GB1601574A - Measuring apparatus employing optical polarisation transducers - Google Patents

Description

(54) IMPROVEMENTS IN OR RELATING TO MEASURING APPARATUS EMPLOYING OPTICAL POLA RISATION TRANSDUCERS (71) We, CENTRAL ELECTRICITY GENERATING BOARD, a British Body Corporate, of Sudbury House, 15 Newgate Street, London, ECIA 7AU, do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed, to be particularly described in and by the following statement:- This invention relates to measuring apparatus employing transducers which, in response to changes in a parameter to be sensed, effect modification of the polarisation of light transmitted through an optical element.

Such transducers find particular application in the determination of current and voltage on high voltage power distribution systems. Transducers making use of magneto-optic, electro-optic and electro-gyration effects employ passive transducers and hence do not require any electrical power at the high voltage.

These transducers offer other advantages, in particular freedom from saturation effects, ability to operate over a very large measurement bandwidth and high sensitivity. Transducers of this nature can be used to measure or sense changes in other parameters apart from magnetic and electric fields, for example they may respond to changes in pressure or temperature or to vibrations.

When linearly polarised light passes through a medium under the influence of a magnetic field, the direction of polarisation will in general be rotated. This phenomenon is known as the Faraday magneto-optic effect. The rotation which occurs is proportional to the line integral of the magnetic field along the propagation path and hence magneto-optic effect devices find immediate application for the measurement of current by sensing the magnetic field.

The electro-optic effect is the name given to the dependence of the linear birefringence on the external electric field. Linear birefringence in a medium is a phenomenon whereby two orthogonal linear directions of polarisation of the light propagating in the medium travel at different velocities.

Optical activity is that phenomenon whereby the direction of polarisation of linearly polarised light propagating through a medium will suffer a rotation. This may thus be described as a circular birefringence, that is a difference between the velocities of right and left-hand circularly polarised light waves. In the absence of external fields, optical activity is characterised by the gyration sensor. It occurs in some isotropic and in some anisotropic materials; it is due to either molecular or crystallographic structural preference for either right or left-handedness. The dependence of this optical activity on an external electric field is referred to as the electro-gyration effect.

The electro-gyration effect may be used for the measurement of electric field.

See for example the Paper "Method for Simultaneous Measurement of Current and Voltage on High Voltage Lines Using Optical Techniques" by Dr. A. J. Rogers, Proc. I.E.E. Vol. 123, No. Oct.1976. Similarly electro-optical effects may also be used for voltage measurement and reference in this connection may be made to the specifications of co-pending Applications Nos. 19073/77 and 19076/77. (Serial No's 1570802 and 1600277).

The measurement techniques described in the above-mentioned publication and co-pending applications make use of means, for example a laser, for transmitting a light beam of predetermined polarisation characteristics through one or more transducer elements which modify the polarisation of the transmitted light in accordance with the magnitude of the magnetic and/or electric fields. The output light then is passed through an analyser, possibly with a quarter wave plate before the analyser, and is then received by a photo-detector, for example a photodiode, which senses the magnitude of the incident light. In some arrangements two photo-detectors may be provided responsive to light transmitted through different transducers and the outputs of these photo-diodes may be combined and processed in order to obtain wanted information. In these arrangements, the incident light intensity on the photo-diode contains wanted information. In practice however the integrity of this information is reduced by a number of unwanted effects. One major problem in this respect is beam intensity noise, that is to say random variations in the beam intensity. Such random variations may occur for example in source characteristics (e.g. laser plasma tube noise) or due to variations in transmissivity of the propagation path or due to weather or as a result of changes in the photo-detector sensitivity. One very particular source of such noise arises due to the fact that there is a variation in sensitivity to light intensity across the detection surface of the photo-diode. Movement of the beam over the surface of a photo-detector will therefore cause changes in the output from this photo-detector.

This is aproblenwhieh is particularLy irnpant4nfree path 4evicsre -wtiere for example a light beam may be directed from the ground to a transducer assembly adjacent to a high voltage conductor and thence back to a receiver on the ground.

Vibration commonly causes lateral motion of the received light spot with respect to the photo-diode surface. The variations in the photo-diode sensitivity across ils aperture area will translate this movement into an unwanted output amplitude variation.

The present invention is directed to providing means discriminating between changes in intensity due to polarisation change and those due to intensity changes or apparent intensity changes.

According to this invention, in a measuring system employing a transducer which effects changes in polarisation characteristics of a light beam passing through an optical transducer element in response to an external parameter and in which the polarisation changes are sensed at a receiver by means including a photodetector providing an electrical output responsive to the intensity of the incident light signal after passage through a polarisation analyser, means are provided for introducing polarisation modulation, detectable at the receiver, in the light beam applied to the transducer element, the modulation being at a frequency sufficiently high that a required lower frequency output from the photo-detector including all the required signal components can be separated from modulation frequency signals by frequency filtering, and, at the receiver, there are provided a low-pass filter to separate a first component from the photo-detector, a bandpass filter and demodulator to separate a second component, the filters having passbands such that only one of the components carries the required signal changes due to polarisation changes effected by the transducer, and means for deriving the required signal from the first and second components.

In one arrangement, the low-pass filter passes signals including the required signals, the amplitude of these signals effectively being multiplied by the unwanted intensity variations. The modulation frequency signal is used to obtain an output at the modulation frequency which is de-modulated, e.g. by a synchronous detector, to provide the second component which, in this case, contains the unwanted intensity variations. The required output can thus be obtained by dividing, e.g. in an analogue divider, the magnitude of the first component by that of the second component.

Conveniently means are provided for demodulating the separated components dependent on the modulation frequency to extract a signal representive of the unwanted intensity changes and this extracted signal is then divided into a low frequency component from the photo-detector which includes the wanted signals and the unwanted intensity signals.

The modulation introduced by the modulating means is polarisation modulation; the modulating means should not produce any intensity variations with components close to the modulation frequency. This is readily achieved however using a magneto-optic or electro-optic device as the modulator.

With this technique a polarisation modulation is imposed upon the interrogating light beam at the point of transmission. In a magneto-optic device, the transducer produces rotation of linear polarisation. In an electro-optic device, the transducer changes the phase relation between the ordinary and extraordinary rays through the transducer. In an electro-gyration effect transducer, the field to be measured alters the activity, that is to say the difference between the velocities of right and left-hand circularly polarised light waves. Depending on the effect to be measured, the modulator would be arranged to produce a suitable type of polarisation change. This, in practice, may readily be achieved by utilising an appropriate form of transducer element through which the light beam is passed and which is subjected to an electric and/or magnetic field at the required modulation frequency. Although reference has been made more particularly to the effect of electric and magnetic fields on transducer elements, this type of measuring device may also be used to measure other parameters such as temperature, pressure and vibration amplitudes which similarly cause changes in the polarisation characteristics of a light beam passed through a transducer element subjected to the parameter.

It is preferred to use modulation of a sinusoidal form to facilitate separation of the wanted and unwanted components by frequency filtering. In some cases however, it is more convenient to effect the modulation by switching means, e.g. by applying a square waveform signal to a magneto-optic or electro-optic modulator.

Consider the case of current measurement using a magneto-optic effect.

Suppose that the transmitted intensity varies with time and is given by I(t). Suppose that the photodiode sensitivity is also a function of time, p(t). The detector output D after the imposed polarisation modulation, transducer action, polarisation analysis and photodetection, will take the form, from the equation: D=2p(t) I(t) {1-2(##s(t)+##m(t))} where a65(t) represents the signal modulation and aRm(t) the imposed modulation.

D consists of three terms, Viz: D=p(t) I(tSp(t) I(t) ##s(t)-p(t) I(t) ##m(t) Suppose now that the time functions extend over the following angular frequency ranges: p(t) D.C. to w,, I(t) D.C. to w a5(t) D.C. to #s 60m(t) t)m Inspection of the three terms which comprise D show that they will now extend over the frequency ranges: (I) -2p(t) I(t): D.C. to (#p+#1) (2) p(t) I(t) ##s(t): D.C. to (Wp+Wi+a)s) (3) p(t) I(t) S4m(t): (#m-#p-#t) to (#m+#p+#1) It is evident that term (3) can be separated from the other two in the frequency domain provided that: (a'mWpa'i) > (p+O)i+a's) i.e. provided that #m > 2(#p+#1)+#s Under this condition, frequency filtering and demodulation of term (3) will yield the quantity: p(t) I(t) Division of terms (I) and (2) by this quantity will provide an output: Do=+##s(t) from which the wanted variation 05(t) may be obtained free from all intensity noise. Only polarisation noise can remain.

From this analysis, it will be seen that, in this case, to obtain the wanted polarisation modulation, the electrical signal output from the detector may be passed through a bandpass filter arranged to accept signals within the frequency band of term 3 above and the output of this filter fed to a demodulator comprising a multiplier where the output of the bandpass filter is multiplied by a signal at the modulation frequency to provide an output which, by frequency filtering, gives the aforementioned quantity p(t) I(t). This signal can then be applied to a divider, conveniently an analogue divider, to form the denominator, the magnitude of which is divided into the magnitude of the output obtained from the detector via a low-pass filter which excludes the terms dependent on the modulation frequency; in other words this low-pass filter passes terms 1 and 2 above. It will be seen that, by this technique, it becomes possible to obtain an output which is free from all intensity noise and has only polarisation noise.

As will be further explained hereinafter; in some cases, it is preferred to separate, from the photo-detector output, a signal containing the modulation frequency components, using a passband wide e-nough to include the sidebands arising from the polarisation variations to be measured so that the demodulated signal includes the required components multiplied by the unwanted intensity variations. The low-pass filter is used to obtain low frequency components representing the unwanted intensity variations without the required higher frequency signals representing the polarisation changes introduced by the transducer. Again an analogue divider can be used to obtain the required signals free from the unwanted intensity variations; in this case the output of the demodulator is divided by the output from the low pass filter.

The following is a description of two embodiments of the invention, reference being made to Figures 1 and 2 of the accompanying drawings, each of which illustrates a noise reduction system for magneto-optic and electro-optic measuring devices.

Referring to Figure 1, there is shown diagrammatically a laser light source 10 producing a light beam which is passed through a polariser 11 to give a linear polarisation. This light beam then passes through a reference modulator 12 which, as described later, gives a sinusoidal change in the angle of the plane of polarisation and thence through a magneto-optic signal modulator 13 constituting a measuring transducer which is located adjacent a current-carrying conductor 14 carrying the current to be measured. The transducer 13 is in the magnetic field produced by that current. The light beam from this transducer 13 is fed through an analyser 15 to a photo-diode 16 forming a photo-detector giving an output having an instantaneous magnitude dependent on the instantaneous plane of polarisation of the output from transducer 13. The polarisation acceptance directions of the polariser 11 and analyser 15 are set at 450 to one another.

A sine-wave oscillator 17 controls a driver 18 which energises the modulator 12 to give a sinusoidal modulation. In this case the modulator 12 may be a magnetooptic device.

At the receiver the output from the detector 16 is separated by two filters 25, 26. The filter 25 is a low-pass filter and filter 26 is a bandpass filter. In this embodiment, the filter 25 passes low frequency components including the required signal components; as explained above, the amplitude of the required components effectively is multiplied by a factor dependent on the unwanted noise. The output from the filter 26 comprising the modulation frequency signals is fed to a multiplier 27 to which is also fed, on lead 19, a signal from the driver 18. The multiplier forms, in effect, a synchronous demodulator giving an output which is fed through a lowpass filter 20 to an analogue divider 21 which divides the output of filter 25 by the output of filter 20 to eliminate the unwanted noise from the signal representing the polarisation modulation which is fed to an output 22.

Figure 2 shows diagrammatically an alternative noise reduction system. In Figure 2 the same reference numerals are used as in Figure 1 to indicate corresponding components. Figure 2 shows diagrammatically a laser light source 10 producing a light beam which is passed through a polariser 11 to give linear polarisation. This light beam is then passed through a modulator 12 to be further described later and thence through a measuring transducer 13 which is located adjacent a current-carrying conductor 14, the signal transducer being in the field produced by that current. This transducer may be, for example, a magneto-optic transducer. The light beam from this signal transducer is fed through an analyser 15 to a photo-detector, for example a photo-diode 16.

In Figure 2, the reference modulator 12 is an electro-optic modulator to switch the direction of linear polarisation of the light beam received by the measuring transducer 13 alternately between two orthogonal transverse directions. This type of modulation may be considered as identical to that produced by an infinite set of rotation modulators of suitable relative amplitude, frequency and phase. In Figure 2, a square wave generator 32 energises the driver 18 at the required modulation frequency. This driver provides the required switching signal to modulator 12. A low-pass filter 28 takes a square wave output from the driver 18 and provides a sinusoidal output at the modulation frequency for feeding to the multiplier 27.

- As in the arrangement of Figure 1, the output from the photo-diode 16 is separated by two filters 25, 26. The filter 25 is a low-pass filter and filter 26 is a bandpass filter. The frequency bands of these filters will be explained in detail later but essentially the filter 25 passes the required low frequency components derived from the unwanted intensity variations whilst excluding those components having frequencies related to the modulation frequency. The latter components including sidebands derived from the wanted signals pass through the bandpass filter 26 to the multiplier 27, in this embodiment an analogue multiplier, which multiplies the output of bandpass filter 26 by the output of the aforementioned filter 28 providing signals of the modulation frequency from the driver 18. This multiplier 27 and filter 28 form a demodulator. The resultant multiplied output in this case contains the required information combined with the unwanted intensity variations and is fed through a low-pass filter 29 as a first input to an analogue divider unit 30. The magnitude of this input is divided, in the unit 30, by the magnitude of the output from the low-pass filter 25 to give an output signal on a lead 31.

The light from the laser is linearly polarised in a direction of 45 to the field direction of the modulator 12. The analyser 15 is set with an acceptance direction either parallel or perpendicular to this field direction. The voltage on the modulator is switched between zero and the half-wave voltage at an angular frequency WA.

The signal received by the detector can be written in the form: I,(t)=cos2 (#R+#s) IL(t)+İA(t) (I) IL(t) is the intensity fluctuation function of the laser and includes any effect due to sensitivity inhomogeneities in the photodiodes. It has frequency components in the range OA eON- IA(t) is the stray light which is received by the detectors. Its highest frequency component will, in general, be WA and also, in general, IA(t) I,(t) s5 is the signal polarisation rotation angle. It has frequency components in the range O####s.

#R is the reference modulation rotation angle.

This can be expressed as fR=45 xf(e)Rt) where f (wRt)=+1 O < #Rt < 180 etc.

=-1 1800 < WAt < 3600 etc.

In terms of Fourier Series

Inserting the expression for for into equation (1) ID(t)=İL(t) { l-f(a't)Sin 20s}+ZIA(t) Taking the case where IA(t) I,(t) and Sin 205~2sS ID(t)=IL(t){sf(WAt)1 Consider the range in frequency space of these two terms: (i) wI,(t): O#####N (ii) IL(t) #s(#Rt): #R-#N-#s####R+#N+#s3#R-#N-#s###3#R+3#N+#s etc.

The magnitudes of the two terms may be found separately so long as #N < #R-#N-#s i.e. R > 2s9"+ S (2) (N.B. The bands in frequency space of the second term will not overlap so long as WR > WN+WS. This condition is automatically fulfilled by the condition above).

In order to isolate the two terms above the apparatus shown in Figure 2 is used. In this apparatus, Filter 25 is a low pass filter set to accept O####N and hence produce directly the term I,(t). Filter 26 is a band pass filter set to accept WRWNWS < a' < WR+WN+WS hence producing a signal of the form l,(t) #s sin WAt.

Filter 28 is a low pass filter set to accept O < lR and hence producing a signal sin WAt. (N.B. This unit 28 also includes a phase adjuster so that the signals from filters 26 and 28 are in phase).

Multiplier 27 is an analogue multiplier which produces the product of the signals from filters 26 and 28. Ideally it should have a bandwidth of 2WA+WN+Ws.

The output signal from multiplier 27 is of the form I,(t) sSin2Wt=+I(t)s1 1-cos 2#Rt} Filter 29 is a low pass filter accepting O####N+#s and produces a signal proportional to I,(t) #s.

Divider 30 is an analogue divider which divides the signals from filter 29 by the output from filter 25 to produce an output on lead 31 proportional to . The bandwidth of the divider 30 should be at least a'N+a'S It is found in practice that the intensity variation which it is desirable to eliminate have frequency components which lie below about 100 kHz. In power supply systems where it may be desired to measure fast transients, it may be required to have a signal bandwidth up to about 500 kHz. Thus a modulation frequency of about I MHz enables the required separation of the components to be achieved.

In the above-described examples, the analyser 15 is set to pass light polarised in a direction at 450 to the plane of polarisation of the light from the light source. If square wave modulation is employed, the plane of polarisation is switched by plus and minus 45". It will be apparent that these are the optimum angles but the systems would be operative if the analyser were at some other angle or if the modulator effected a different rotation of the plane of polarisation.

WHAT WE CLAIM IS: 1. A measuring system employing a transducer which effects changes in polarisation characteristics of a light beam passing through an optical transducer element in response to an external parameter and in which the polarisation changes are sensed at a receiver by means including a photo-detector providing an electrical output responsive to the intensity of the incident light signal after passage through a polarisation analyser, wherein means are provided for introducing polarisation modulation, detectable at the receiver, in the light beam applied to the transducer element, the modulation being at a frequency sufficiently high that a required lower frequency output from the photodetector including all the required signal components can be separated from modulation frequency signals by frequency filtering, and wherein, at the receiver, there are provided a low-pass filter to separate a first component from the photo-detector, a bandpass filter and demodulator to separate a second component, the filters having passbands such that only one of the components carries the required signal changes due to polarisation changes effected by the transducer, and means for deriving the required signal from the first and second components.

2. A measuring system as claimed in claim 1 wherein the low-pass filter is arranged to pass signals including the required signals, and wherein the output of the bandpass filter is de-modulated to provide the second component which is representative of unwanted intensity variations, and wherein the means for deriving the required signal comprises means for dividing the magnitude of the first component by that of the second component.

3. A measuring system as claimed in claim 1 wherein said bandpass filter and demodulator is arranged to separate, from the photo-detector output, a signal containing the modulation frequency components, using a passband wide enough to include the sidebands arising from the polarisation variations to be measured so

**WARNING** end of DESC field may overlap start of CLMS **.

Claims (15)

**WARNING** start of CLMS field may overlap end of DESC **. The magnitudes of the two terms may be found separately so long as #N < #R-#N-#s i.e. R > 2s9"+ S (2) (N.B. The bands in frequency space of the second term will not overlap so long as WR > WN+WS. This condition is automatically fulfilled by the condition above). In order to isolate the two terms above the apparatus shown in Figure 2 is used. In this apparatus, Filter 25 is a low pass filter set to accept O####N and hence produce directly the term I,(t). Filter 26 is a band pass filter set to accept WRWNWS < a' < WR+WN+WS hence producing a signal of the form l,(t) #s sin WAt. Filter 28 is a low pass filter set to accept O < lR and hence producing a signal sin WAt. (N.B. This unit 28 also includes a phase adjuster so that the signals from filters 26 and 28 are in phase). Multiplier 27 is an analogue multiplier which produces the product of the signals from filters 26 and 28. Ideally it should have a bandwidth of 2WA+WN+Ws. The output signal from multiplier 27 is of the form I,(t) sSin2Wt=+I(t)s1 1-cos 2#Rt} Filter 29 is a low pass filter accepting O####N+#s and produces a signal proportional to I,(t) #s. Divider 30 is an analogue divider which divides the signals from filter 29 by the output from filter 25 to produce an output on lead 31 proportional to . The bandwidth of the divider 30 should be at least a'N+a'S It is found in practice that the intensity variation which it is desirable to eliminate have frequency components which lie below about 100 kHz. In power supply systems where it may be desired to measure fast transients, it may be required to have a signal bandwidth up to about 500 kHz. Thus a modulation frequency of about I MHz enables the required separation of the components to be achieved. In the above-described examples, the analyser 15 is set to pass light polarised in a direction at 450 to the plane of polarisation of the light from the light source. If square wave modulation is employed, the plane of polarisation is switched by plus and minus 45". It will be apparent that these are the optimum angles but the systems would be operative if the analyser were at some other angle or if the modulator effected a different rotation of the plane of polarisation. WHAT WE CLAIM IS:

1. A measuring system employing a transducer which effects changes in polarisation characteristics of a light beam passing through an optical transducer element in response to an external parameter and in which the polarisation changes are sensed at a receiver by means including a photo-detector providing an electrical output responsive to the intensity of the incident light signal after passage through a polarisation analyser, wherein means are provided for introducing polarisation modulation, detectable at the receiver, in the light beam applied to the transducer element, the modulation being at a frequency sufficiently high that a required lower frequency output from the photodetector including all the required signal components can be separated from modulation frequency signals by frequency filtering, and wherein, at the receiver, there are provided a low-pass filter to separate a first component from the photo-detector, a bandpass filter and demodulator to separate a second component, the filters having passbands such that only one of the components carries the required signal changes due to polarisation changes effected by the transducer, and means for deriving the required signal from the first and second components.

2. A measuring system as claimed in claim 1 wherein the low-pass filter is arranged to pass signals including the required signals, and wherein the output of the bandpass filter is de-modulated to provide the second component which is representative of unwanted intensity variations, and wherein the means for deriving the required signal comprises means for dividing the magnitude of the first component by that of the second component.

3. A measuring system as claimed in claim 1 wherein said bandpass filter and demodulator is arranged to separate, from the photo-detector output, a signal containing the modulation frequency components, using a passband wide enough to include the sidebands arising from the polarisation variations to be measured so

that the demodulated signal includes the required components multiplied by the unwanted intensity variations and wherein the low-pass filter is arranged to provide low frequency components representing the unwanted intensity variations without the required higher frequency signals representing the polarisation changes introduced by the transducer and wherein said means for deriving the required signal comprises means for dividing the output of the bandpass filter and demodulator by the output from the low-pass filter.

4. A measuring system as claimed in either claim 2 or claim 3 wherein said means for deriving the required signal comprises an analogue divider.

5. A measuring device as claimed in any of the preceding claims wherein said demodulator for providing said second component comprises a synchronous detector with a reference input at the modulation frequency for demodulating the output of the bandpass filter.

6. A measuring device as claimed in any of claims 1 to 5 wherein said transducer is a magneto-optic device.

7. A measuring device as claimed in any of claims 1 to 5 wherein said transducer is an electro-optic device.

8. A measuring device as claimed in any of claims 1 to 5 wherein said transducer is an electro-gyration effect transducer.

9. A measuring system as claimed in any of the preceding claims wherein said means for introducing polarisation modulation comprise a modulator effect modulation of a sinusoidal form at a predetermined frequency.

10. A measuring system as claimed in any of claims 1 to 8 wherein said means for introducing polarisation modulation comprise a modulator arranged to effect square-wave modulation by polarisation switching at a predetermined frequency.

I 1. A measuring system as claimed in either claim 9 or claim 10 wherein said modulator comprises a magneto-optic device in the light path to the transducer.

12. A measuring system as claimed in either claim 9 or claim 10 wherein said modulator comprises an electro-optic device in the light path to the transducer.

13. A measuring system as claimed in either claim 11 or 12 and having a light source providing plane polarised light polarised in a predetermined plane and wherein the receiver comprises an analyser through which the received light is passed to said photo-detector, said analyser being set to pass light polarised at 450 to said predetermined plane.

14. A measuring system as claimed in claim 13, as appendant to claim 10, wherein said modulator is arranged to effect square-wave modulation switching the plane of polarisation by plus and minus 45 .

15. A measuring system substantially as hereinbefore described with reference to Figure 1 or Figure 2 of the accompanying drawings.