Fibre-optic measurement device, rate gyro, and inertial stabilisation and navigation unit

Abstract

A fiber-optic measurement device (10) includes a SAGNAC ring interferometer (20) having a proper frequency f.sub.p, a detector (14) and a modulation chain (30) generating a phase-shift modulation φ.sub.m(t) between the two counter-propagating waves (24, 25) propagating in the ring interferometer. The device aims to reduce measurement faults due to the linearity defects in the modulation chain of such a measurement device with optical fiber. For this reason, the fiber-optic measurement device reduces the amplitude of the phase-shift modulation φ.sub.m(t) which is the sum of a first biasing phase-shift modulation component φ.sub.b1(t) and a first counter-reaction phase-shift modulation component φ.sub.cr1(t), the phase-shift modulation φ.sub.m(t) falling or rising by twice the amplitude of the first biasing phase-shift modulation component φ.sub.b1(t). A rate gyro including such a measurement device and an inertial stabilization or navigation unit including at least one such rate gyro are also described.

Claims

1. A fibre-optic measurement device (10) in which a parameter to be measured generates a phase difference Δφ.sub.p between two counter-propagating waves (24, 25), comprising: a light source (11), a fiber-optic SAGNAC ring interferometer (20), including a coil (21) and a splitting element (23), in which said two counter-propagating waves (24, 25) propagate, said ring interferometer (20) having a proper frequency f.sub.p, an electromagnetic radiation detector (14), receiving the luminous power exiting from said ring interferometer (20) and delivering a modulated electrical signal representative of the luminous power, which is function of the total phase difference Δφ.sub.t between said two counter-propagating waves (24, 25) at the output of said ring interferometer (20), a modulation chain (30) adapted to modulate said luminous power exiting from said ring interferometer (20), said modulation chain (30) including at least one phase modulator (33) placed in said ring interferometer (20) and adapted to generate at the output of said modulation chain (30) a phase-shift modulation φ.sub.m(t), introducing between said two counter-propagating waves a phase-difference modulation Δφ.sub.m(t) such that: Δφ.sub.m(t)=φ.sub.m(t)−φ.sub.m(t−Δτ.sub.g), Δτ.sub.g=1/(2 f.sub.p) being the transit time difference between said two counter-propagating waves (24, 25) determined between said phase modulator (33) and said splitting element (23), and an electronic module (100) comprising: i) signal processing means (110) including: an analog/digital converter (111) digitizing said modulated electrical signal received from the detector (14) and representative of said luminous power received by said detector (14) to deliver a digital electrical signal, and a digital processing unit (112) adapted to process said digital electrical signal to deliver a signal function of said phase difference Δφ.sub.p and of said parameter to be measured, ii) a biasing module (130) providing a first biasing signal, producing at the output of the modulation chain (30), a first, square pulse-wave, biasing phase-shift modulation component φ.sub.b1(t) of amplitude π/a.sub.1,a.sub.1 being a non-zero real number, periodic at a first biasing modulation frequency f.sub.b1 such that f.sub.b1=(2k.sub.1+1)f.sub.p, k.sub.1 being a natural number and f.sub.p being the proper frequency, iii) feedback means (120) adapted to process said signal function of said phase difference Δφ.sub.p to generate a first feedback signal, producing at the output of the modulation chain (30), a first, stair-step, feedback phase-shift modulation component φ.sub.cr1(t), each step having a duration Δτ.sub.g/(2k.sub.1+1), said first feedback phase-shift modulation component φ.sub.cr1(t) introducing between said two counter-propagation waves (24, 25) a first feedback phase-difference modulation component Δφ.sub.cr1(t)=φ.sub.cr1(t)−φ.sub.cr1(t−Δτ.sub.g) that is function of said phase difference Δφ.sub.p, and iv) a control module (140) for controlling said modulation chain (30), the control module (140) receiving said first biasing signal from the biasing module (130) and said first feedback signal from said feedback means (120), the control module (140) being adapted to process said first biasing signal and said first feedback signal to deliver at least one first control signal at the input of said modulation chain (30), the at least one first control signal producing at the output of the modulation chain (30) a first phase-shift modulation component φ.sub.m1(t) that is the phase sum of said first biasing phase-shift modulation component φ.sub.b1(t) and of said first feedback phase-shift modulation component φ.sub.cr1(t), such that φ.sub.m1(t)=φ.sub.b1(t)+φ.sub.cr1(t), wherein the control module (140) is arranged so that said first phase-shift modulation component φ.sub.m1(t) operates a transition of twice the amplitude of the first biasing phase-shift modulation component φ.sub.b1(t) when the level of said first phase-shift modulation component φ.sub.m1(t) exceeds the amplitude of the first biasing phase-shift modulation component φ.sub.b1(t).

2. The fibre-optic measurement device (10) according to claim 1, wherein said first feedback phase-shift modulation component φ.sub.cr1(t) has stair steps of height −Δφ.sub.p/(2k.sub.1+1), such that said first feedback phase-difference modulation component Δφ.sub.cr1(t) is such that Δφ.sub.cr1(t)=−Δφ.sub.p, to compensate for said phase difference Δφ.sub.p due to the parameter to be measured.

3. The fibre-optic measurement device (10) according to claim 2, wherein said biasing module (130) is adapted to generate a second biasing signal, producing at the output of the modulation chain (30) a second component of biasing phase-shift modulation φ.sub.b2(t), said second biasing phase-shift modulation component φ.sub.b2(t) being: a square pulse-wave modulation of amplitude π/a.sub.2,a.sub.2 being a non-zero real number different from a.sub.1, periodic at a second biasing modulation frequency f.sub.b2 such that f.sub.b2=(2k.sub.2+1)f.sub.p, k.sub.2 being a natural number such that (2k+1) and (2k.sub.2+1) are multiples of each other, and f.sub.p being the proper frequency, in quadrature relative to the first biasing phase-shift modulation component φ.sub.b1(t).

4. The fibre-optic measurement device (10) according to claim 3, wherein a.sub.1=1.

5. The fibre-optic measurement device (10) according to claim 4, also comprising a gain-control module (150) that controls the gain of said modulation chain (30) allowing to keep adjusted the transfer function of said modulation chain (30).

6. The fibre-optic measurement device (10) according to claim 3, wherein a.sub.2=1.

7. The fibre-optic measurement device (10) according to claim 6, also comprising a gain-control module (150) that controls the gain of said modulation chain (30) allowing to keep adjusted the transfer function of said modulation chain (30).

8. The fibre-optic measurement device (10) according to claim 3, also comprising a gain-control module (150) that controls the gain of said modulation chain (30) to keep adjusted the transfer function of said modulation chain (30).

9. The fibre-optic measurement device (10) according to claim 3, wherein k.sub.2=0.

10. The fibre-optic measurement device (10) according to claim 1, wherein, a.sub.1=1, said first feedback phase-shift modulation component φ.sub.cr1(t) has stair steps of height [a.sub.2/(a.sub.2−1)][−Δφ.sub.p/(2k.sub.1+1)], a.sub.2 being a real number strictly higher than a.sub.1=1, said biasing module (130) is adapted to generate a second biasing signal producing at the output of the modulation chain (30) a second biasing phase-shift modulation component φ.sub.b2(t), said second biasing phase-shift modulation component φ.sub.b2(t) being: a square pulse-wave modulation of amplitude π/a.sub.2, periodic at a second biasing modulation frequency f.sub.b2 such that f.sub.b2=f.sub.b1=(2k.sub.1+1) f.sub.p, f.sub.b1 being the first biasing modulation frequency and f.sub.p being the proper frequency, and in lagging quadrature relative to the first biasing phase-shift modulation component φ.sub.b1(t), said feedback means (120) are adapted to generate a second feedback signal, producing at the output of the modulation chain a second feedback phase-shift modulation component φ.sub.cr2(t), said second feedback phase-shift modulation component φ.sub.cr2(t) being: a stair-step modulation, each step having a duration Δτ.sub.g/(2k.sub.1+1), and a height [1/(a.sub.2−1)][−Δφ.sub.p/(2k.sub.1+1)], in lagging quadrature relative to the first feedback phase-shift modulation component φ.sub.cr1(t), and said second feedback phase-shift modulation component φ.sub.cr2(t) introducing a second feedback phase-difference modulation component Δφ.sub.cr2(t)=φ.sub.cr2(t)−φ.sub.cr2(t−Δτ.sub.g) between said two counter-propagating waves (24, 25), such that the difference between the first feedback phase-difference modulation component Δφ.sub.cr1(t) and the second feedback phase-difference modulation component Δφ.sub.cr2(t) compensates for the phase difference Δφ.sub.p said control module (140) is adapted to process said second biasing signal and said second feedback signal to deliver at least one second control signal at the input of said modulation chain (30), producing at the output of the modulation chain (30) a second phase-shift modulation component φ.sub.m2(t) that is the sum of said second biasing phase-shift modulation component φ.sub.b2(t) and said second feedback phase-shift modulation component Δφ.sub.cr2(t), so that φ.sub.m2(t)=φ.sub.b2(t)+φ.sub.cr2(t), and the control module (140) is arranged so that said second phase-shift modulation component φ.sub.m2(t) operates a transition of twice the amplitude of the second biasing phase-shift modulation component φ.sub.b2(t) when its level exceeds the amplitude of the second biasing phase-shift modulation component φ.sub.b2(t), the phase-shift modulation φ.sub.m(t) being equal to the difference between the first phase-shift modulation component φ.sub.m1(t) and the second phase-shift modulation component φ.sub.m2(t), so that φ.sub.m(t)=φ.sub.m1(t)−φ.sub.m2(t).

11. The fibre-optic measurement device (10) according to claim 1, wherein k.sub.1=0.

12. The fiber-optical measurement device (10) according to claim 1, wherein said fiber-optic measurement device is a gyrometer, the parameter to be measured being a component of the rotational speed of the ring interferometer (20).

Description

BRIEF DESCRIPTION OF THE DRAWING FIGURES

(1) Embodiments of the invention will be described in detail with reference to the drawings in which:

(2) FIG. 1 shows an ascending stair-step feedback phase-shift modulation φ.sub.cr0 according to the prior art;

(3) FIG. 2 shows a pulse-wave biasing phase-shift modulation φ.sub.b0 according to the prior art;

(4) FIG. 3 shows a phase-shift modulation φ.sub.m0 according to the prior art, which is the sum of the feedback phase-shift modulation φ.sub.cr0 of FIG. 1 and of the biasing phase-shift modulation φ.sub.b0 of FIG. 2;

(5) FIG. 4 shows an ascending stair-step feedback phase-shift modulation φ.sub.cr0 according to the prior art, which falls down by 2π at the instant t=t.sub.R;

(6) FIG. 5 shows a pulse-wave biasing phase-shift modulation φ.sub.b0 according to the prior art, which passes from a high level to a low level at the instant t=t.sub.R;

(7) FIG. 6 shows a phase-shift modulation φ.sub.m0 according to the prior-art, which is the sum of the feedback phase-shift modulation φ.sub.cr0 of FIG. 4 and of the biasing phase-shift modulation φ.sub.b0 of FIG. 5;

(8) FIG. 7 shows a pulse-wave biasing phase-shift modulation φ.sub.b0 according to the prior art, which passes from a low level to a high level at the instant t=t.sub.R;

(9) FIG. 8 shows a phase-shift modulation φ.sub.m0 according to the prior art, which is the sum of the feedback phase-shift modulation φ.sub.cr0 of FIG. 4 and of the biasing phase-shift modulation φ.sub.b0 of FIG. 7;

(10) FIG. 9 shows a so-called “4-state” biasing phase-shift modulation φ.sub.b0 according to the prior art, which passes from a low level to a high level at the instant t=t.sub.R;

(11) FIG. 10 shows a phase-shift modulation φ.sub.m0 according to the prior art, which is the sum of the feedback phase-shift modulation φ.sub.cr0 of FIG. 4 and of the biasing phase-shift modulation φ.sub.b0 of FIG. 9;

(12) FIG. 11 shows a so-called “4-state” biasing phase-shift modulation φ.sub.b0 according to the prior art, which passes from a high level to a low level at the instant t=t.sub.R;

(13) FIG. 12 shows a phase-shift modulation φ.sub.m0 according to the prior art, which is the sum of the feedback phase-shift modulation φ.sub.cr0 of FIG. 4 and of the biasing phase-shift modulation φ.sub.b0 of FIG. 11;

(14) FIG. 13 shows a schematic view of the measurement device according to the prior art;

(15) FIG. 14 shows a functional diagram representing the different means implemented in the measurement device according to the invention;

(16) FIG. 15 shows a first biasing phase-shift modulation component φ.sub.b1(t) as a function of time t;

(17) FIG. 16 shows a first feedback phase-shift modulation component φ.sub.cr1(t) as a function of time t, in the form of an ascending stair-step ramp;

(18) FIG. 17 shows a first phase-shift modulation component φ.sub.m1(t) as a function of time t, which is the sum of the first biasing phase-shift modulation component φ.sub.b1(t) of FIG. 15 and of the first feedback phase-shift modulation component φ.sub.cr1(t) of FIG. 16;

(19) FIG. 18 shows a first feedback phase-shift modulation component φ.sub.cr1(t) as a function of time t, in the form of a descending stair-step ramp;

(20) FIG. 19 shows a first phase-shift modulation component φ.sub.m1(t) as a function of time t, which is the sum of the first biasing phase-shift modulation component φ.sub.b1(t) of FIG. 15 and of the first feedback phase-shift modulation component φ.sub.cr1(t) of FIG. 18;

(21) FIG. 20 shows the first phase-difference modulation component Δφ.sub.m1(t) corresponding to the first phase-shift modulation component φ.sub.m1(t) of FIG. 17;

(22) FIG. 21 shows the first phase-difference modulation component Δφ.sub.m1(t) corresponding to the first phase-shift modulation component φ.sub.m1(t) of FIG. 19;

(23) FIG. 22 shows the total phase difference Δφ.sub.t(t) in a first embodiment of the invention, the luminous power received by the detector at the output of the interferometer and the corresponding modulated electrical signal;

(24) FIG. 23 shows a functional diagram showing the different means implemented in the measurement device according to the invention and comprising means for controlling the transfer function of the modulation chain;

(25) FIG. 24 shows a first phase-shift modulation component φ.sub.m1(t) in a second embodiment of the invention;

(26) FIG. 25 shows a second biasing phase-shift modulation component φ.sub.b2(t) of levels +π/8 and −π/8, in a second embodiment of the invention;

(27) FIG. 26 shows the phase-shift modulation φ.sub.m(t) in the second embodiment, resulting from the summing of the first phase-shift modulation component φ.sub.m1(t) of FIG. 24 and of the second biasing phase-shift modulation component φ.sub.b2(t) of FIG. 25;

(28) FIG. 27 shows the phase-difference modulation Δφ.sub.m(t) corresponding to the phase-shift modulation φ.sub.m(t) of FIG. 26;

(29) FIG. 28 shows the total phase difference Δφ.sub.t(t) in the second embodiment of the invention, the luminous power received by the detector at the output of the interferometer and the corresponding modulated electrical signal when the transfer function of the modulation chain is correctly adjusted;

(30) FIG. 29 shows the total phase difference Δφ.sub.t(t) in the second embodiment of the invention, the luminous power received by the detector at the output of the interferometer and the corresponding modulated electrical signal when the first feedback phase-difference modulation component Δφ.sub.cr1(t) does not compensate exactly for the phase difference Δφ.sub.p due to the parameter to be measured and when the transfer function of the modulation chain is correctly adjusted;

(31) FIG. 30 shows the total phase difference Δφ.sub.t(t) in the second embodiment of the invention, the luminous power received by the detector at the output of the interferometer and the corresponding modulated electrical signal when the transfer function of the modulation chain is incorrectly adjusted;

(32) FIG. 31 shows a first phase-shift modulation component φ.sub.m1(t) in a third embodiment of the invention;

(33) FIG. 32 shows a second biasing phase-shift modulation component φ.sub.b2(t) of levels +π/2 and −π/2, in the third embodiment of the invention;

(34) FIG. 33 shows the phase-shift modulation φ.sub.m(t) in the third embodiment, resulting from the summing of the first phase-shift modulation component φ.sub.m1(t) of FIG. 31 and of the second biasing phase-shift modulation component φ.sub.b2(t) of FIG. 32;

(35) FIG. 34 shows the phase-difference modulation Δφ.sub.m(t) corresponding to the phase-shift modulation φ.sub.m(t) of FIG. 33;

(36) FIG. 35 shows the total phase difference Δφ.sub.t(t) in the third embodiment of the invention, the luminous power received by the detector at the output of the interferometer and the corresponding modulated electrical signal when the transfer function of the modulation chain is correctly adjusted;

(37) FIG. 36 shows a first phase-shift modulation component φ.sub.m1(t) in a fourth embodiment of the invention;

(38) FIG. 37 shows a second phase-shift modulation component φ.sub.m2(t) in a fourth embodiment of the invention;

(39) FIG. 38 shows the phase-shift modulation φ.sub.m(t) in the fourth embodiment, resulting from the difference between the first phase-shift modulation component φ.sub.m1(t) of FIG. 36 and the second phase-shift modulation component φ.sub.m2(t) of FIG. 37;

(40) FIG. 39 shows the phase-difference modulation Δφ.sub.m(t) corresponding to the phase-shift modulation φ.sub.m(t) of FIG. 38;

(41) FIG. 40 shows the total phase difference Δφ.sub.t(t) in the fourth embodiment of the invention, the luminous power received by the detector at the output of the interferometer and the corresponding modulated electrical signal when the transfer function of the modulation chain is correctly adjusted.

DETAILED DESCRIPTION OF THE INVENTION

(42) FIG. 13 shows a fibre-optic measurement device 10 according to the prior art, of the type in which a parameter to be measured generates a phase difference Δφ.sub.p between two waves.

(43) The fibre-optic measurement device 10 first includes a light source 11 herein comprising a laser diode.

(44) As a variant, the light source may comprise for example a super-luminescent diode or a doped-fibre light source of the ASE (“Amplified Spontaneous Emission”) type.

(45) The device 10 also comprises a first splitting element 12. This first splitting element 12 is herein a semi-reflective plate having a transmittance of 50% and a reflectance of 50%.

(46) As a variant, the splitting element may be, for example, a −3-decibel 2×2 coupler or an optical circulator.

(47) The luminous wave emitted by the light source 11 is hence transmitted in part by the first splitting element 12 towards an optical filter 13 at the output of which the luminous wave has been filtered. The optical filter 13 preferably includes a polarizer and a spatial filter. This spatial filter is herein a single-mode optical fiber, preferably of the polarization-maintaining type.

(48) The device 10 also includes a SAGNAC ring interferometer 20 comprising a fiber-optic coil 21 wound around itself. It is herein an optical fiber, preferably of the single-mode and polarization-maintaining type.

(49) This SAGNAC ring interferometer 20 also comprises a second splitting element 23 allowing to split the wave exiting from the optical filter 13 into two counter-propagating waves 24, 25 on the two arms of the ring interferometer 20, these two arms defining two optical paths 24A and 25A. The second splitting element 23 is herein a semi-reflective plate having a transmittance of 50% and a reflectance of 50%.

(50) The second splitting element 23 also allows to recombine the two counter-propagating waves 24, 25 at the output of the ring interferometer 20.

(51) As a variant, the second splitting element may be, for example, a −3-decibel 2×2 coupler or a “Y”-junction in integrated optics.

(52) The two counter-propagating waves 24, 25 then pass through the optical filter 13 and are reflected by the first splitting element 12 towards an electromagnetic radiation detector 14.

(53) This detector 14 is a semi-conductor photodiode.

(54) The detector 14 is sensitive to the luminous power P received, which is herein function of the interference state between the two counter-propagating waves 24, 25 during their recombination at the output of the SAGNAC ring interferometer 20. It hence delivers an electrical signal that is representative of the total phase difference Δφ.sub.t between the two counter-propagating waves 24, 25.

(55) It is known that, for a SAGNAC ring interferometer 20, the luminous power P(Δφ.sub.t) received by the detector as a function of the total phase difference Δφ.sub.t is a cosine function of this total phase difference Δφ.sub.t, i.e. the following relation is satisfied: P(Δφ.sub.t)=P0[1+cos(Δφ.sub.t)].

(56) It will be seen in the following of the description that this electrical signal is a modulated electrical signal.

(57) The device 10 also includes a modulation chain 30 comprising a digital/analog converter 31, an amplifier 32 and a phase modulator 33.

(58) The digital/analog converter 31 processes a digital control signal delivered by the electronic means 100, the decomposition of this signal being described in detail hereinafter. The digital/analog converter 31 delivers as an output an analog control signal.

(59) The amplifier 32 then processes this analog control signal to deliver a modulation control voltage V.sub.m(t) to the phase modulator 33.

(60) The phase modulator 33 is placed in the ring interferometer 20 and is thus also a part thereof. It herein advantageously comprises a pair of elementary phase modulators 33A, 33B mounted in “push-pull” configuration, placed respectively at each of the ends of the arms of the SAGNAC ring interferometer 20.

(61) It is known that the “push-pull” mounting allows to eliminate the response of the phase modulator 33 and the non-linearities of even order (2.sup.nd order, 4.sup.th order, etc. . . . ). Hence, the modulation chain 30 has an odd non-linear response with for main components the component of 1.sup.st order (linear component) and the component of 3.sup.rd order.

(62) The elementary phase modulator 33A, 33B are herein of the so-called “Pockels effect”, electro-optical type, in proton-exchange lithium-niobate integrated optics.

(63) The phase modulator 33 allows, when the time-dependant control modulation voltage V.sub.m(t) is applied at the input thereof, to generate a proportional phase-shift modulation φ.sub.m(t), and thus with the same time dependency, in a luminous wave passing through it at the given instant t in one direction or another.

(64) In the case of the SAGNAC ring interferometer 20 shown in FIG. 13, the transit-time difference of the counter-propagating waves 24, 25 along the two optical paths 24A, 25A between the phase modulator 33 and the second splitting element 23 is denoted Δτ.sub.g.

(65) Hence, the phase-shift modulation φ.sub.m(t) generated by the phase modulator 33 controlled by the modulation control voltage V.sub.m(t) introduces between the two counter-propagating waves 24, 25, a phase-difference modulation Δφ.sub.m(t) such that: Δφ.sub.m(t)=φ.sub.m(t)−φ.sub.m(t−Δτ.sub.g).

(66) The transit-time difference Δτ.sub.g also defines a proper frequency f.sub.p of the SAGNAC ring interferometer 20 by the relation: f.sub.p=1/(2Δτ.sub.g).

(67) This proper frequency f.sub.p thus depends on the length of the coil 21 in the SAGNAC ring interferometer 20. With the fiber-optic coil 21 used herein, a coil having a length of 1 kilometer, the proper frequency f.sub.p of the SAGNAC ring interferometer 20 is of about 100 kilohertz (kHz), corresponding to a transit-time difference Δτ.sub.g of 5 microseconds (μs).

(68) The luminous power P(Δφ.sub.t) received by the detector 14 is also modulated and the electrical signal delivered by the detector 14 will thus be a modulated electrical signal, examples of which will be given hereinafter.

(69) This modulated electrical signal is transmitted to electronic means 100 that process it to deliver a signal function of the phase difference Δφ.sub.p and of the parameter to be measured.

(70) For that purpose, the electronic means 100 comprise signal processing means 110, as shown in FIG. 14. These signal processing means 110 include an analog/digital converter 111 digitizing the modulated electrical signal provided by the detector 14 to deliver a digital electrical signal.

(71) This digitization operation is performed at a synchronization frequency fixed by the clock 101. The synchronization frequency of the clock 101 is preferably a multiple of the proper frequency f.sub.p of the SAGNAC ring interferometer 20.

(72) The signal processing means 110 also comprise a digital processing unit 112 configured to process the digital electrical signal provided at the output of the analog/digital converter 111. The digital processing unit 112 also includes a digital demodulator, a control-loop digital filter fed with a first demodulated digital signal exiting from the digital demodulator and a register.

(73) The digital processing unit 112 delivers a signal function of the phase difference Δφ.sub.p and of the parameter to be measured for any desired external use.

(74) The electronic means 100 also control in return the modulation chain 30.

(75) For that purpose, the electronic means 100 include, on the one hand, biasing means 130 and, on the other hand, feedback means 120.

(76) On one side, the biasing means 130 generate a first biasing signal producing at the output of the modulation chain a first biasing phase-shift modulation component φ.sub.b1(t) as shown in FIG. 15.

(77) This first biasing phase-shift modulation component φ.sub.b1(t) is a pulse-wave modulation having herein a high level of value π/2a.sub.1 (a.sub.1 being a non-zero real number) and a low level of value −π/2a.sub.1.

(78) This modulation is hence: square: the duration of the high level is herein equal to the duration of the low level, and of amplitude π/a.sub.1, the amplitude being defined as the distance between the high level (π/2a.sub.1) and the low level (−π/2a.sub.1) of the modulation, i.e. π/2a.sub.1−(−π/2a.sub.1)=2(π/2a.sub.1)=π/a.sub.1.

(79) As illustrated in FIG. 15, it is advantageous that the first phase-shift modulation component φ.sub.m1(t) is centred about zero. Indeed, this allows to reduce the effects of the non-linearities of the modulation chain 30, in particular thanks to the “push-pull” mounting of the phase modulator 33.

(80) Nevertheless, according to the invention, an equivalent result would be obtained with a non-centred modulation.

(81) Furthermore, the first biasing phase-shift modulation component φ.sub.b1(t) is a periodic modulation at a first biasing modulation frequency f.sub.b1, which is herein such that f.sub.b1=f.sub.p, f.sub.p being the proper frequency of the SAGNAC ring interferometer 20. The period of the first biasing phase-shift modulation component φ.sub.b1(t) is thus equal to 2Δτ.sub.g and the duration of each of the high and low levels is equal to Δτ.sub.g.

(82) Generally, the first biasing modulation frequency f.sub.b1 may be such that f.sub.b1=(2k.sub.1+1)f.sub.p, k.sub.1 being a natural number and f.sub.p being the proper frequency.

(83) The biasing means 130 are configured so as to generate a first biasing signal at precise instants, synchronized by the frequency of the clock 101.

(84) The biasing means 130 allows to displace the operating point of the measurement device 10 so as to use it with the best possible sensitivity.

(85) On the other side, the feedback means 120 process the signal function of the phase difference Δφ.sub.p to generate a first feedback signal, producing at the output of the modulation chain a first feedback phase-shift modulation component φ.sub.cr1(t) as shown in FIG. 16.

(86) This first feedback phase-shift modulation component φ.sub.cr1(t) is a stair-step modulation, each stair step having herein a duration Δτ.sub.g.

(87) Generally, and when the first biasing modulation frequency f.sub.b1 is such that f.sub.b1=(2k.sub.1+1)f.sub.p, the first feedback phase-shift modulation component φ.sub.cr1(t) has stair steps of duration Δτ.sub.g/(2k.sub.1+1).

(88) The first feedback phase-shift modulation component φ.sub.cr1(t) is in phase with the first biasing phase-shift modulation component φ.sub.b1(t), i.e. the first feedback phase-shift modulation component φ.sub.cr1(t) passes from one step to another upon the passage of the first biasing phase-shift modulation component φ.sub.b1(t) from one level to another.

(89) Besides, as illustrated in FIG. 16, the first feedback phase-shift modulation component φ.sub.cr1(t) has stair steps of height φ.sub.s function of said phase difference Δφ.sub.p. The level of each step being herein higher than the previous one, it is hence referred to an ascending ramp for the first feedback phase-shift modulation component φ.sub.cr1(t).

(90) FIG. 18 shows an example of the first feedback phase-shift modulation component φ.sub.cr1(t), which is a descending ramp, where the level of each step is lower than the previous one. It will be seen in the following of the description in which case the ramp is ascending and in which case it is descending.

(91) The first feedback phase-shift modulation component φ.sub.cr1(t) introduces, between the two counter-propagation waves 24, 25, a first feedback phase-difference modulation component Δφ.sub.cr1(t)=φ.sub.cr1(t)−φ.sub.cr1(t−Δτ.sub.g)=φ.sub.s that is a function of the phase difference Δφ.sub.p.

(92) The feedback means 120 herein include an accumulator.

(93) The electronic means 100 further include control means 140 for controlling the modulation chain 30, which process the first biasing signal and the first feedback signal to deliver at least one first control signal at the input of the modulation chain 30.

(94) This first control signal produces at the output of the modulation chain 30 a first phase-shift modulation component φ.sub.m1(t), which is the phase sum of the first biasing phase-shift modulation component φ.sub.b1(t) and the first feedback phase-shift modulation component φ.sub.cr1(t), such that φ.sub.m1(t) satisfies the relation: φ.sub.m1(t)=φ.sub.b1(t)+φ.sub.cr1(t).

(95) The control means 140 have two inputs and one output. At the input, the control means 140 receive on the one hand the first feedback signal and on the other hand the first biasing signal. These signals are then processed by the control means 140. At the output, the control means 140 deliver the first control signal, which is then transmitted to the modulation chain 30.

(96) Hence controlled, the modulation chain 30 generates a first phase-shift modulation component φ.sub.m1(t) via the phase modulator 33. A first phase-difference modulation Δφ.sub.m1(t) is then introduced between the two counter-propagating waves 24, 25 propagating in the SAGNAC ring interferometer 20.

(97) The different particular embodiments of the invention allowing to limit the effects of the non-linearities of the modulation chain 30 on the measurement of the phase difference Δφ.sub.p and of the parameter to be measured will now be described.

1st Embodiment

(98) In this first particular embodiment of the invention, the first biasing modulation component φ.sub.b1(t) 1500 is that shown in FIG. 15, and the first feedback modulation component φ.sub.cr1(t) 1600 is that shown in FIG. 16.

(99) According to the invention, the control means 140 for controlling the modulation chain 30 are arranged in such a manner that the first phase-shift modulation component φ.sub.m1(t) operates a transition of twice the amplitude of the first biasing phase-shift modulation component φ.sub.b1(t), i.e. 2π/a.sub.1, when its level exceeds the amplitude of the first biasing phase-shift modulation component φ.sub.b1(t), i.e. π/a.sub.1.

(100) This may be understood at the light of FIG. 17, on which has been shown the first phase-shift modulation component φ.sub.m1(t) 1700, which as the form of an ascending pulse-wave modulation, due to the ascending ramp of FIG. 16.

(101) It can be observed in FIG. 17 that, at the instant t=t.sub.R, the first phase-shift modulation component φ.sub.m1(t) had to pass from a low level before falling down 1701 to a high level 1702 shown in dash line.

(102) However, as this high level 1702 is such that it slightly exceeds the value π/a.sub.1, i.e. the value of the amplitude of the first biasing phase-shift modulation component φ.sub.b1(t) 1500, the control means 140 make the phase-shift modulation component φ.sub.m1(t) 1700 fall down to bring the high level 1702 to the low level after falling down 1703. The amplitude 1704 of this falling down is hence herein equal to 2π/a.sub.1, i.e. twice the amplitude of the first biasing phase-shift modulation component φ.sub.b1(t) 1500.

(103) So formed, the first phase-shift modulation component φ.sub.m1(t) 1700 has a maximal amplitude 1705 equal to 2π/a.sub.1. The amplitude is hence lowered and the excursion on the modulation chain 30 is reduced, limiting the effects of the non-linearities of this modulation chain 30.

(104) According to an advantageous characteristic of this first embodiment of the invention, the first feedback phase-shift modulation component φ.sub.cr1(t) 1600 of FIG. 16 has stair steps of height φ.sub.s such that: φ.sub.s=−Δφ.sub.p, in such a manner that the feedback phase-difference modulation Δφ.sub.cr1(t) is equal to −Δφ.sub.p, to compensate for the phase difference Δφ.sub.p due to the parameter to be measured.

(105) This allows in particular to make the device 10 operate in closed loop so as to reach a good linearity and stability of the measurement of the parameter generating the phase difference Δφ.sub.p.

(106) In the case where the phase difference Δφ.sub.p due to the parameter to be measured is positive, the height φ.sub.s of the stair steps of the first feedback phase-shift modulation component φ.sub.cr1(t) 1800 is negative, which means that this first component is a descending ramp, as shown in FIG. 18.

(107) The first phase-shift modulation component φ.sub.m1(t) 1900 then takes the form of a descending pulse-wave modulation (see FIG. 19).

(108) It can be observed in FIG. 19 that, at the instant t=t.sub.R, the first phase-shift modulation component φ.sub.m1(t) 1900 had to pass from a high level before falling down 1901 to a low level 1902 shown in dash line.

(109) However, as this low level 1902 is such that it slightly exceeds, in absolute value, the value π/a.sub.1, i.e. the value of the amplitude of the first biasing phase-shift modulation component φ.sub.b1(t) 1500, the control means 140 make the first phase-shift modulation component φ.sub.m1(t) 1900 rise up, to bring the low level 1902 to the high level after rising up 1903. The amplitude 1904 of this rising up is hence equal to 2π/a.sub.1, i.e. twice the amplitude of the first biasing phase-shift modulation component φ.sub.b1(t) 1500.

(110) So formed, the first phase-shift modulation component φ.sub.m1(t) still has a maximum amplitude 1905 equal to 2π/a.sub.1.

(111) By comparing the cases of FIGS. 17 and 19, corresponding, respectively, to the case of an ascending ramp and to the case of a descending ramp, it is understood that the first phase-shift modulation component φ.sub.m1(t) is caused to fall down or to rise up when the value of the level it should have reached without falling down (or rising up) is higher, in absolute value, than twice the amplitude of the first biasing phase-shift modulation component φ.sub.b1(t).

(112) Moreover, it will be noted that, in this first embodiment of the invention, the falling down (or the rising up) of the first phase-shift modulation component φ.sub.m1(t) is such that the latter has a maximum amplitude, between its highest level and its lowest level, lower than or equal to twice the amplitude of the first biasing phase-shift modulation component φ.sub.b1(t), i.e. 2π/a.sub.1.

(113) Hence, the amplitude of the phase-shift modulation is reduced with respect to the case in which a falling down (or rising up) by 2π is made on the first feedback phase-shift modulation component φ.sub.cr1(t), independently of its sum with the first biasing phase-shift modulation component φ.sub.b1(t). Indeed, in this case, the maximal amplitude would be of about 2π+π/a.sub.1.

(114) It has be shown in FIGS. 20 and 21, respectively, the first phase-difference modulation components Δφ.sub.m1(t) introduced between the counter-propagating waves 24, 25 and resulting from the generation of the first phase-shift modulation components φ.sub.m1(t) of FIGS. 17 and 19, respectively, by the modulation chain 30 controlled by the control means 140.

(115) In the case of an ascending ramp (i.e. φ.sub.s>0, case of FIG. 20), the first phase-difference modulation component Δφ.sub.m1(t) 2000 is a pulse-wave modulation, having a high level of value +π/a.sub.1+φ.sub.s and a low level of value −π/a.sub.1+φ.sub.s, these extreme levels being hence symmetrical with respect to the level line of ordinate φ.sub.s.

(116) The first phase-difference modulation component Δφ.sub.m1(t) 2000 has a total amplitude equal to 2π/a.sub.1.

(117) Moreover, it can be observed in FIG. 20 that the first phase-difference modulation component Δφ.sub.m1(t) 2000 is alternately at its high level and at its low level, until the instant t=t.sub.R of the falling down of the first phase-shift modulation component φ.sub.m1(t) 1700 (see FIG. 17), where it remains at the same low level 2001, instead of getting to the high level 2002. Next, the first phase-difference modulation component Δφ.sub.m1(t) 2000 resumes its alternations between its high level and its low level, and this until the next falling down.

(118) Likewise, in the case of a descending ramp (i.e. φ.sub.s<0, case of FIG. 21), the first phase-difference modulation component Δφ.sub.m1(t) 2100 is a pulse-wave modulation, having a high level of value +π/a.sub.1+φ.sub.s and a low level of value −π/a.sub.1+φ.sub.s, these extreme levels being thus symmetrical with respect to the level line of ordinate φ.sub.s (herein φ.sub.s<0).

(119) The first phase-difference modulation component has a total amplitude Δφ.sub.m1(t) 2100 equal to 2π/a.sub.1.

(120) Moreover, it can be observed in FIG. 21 that the phase-shift modulation component Δφ.sub.m1(t) 2100 is alternately at its high level and at its low level, until the instant t=t.sub.R of the rising up of the phase-shift modulation component φ.sub.m1(t) 1900 (see FIG. 19), where it remains at the same high level 2101, instead of getting to the low level 2102. Next, the first phase-difference modulation component Δφ.sub.m1(t) 2100 resumes its alternations between its high level and its low level, until the next rising up.

(121) FIG. 22 is an example of the first embodiment of the invention for which a.sub.1=4/3. Hence, in FIG. 22, it has been shown the total phase difference Δφ.sub.t 2200 between the two counter-propagating waves 24, 25 propagating in the SAGNAC ring interferometer 20. The total phase difference Δφ.sub.t is herein the sum of the phase difference Δφ.sub.p due to the parameter to be measured and of the first phase-difference modulation component Δφ.sub.m1: Δφ.sub.t=Δφ.sub.p+Δφ.sub.m1.

(122) For this example, the first phase-shift modulation component φ.sub.m1(t) is formed based on: a first biasing phase-shift modulation component φ.sub.b1(t) of levels +3π/8 (=π/2a.sub.1) and −3π/8 (=−π/2a.sub.1), and a first feedback phase-shift modulation component φ.sub.cr1(t) such that the step value φ.sub.s compensates for the phase difference Δφ.sub.p of the parameter to be measured, so that φ.sub.s=−Δφ.sub.p.

(123) Hence, the first phase-difference modulation component Δφ.sub.m1(t) introduced between the two counter-propagative waves 24, 25 from the generation of the first phase-shift modulation component φ.sub.m1(t) by the modulation chain 30 is similar to that shown in FIG. 22. This is a pulse-wave modulation, of symmetrical extreme levels, of values +3π/4 (high level 2201) and −3π/4 (low level 2202), its total amplitude 2203 being equal to 3π/2.

(124) As explained hereinabove for the first phase-difference modulation component Δφ.sub.m1(t), the total phase difference Δφ.sub.t is alternately at its high level and at its low level, until the instant t=t.sub.R of the falling down, where it remains at the same level 2204, instead of getting to the high level 2205 shown in dash line.

(125) It has also been shown in FIG. 22 the response of the ring interferometer 20, i.e. the luminous power received P(Δφ.sub.t) 2206 by the detector at the output of the interferometer. This received luminous power is cosine waveform: P(Δφ.sub.t)=P.sub.0[1+cos(Δφ.sub.t)], and has a maximum equal to 2P.sub.0 when the total phase difference Δφ.sub.t is null (Δφ.sub.t=0).

(126) The total phase difference Δφ.sub.t being modulated as described above, the luminous power P(Δφ.sub.t) 2206 received by the detector 14 is modulated according to two distinct modulation states: a state E1 for which Δφ.sub.t=Δφ.sub.1=3π/4, and a state E2 for which Δφ.sub.t=Δφ.sub.2=−3π/4.

(127) As can be seen in FIG. 22, the luminous power P(Δφ.sub.t) 2206 received by the detector 14 is the same in the two modulation states E1 and E2. Indeed, the received luminous power P(Δφ.sub.t) 2206 being a cosine function of the total phase difference Δφ.sub.t, the following relation is satisfied: P(Δφ.sub.1)=P(Δφ.sub.2).

(128) The detector 14 then delivers a modulated electrical signal S(t) 2207 such as shown in FIG. 22. This modulated electrical signal S(t) 2207 takes sequentially two values S1 and S2 associated with the two modulation states E1 and E2, respectively, of the modulated total phase difference Δφ.sub.t.

(129) It will also been observed in FIG. 22 that the modulated electrical signal S(t) 2207 exhibits peaks 2208 corresponding to the transitions from the modulation state E1 to the modulation state E2 (and vice versa), when the received luminous power P(Δφ.sub.t) 2206 passes by a maximum at the value Δφ.sub.t=0, this maximum having the value 2P.sub.0.

(130) These peaks 2208 are cumbersome insofar as they introduce unwanted defects in the modulated electrical signal S(t) 2207.

(131) The modulated electrical signal S(t) 2207 is then digitized by the analog/digital converter 111, which delivers and transmits a digital electrical signal to the digital processing unit 112.

(132) It is important to note herein that the falling down of the first phase-shift modulation component φ.sub.m1(t) at the instant t=t.sub.R creates no defect in the modulated electrical signal S(t) 2207 insofar as the total phase difference Δφ.sub.t does not change during this falling down. Hence, the falling down does not disturb the measurement.

2nd, 3rd and 4th Embodiments of the Invention

(133) FIGS. 23 to 40 relates to three particular embodiments of the invention in the total phase difference Δφ.sub.t between the two counter-propagating waves 24, 25 is no longer modulated according to two modulation states but according to four modulation states. It will be seen in particular how these particular embodiments allow to control the modulation chain 30.

(134) Indeed, advantageously, in these second, third and fourth embodiments of the invention, it is known that the electronic means 100 of the fibre-optic measurement device 10 include gain-control means 150 such as shown in FIG. 23. These gain-control means 150 keep adjusted the transfer function of the modulation chain 30, this transfer function electronically characterizing the modulation chain 30 between its input and its output.

(135) This transfer function characterizes the response of the modulation chain 30: it corresponds to the ratio between the value (in radians) of the phase shift effectively generated by the modulation chain 30 via the phase modulator 33 and the value (with no unit) of the digital control signal transmitted to the modulation chain 30.

(136) The gain control means 150 comprise another digital processing unit (not shown) using the digital electrical signal delivered by the analog/digital converter 111 so as to provide a signal function of the transfer function of the modulation chain 30.

(137) This later signal is filtered by a digital control-loop integrator filter which feeds another digital/analog converter controlling the variable gain G of the amplifier 32 or the reference analog voltage of the digital/analog converter 31.

(138) Hence, the transfer function of the modulation chain 30 is kept correctly adjusted, as well as the modulation control voltage delivered by the amplifier 32 to the phase modulation 33.

(139) It is meant by this that a given value of the digital control signal at the input of the modulation chain 30 will always give the same value (in radians) of phase-shift modulation φ.sub.m generated by the phase modulator 33, and hence the same value (in radians) of phase-difference modulation Δφ.sub.m introduced between the two counter-propagating waves 24, 25 in the SAGNAC ring interferometer 20.

(140) It will be understood at the light of the following description how operate the control means 150 and how they allow to keep correctly adjusted the transfer function of the modulation chain 30.

2nd Embodiment

(141) It will now be detailed, in FIGS. 24 to 30, a second embodiment of the invention, in which: a.sub.1=1 and k.sub.1=0, i.e. the first biasing phase-shift modulation component φ.sub.b1(t) is a square pulse-wave modulation of amplitude π between a high level +π/2 and a low level −π/2 and periodic at a first biasing modulation frequency f.sub.b1 equal to the proper frequency f.sub.p of the SAGNAC ring interferometer 20, and Δφ.sub.p=−π/20, so that the first feedback phase-shift modulation component φ.sub.cr1(t) is an ascending stair-step modulation of height φ.sub.s=−Δφ.sub.p=π/20 compensating for the phase difference Δφ.sub.p due to the parameter to be measured.

(142) Hence, according to the invention, the first phase-shift modulation component φ.sub.m1(t) is such as shown in FIG. 24. It has an ascending pulse-wave modulation shape, which falls down by 2π at the instant t=t.sub.R, and has a maximum amplitude lower than 2π.

(143) Furthermore, in this second embodiment of the invention, the biasing means 120 generate a second biasing signal producing at the output of the modulation chain 30 a second biasing phase-shift modulation component φ.sub.b2(t) 2500, this second biasing phase-shift modulation component φ.sub.b2(t) 2500 being (cf. FIG. 25): a square pulse-wave modulation of amplitude π/a.sub.2=π/4, with herein a.sub.2=4≠a.sub.1=1, between a high level +π/8 and a low level −π/8, periodic at a second biasing modulation frequency f.sub.b2 such that f.sub.b2=(2k.sub.2+1)f.sub.p=f.sub.p, with k.sub.2=k.sub.1=0 so that (2k.sub.1+1)=1 and (2k.sub.2+1)=1 are multiple of each other, in quadrature relative to the first biasing phase-shift modulation component φ.sub.b1(t).

(144) It will be defined herein that the first biasing phase-shift modulation component φ.sub.b1(t) and the second biasing phase-shift modulation component φ.sub.b2(t) 2500 are in quadrature when a transition of the first biasing phase-shift modulation component φ.sub.b1(t) from an extreme level to another occurs at equal distance from two successive zeros of the second biasing phase-shift modulation component φ.sub.b2(t) 2500.

(145) It has been shown in FIG. 26 the phase-shift modulation φ.sub.m(t) 2600 resulting from the sum of the first phase-shift modulation component φ.sub.m1(t) 2400 and of the second biasing phase-shift modulation component φ.sub.b2(t) 2500.

(146) It can be observed in FIG. 26 that, taking into account that the first feedback phase-shift modulation component φ.sub.cr1(t) is herein an ascending modulation, the first phase-shift modulation component φ.sub.m1(t) 2400 falls down by 2π according to the above-described rule, at the falling down instant t=t.sub.R. Likewise, the resulting phase-shift modulation φ.sub.m(t) 2600 also falls down at the same instant t=t.sub.R.

(147) As can be seen in FIG. 26, the maximum amplitude 2601 of the phase-shift modulation φ.sub.m(t) 2600 is herein equal to 2π+π/4.

(148) Hence generated by the phase modulator 33 of the modulation chain 30, the phase-shift modulation φ.sub.m(t) 2600 introduces between the two counter-propagating waves 24, 25 at the output of the SAGNAC ring interferometer 20 a phase-difference modulation Δφ.sub.m(t)=φ.sub.m(t)−φ.sub.m(t−Δπ.sub.g) as shown in FIG. 27.

(149) It can be noted that this phase-difference modulation Δφ.sub.m(t) 2700 oscillates between 4 different levels: two high levels: a first level or level 1 when Δφ.sub.m(t)=π−π/4+φ.sub.s a second level or level 2 when Δφ.sub.m(t)=π+π/4+φ.sub.s, and two low levels: a third level or level 3 when Δφ.sub.m(t)=−π+π/4+φ.sub.s a fourth level or level 4 when Δφ.sub.m(t)=−π−π/4+φ.sub.s.

(150) As can be seen in FIG. 27, the phase-difference modulation Δφ.sub.m(t) 2700 is hence shifted towards the high side of the value φ.sub.s=−Δφ.sub.p (=π/20 herein) as seen above.

(151) It can also be noted that, at the instant t=t.sub.R of the falling down of the phase-shift modulation φ.sub.m(t), the phase-difference modulation Δφ.sub.m(t) does not rise up (dash line 2701) but remains at its fourth level 2702 during a step duration, to thereafter rise up at the instant t.sub.R+Δτ.sub.g towards the high levels 1 and 2, the phase-difference modulation Δφ.sub.m(t) being symmetrical relative to the falling down instant (that is to say that Δφ.sub.m(t.sub.R+t)=Δφ.sub.m(t.sub.R−t)). The alternation between high and low levels then continues until the next falling down.

(152) To sum up, in the second particular embodiment of the invention, it has been seen on the one hand that the parameter to be measured introduces between the two counter-propagating waves a phase difference Δφ.sub.p=−π/20 at the output of the SAGNAC ring interferometer 20. On the other hand, it has been seen that the counter-propagating means 120 generate a first feedback signal to produce at the output of the modulation chain 30 a first stair-step feedback modulation component φ.sub.cr1(t) of step height φ.sub.s=π/20 so as to introduce between the two counter-propagating waves 24, 25 a first feedback phase-difference modulation component Δφ.sub.cr1(t) compensating for the phase difference Δφ.sub.p due to the parameter to be measured.

(153) Hence, it has been shown in FIG. 28: the total phase difference Δφ.sub.t(t) 2801 between the two counter-propagating waves 24, 25 at the output of the ring interferometer 20, this total phase difference Δφ.sub.t(t) being such that: Δφ.sub.t(t)=Δφ.sub.p+Δφ.sub.m(t), the luminous power P(Δφ.sub.t) 2802 received by the detector 14 as a function of the total phase difference Δφ.sub.t(t) 2801, and the modulated electrical signal S(t) 2803 delivered by the detector 14 as a function of time t.

(154) As the total phase difference Δφ.sub.t(t) is the sum of the phase difference Δφ.sub.p due to the parameter to be measured and of the phase-difference modulation Δφ.sub.m(t), it will first be noted in FIG. 28 that the curve 2801 showing the total phase difference Δφ.sub.t(t) as a function of time t corresponds to the curve 2700 of FIG. 27 showing the phase-difference modulation Δφ.sub.m(t), this curve being shifted towards the abscissa axis of the value of the phase difference Δφ.sub.p due to the parameter to be measured.

(155) Hence, the total phase-difference modulation Δφ.sub.t(t) 2801 has sequentially four different levels defining four different modulation states, which are: State E1 when the total phase-difference modulation Δφ.sub.t(t) is Δφ.sub.1=π−π/4=3π/4, State E2 when the total phase-difference modulation Δφ.sub.t(t) is Δφ.sub.2=π+π/4=5π/4, State E3 when the total phase-difference modulation Δφ.sub.t(t) is Δφ.sub.3=−π+π/4=−3π/4 (=Δφ.sub.2−2π), State E4 when the total phase-difference modulation Δφ.sub.t(t) is Δφ.sub.4=−π−π/4=−5π/4 (=Δφ.sub.1−2π).

(156) The luminous power P(Δφ.sub.t) 2802 received by the detector 14 is hence modulated following these four distinct modulation states and the modulated electrical signal S(t) 2803 delivered by the detector 14 takes sequentially four values S1, S2, S3, and S4 associated with the four modulation states E1, E2, E3, and E4, respectively, of the total phase-difference modulation Δφ.sub.t(t).

(157) As can be seen in FIG. 28, the luminous power P(Δφ.sub.t) 2802 received by the detector 14 in the four modulation states E1 to E4 is the same. Indeed, the received luminous power P(Δφ.sub.t) 2802 being a cosine function of the total phase difference Δφ.sub.t 2801, the following relation is satisfied: P(Δφ.sub.1)=P(Δφ.sub.4) and P(Δφ.sub.2)=P(Δφ.sub.3). Moreover, the states E1 and E2 (respectively the states E3 and E4) being symmetrical relative to π (respectively relative to −π), the following relations are also satisfied: P(Δφ.sub.1)=P(Δφ.sub.2) and P(Δφ.sub.3)=P(Δφ.sub.4).

(158) The detector 14 then delivers a modulated electrical signal S(t) 2803 such as shown in FIG. 28. This modulated electrical signal S(t) 2803 takes sequentially the four values S1, S2, S3, and S4 associated with the four modulation states E1, E2, E3, and E4, respectively. Taking into account what has been explained regarding the luminous power P(Δφ.sub.t) 2802 received by the detector 14 in the four modulation states E1 to E4, these four values S1, S2, S3, and S4 taken by the modulated electrical signal S(t) 2803 are herein all identical S1=S2=S3=S4.

(159) In particular, as for the first embodiment of the invention described hereinabove, it will also be noted that the falling down 2701 of the phase-difference modulation Δφ.sub.m(t) 2700 at the instant t=t.sub.R (see FIG. 27) creates no defect in the modulated electrical signal S(t) 2803 insofar as the total phase difference Δφ.sub.t 2801 does not change during this falling down (Δφ.sub.m(t) remaining at its low level 2702). Hence, the falling down does not disturb the measurement.

(160) It is nevertheless to be noted in FIG. 28 that the modulated electrical signal S(t) has peaks 2804 corresponding alternately to the transitions from the modulation state E1 to the modulation state E4 and from the modulation state E3 to the modulation state E2, when the received luminous power P(Δφ.sub.t) 2802 passes by a maximum at the value Δφ.sub.t=0, this maximum having for value 2P.sub.0.

(161) These peaks 2804 are cumbersome insofar as they introduce unwanted defects in the modulated electrical signal S(t) 2803.

(162) From the preceding situation, described in FIG. 28, where the parameter to be measured generates a phase difference Δφ.sub.p that is exactly compensated thanks to the first feedback phase-difference modulation component Δφ.sub.cr1(t) introduced between the two counter-propagating waves 24, 25 in the SAGNAC ring interferometer 20 by the first, stair-step, feedback phase-shift modulation component Δφ.sub.cr1(t), the situation passes to that described in FIG. 29, where the phase difference Δφ.sub.p is not exactly compensated by the first feedback phase-difference modulation component Δφ.sub.cr1(t).

(163) This situation can, for example, occur when the parameter to be measured varies abruptly, so that the phase difference Δφ.sub.p also varies abruptly. In this case, it is necessary to wait for a few clock times 101 to come back to the situation of FIG. 27.

(164) It will be considered in this example that the phase difference Δφ.sub.p generated by the parameter is not exactly compensated for, so that the total phase difference Δφ.sub.t is increased by the value π/16.

(165) This can be shown in FIG. 29 by shifting the curve 2901 representing the total phase difference Δφ.sub.t by the value π/16. This shift causes a change of the four modulation states on which is modulated the signal received by the detector 14, which is function of the luminous power P(Δφ.sub.t) 2902 received by the latter.

(166) The four levels of the total phase-difference modulation Δφ.sub.t 2901 associated with the four modulation states are hence now: For the state E1: Δφ.sub.t=Δφ.sub.1+π/16=3π/4+π/16=13π/16 For the state E2: Δφ.sub.t=Δφ.sub.2+π/16=5π/4+π/16=21π/16 For the state E3: Δφ.sub.t=Δφ.sub.3+π/16=−3π/4+π/16=−11π/16 For the state E4: Δφ.sub.t=Δφ.sub.4+π/16=−5π/4+π/16=−19π/16.

(167) Hence, as can be seen in FIG. 29, the luminous power P(Δφ.sub.t) 2902 received by the detector 14 in the modulation states E1 and E4 is lower, and that received in the modulation states E2 and E3 is higher.

(168) The detector 14 then delivers a modulated electrical signal S(t) 2903 such as shown in FIG. 29. This modulated electrical signal S(t) 2903 takes sequentially the four values S1, S2, S3 and S4 associated with the four modulation states E1, E2, E3 and E4, respectively. These four values S1, S2, S3 and S4 taken by the modulated electrical signal S(t) 2903 are herein identical two by two: S1=S4 and S2=S3.

(169) The modulated electrical signal S(t) 2903 is then digitalized by the analog/digital converter 111 that delivers and transmits a digital electric signal to the digital processing unit 112.

(170) This digital electrical signal is also modulated and takes four digital values Σ1, Σ2, Σ3, and Σ4 according to the four modulation states E1, E2, E3, and E4 of the total phase-difference modulation Δφ.sub.t 2901.

(171) The digital processing unit 112 demodulates the digital electrical signal in phase with the second biasing phase-shift modulation component φ.sub.b2(t) (cf. FIG. 25) independently of the first phase-shift modulation component φ.sub.m1(t) (cf. FIG. 26).

(172) It is meant by this that the digital processing unit 112 delivers a first demodulated digital signal Σ.sub.p based on the 4 digital values Σ1, Σ2, Σ3, and Σ4 associated with the 4 modulation states E1, E2, E3, and E4, respectively, by performing a calculation operation of the type: Σ.sub.p=−Σ1+Σ2+Σ3−Σ4 where the weight of each digital value in the previous expression depends on the sign of the second biasing phase-shift modulation component φ.sub.b2(t), in the modulation state associated with this digital value, but does not depend on the level of the first phase-shift modulation component φ.sub.m1(t), in this modulation state.

(173) The digital processing unit 112 hence produces a first demodulated digital signal Σ.sub.p depending on the phase-shift Δφ.sub.p and representative of the value of the parameter to be measured in the SAGNAC ring interferometer 20.

(174) In a closed-loop operation, the first demodulated digital signal Σ.sub.p serves as an error signal to control the total phase difference Δφ.sub.t to zero by compensating for the non-reciprocal phase-shift Δφ.sub.p with the opposite phase-shift Δφ.sub.cr1 introduced by the phase modulator 33 controlled by the feedback means 120.

(175) This phase-shift Δφ.sub.cr1 being generated through the same modulation chain 30 as the biasing modulation φ.sub.b1, the control of the modulation chain 30, whose operation is detailed hereinafter, hence allows to have a steady and controlled measurement of Δφ.sub.cr1, and hence finally of Δφ.sub.p, which is opposite thereto and which is the parameter that is desired to be measured.

(176) FIG. 30 shows the case of a fibre-optic measurement device 10 according to the second embodiment of the invention, in which the transfer function of the modulation chain 30 in incorrectly adjusted.

(177) In practice, the transfer function, which depends on the characteristics of both the digital/analog converter 31 via its analog reference voltage and the amplifier 32 via its variable gain G, may undergo variations as a function of the measurement conditions, for example the operating temperature of the device 10 or the electrical drift of certain electronic components of the electronic means 100. Generally, the parameters influencing the transfer function cause low and slow variations of the latter, so that the gain control means 150 operate easily and rapidly so as to keep adjusted the transfer function of the modulation chain 30.

(178) The fact that the transfer function of the modulation chain 30 is incorrectly adjusted translates at the level of the total phase difference Δφ.sub.t by a dilatation of the curve of FIG. 28 representing the total phase Δφ.sub.t so that the total phase difference Δφ.sub.t 3001 is similar to that shown in FIG. 30.

(179) Hence, this dilatation (herein of ratio 16/15) causes a change of the four modulation states E1, E2, E3, and E4, on which is modulated the signal received by the detector 14, which is function of the received luminous power P(Δφ.sub.t) 3002 at the output of the SAGNAC ring interferometer 20.

(180) The four levels of the total phase-difference modulation Δφ.sub.t 3001 associated with the four modulation states in the example of FIG. 30 are hence: For the state E1: Δφ.sub.t=(16/15).Math.Δφ.sub.1=4π/5 For the state E2: Δφ.sub.t=(16/15).Math.Δφ.sub.2=4π/3 For the state E3: Δφ.sub.t=(16/15).Math.Δφ.sub.3=−4π/5 For the state E4: Δφ.sub.t=(16/15).Math.Δφ.sub.4=−4π/3.

(181) Hence, the luminous power P(Δφ.sub.t) 3002 received by the detector 14 in the modulation states E1 and E3 is identical, but lower than the received luminous power when the transfer function of the modulation chain 30 is correctly adjusted, as in FIGS. 28 and 29.

(182) Likewise, the luminous power P(Δφ.sub.t) 3002 received by the detector 14 in the modulation states E2 and E4 is identical, but higher than the received luminous power when the transfer function of the modulation chain 30 is correctly adjusted, as in FIGS. 28 and 29.

(183) The detector 14 then delivers a modulated electrical signal S(t) 3003 such as shown in FIG. 30. This modulated electrical signal S(t) 3003 takes sequentially four values S1, S2, S3, and S4 associated with the four modulation states E1, E2, E3, and E4, respectively. These four values are herein identical two by two: S1=S3 and S2=S4.

(184) The four values Σ1, Σ2, Σ3, and Σ4 of the digital electrical signal associated with the four modulation states E1, E2, E3 and E4, respectively, being also identical two by two, with Σ1=Σ3 and Σ2=Σ4, the first demodulated digital signal Σ.sub.p, calculated by the operation Σ.sub.p=−Σ1+Σ2+Σ3−Σ4, is hence zero.

(185) Besides, the digital electrical signal delivered by the analog/digital converter 111 is transmitted to the gain control means 150 such as shown in FIG. 23.

(186) The gain control means 150 demodulate the digital electric signal so as to provide a signal function of the transfer function of the modulation chain 30.

(187) More precisely, the other digital processing unit of the gain control means 150 operate a calculation operation of the type: Σ.sub.G=Σ1−Σ2+Σ3−Σ4, so as to produce a second demodulated digital signal Σ.sub.G independent of the phase difference Δφ.sub.p generated by the parameter to be measured, but significant of the transfer function of the modulation chain 30.

(188) In particular, in the case shown in FIG. 30, the second demodulated digital signal Σ.sub.G is non-zero, the transfer function of the modulation chain 30 being incorrectly adjusted.

(189) The second demodulated digital signal Σ.sub.G then serves as an error signal for a control loop of the transfer function of the modulation chain 30.

(190) For that purpose, the second demodulated digital signal Σ.sub.G is filtered by a control-loop digital integrator filter that then feeds the digital/analog converter 31 to control its analog reference voltage or the amplifier 32 to control its variable gain G.

(191) Hence, the transfer function of the modulation chain 30 is kept correctly adjusted between the value of the digital control signal and the value of the phase-shift modulation effectively applied by the phase modulator 33.

(192) It will be observed that, in the case of FIGS. 28 and 29, the second demodulated digital signal Σ.sub.G is zero because the transfer function of the modulation chain 30 is correctly adjusted.

(193) Indeed, in this case: Σ1=Σ4, the received luminous power P(Δφ.sub.t) received in the state E1 and in the state E4 being the same, and Σ2=Σ3, the received luminous power P(Δφ.sub.t) received in the state E2 and in the state E3 being the same.

3rd Embodiment

(194) A third embodiment of the invention will now be detailed with reference to FIGS. 31 to 35. In this embodiment: a.sub.1=4 and k.sub.1=0, i.e. the first biasing phase-shift modulation component φ.sub.b1(t) is a square pulse-wave modulation of amplitude π/4 between a high level +π/8 and a low level −π/8 and periodic at a first phase-shift modulation frequency f.sub.b1 equal to the proper frequency f.sub.p of the SAGNAC ring interferometer 20, and Δφ.sub.p=−π/40 so that the first feedback phase-shift modulation component φ.sub.m1(t) is an ascending stair-step modulation of step height φ.sub.s=−Δφ.sub.p=π/40 compensating for the phase difference Δφ.sub.p due to the parameter to be measured.

(195) Hence, according to the invention, the first phase-shift modulation component φ.sub.m1(t) 3100 is such as shown in FIG. 31. It has an ascending pulse-wave modulation shape, falls down by 2π/4 (=π/2) at the instant t=t.sub.R, and has a maximum amplitude lower than 2π/4 (=π/2).

(196) Furthermore, in this third embodiment of the invention, the biasing means 120 generate a second biasing signal producing at the output of the modulation chain 30 a second biasing phase-shift modulation component φ.sub.b2(t) 3200, this second biasing phase-shift modulation component φ.sub.b2(t) 3200 being (cf. FIG. 32): a square pulse-wave modulation of amplitude π/a.sub.2=π, with herein a.sub.2=1≠a.sub.1=4, between a high level +π/2 and a low level −π/2, periodic at a second phase-shift modulation frequency f.sub.b2 such that f.sub.b2=(2k.sub.2+1)f.sub.p=f.sub.p, with k.sub.2=k.sub.1=0 so that (2k.sub.1+1)=1 and (2k.sub.2+1)=1 are multiple of each other, in quadrature with respect to the first biasing phase-shift modulation component φ.sub.b1(t).

(197) It has been shown in FIG. 33 the phase-shift modulation φ.sub.m(t) 3300 resulting, for the third embodiment, from the sum of the first phase-shift modulation component φ.sub.m1(t) 3100 and of the second phase-shift modulation component φ.sub.b2(t) 3200.

(198) It can be seen in FIG. 33 that, taking into account that the first feedback phase-shift modulation component φ.sub.cr1(t) is an ascending modulation, the first phase-shift modulation component φ.sub.m1(t) 3100 falls down by 2π/4 according to the above-described rule, at the falling down instant t=t.sub.R. Likewise, the resulting phase-shift modulation φ.sub.m(t) 3300 also falls down at the same instant t=t.sub.R.

(199) As can be seen in FIG. 33, the maximum amplitude 3301 of the phase-shift modulation φ.sub.m(t) 3300 is herein equal to 2(π/4)+π.

(200) Hence generated by the phase modulator 33 of the modulation chain 30, the phase-shift modulation φ.sub.m(t) introduces between the two counter-propagating waves 24, 25 at the output of the SAGNAC ring interferometer 20 a phase-difference modulation Δφ.sub.m(t)=φ.sub.m(t)−φ.sub.m(t−Δφ.sub.g) such as shown in FIG. 34.

(201) It can be observed, for this third embodiment, that this phase-difference modulation Δφ.sub.m(t) 3400 oscillates between 4 different levels: two high levels: a first level or level 1 when Δφ.sub.m(t)=π−π/4+φ.sub.s a second level or level 2 when Δφ.sub.m(t)=π+π/4+φ.sub.s, and two low levels: a third level or level 3 when Δφ.sub.m(t)=−π+π/4+φ.sub.s a forth level or level 4 when Δφ.sub.m(t)=π−π/4+φ.sub.s.

(202) As can be seen in FIG. 34, the phase-difference modulation Δφ.sub.m(t) 3400 is hence shifted towards the high side of the value φ.sub.s=−Δφ.sub.p (=π/40 herein) as seen above.

(203) Moreover, it can be observed that, at the instant t=t.sub.R of the falling down of the phase-shift modulation φ.sub.m(t), the phase-difference modulation Δφ.sub.m(t) does not rise up (dash line 3401) but remains at its first level 3402 during half a step duration, to then fall down at the instant t.sub.R+(Δτ.sub.g/2) towards the third level 3403, the phase-difference modulation Δφ.sub.m(t) being symmetrical with respect to the instant of the falling down (i.e. that Δφ.sub.m(t.sub.R+t)=Δφ.sub.m(t.sub.R−t)). The alternation between high levels and low levels then continue until the next falling down.

(204) To sum up, in the third particular embodiment of the invention, it has been seen on the one hand that the parameter to be measured introduces between the two counter-propagating waves a phase difference Δφ.sub.p=−π/40 at the output of the SAGNAC ring interferometer 20. On the other hand, it has been seen that the feedback means 120 generate a first feedback signal to produce at the output of the modulation chain 30 a first stair-step feedback phase-shift modulation component φ.sub.cr1(t) of step height φ.sub.s=π/40 so as to introduce between the two counter-propagating waves 24, 25 a first feedback phase-difference modulation component Δφ.sub.cr1(t) compensating for the phase difference Δφ.sub.p due to the parameter to be measured.

(205) For the third embodiment of the invention, it has hence been shown in FIG. 35: the total phase difference Δφ.sub.t(t) 3501 between the two counter-propagating waves 24, 25 at the output of the ring interferometer 20, this total phase difference Δφ.sub.t(t) being such that: Δφ.sub.t(t)=Δφ.sub.p+Δφ.sub.m(t), the luminous power P(Δφ.sub.t) 3502 received by the detector 14 as a function of the total phase difference Δφ.sub.t(t), and the modulated electrical signal S(t) 3503 delivered by the detector 14 as a function of time t.

(206) As the total phase difference Δφ.sub.t(t) 3501 is the sum of the phase difference Δφ.sub.p due to the parameter to be measured and of the phase-difference modulation Δφ.sub.m(t) 3400, it will also be observed in FIG. 35 that the curve 3501 representing the total phase difference Δφ.sub.t(t) as a function of time t corresponds to the curve 3400 of FIG. 34 representing the phase-difference modulation Δφ.sub.m(t), this curve being shifted towards the abscissa axis by the value of the phase difference Δφ.sub.p due to the parameter to be measured.

(207) Hence, the total phase-difference modulation Δφ.sub.t(t) 3501 has sequentially four different levels defining four different modulations states, which are: State E1 when the total phase-difference modulation Δφ.sub.t(t) is Δφ.sub.1=π−π/4=3π/4, State E2 when the total phase-difference modulation Δφ.sub.t(t) is Δφ.sub.2=π+π/4=5π/4, State E3 when the total phase-difference modulation Δφ.sub.t(t) is Δφ.sub.3=−π+π/4=−3π/4 (=Δφ.sub.2−2π), State E4 when the total phase-difference modulation Δφ.sub.t(t) is Δφ.sub.4=−π−π/4=−5π/4 (=Δφ.sub.1−2π).

(208) The luminous power P(Δφ.sub.t) 3502 received by the detector 14 is hence modulated according to the four distinct modulation states and the modulated electric signal S(t) 3503 delivered by the detector 14 takes sequentially four values S1, S2, S3, and S4 associated with the four modulation states E1, E2, E3, and E4, respectively, of the total phase-difference modulation Δφ.sub.t(t) 3501.

(209) As can be seen in FIG. 28, the luminous power P(Δφ.sub.t) 3502 received by the detector 14 in the four modulation states E1 to E4 is the same. Indeed, the received luminous power P(Δφ.sub.t) 3502 being a cosine function of the total phase difference Δφ.sub.t 3501, the following relations are satisfied: P(Δφ.sub.1)=P(Δφ.sub.4) and P(Δφ.sub.2)=P(Δφ.sub.3). Moreover, the states E1 and E2 (respectively the states E3 and E4) being symmetrical with respect to π (respectively with respect to −π), the following relations are also satisfied: P(Δφ.sub.1)=P(Δφ.sub.2) and P(Δφ.sub.3)=P(Δφ.sub.4).

(210) The detector 14 then delivers a modulated electric signal S(t) 3503 as shown in FIG. 35. This modulated electric signal S(t) 3503 takes sequentially the four values S1, S2, S3, and S4 associated with the four modulation states E1, E2, E3, and E4, respectively. Taking into account what have been explained regarding the luminous power P(Δφ.sub.t) 3502 received by the detector 14 in the four modulation states E1 to E4, these four values S1, S2, S3, and S4 taken by the modulated electrical signal S(t) 3503 are herein all identical: S1=S2=S3=S4.

(211) As herein-above, it can be observed that the falling down of the phase-difference modulation Δφ.sub.m(t) 3400 at the instant t=t.sub.R (see FIG. 34) creates no defect in the modulated electrical signal S(t) 3503 of FIG. 35 insofar as the total phase difference Δφ.sub.t 3501 does not change during this falling down (Δφ.sub.m(t) remaining at its high level 3402). Hence, the falling down does not disturb the measurement.

4th Embodiment

(212) A fourth embodiment of the invention will now be detailed with reference to FIGS. 36 to 40. In this embodiment: a.sub.1=1 and k.sub.1=0, i.e. the first biasing phase-shift modulation component φ.sub.b1(t) is a square pulse-wave modulation of amplitude π between a high level +π/2 and a low level −π/2 and periodic at a first biasing modulation frequency f.sub.b1 equal to the proper frequency f.sub.p of the SAGNAC ring interferometer 20, and Δφ.sub.p=−3π/40 so that the first feedback phase-shift modulation component φ.sub.cr1(t) is an ascending stair-step modulation of step height φ.sub.s1=[a.sub.2/(a.sub.2−1)](−Δφ.sub.p)=π/10 with a.sub.2=4≠a.sub.1=1.

(213) Hence, in this fourth embodiment of the invention, the first phase-shift modulation component φ.sub.m1(t) 3600 is such as shown in FIG. 36. It has an ascending pulse-wave modulation shape, falls down by 2π at the instant t=t.sub.R, and has a maximum amplitude lower than 2π.

(214) Moreover, in this fourth embodiment of the invention, the biasing means 120 generate a second biasing signal producing at the output of the modulation chain 30 a second biasing phase-shift modulation component φ.sub.b2(t), this second biasing phase-shift modulation component φ.sub.b2(t) being: a square pulse-wave modulation of amplitude π/a.sub.2=π/4 (hence a.sub.2=4, as described above) between a high level +π/8 and a low level −π/8, periodic at a second biasing modulation frequency f.sub.b2 such that f.sub.b2=f.sub.b1=(2k.sub.1+1)f.sub.p=f.sub.p, because herein k.sub.1=0 (see above), in lagging quadrature relative to the first biasing phase-shift modulation component φ.sub.b1(t).

(215) It will be defined herein that the second biasing phase-shift modulation component φ.sub.b2(t) is in lagging quadrature relative to the first biasing phase-shift modulation component φ.sub.b1(t) when a transition of the first biasing phase-shift modulation component φ.sub.b1(t) from a high level to a low level occurs when the second biasing phase-shift modulation component φ.sub.b2(t) is at a high level. By symmetry, a transition of the first biasing phase-shift modulation component φ.sub.b1(t) from a low level to a high level occurs when the second biasing phase-shift modulation component φ.sub.b2(t) is at a low level.

(216) In this fourth embodiment of the invention, the feedback means 120 generate a second feedback signal, producing at the output of the modulation chain a second feedback phase-shift modulation component φ.sub.cr2(t) which is: a stair-step modulation, each step having a duration Δτ.sub.g and a height φ.sub.s2=[1/(a.sub.2−1)](−Δφ.sub.p)=(1/3).Math.(3π/40)=π/40, in lagging quadrature relative to the first feedback phase-shift modulation component φ.sub.cr1(t).

(217) So generated, the second feedback phase-shift modulation component φ.sub.cr2(t) introduces a second feedback phase-difference modulation component Δφ.sub.r2(t)=φ.sub.cr2(t)−φ.sub.cr2(t−Δτ.sub.g) between the two counter-propagating waves 24, 25 propagating in the SAGNAC ring interferometer 20.

(218) The heights φ.sub.S1 and φ.sub.s2 of the first and second feedback phase-shift modulation components φ.sub.cr1(t) and φ.sub.cr2(t) are such that the difference between the first feedback phase-difference modulation component Δφ.sub.cr1(t) and the second feedback phase-difference modulation component Δφ.sub.cr2(t) compensates for the phase difference Δφ.sub.p.

(219) Indeed, the feedback phase-difference modulation Δφ.sub.cr(t), defined as the difference between the first feedback phase-difference modulation component Δφ.sub.cr1(t) and the second feedback phase-difference modulation component Δφ.sub.2(t), is a stair-step modulation, each step having a height φ.sub.s=φ.sub.s1−φ.sub.s2=π/10−π/40=3π/40=Δφ.sub.p.

(220) Besides, in this fourth embodiment, the control means 140 for controlling the modulation chain 30 process the second biasing signal and second feedback signal to deliver at least one second control signal at the input of the modulation chain 30.

(221) This second control signal then produces at the output of the modulation chain 30 a second phase-shift modulation component φ.sub.m2(t) which is the sum of the second biasing phase-shift modulation component φ.sub.b2(t) and the second feedback phase-shift modulation component φ.sub.cr2(t), so that: φ.sub.m2(t)=φ.sub.b2(t)+φ.sub.cr2(t).

(222) Furthermore, according to this fourth embodiment, the control means 140 are arranged so that the second phase-shift modulation component φ.sub.m2(t) operates a transition of twice the amplitude of the second biasing phase-shift modulation component φ.sub.b2(t), i.e. herein 2π/a.sub.2=2π/4=π/2 (with a.sub.2=4), when its level exceeds the amplitude of the second biasing phase-shift modulation component φ.sub.b2(t), i.e. π/a.sub.2=π/4.

(223) Hence, it has been shown in FIG. 37 the curve representing this second phase-shift modulation component φ.sub.m2(t) 3700. Being formed similarly to the first phase-shift modulation component φ.sub.m1(t) 3600, the second phase-shift modulation component φ.sub.m2(t) 3700 is also a pulse-wave modulation, herein ascending. This modulation falls down at the instant t=t′.sub.R=t.sub.R+Δτ.sub.g/2, φ.sub.m2(t) being in lagging quadrature relative to φ.sub.m1(t).

(224) Mathematically, the relation that links the first phase-shift modulation component φ.sub.m1(t) 3600 to the second phase-shift modulation component φ.sub.cr2(t) 3700 is herein the following: φ.sub.m2(t)=(1/a.sub.2).Math.φ.sub.m1(t−Δτ.sub.g/2).

(225) Generally, in this fourth embodiment, when: the first phase-shift modulation component φ.sub.m1(t) is a pulse-wave modulation, the width of each pulse of which is equal to Δτ.sub.g/(2k.sub.1+1), the first biasing phase-shift modulation component φ.sub.b1(t) being a periodic modulation at a first biasing modulation frequency f.sub.b1 such that f.sub.b1=(2k.sub.1+1)f.sub.p (f.sub.p=1/2Δτ.sub.g being the proper frequency of the ring interferometer 20), and the second phase-shift modulation component φ.sub.m2(t) is a pulse-wave modulation, the width of each pulse of which is equal to Δτ.sub.g/(2k.sub.1+1), the second biasing phase-shift modulation component φ.sub.b1(t) being a periodic modulation at a second biasing modulation frequency f.sub.b2 such that f.sub.b2=f.sub.b1=(2k.sub.1+1)f.sub.p,
the mathematical relation that links the first phase-shift modulation component φ.sub.m1(t) to the second phase-shift modulation component φ.sub.m2(t) is then the following: φ.sub.m2(t)=(1/a.sub.2).Math.φ.sub.m1(t−[Δτ.sub.g/2(2k.sub.1+1)]).

(226) A phase-shift modulation φ.sub.m(t) is then defined as the difference between the first phase-shift modulation component φ.sub.m1(t) and the second phase-shift modulation component φ.sub.m2(t) such that φ.sub.m(t)=φ.sub.m1(t)−φ.sub.m2(t).

(227) It has hence been shown in FIG. 38 the resulting phase-shift modulation φ.sub.m(t) 3800 for this fourth embodiment.

(228) It can be observed in this figure that the phase-shift modulation φ.sub.m(t) 3800 is an ascending, “four state” modulation, having: a falling down at the instant t=t.sub.R of the falling down of the first phase-shift modulation component φ.sub.m1(t) 3600, this falling down having the same amplitude, i.e. 2π, and a rising up at the instant t=t′.sub.R=t.sub.R+Δτ.sub.g/2 of the falling down of the second phase-shift modulation component φ.sub.m2(t) 3700, this rising up having the same amplitude as the falling down, i.e. 2π/4.

(229) The phase-shift modulation φ.sub.m(t) 3800 operates a rising up at the instant t=t′.sub.R rather than a falling down as the second phase-shift modulation component φ.sub.m2(t) 3700 because the contribution of the second phase-shift modulation component φ.sub.m2(t) 3700 is added negatively to the phase-shift modulation φ.sub.m(t) 3800, being equal to φ.sub.m1(t)−φ.sub.m2(t).

(230) So formed, the phase-shift modulation φ.sub.m(t) 3800 has a maximum amplitude strictly lower than 2π, as can be verified in FIG. 38. The excursion of the modulation chain is hence limited and the effects of the non-linearities of the latter are reduced.

(231) So generated by the phase modulator 33 of the modulation chain 30, the phase-shift modulation φ.sub.m(t) 3800 introduces between the two counter-propagating waves 24, 25 at the output of the SAGNAC ring interferometer 20 a phase-difference modulation Δφ.sub.m(t)=φ.sub.m(t)−φ.sub.m(t−Δτ.sub.g) such as shown in FIG. 39.

(232) It is observed, for this fourth embodiment, that the phase-difference modulation Δφ.sub.m(t) 3900 also oscillates between 4 different levels: two high levels: a first level when Δφ.sub.m(t)=π−π/4+φ.sub.s, a second level when Δφ.sub.m(t)=π+π/4+φ.sub.s, and two low levels: a third level when Δφ.sub.m(t)=−π+π/4+φ.sub.s a fourth level when Δφ.sub.m(t)=−π−π/4+φ.sub.s.

(233) As can be seen in FIG. 39, the phase-difference modulation Δφ.sub.m(t) 3900 is hence shifted towards the high side of the value φ.sub.s=−Δφ.sub.p (=3π/40 herein) as seen above. Moreover, its amplitude is herein equal to 2π+2(π/4)=2π+π/2.

(234) Besides, it can be observed that at the instant t=t.sub.R of the falling down of the phase-shift modulation φ.sub.m(t) 3800, the phase-difference modulation Δφ.sub.m(t) 3900 does not rise up but remains at its third level 3901 during half a step duration Δτ.sub.g/2. Then, at the instant t=t′.sub.R=t.sub.R+Δτ.sub.g/2 of the falling down of the second phase-shift modulation component φ.sub.m2(t), the phase-difference modulation Δφ.sub.m(t) still remains at its third level 3902 during half a step duration Δτ.sub.g/2.

(235) The phase-difference modulation Δφ.sub.m(t) then rises up towards the levels to resume its alternations between high levels and low levels until the next transition.

(236) For the fourth embodiment of the invention, it has finally been shown in FIG. 40: the total phase difference Δφ.sub.t(t) 4001 between the two counter-propagating waves 24, 25 at the output of the ring interferometer 20, this total phase difference Δφ.sub.t(t) 4001 being such that: Δφ.sub.t(t)=Δφ.sub.p+Δφ.sub.m(t), the luminous power P(Δφ.sub.t) 4002 received by the detector 14 as a function of the total phase difference Δφ.sub.t(t) 4001, and the modulated electric signal S(t) 4003 delivered by the detector 14 as a function of time t.

(237) As hereinabove, it can be observed that neither the falling down at the instant t=t.sub.R, nor the rising up at the instant t=t′.sub.R of the phase-difference modulation Δφ.sub.m(t) 3900 (see FIG. 39) creates a defect in the modulated electric signal S(t) 4003 of FIG. 40, insofar as the total phase difference Δφ.sub.t 4001 does not change during this falling down or this rising up (Δφ.sub.m(t) remaining at its third level 3901 and 3902). Hence, neither the falling down, nor the rising up, disturbs the measurement.

(238) The measurement device according to the invention is particularly well adapted to the realization of a gyrometer. In this case, the parameter to be measured is a component of the rotational speed of the ring interferometer.

(239) This gyrometer hence advantageously enters into the making of navigation or inertial-stabilization systems.

(240) Such an arrangement is also well adapted to the realization of a device for measuring magnetic fields or electric currents, using advantageously the FARADAY effect.