EP0328611A1

EP0328611A1 - Signal processing apparatus

Info

Publication number: EP0328611A1
Application number: EP19880907353
Authority: EP
Inventors: Mark Adrian Vincent Chapman
Original assignee: Renishaw PLC
Current assignee: Renishaw PLC
Priority date: 1987-08-22
Filing date: 1988-08-19
Publication date: 1989-08-23
Also published as: WO1989002062A1; GB8719878D0; JPH02500218A

Abstract

Signal processing apparatus is described for processing a plurality of analogue signals from the detectors, for exam ple, of a laser interferometer system. As can be seen in Fig. 3 of the drawings the apparatus uses one or more analogues to digital converters (AD1, AD2) to digitise the signals, and a look-up table or tables (LUT1, LUT2) into which are pre-pro grammed information such as signal strength data and interpolated position data. The digitised signals from the converters are passed to the input addresses of both of the look-up tables, which output the corresponding signal strength and posi tion data extremely rapidly. The outputs of the look-up tables are latched and subsequently read by a micro-processor which provides the required information. The system is very flexible and extremely fast thus providing effectively continu ous monitoring of the detector signals.

Description

SIGNAL PROCESSING APPARATUS

The present invention relates to signal processing apparatus.

The invention has particular but not exclusive relevance to apparatus which can be used for extracting information from a plurality of a.c. outputs of a detector system.

A problem which occurs in a laser interferometer system is that if the laser beam is broken for any reason while the machine on which a measuring operation is being performed is moving, the interferometer counters will miss one or more counts, so that when the beam is restored, all subsequent readings will be in error. Therefore it is essential to be able to determine, effectively on a continuous basis, the signal strength coming from the detector system for the interferometer. The detector signals may be in the form of a plurality of a.c. signals in phase quadrature.

Electrical circuits are known for the measurement of the strength of an a.c. signal and which involve the measurement of the peak amplitude, or the mean amplitude, of the waveform using various combinations of resistors, — diodes and capacitors.

However these circuits can only monitor relatively slowly changing amplitudes depending upon the response time of the circuit. These circuits therefore are not suitable for high speed monitoring of signal strength such as to provide effectively continuous monitoring of signal strength as required in a laser interferometer. Another disadvantage of such known circuits is that they do not work at all if the frequency of the signal being monitored drops to zero.

Another requirement of a laser interferometer is to be able to interpolate between the fringe counts provided by the counters in order to increase the resolution of the in erferometer.

It is known from UK Patent No. 2,000,589 to interpolate the readings provided by a scale and read-head system by using an analogue to digital converter which produces digital signals from the a.c. signals produced by photo-detectors in the read-head, and to pass the digital signals to a micro-processor for interpolating between the readings. Such digital micro-processor although very much quicker than the previous analogue system is still not fast enough for application in a laser interferometer in which the fringe counting rate may reach several million counts per second.

The present invention as defined in any one or more of the appended claims, in one aspect allows relatively high speed monitoring of the strength of the a.c. outputs of a detector sytem, and in another aspect allows high speed interpolation of such output signals.

The term "relatively high speed" as used throughout this specification should be understood to mean capable of processing several million readings per second so as to provide effectively continuous processing. In another aspect, the invention as defined in one or more of the appended claims, provides means for the high speed determination of variations of signal strength, and simultaneously for high speed interpolation between fringe counts provided by the counters linked to the detector system of a laser interferometer. By this means it becomes possible not only to know that the laser beam has been broken, so that all subsequent readings will be erroneous, but also at precisely which reading the beam was broken.

The invention may however, be used with a variety of detector systems which produce electrical signals in response to variations of a parameter being measured, and is not limited to a laser interferometer as described above. For example, the invention may be used in other optical or magnetic measuring systems.

Neither is the invention limited to the type of information to be extracted from the detector signals, (eg interpolation, or signal strength monitoring) and may for example, depending on the nature of the detector outputs, provide means for determining other parameters of the signals.

Examples of the invention will now be more particularly described with reference to the accompanying drawings in which:

Fig. 1 is a block diagram of the main components of the signal processing apparatus of the present invention.

Fig. 2 is an input/output diagram for the converter.

Fig. 3 is a more detailed block diagram of the signal processing apparatus shown in Fig. 1 including some additonal components specifically for use in a laser interferometer embodiment. The invention as illustrated in Figs. 1 and 2 is described with reference to the requirements of a laser interferometer but it is to be understood that the invention in its broadest aspects is not limited as to the origin of the signals being processed or the information to be obtained therefrom.

Referrring now to Fig. 1, three photo-diode detectors A,B and C are shown which form part of a detector system for a laser interferometer which may be of known type and the constructional details of which are not described in detail.

The three detectors, in known manner, produce a.c. voltage signals in quadrature from the interference fringes in the light beams falling on them. The a.c. signals however, will include a d.c. level, and this is removed by subtraction devices SI and S2, known per se, to produce two quadrature a.c. voltage signals (referred to as the sine and cosine signals) of amplitude A with no d.c. offset. The two a.c. signals are respectively of the form A Sin θ and A Cos θ, where A is the amplitude of the signal and i a measure of the beam strength.

In order to provide the necessary high speed operation, the invention provides two flash analogue to digital converters AD1 and AD2 which respectively receive the sine and cosine voltage signals and convert them to digital form, to produce binary numbers representing the instantaneous voltages of the signals.

The range of the converter outputs depends on the resolution of the converter used, so that, in the present example where a 7-bit conversion is used, the converter output is an integer in the range 0 to 127. As can be seen in Fig. 2 the arrangement is such that for an input to the converter of -V volts the output is 0, and for an input of +V the output is 127.

Thus, in this example, the two converter outputs which can be designated x and y, are of the form x = k cos θ + 63.5 and y = k sin θ + 63.5 where k is a new constant representative of the amplitude in bits of the input waveform at any instant, and 63.5 is the half scale offset value introduced by the conversion.

Re-arranging these expressions gives k cos θ = x - 63.5 1 k sin θ = y - 63.5 . . . . . . . 2

so that the values of θ or k can be determined from the values of x and y which are the outputs of the converters, k giving a measure of beam strength and θ being the phase angle of the signals from the detectors.

One problem which can be solved using the invention is that of high speed monitoring of the amplitude of the interference signal received by the detectors in an interferometer.

As explained above it is a requirement in a laser interferometer that a warning signal is generated should the laser beam be broken by some object obstructing its path during a measuring operation, so that the resulting inaccurate measurements can be discarded and re-taken. A low signal strength warning, or an excessively high signal strength warning, is also of benefit to an operator giving an indication of possible problems to come. Detector signal strength monitoring is also useful for example, in setting up the laser interferometer when the interferometer is stationary prior to taking measurements, where the correct alignment of the laser is indicated by -6- detection of the maximum signal strength on the detector system.

However, the conventional circuits for measuring signal amplitude mentioned earlier are not fast enough for high speed monitoring and thus cannot detect very short duration breaks of the laser beam, such as occur for example if a chip from an operating machine cuts the laser beam.

Also, in the case where the laser used in the interferometer outputs a single frequency beam, the conventional methods cannot determine changes in signal strength when the interferometer parts are stationary, since the signal frequency falls to zero and each detector outputs a constant, but not necessarily equal voltage at that time. However, because the detectors are receiving parts of a combined beam in which certain levels of constructive and destructive interference are present, there will still be a ^"constant phase difference between the detector outputs.

The invention makes use of this fact and provides an EPROM look-up table LUT into which the values of k have previously been programmed using the expression;

k = (y - 63.5)² + (X - 63.5)² . . . . . 3 which is generated from the simple equation

sin² θ + cos² θ - 1 4 so that k² sin² θ + k² cos² θ = k² . . . . . . . 5 from which k is determined for the particular converters used by substitution of k sine and k cosθ from equations 1 and 2 above.

The look-up table consists of an array of data which has previously been calculated, and can be visualised as a two di ensional table having rows and columns for values of x and y respectively, with the value of k for any combination of values of x and y being shown in the square at the intersection of the respective row and column for x and y. The output of the table is the information which has been previously programmed into each square of the table.

The x and y outputs of the two converters AD1 and AD2 are passed to the address lines of the look-up table from which the value of k is read at sufficiently high speed to enable effectively continuous monitoring of the signal strength to take place. The look-up table includes a standard memory which produces an 8-bit output which is passed to a micro-processor MP which is programmed to produce information in the required form from the output of the look-up table.

Not all of the 8 bits need to be encoded with the signal strength, some bits can be programmed to flag fault conditions, such as signal strength low, high or lost altogether. These flags are included in the information contained in the 8-bit information code as the look up table locations are programmed and are set according to the value of k at that location. For example, if the signal strength falls to zero then both k sinθ and k cosθ become zero so that both x and y are 63.5. The appropriate squares in the look-up table are therefore programmed so that five of the bits show a signal strength code of zero, the signal low, and signal lost bits are coded with a 1 or zero whichever is chosen to indicate a fault condition, and the signal high is coded oppositely to indicate a normal condition.

This then provides electrical signals as the output of the look-up table which can be continuously monitored electronically to detect momentary signal faults of much shorter duration than could be detected by a micro¬ processor alone even if the micro-processor was dedicated to doing nothing other than reading the output of the table.

An implementation of this is described with reference to Fig.3.

It can be seen from expression 3 that the value of k can be determined independently of changes in the value of θ in the two signals k sin θ and k cos θ, and therefore enables any changes in the value of k to be determined even when the interferometer is stationary and the value of θ is constant, i.e. when the frequency of the two signals drops to 0.

In the case of an interferometer the value of θ derived from the signals of the photo-diode detectors is a measure of the parameter being measured by the interferometer, e.g. distance moved by a moving element of a machine.

In making a distance measurement a complete cycle of the detector output, i.e. θ changing from 0° to 360° usually represents a movement of the moving element of the machine of one half of the wavelength of the light used in the interferometer. This assumes that the light beam from the laser of the interferometer traverses the distance to the moving element of the machine twice, i.e. there and back, so that the optical path length change is twice the movement of the machine element. For more accurate measurements e.g. down to nanometres of movement -it is necessary to interpolate between the values of θ =0° and θ = 360°. This can be done by calculation by some form of computing means-/ such as -a -micro-processor.

However, the invention proposes the use of an EPROM look-up table to evaluate θ from the outputs of the two converters as follows: since Tan θ = k sin θ k cos θ

θ = ARCTAN k sin θ k cos θ

substituting for k cos θ and k sin θ from expressions 1 and 2 gives:

θ = ARCTAN y - 63.5

X - 63.5 6

It can be seen from this expression that the value of θ can be determined independently of the value of k, thus it is possible to correctly evaluate the interpolation angle θ even under conditions of varying signal strength.

The values of θ are pre-programmed into the look-up table and by using the x and y outputs of the converters to address the rows and columns of the look-up table, the corresponding values of θ can be read from the table at high speed to provide the required interpolation of the distance moved by the machine part. This has the advantage of operating at much higher speed than a micro-processor and thus allows the apparatus to operate, for example, in a high speed, very high resolution servo loop position control sytem.

Since the outputs of available look-up tables are presently 8-bit outputs, this enables the range of values of θ from 0° to 360° to be interpolated as the machine element moves by a factor of I/256₇- but since the^~optical path length change is twice the movement of the machine element the actual movement of the machine element can be interpolated down to a factor of ₅₁₂ ^{of the} wavelength -10- of light emanating from the laser beam. In practice this gives interpolation down to 1.24 nanometres.

Clearly the resolution of the converter and the size of the look-up table need not be same for signal strength monitoring as it is for interpolation. The choice of converter resolution and the size of the look-up table will depend on the amount of information required.

The speed with which the apparatus of the present invention provides information can be used beneficially in connection with the beam strength monitoring embodiment to provide automatic gain control for the signals at the subtraction devices SI and S2 to boost or reduce the signal level as appropriate.

Fig. 3 shows a more detailed embodiment of the invention in which additional refinements are included. Those parts of this embodiment which are identical to the parts shown in Fig. 1 are given the same reference numerals.

The detectors A, B and C are photo-diode detectors which are arranged to receive differently polarised parts of the interfering combined beam entering the detector system. The arrangment may, for example, be as disclosed and claimed in our co-pending UK Patent Application No. 8718803.

In that arrangement each detector A, B and C is exposed only to light passing through a polaroid immediately in front of it, the polarisation states of which are respectively at 0°, 45° and 90° relative to one another. Thus the signals produced by the detectors A, B and C are sine waves in quadrature having the characteristics:

K sin (θ + 0°) + a d.c. offset K sin (θ + 90°) + a d.c. offset and K sin (θ + 180°) + a d.c. offset

where K is a constant.

From these three signals the subtractors SI and S2 produce the two sine and cosine signals A Sin θ and A Cos θ which form the inputs to the converters AD1 and AD2.

From Fig. 3 it can be seen that the outputs from the two converters are passed to two look-up tables LUT1 and LUT2. The 8-bit outputs of the two look-up tables are passed via two BUS interfaces Bl and B2 to the input of the micro-processor MP. Table LUT1 is a composite having two parts, one of which PI, contains values of the a.c. signal strength of the detectors A, B and C and the other one of which P2, contains values of the d.c. levels in detectors A and C. The output from table LUT1 is switched as described later to provide either of these values.

The BUS interfaces perform the function of simultaneously latching the instantaneous readings of the two look-up tables so that when addressed sequentially by the micro-processor, the values of the signal strength and the interpolated distance reading can be provided for the same instant in time. This combined use of two look-up tables addressed from the same two digitised sine and cosine signals enables the interpolation angle θ and the signal strength, to be simultaneously evaluated. In addition, however, it enables the operator of the interferometer to know not only at any instant that the laser beam has been broken, and therefore certain readings will be erroneous, but also to know which was the last correct reading. This saves the operator from having to repeat all of the previous readings after receiving a warning signal from the micro-processor that the beam has been broken. Figure 3 also shows electronic monitoring of one of the output bits from the signal strength look-up table as previously described. This bit has been programmed to flag signal loss. When momentary signal loss occurs a pulse is generated on signal line LI which triggers a flip flop device F, which then changes state to indicate that a momentary loss of signal has occured. The output from flip flop F is periodically read by the micro-processor unit via bus interface Bl, thus allowing the detection of beam breakages of much shorter duration than would be possible using the MPU alone. The flip flop F can be reset by the MPU using line L2. Gating circuitry (not shown) ensures that the flip flop is not triggered during the transition between successive data valid output states from the look-up table.

As explained above, any variation in signal strength can be determined even when the interferometer is stationary, which assists in setting up the interferometer components. There are occasions, however, when the component being aligned does not involve the interferometer but may be a simple reflector which reflects the laser beam onto the detector. In these circumstances there will be no a.c. interference signal to provide a phase difference in the detector signals, so that subtracting signal B from signals A and C in subtractors SI and S2 respectively, could eliminate the d.c. level as well and leave no signal. Another complication arises where the laser beam sends out polarised light, in that different optical components have different effects on the polarisation. Thus looking at the beam strength hitting a single detector behind a polaroid would provide a variety of signal strengths, even down to zero, if orientation of the polaroid did not match that of the returning light beam.

To overcome this problem it is necessary to look at the signals from two detectors having polaroids in front of them polarised in different orthogonal orientations. The values programmed into the appropriate look-up table may then be the sum of the two detector signals divided by two, or the square root of the sum of the squares of the signals.

In order to enable the d.c. levels of the two detectors A and C to be monitored, the apparatus is modified by the introduction of a switch Tl in the line from the detector B, which removes the signal from detector B from the inputs to the subtracters SI and S2.

By this means the converters receive signals relating to the d.c. levels of detectors A and C produced from the laser beam. Simultaneously with the switching out of detector B, a second switch T2 is operated by the micro-processor unit to cause the address lines from the converter AD1 to address the second part P2 of the look-up table LUT1, in which is stored pre-programmed beam strength data.

Thus the invention also provides the capability to monitor beam strength of a non-interfering beam, which capability is of value in setting up certain parts of the apparatus without requiring the use of the interferometer component.

A further feature which enhances the usefulness of the interferometer is the use of one of the bits from each of the converters AD1 and AD2 to drive an up/down direction decoder and fringe counting circuit (not shown) . The top bits of the two converter outputs change state as the sine and cosine signals pass through zero, and thus they can be used as two logic input signals in quadrature for the decoder and counter.

Also, since many standard machine controllers for coordinate measuring machines or machine tools require quadrature input signals from their measurement scales, these two logic signals in quadrature from the converters can be passed to a machine interface in place of the signals from the machine scale readers thus allowing the laser interferometer to be used as a measurement scale for the machine giving a basic resolution ofγ8. The resolution can be improved by re-programming the look-up table to include the codes needed to generate higher resolution quadrature signals and passing these to the machine controller. Alternatively the look-up table could be programmed to include grey codes if required by a particular machine controller.

The above-described method of improving resolution by programming codes into the look-up table can be adopted in apparatus other than a laser interferometer system.

In a further modification for improving the accuracy of the apparatus, the figures stored in the look-up tables may be corrected for errors whereby corrected figures can be output directly. Errors may arise, for example, from the characteristics of the optical components not being exact, non-linearity of the photo-diode detectors, or the detector signals not being exactly in quadrature.

The errors will however be constant for any given combination of detectors and optical devices so that once measured they can be allowed for. The errors can be determined either by physical calibration of the system or by analysis of the Lissajous figure distortion observed by plotting k sin θ against k cos θ. The data to be stored in the look-up table is then calculated using corrected values provided from the error analysis.

Alternatively where the error is not constant due to a change of optical components or their alignments, the error data derived by analysis of the lissajous figure can be used by the micro-processor to dynamically correct values read from the look-up table using an algorithm which is updated according to the distortion of the lissajous figure at any given time.

Hence it can be seen that the invention provides a very rapid and highly flexible system for processing signals which is significantly faster than a micro-processor alone. The system can deal with massive flows of information giving what is effectively continuous monitoring of the information.

Claims

1. Signal processing apparatus for processing two or more analogue signals comprising: analogue to digital converter means for digitising each of the signals to provide digitised output signals, characterised by one or more pre-programmed look-up tables to the input addresses of each of which are passed one or more of the digitised output signals from at least two convertors, and which produces an output in accordance with the information stored within the table, and means arranged to receive the output from the look-up table or tables and which provides information in the described form from said output.

2. Signal processing apparatus according to claim 1

" characterised in that the means receiving the outputs from the look-up table or tables is a computer.

3. Signal processing apparatus according to claim 2 characterised in that means are provided to latch the outputs of the look-up table or tables and the latched values are received by the computer.

4. Signal processing apparatus according to claim 3 characterised in that the look-up table or tables contain different pre-programmed information in different locations therein, and the outputs of the converters are each passed to address.lines of the different areas of the table.

5. Signal processing apparatus according to claim 1 characterised in that the signals being processed are detector signals from the detectors of a position measurement system.

6. Signal processing apparatus according to claim 5 characterised in that the detector signals include at least a.c. components the phase relationships of which are known.

7. Signal processing apparatus according to claim 6 characterised in that said detector signal include a d.c. offset value.

8. Signal processing apparatus according to claim 6 characterised in that the look-up table or tables are pre-programmed with signal strength data, based on the amplitudes of the a.c. components of the signals.

9. Signal processing apparatus according to claim 6 characterised in that the look-up table or tables are pre-programmed with interpolated position data.

10. Signal processing apparatus according to claim 6 characterised in that the look-up table or tables are pre-programmed with both signal strength data and interpolated position data.

11. Signal processing apparatus according to claim 6 characterised in that the look-up table or tables are pre-programmed with grey codes or high resolution quadrature codes.

12. Signal processing apparatus according to claim 7 characterised in that the look-up table or tables are pre-programmed with signal strength data based on the d.c. offset values of the signals.

13. Signal processing apparatus according to any one of claims 5 to 12 characterised in that the means receiving the outputs of the look-up table or tables is a computer.

14. Signal processing apparatus according to any one of claims 5 to 12 characterised in that the detector system forms part of a laser interferometer system.

15. Signal processing apparatus according to claim 13 characterised in that the detector system forms part of a laser interferometer system.

16. Signal processing apparatus according to claim 7 and wherein the detector system produces three signals each having equal d.c. offsets, subtracter means are provided for subtracting two different pairs of said signal to produce a.c. signals with no d.c. offset, characterised in that the look-up table or tables are pre-programmed with signal strength data based on the d.c. offset values, switch means are provided for selectively removing the one of the signals from each pair from the inputs of the subtracting means when the a.c. components of the detector signals are zero, whereby the output of the subtracting means relates only to the d.c. offset values, the outputs of the subtracting means form the inputs of the analogue to digital converters, and the outputs of the converters are passed to the input addresses of the look-up table corresponding to the location of the signal strength data based on the d.c offset values of the signals.

17. Signal processing apparatus for processing a plurality of signals from a detector system of a laser interferometer comprising: means for generating a plurality of a«c. signals of known phase relationship from the detector signals, means for digitising each of the a.c. signals, and for providing digitised output signals representative of the a.c. signals, a first pre-programmed look-up table containing data relating to detector signal strength arranged to receive the digitised output signals and to produce an output indicative of signal strength, a second pre-programmed look-up table containing interpolated position data arranged to receive the digitised output signals and to produce an output indicative of said interpolated position means for passing the digitised signals to the inputs of both of the look-up tables simultaneously, latch means for capturing and holding the outputs of the two look-up tables, and a micro-processor for addressing the latch means to obtain the outputs of the two look-up tables which correspond to the same instant in time, and for producing an output indicative of both signal strength and interpolated position data of the interferometer system for the same instant.

18. Signal processing apparatus according to claim 2 or claim 13 characterised in that the computer is programmed to determine any error in the output, from the look-up table dependent on changes in operating conditions from those under which the information pre-programmed into the look-up table was calculated, and to correct the values programmed into the look-up table.