Optical Fiber Gyroscope With (3.times.3) Directional Coupler

It has long been known that Sagnac effect may be utilized to detect rotation rate of various obJects. See references: e.g., Post review of Modern Physics, Vol. 39 p. 475 1967; and V. Vali, R. W. Shorthill and M. F. Berg, Applied Optics Vol. 15, p. 1099, 1977. Also, refer to U.S. Pat. Nos. 3,102,953, and 4,013,365 for disclosures relating to apparatus for measuring angular rotation. The Sagnac gyroscope has been developed vigorously, especially since low loss long length single mode optical fibers became available.

One of the technical difficulties in optical gyroscope technology stems from the fact that unless .pi./2 non-reciprocal phase bias scheme is utilized, the gyro sensitivity is at its minimum near zero rotation rate. And most of the gyro applications deal with measurement of near zero rotation rate. There are various ways of implementing the .pi./2 nonreciprocal phase bias in the optical fiber loop in order to overcome the problem. However, they add to the gyro system additional optical noise system complication, and device size and cost. The present invention provides an optical fiber gyroscope with a three fiber directional coupler for coupling light into opposite ends of a fiber loop to traverse the loop in mutually opposite directions and for receiving the light from opposite respective fiber ends and combining it in a manner whereby optical output is at quadrature or most sensitive to change at near zero rotation of the loop.

Accordingly, it is the purpose of the present invention to provide an optical fiber gyroscope that has near maximum sensitivity at a near zero rotation rate, and thus does not need the .pi./2 nonreciprocal phase bias.

This purpose is achieved by recognizing that all the known gyroscopes use beam splitters that have two input ports and two output ports. Specifically, the purpose of the present invention is achieved by using as the beam splitter a directional coupler (herein designated a (3.times.3) directional coupler) having three input ports (three ports on the input side) and three output ports. The first input port among the three input ports receives an optical input beam. The second and the third input ports are disposed symmetrically with respect to the first input port. Thus, the second and the third ports receive optical beams of the same magnitude and phase from the first input port through evanescent coupling therebetween. The beams from the second and third output ports are directed into the ends of an optical fiber which is arranged to define an optical gyrsocope loop. Since the second and third ports are symmetric and indistinguishable, the final gyro loop outputs back to the second and the third outputs are the same unless the gyro is in rotation. This means that the gyro sensitivity is near maximum at zero or near zero rotation rate. As a result, an optical gyroscope using a (3.times.3) directional coupler as a beam splitter does not need a .pi./2 phase bias in the gyro loop in order to maximize gyro sensitivity at very small rotation rates.

Additional advantage and features will become apparent as the subJect invention becomes understood by reference to the following detailed description when considered in conjunction with the accompanying drawings.

FIG. 1 shows the relationship between optical gyroscope output P.sub.I and P.sub.II relative to rotation rate in the conventional gyroscope without .pi./2 nonreciprocal phase bias and illustrates that gyro sensitivity dP.sub.I /d.OMEGA. and dP.sub.II /d.OMEGA. vanishes at .OMEGA.=0.

FIGS. 2 and 3 are schematic illustrations of conventional optical gyroscopes using two beam splitters, each with two inlets and two outlets.

FIG. 4 is a schematic representation of the present invention wherein an optical gyroscope uses as the beam splitter a directional coupler with three inlets and three outlets, i.e. a (3.times.3) directional coupler.

FIG. 5 is a cross-sectional illustration of optical coupling between three equispaced optical fibers as in the directional coupler of FIG. 4.

FIG. 6 shows the relationship between outputs P.sub.II (and P.sub.III) of the gyroscope shown in FIG. 5 relative to rotation rate .OMEGA..

First a background review will be made regarding the phase bias problem of the conventional optical gyroscope. FIG. 1 shows, for example, the relationship between gyro output P.sub.I and rotation rate .OMEGA.. As seen, the gyro sensitivity being defined by dP.sub.I /d.OMEGA. equal to zero at .OMEGA.=0. The rate change to either side of .OMEGA.=0 causes only minimal amplitude change along the output curve. Since many gyro applications involve the measurement of very small rotation rates. it is necessary to use so called nonreciprocal phase bias to move the initial (.OMEGA.=0) operating point A' to A, as depicted in FIG. 1, for increased sensitivity.

Such a problem occurs in the conventional optical gyroscope as shown, for example, in FIGS. 2 and 3. In Fig. 2, input beam P.sub.in is split into two by a directional coupler 1. The two split beams make round trips along the closed fiber loop 2, one propagating clockwise and the other propagating counter clockwise, and recombine at directional coupler 1. In FIG. 3 an equivalent optical gyroscope is shown in which bulk optics are used instead of the all-guided wave optical components as illustrated in FIG. 2. Input beam P.sub.in is split by a bulk beam splitter 3. The split beams are folded by mirrors 4, 5 and 6 to make round trips in opposite directions.

It is worthwhile to notice that the beam splitters 1 and 3 in FIGS. 2 and 3, respectively, have two inlets and two outlets. Thus, it is always possible to distinguish one inlet from the other inlet because one receives the input beam while the other does not.

FIG. 4 is a schematic representation of the present invention in which a directional coupler 10 with three inlets and three outlets is used as the beam splitter. With respect to the center inlet waveguide 12, the two other waveguides 14 and 16 are indistinguishable. There is illustrated in FIG. 5 a typical cross-sectional view of optical fibers defining directional coupler 10 in FIG. 4. Three equidistant or equispaced fibers a, b and c are illustrated, in which the coupling coefficients K.sub.ab (between the fibers a and b) and K.sub.ac (between the fibers a and c) are same. Assuming a beam P.sub.in is launched into optical fiber a, for example, it will be coupled by its evanescent field equally into optical fibers b and c as illustrated b K.sub.ab and K.sub.ac. Optical fibers a, b and c, which define waveguides, may be disposed parallel or mutually twisted about each other as they traverse some length L (Fig. 4) across directional coupler 10 sufficient to establish evanescent coupling. Light from optical fibers b and c may be coupled into respective ends of optical fiber loop 20. Light introduced into respective ends of loop 20 propagate therethrough in mutually opposite directions and emerge at opposite respective ends to enter optical fiber waveguides b and c, this time propagating in the reverse direction. Again, optical fibers b and c define waveguides of close proximity, as illustrated in FIG. 5, to allow mutual interference of the beams at directional coupler 10.

As shown in FIG. 4, directional coupler 10 comprises three optical waveguides, such as optical fibers, which lie in coextensive adjacency whereby laser light transported through waveguide 12 is equally coupled by evanescent coupling into equidistant waveguides 14 and 16. Waveguide 12 is relatively short. Waveguides 14 and 16 may also be of short lengths, in which case it is necessary to couple light from their ends into respective ends of the optical fiber forming loop 20. However, waveguides 14 and 16 may be integral extensions of the optical fiber waveguide defining loop 20. Beams returning from round trips through loop 20 are directed through lengths of waveguides 14 and 16 in close adJacency for mixing for mutual interference prior to separation as P.sub.II and P.sub.III, as shown in FIG. 4. Thereafter, beam outputs P.sub.II and P.sub.III are directed from ends of respective waveguides onto detector means (not illustrated) where the outputs P.sub.II and P.sub.III are detected and compared to provide an indication of loop rotation. Waveguides 12, 14 and 16 may be fixedly carried in a substrate not illustrated to maintain their relative spacings from each other. When optical fibers define the waveguide, they may be disposed either parallel or mutually twisted in a substrate such as a suitable liquid or set epoxy. Otherwise, the waveguides may be diffused, for example, into a substrate at controlled locations and depths such that their equidistances are maintained as defined in FIG. 5.

If the coupling coefficients K.sub.ab and K.sub.ac are same, P.sub.II and P.sub.III will be same if .OMEGA.=0, since they are indistinguishable. This is shown in FIG. 6 in which the curves P.sub.II, P.sub.III versus rotation rate are shown. While the sensitivity of dP.sub.I /d.OMEGA. (or dP.sub.II /d.OMEGA.) is zero at .OMEGA.=0 in FIG. 1, the sensitivity of dP.sub.II /d.OMEGA. (or dP.sub.III /d.OMEGA.) is not zero at .OMEGA.=0 in FIG. 6.

The mathematical analysis to derive the curves in FIG. 6 were published in the Journal of Applied Physics 52, (6), June 1981 in an article entitled "Optical Fiber Interferometers with (3.times.3) Directional Couplers; Analysis," and also in Applied Physics Letters, Vol. 37, p. 869 Nov. 15, 1980 entitled "Fiber-Optic Gyroscope with (3.times.3) Directional Coupler".

Directional couplers 1 and 10, as illustrated in FIGS. 2 and 4, respectively, employ evanescent field coupling between two or more optical fibers in adjacency. Such coupling is more fully described in U.S. Pat. No. 4,264,126 issued Apr. 28, 1981 to Sang K. Sheem. The couplers perform a beam splitting function and are used instead of the more conventional bulk-type partially reflective mirrors.

The outputs from a conventional optical fiber gyroscope are illustrated in both FIGS. 1 and 6 (solid line) without a .pi./2 phase bias in the optical fiber loop. When the gyroscope is in a stationary state (.OMEGA.=0), its output intensity sits at the valley of the curve where sensitivity d P /d.OMEGA. is zero. This is illustrated by point A in FIGS. 1 and 6. Since the rotation rate to be detected by a gyroscope is usually small, the output variation at the valley position is minimal unless a nonreciprocal .pi./2 phase bias is purposely introduced between the two oppositely propagating waves to move the operating point up from point A to point A'. The present invention accomplishes this objective by using a (3.times.3) directional coupler 10 as the beam splitter which is illustrated in FIG. 4. In this figure, an optical beam P.sub.in from a laser (not illustrated) is launched into a port defined by the end of optical fiber 12, from which it is distributed to optical fiber 14 and 16 and among output ports P.sub.1, P.sub.2 and P.sub.3. Beams P.sub.2 and P.sub.3 make round trips in opposite directions around loop 20 and reenter directional coupler 10 via fibers 14 and 16 to provide outputs P.sub.II and P.sub.III. Optical fiber 12 is not available as an output. The directional coupler mixes the returning fields over the three waveguides (optical fibers 12, 14 and 16) through evanescent field coupling. As shown in FIGS. 6, output P.sub.II and output P.sub.III are near quadrature (point B) at .OMEGA.=0. Comparison of P.sub.II and P.sub.III (dashed and dash-dotted lines) may be made in FIGS. 6 with the output (solid line) of the conventional gyroscopes known in the prior art. The advantage is obvious for at .OMEGA.=0 the gyroscope inherently exhibits maximum sensitivity. A slight later change in causes an appreciable rise or fall along sensitivity curves P.sub.II and P.sub.III.

The described (3.times.3) directional coupler of three input and three output ports operates automatically near maximum sensitivity at zero rotation. Therefore there is no need for a nonreciprocal phase bias in the fiber loop which would add noise and complexity to the system.

The coupling coefficients K.sub.ab and K.sub.ac in FIG. 5 may not be exactly same, but differ slightly. Then P.sub.II and P.sub.III in FIG. 6 will not be exactly same at .OMEGA.=0; nontheless, the sensitivities dP.sub.II /d.OMEGA. and dP.sub.III /d.OMEGA. will not be zero at .OMEGA.=0.

While the invention has been particularly shown and described with reference to a specific embodiment, it will be understood by those skilled in the art that various changes in form and detail may be made thereto without departing from the spirit and scope of the invention which is limited only by the scope of the claims annexed hereto.