a second layer of the printed circuit board 40. Another pair of tracks 45 carry the quadrature signal onwards to a quadrature output 46 of the printed circuit board 40, and which in turn would connect into the electrical cable 3 and onwards to the laser unit 2. Of course, in practice the printed circuit board 40 would comprise many tracks other than those depicted in Figure 15, and these tracks would have an appearance similar to that shown in the representative track layout of Figure 16. The track layout would typically be designed using software which helps to route the tracks in a space efficient manner, and accordingly the layout would typically consist of tracks that run parallel or orthogonal to one another with an equal spacing therebetween, and with bends at either 45° or 90° in order to maintain an ordered track layout. The use of multiple layers, connected by vias, is of course not essential but does typically enable the printed circuit board designer to route tracks and connect components in a more space-efficient manner.

[0094] In the schematic printed circuit board 40 shown in Figure 15, the track carrying the cosine signal has a representative length of 30 (in arbitrary length units) between the quadrature signal output 42 and the via 44, and a representative length of 11 between the via 44 and the quadrature output 46, giving a total representative length of 41. On the other hand, the track carrying the sine signal has a representative length of 35 between the quadrature signal output 42 and the via 44, and a representative length of 11 between the via 44 and the quadrature output 46, giving a total representative length of 46. The sine signal therefore has to travel along a greater track length than the cosine signal, and this leads to an additional phase delay for the sine signal. Accordingly, by the time it reaches the quadrature output 46, the sine signal is lagging the cosine signal by more than the intended 90°.

[0095] As described above, the present applicant has appreciated that this phase shift leads to a measurement error. Neither the presence of nor the reason for this measurement error has been appreciated previously, perhaps because it has not been significant enough to be detectable. However, the present applicant has been striving for ever increasing accuracy and dealing with applications with ever increasing velocity requirements (in relation to movement of the target), and as part of this development work has now understood that the phase shift introduced between the sine and cosine signals is a significant source of error in these types of application.

[0096] To address the problem described above, careful length-matching of electrical tracks (or electrical pathways in general) is employed in an embodiment of the present invention to maintain the correct phase shift between the constituent signals in the detector head 4. By ensuring that the electrical pathways that carry the constituent signals have the same length wherever possible and practical, this helps to avoid introducing a measurement error in the positional data that is derived from the quadrature signal.

[0097] As noted above, the track carrying the cosine signal in Figure 15 has a total representative length of 41, whereas the track carrying the sine signal has a total representative length of 46. This is how the PCB design software has, in the absence of any other constraints, chosen to lay out these tracks because it has done so in order to create the most space efficient layout bearing in mind all of the other tracks and components that are also present on the printed circuit board.

[0098] The track layout used in the printed circuit board 40 shown in Figure 17 is similar to that of Figure 15, but with a crucial change to the routing of the track 45 carrying the cosine signal. Because the track carrying the sine signal in Figure 15 is longer than that carrying the cosine signal by a representative length of 5, with the printed circuit board 40 of Figure 17 the track 45 which carries the cosine signal between the via 44 and the output 46 has been increased by 5, from 11 to 16. Therefore, the representative length of both tracks is now equal at 46. Of course, this length matching could also have been achieved by lengthening track 43 which carries the cosine signal by 5, or by shortening either of the tracks 43 and 45 which carries the sine signal, or a combination of these changes.

[0099] Because the tracks carrying the constituent signals have been arranged in an embodiment of the present invention to have the same length, the sine signal lags the cosine signal by the expected 90° not only at the output 42 from the circuitry 41 (which generates the quadrature signal) but also at the output 46 of the printed circuit board 40 itself, as represented in Figure 17. Accordingly, the measurement error that would have been introduced into the derived positional data due to the additional phase shift shown at the output 46 in Figure 15 has been avoided by adopting length matching of tracks as shown in Figure 17.

[0100] The length matching feature, which is the extension of the cosine track 45 compared to how it would have been laid out by the design software without the length matching constraint, is denoted as M in Figure 17. Such a feature M would normally be easily discernible by a visual inspection of an actual printed circuit board. For example, it would be reasonable to conclude that there would be no reason for the “detour” taken by the cosine track 45 in Figure 17 other than for length matching purposes because it is clear that it would have been far more space efficient to adopt the layout as shown in Figure 15. Length matching of tracks has been described above in relation to the printed circuit board 40 that is used in the detector head 4, and this ensures that the correct phase difference between the components of the analogue quadrature signal is maintained within the printed circuit board 40 through to the quadrature output 46. But it will be understood that it would be of benefit to apply the same concept to the electrical signal pathways that carry the analogue quadrature signal onwards through the laser encoder system 1. In this respect, it can be useful to consider the signal pathways of the laser encoder system 1 as a whole, with the ideal being to have the length of the electrical pathway for the constituent sine signal the same as the length of the electrical pathway for the constituent cosine signal.

[0101] This is illustrated schematically in Figure 18, which shows that length matching M has been implemented not only in the detector head 4 (which is the origin of the analogue quadrature signal) but also in the electrical cable 3 which carries the analogue quadrature signal to the laser unit 2. This can be achieved by careful matching of the electrical cores 47 in the cable 3 that carry the respective components of the analogue quadrature signal. Length matching M is also implemented in the laser unit 2, for example using a similar approach to that shown in Figure 17 in the layout of the relevant tracks that carry the quadrature signal components in the printed circuit board in the laser unit 2. As illustrated in Figure 18, by length matching the electrical pathways it is ensured that the phase difference between the quadrature signal components is preserved.

[0102] Length matching of these pathways is ideally carried out at least up to the point where the quadrature analogue signal is digitised and interpolated by a signal conversion unit 60. The role of the signal conversion unit 60 will be discussed in more detail below. The signal conversion unit 60 is shown in Figure 18 as forming part of the laser unit 2, which would be the case if the laser unit 2 is configured to output a digital quadrature signal (see above) but if the laser unit 2 is configured to output an analogue quadrature signal (without conversion) then the signal conversion unit 60 would for example be in the interface 6 (for example the RPI20 or RPI30 parallel interface mentioned above) with the laser unit 2 being another component in the chain rather than the end of it.

[0103] Figures 19 and 20 will assist in a further understanding regarding the role of the signal conversion unit 60. Figure 19 is based on Figure 18 but does not show the signal conversion unit 60 as being part of the laser unit 2 (because it could for example be a functional component of the interface 6 as mentioned above), and also shows an overall signal pathway 49 which carries the analogue quadrature signal AQ from the detection and processing circuitry 41 to the signal conversion unit 60. Figure 19 also schematically shows the core functional components of the signal conversion unit 60. In this respect, the analogue constituent signals received at the input of the signal conversion unit 60 are converted into digital form by A / D converters 61 and the digitised versions of these signals are then passed through an interpolator algorithm (IA) 63 to generate a pair of digitised and interpolated output signals, which are illustrated schematically at the output of the signal conversion unit 60 as a digital quadrature signal DQ. For example, the interpolator algorithm 63 could base the interpolation on the most significant bit (MSB) of the digitised signal from the A / D converter 61, such that the interpolated outputs would follow the zero crossings of the sine and cosine signals. The digital quadrature signal DQ is passed to a positional data deriving unit 70 which is operable to derive positional data PD from the digital quadrature signal DQ.

[0104] To illustrate the digitisation and interpolation graphically, Figure 20 shows the analogue quadrature signal AQ (comprising a pair of cosine and sine signals), and below that shows two example conversions into digital quadrature signals DQ1 and DQ5. For the first conversion an interpolation factor of 1 is used by the interpolator algorithm 63, so that the digital quadrature signal DQ1 (which comprises a pair of digital signals Al, Bl corresponding respectively to the original cosine and sine signals) has the same frequency as the original analogue quadrature signal AQ. For the second conversion an interpolation factor of 5 is used by the interpolator algorithm 63, so that the digital quadrature signal DQ5 has a frequency that is five time that of the original analogue quadrature signal AQ.

[0105] The A and B signals (marked as Al, Bl and A5, B5 in Figure 20) are used to drive a digital counter which may form the basis of the positional data derived by the positional data deriving unit 70. Accordingly, the spatial resolution in the positional data PD derived from the DQ5 signal will be five times that derived from the DQ1 signal, so that smaller movements of the target optic 8 can ultimately be measured. The resolution of the DQ1 signal is represented as R1 in Figure 20, and similarly the resolution of the DQ5 signal is represented as R5. The signal conversion unit 60 is sometimes just referred to as an interpolator.

[0106] It is preferable to apply length matching to the analogue sine and cosine signals all the way to the digitiser (i.e. to the signal conversion unit 60 of Figures 18 and 19). The interpolation process effectively performs an arctangent (or inverse tangent) operation to calculate the angle and therefore position, and any phase deviation causes an error in this calculation. This is particularly important for applications where sub-nanometre accuracies are required, for example when using a high precision interpolator as provided by the RPI20 or RPI30 parallel interface (mentioned above), both of which interfaces receive an analogue quadrature signal from the laser unit 2. If the digital quadrature (A / B) signal is generated by the laser unit 2 itself, length matching of the digital quadrature signal itself is not particularly critical and rather it just becomes a digital signal integrity issue that can be handled in a known way. In this respect, the exact position of the A / B edges is not typically critical so long as the positional data deriving unit 55 can sample the four states correctly.

[0107] Length matching of printed circuit board tracks has been considered previously in high-speed digital systems (involving clock rates of the order of hundreds of MHz or GHz) to ensure that multiple signals arrive at the same time. However, the use of length matching has not been previously proposed in the context of a laser encoder system such as described herein, and for relatively low speed analogue systems (of the order of 10 to 30 MHz). As set out above, in the particular context of an analogue quadrature signal that is used to describe or encode position, the present applicant has determined that (particularly in some scenarios) the precision of the interpolated position within one cycle can be critically dependent on the accuracy of the phase alignment of the sine and cosine signals (though variations in their offset and amplitude will also have an impact). The difference in time it takes for sine and cosine signals to reach the digitisation generates a phase error which becomes more significant as the frequency of signal (velocity) increases. For example, a target (maximum) phase error of 0.2° at 2 m / s (12.66 MHz) amounts to a target (maximum) delay difference of about 44 ps, which is into the realms of high-speed signal length matching, which has not previously been proposed in the context of a laser encoder system.

[0108] In practice it is not usually required to make the lengths of the two signal pathways precisely equal, so instead it can be considered that an embodiment of the present invention should aim to make the two lengths sufficiently close to one another to bring the phase difference at a reference point to within a predetermined acceptable range of the expected phase difference. The expected phase difference in the above-described embodiment is 90°. The reference point could for example be the point where the analogue quadrature signal is digitised and interpolated by the signal conversion unit 60. The acceptable range for the phase difference might for example be within 5° of the expected phase difference, or more preferably within 1° of the expected phase difference, or more preferably still within 0.5° or even 0.2° of the expected phase difference; this can be referred to as the phase error.

[0109] These issues will now be discussed in more detail, in order to gain a better understanding of what sorts of factors would typically be taken into account when determining the degree of length matching that might required depending on the circumstances.

[0110] The maximum acceptable phase error referred to above can itself be considered as dependent on the maximum acceptable sub -divisional error (SDE) and maximum target velocity for the application concerned, so those values could instead be considered as a basis for determining what length matching is required. The phase error can be converted to a timing error which can itself be converted into a length matching requirement.

[0111] For example, with a 158 nm Lissajous the contribution to SDE from phase error will be approximately 1 nm for a phase error of 5° and approximately 0.1 nm for a phase error of 0.5°. In this respect, an approximation or rule of thumb for the overall SDE (in nm), taking into account offset, amplitude and phase error, can be determined based on the following expression:

[0112] SDE in nm = (0.25 nm x offset in %)

[0113] + (0.125 nm x amplitude mismatch in %)

[0114] + (0.22 nm x phase error in degrees).

[0115] Therefore, an approximate relationship between the phase error PE (in degrees) and the sub -divisional error SDE for the phase contribution (in nm) alone is as follows:

[0116] SDE = 0.22 x PE

[0117] PE = 4.55 x SDE

[0118] Therefore, considering just the phase contribution to SDE and with a phase error PE of 5°, the contribution to SDE due to phase is 0.22 nm x 5 = 1.1 nm. A more precise value for the phase contribution to SDE can be determined by plotting [atan2(cos(0), sin(9 + PE)) - atan2(cos(0), sin(0))] x [158 nm / 360°]. Putting in a phase error PE of 5° and plotting over the full cycle, calculating (max - min) / 2 gives an SDE value of 1.1 nm, which is a close match to the above approximation. A more precise value for the contribution to SDE from amplitude mismatch and offset can similarly be obtained based on the expression [atan2((CA x cos(0)) + CO, (SA x sin(9 + PE)) + SO) - atan2(cos(0), sin(9))] x [158 nm / 360°] where CA is the cosine amplitude, CO is the cosine offset, SA is the sine amplitude and SO is the sine offset.

[0119] Returning to the example of a 158 nm Lissajous and a phase error (between sine and cosine signals) of 0.5° (which is equivalent to a phase contribution to SDE of 0.1 nm) and a target that is moving at a speed of 1 m / s, let us now consider what time delay (between sine and cosine signals) this phase error of 0.5° would represent. One Lissajous is 158nm, so it takes (1 m / 158 nm) or 6.329 million Lissajous to cover a length of 1 m. Each Lissajous is one sine / cosine cycle, so if the target is travelling at 1 m / s then that amounts to 6.329 million cycles per second or 6.329 MHz. Therefore, based on the standard wave equation (v = x f) or (k = v / f) it can be determined that the period (or wavelength X amounts to 158 ns, i.e. 1 Lissajous takes 158 ns. A quicker derivation is that 1 m / s is 1 nm / ns, so a full 158 nm cycle takes 158 ns. Since a full Lissajous (i.e. 360°) amounts to 158 ns, the time delay associated with just 0.5° of the full 360° amounts to (0.5° / 360°) x 158 ns = 0.2194 ns = 219.4 ps. Therefore, for a target moving at a speed of 1 m / s, a phase delay of 0.5° amounts to a time delay of 219.4 ps, and at 2 m / s the time delay would be 109.7 ps.

[0120] It would be useful to distil the above into a simple expression for the time delay based on the maximum acceptable SDE (for the phase contribution) and the maximum speed of the target. Firstly, based on the above explanations, the time delay TD (in ns) can be expressed as (PE / 360) x (LJ / TS) where PE is the phase error (in degrees) and LJ is the Lissajous (in nm) and TS is the speed of the target (in m / s). Secondly, the PE can be expressed as (SDE / 0.2) based on the above- mentioned rule of thumb, where SDE is the sub -divisional error for the phase contribution (in nm). Substituting the expression for PE into that for TD, we get:

[0121] TD = (SDE / 0.2 / 360) x (LJ / TS) = (SDE x LJ) / (72 x TS).

[0122] So, this provides a limit on the time delay TD (between sine and cosine signals) in order to keep within the required sub -divisional error SDE for the given speed S of the target and for the given Lissajous LJ. Now can consider how this limit on the time delay will translate into a length matching requirement on the signal pathways mentioned above. When working out propagation delays in circuits for high-speed design, the circuit designer (or circuit design software) will typically make an assumption that there will be a nominal delay introduced for each metre of signal pathway. For example, the typical propagation delay for a category 5e UTP cable is around 5 ns per meter, i.e. there is a 5 ns delay introduced for each metre travelled. The exact value will depend on the type of cable, and will differ slightly for the tracks on a PCB, and this is turn will depend on the track width and so on. For the purposes of the example presented in the present application, it will be assumed that the value is a nominal 5 ns / m along the whole of the signal pathway under consideration.

[0123] Therefore, using the above example, at a target speed of 1 m / s the time delay was determined to be 219.4 ps, which is the maximum acceptable time delay between the sine and cosine signals that would result from a path length difference travelled by these signals. This amount to a length matching requirement of (219.4 ps / 5 ns / m) = 43.9 mm, which is the maximum acceptable difference in path length between the sine and cosine signals. And at 2 m / s the time delay was 109.7 ps, which would amount to a length matching requirement of 21.9 mm. By way of example, even when considering only the PCB used in the laser unit 2, the sine and cosine channels might be separated by, say, 15 mm from the connectors, so this would already give a difference in path length of 30 mm (15 mm at the input and 15 mm at the output) without any length matching measures being employed. This is already over the length matching requirement of 21.9 mm determined above for a maximum acceptable SDE contribution of 0.1 nm and a target speed of 2 m / s. It will therefore be appreciated that path length differences can very quickly accumulate in the absence of length matching techniques being explicitly employed. It will be understood that the nominal value of 5 ns / m for the propagation delay per metre will likely be at the faster end of what would be expected in practice (i.e. with a value at the lower end of the expected range), so that length derived matching requirement will accordingly be less stringent than might actually be required in practice. With a value more towards the middle or the slower end of the expected range (i.e. with a value greater than 5 ns / m) the derived length matching would accordingly be more stringent. In addition, since phase error is only one contribution to SDE (see above), the length matching requirements would likely be made more stringent to account for those contributions too (for example based on the 0.2° example mentioned above).

[0124] Both parts of the above analysis can be combined into a more general overall expression for the length matching requirement LM (in mm), based on a nominal delay factor DF (in ns / m), as follows: LM = 1000 x TD / DF. Substituting the above expression for TD into this we get:

[0125] LM = (125 x LJ x SDE) / (9 x TS x DF) where:

[0126] LJ is the Lissajous length (in nm);

[0127] SDE is the maximum acceptable phase-related sub -divisional error (in nm); TS is the maximum expected speed of target (in m / s); and DF is the assumed signal pathway delay factor (in ns / m).

[0128] For example, if it is decided that the maximum acceptable sub -divisional error SDE for the application concerned is 0.1 nm and that the maximum expected target speed TS is 2 m / s, and with a Lissajous LJ of 158 nm and an assumed signal pathway delay factor DF of 5 ns / m, the above expression gives LM = (125 * 158 x 0.1) / (9 x 2 x 5) = 21.9 mm, which agrees with the previous value based on the same starting points.

[0129] It will be appreciated that the above approach to determining the degree of length matching that will be required in order to achieve the required benefit (which is to keep the SDE within acceptable limits), and that other approaches can equally well be taken instead, for example based on a different set of targets, constraints and assumptions.

[0130] Although the required length matching is ideally applied over the whole signal path length from the detection and processing circuitry 41 all the way to the A / D converters 61 in the signal conversion unit 60 (see Figure 19), i.e. taking into account all PCBs, connectors, internal and external cables and so on, it will be appreciated that sufficient benefit might still be obtained by length matching over only part of the overall signal path.

Claims

CLAIMS1. A laser encoder system comprising a detector head which is adapted to generate a detection signal from which positional data relating to the position of a target is derivable, the detection signal being an analogue quadrature signal comprising first and second constituent signals having a predetermined phase difference between them, wherein first and second electrical pathways respectively carrying the constituent signals within the laser encoder system are length matched to maintain the phase difference between the first and second constituent signals within a predetermined acceptable range.

2. A system as claimed in claim 1, comprising a laser unit that is connected in use to the detector head to provide laser light to the detector head.

3. A system as claimed in claim 2, wherein the laser unit is connected in use to the detector head via an optical fibre assembly to provide the laser light to the detector head via the optical fibre assembly.

4. A system as claimed in claim 1, 2 or 3, wherein the detection signal is an interference signal resulting from interference between a reference laser beam and a measurement laser beam, with the measurement laser beam being directed onto and at least partially reflected back from the target.

5. A system as claimed in any preceding claim, wherein length matching is employed in relation to first and second electrical tracks within a printed circuit board carrying the first and second constituent signals respectively.

6. A system as claimed in claim 5, wherein length matching is employed at least in relation to a printed circuit board in the detector head.

7. A system as claimed in any preceding claim, wherein length matching is employed in relation to first and second electrical cores within an electrical cable carrying the first and second constituent signals respectively.

8. A system as claimed in claim 7, when dependent on claim 2, wherein length matching is employed at least in relation to an electrical cable connected between the detector head and the laser unit.

9. A system as claimed in any preceding claim, wherein the first and second electrical pathways are length matched at least until a point at which the analogue quadrature signal is digitised and interpolated by a signal conversion unit.

10. A system as claimed in claim 9, wherein length matching is employed in relation to electrical pathways in the signal conversion unit, such as electrical tracks formed on a printed circuit board.

11. A system as claimed in claim 9 or 10, when dependent on claim 2, wherein the signal conversion unit is or forms part of or is implemented as a functional unit of the laser unit.

12. A system as claimed in any one of claims 9 to 11, wherein the signal conversion unit is or forms part of or is implemented as a functional unit of an interface component.

13. A system as claimed in any one of claims 9 to 12, wherein the signal conversion unit is arranged to output the digitised and interpolated signal to a positional data deriving unit, with the positional data deriving unit being adapted to derive positional data from the digitised and interpolated signal.

14. A system as claimed in claim 13, wherein the positional data deriving unit is or forms part of or is implemented as a functional unit of an interface component or controller.

15. A system as claimed in any preceding claim, wherein the predetermined phase difference is 90°.

16. A system as claimed in any preceding claim, wherein the detection signalrelates to a distance between the detector head and a target.

17. A system as claimed in any preceding claim, wherein the first and second constituent signals are sinusoidal signals.

18. A system as claimed in any preceding claim, wherein the first and second constituent signals are cosine and sine signals respectively.

19. A system as claimed in any preceding claim, wherein the first and second constituent signals have a frequency of less than 500 MHz, more preferably less than 100 MHz, more preferably less than 60 MHz, more preferably less than 30 MHz.

20. A system as claimed in any preceding claim, wherein the predetermined acceptable range is dependent on a predetermined acceptable sub -divisional error in the derived positional data which is attributable to the phase difference.

21. A system as claimed in claim 20, wherein the predetermined acceptable range is dependent on a speed of the target.

22. A method of arranging electrical pathways for a laser encoder system as claimed in any preceding claim, comprising length matching the first and second electrical pathways carrying the constituent signals within the laser encoder system to maintain the phase difference between the first and second constituent signals within a predetermined acceptable range.

23. A method as claimed in claim 22, when dependent on claim 9, comprising length matching the first and second electrical pathways at least until a point at which the analogue quadrature signal is digitised and interpolated by the signal conversion unit.

24. A method as claimed in claim 22 or 23, when dependent on claim 20, wherein the degree of length matching between the first and second electrical pathways is determined in dependence upon a predetermined acceptable sub-divisional error in the derived positional data which is attributable to the phase difference.

25. A method as claimed in claim 22, 23 or 24, when dependent on claim 21, wherein the degree of length matching between the first and second electrical pathways is determined in dependence upon a speed of the target.

26. A method as claimed in any one of claims 22 to 25, wherein the predetermined acceptable range is within 1° of the predetermined phase difference.

Images