Fiber Ring Interferometer Using a Kerr Medium Optical Fiber Loop

The invention pertains to Sagnac type optical fiber interferometers. More particularly, the invention pertains to an improved Sagnac type interferometer in which the effect of noise and other interference on the accuracy of the interferometer is minimized.

A Sagnac interferometer is a device which can be used, among other things, to detect rotational motion and velocity. In a Sagnac interferometer, coherent light (that is, light of a single wavelength and a single phase) is split into two beams and introduced into opposite ends of a loop of optical fiber cable. The loop need not be circular, but it is preferably bent such that the two ends of the loop are at the same location. The light introduced at each end of the loop propagates through the loop and appears at the opposite end. Since the two light beams were in phase when introduced to the loop and since they travel over the same path, i.e., distance, when they reach the respective opposite ends of the cable, they should still be in phase with each other if the loop is stationary. However, if the loop rotates while the light travels through it, the light traveling in the direction in which the loop is rotated (e.g., clockwise) travels a greater distance before reaching the opposite end of the loop than the light traveling in the opposite direction through the loop. If the loop is formed in a circle of radius r, the distance each beam travels through the stationary loop is 2.pi.r and the amount of time, t, that the light takes to travel through the stationary loop is 2.pi.r/c, where c is the speed of light through the optical fiber material. If the loop rotates while each of the light beams are traveling through it, the light beam traveling in the direction of rotation of the loop does not travel over a distance 2.pi.r, but rather travels over a distance 2.pi.r+.DELTA..sub.d1, where .DELTA..sub.d1 =.theta..sub.1 r and .theta..sub.1 r is the angle traversed by the loop during the period that the light beam traveling in the direction of rotation travels through the loop. The light traveling through the loop in the direction opposite to the direction of rotation travels a distance 2.pi.-.DELTA..sub.d2, where .DELTA..sub.d2 =.theta..sub.2 r and .theta..sub.2 is the angle traversed by the loop during the period that the light beam traveling in the opposite direction of rotation travels through the loop. .theta..sub.1 and .theta..sub.2 are equal if the loop stops rotating before either of the oppositely directed beams of light reach the respective opposite ends of the cable. However, if the loop rotates continuously during the period that the two light beams are traveling through the loop, then .theta..sub.1 is slightly larger than .theta..sub.2 because the loop continues to rotate during the period after the beam traveling in the direction opposite to the direction of rotation of the loop exits the opposite end of the loop, but before the beam traveling in the direction of rotation exits the end of the loop. In any event, the difference in distance traveled by the two light beams to reach the respective opposite ends of the loop is .DELTA..sub.d1 +.DELTA..sub.d2 or approximately 2.DELTA.d where .DELTA.d is the average of .DELTA..sub.d1 and .DELTA..sub.d 2.

The difference in distance traveled of 2.DELTA..sub.d causes the light exiting from the opposite ends of the loop to no longer be in phase. The phase difference, .DELTA..sub.p, is a direct function of the distance .DELTA..sub.d and is given by the equation, .DELTA..sub.p =2 .pi..DELTA..sub.d /.lambda., where .lambda. is the wavelength of the light. An interferometric detector determines the phase difference of the two beams based on the interference pattern of the two beams which directly indicates the direction and velocity of rotation of the loop.

In an actual Sagnac interferometer, the optical fiber cable is looped many times around a cylinder in order to increase the distance traveled; however, the concept is exactly the same as that described above with respect to a single loop.

Discrete light pulses are typically used in a Sagnac interferometer, however, continuous wave light beams may also be used.

Since Sagnac interferometers detect extremely small phase differences, accuracy can be greatly compromised by even small signal degradation in the system. As light beams travel over distance, the beams are subject to interference and are generally degraded. Such degradation is not of concern in the optical fiber loop where the two light beams travel over the same path. Such beams are subject to the same interference and degradation and thus the effect of such degradation is cancelled out when the two beams are compared. However, in a typical Sagnac interferometer, when the two light beams exit from the opposite ends of the fiber optic loop, they travel over different paths before reaching the detector and are thus subject to different interference and degradation. This may cause an unwanted relative phase shift between the two beams and lead to inaccuracies in the operation and a general decrease in sensitivity of the interferometer.

Accordingly, it is an object of the present invention to provide an improved Sagnac interferometer.

It is a further object of the present invention to provide a Sagnac interferometer which is immune to interference and degradation of the two light signals.

It is another object of the present invention to provide a more accurate and sensitive Sagnac interferometer.

The various objects and goals are met in the present invention by providing a means by which the two light beams exiting from opposite ends of the optical fiber cable of a Sagnac interferometer are allowed to travel over the same path to the interferometric detector. In the invention, a single polarized light beam is split by a polarizing beam splitter into two orthogonally polarized coherent beams. These light beams are introduced into opposite ends of, and travel in opposite directions around, a polarization maintaining optical fiber loop. A 90.degree. twist is imparted to the cable. Upon exiting the opposite ends of the loop, each of the beams again strike the polarizing beam splitter and, since their polarizations have been shifted 90.degree. by the twist in the cable, are recombined into a single beam.

The combined light beam is passed through a birefringent quarter-wave plate which converts linear polarized light to circular polarized light. The two distinct polarizations which comprise the combined beam are converted to circular polarizations which are also distinct from each other. The beam is then forwarded to a second polarizing beam splitter, which splits the beam into two beams of distinct intensities depending upon the particular circular polarization of the combined beam, which, in turn, depends upon the relative phase difference of the two beams which were combined to form the beam. The two separated light beams are immediately introduced to a balanced detector. The balanced detector detects the difference in intensity between the two light beams. The difference in power indicates the loop's rotational direction and velocity. The interferometer can be mounted, for instance, in a boat or plane and be used as an optical fiber gyro for detecting rotation of the boat or plane.

FIG. 1 is a simplified diagram of a typical prior art Sagnac interferometer.

FIG. 2 is a simplified diagram of a preferred embodiment of the interferometer of the present invention.

FIGS. 3A-3G are graphs of the polarization vectors of various exemplary light beams in various stages in the interferometer of the present invention.

FIG. 1 is a simplified diagram of a prior art Sagnac interferometer. A light source 12 produces a beam of light 14 which is directed towards a 50/50 beam splitter 16. The beam splitter 16 passes half of the light into end 18a of an optical fiber loop 18 and reflects the other half of the light into end 18b of the loop 18.

Beam splitter 16 splits the beam, much like a two way mirror, by reflecting half of the incident light and allowing half of the incident light to pass through. The beam is typically split regardless of the particular properties of the incident beam. That is, there is no particular property or properties of the light which dictate which light rays pass through the beam splitter and which are reflected off the beam splitter.

The beam directed into end 18a (hereinafter beam a) travels in the direction of arrow 20 through the fiber optic loop 18. The beam which is directed towards end 18b (hereinafter beam b) travels around the loop in the opposite direction, as illustrated by arrow 22. Beam a exits the loop from end 18b and is directed to interferometric detector 26. Beam b exits the loop from end 18a and is directed to interferometric detector 26. Various means are well known to those skilled in the art for directing light beams. Such means might include mirrors, waveguides or additional optical fiber cables. However, FIG. 1 does not show such well known light directing means in order not to obfuscate the invention.

As discussed above, if the loop 18 was not in motion as beams a and b passed through it, then beams a and b, which were in phase upon entering the loop 18, will still be in phase when they reach balanced detector 26, if there was no undesired signal degradation. In such a case, the phases of beams a and b will overlap causing constructive interference over the entire wavelength spectrum in detector 26, thus indicating no phase shift and, consequently, that the loop was stationary. However, if the loop was rotating, beams a and b will not be in phase upon exiting loop 18 and detector 26 will detect a different interference pattern. As described above, the interference pattern can be used to determine the phase shift between beams a and b and thus the direction and velocity of rotation of the loop.

As seen in FIG. 1, beams a and b travel over different space once they exit from the optical fiber loop 18. In fact, beams a and b must travel over different paths because, if recombined, there will be no way to distinguish the phase properties of beam a from the phase properties of beam b in detector 26. Since they travel through different paths, beams a and b may be subject to different noise and other types of degradation which might cause a relative phase shift between the two beams. Any relative phase shift caused outside of the fiber optic loop 18 is undesirable because the phase difference between beam a and beam b would no longer be due solely to the rotation of the optical fiber loop, thus causing inaccurate results in detector 26. As one particular example, a temperature difference between the paths over which beams a and b travel would cause a relative phase shift between the two beams. Since the typical phase shift which occurs in a fiber optic gyro is extremely small, very minor errors, even on the order of a factor of 10.sup.-10, create noticeable inaccuracy.

FIG. 2 is a simplified illustration of a fiber optic interferometer of the present invention. A light source 30, which may, for example, be a laser, produces a light beam 32. Light beam 32 is preferably polarized by a polarizer 33 and more preferably linearly polarized to have a polarization vector of 45.degree. as shown in the polarization vector graph of FIG. 3A.

Polarized light is light in which the fields of all the light rays are aligned. In particular, a light wave comprises an electric and a magnetic field oscillating perpendicular to each other and to the direction the wave is traveling. Polarization, in general, exists where there is an alignment relationship between all the light waves comprising a beam of light. Linear polarization is the specific situation where the beam comprises waves all of which have their electric fields in the same plane. (Since the electric and magnetic fields of a given wave are always perpendicular, if their electric fields are aligned in a first plane, their magnetic fields are necessarily aligned in a second plane which is perpendicular to the first plane). Circular or elliptic polarization exists where the fields of the various light waves comprising a beam change regularly along the light wave.

Beam 32 is directed through a polarizing beam splitter 34. Preferably, beam splitter 34 splits the beam polarized along the 45.degree. axis creating two light beams of equal intensity which are linearly, orthogonally polarized (hereinafter beams c and d), as illustrated by FIG. 3B. The beam splitter 34 reflects the component of the incident beam 32 which is linearly polarized in a first orthogonal direction (e.g., the parallel direction) and transmits the portion of the incident beam 32 which is linearly polarized in the other orthogonal direction (e.g., the perpendicular direction).

Reflected beam c enters fiber optic loop 36 at end 36a while transmitted beam d enters fiber optic loop at the opposite end 36b. Beam c travels around the fiber optic loop 36 in the direction of arrow 38 while beam d travels around the fiber optic loop 36 in the direction of arrow 40. Fiber optic loop 36 is constructed of a linear polarization maintaining fiber such that the polarizations of beams c and d are preserved in the loop 36. In the fiber loop, beams c and d are time reversed from each other.

Normal optical fibers are insensitive to polarization and the polarization properties of the waves in the fiber can change. Unless the optical fiber is very short, any polarization properties which a light beam entering the fiber might have had will be lost by the time it reaches the other end of the fiber. Polarization maintaining fibers, however, maintain the polarization of the light that originally entered the fiber with the eigen polarization of the particular fiber. Such fibers have different properties for light of different polarizations. For instance, currently available linear polarization maintaining fibers can maintain the polarization properties of linearly polarized light introduced into the fiber at one of two orthogonal eigen polarizations of the fiber. The fiber can maintain light entering the fiber at only one of its eigen polarizations at a time. In such a fiber, the two orthogonal linear polarizations are isolated from each other as the light travels along the fiber. Accordingly, when light emerges from the opposite end of a linear polarization maintaining fiber, it has the same polarization relative to the end of the fiber at which it entered.

In the present invention, ends 36a and 36b of the fiber optic cable are twisted 90.degree. with respect to one another such that light which enters one end in a particular polarization, e.g., parallel polarization, exits from the opposite end shifted 90.degree., e.g., perpendicular polarization. Accordingly, when beam c, which entered fiber optic loop 36 at end 36a in the parallel polarization exits from end 36b, and again strikes polarizing beam splitter 34, instead of being reflected, this time it will be transmitted. Likewise, beam d, which comprises the light of perpendicular polarization of original beam 32 which was transmitted through polarizing beam splitter 34, exits from end 36a in parallel polarization. Since it is in parallel polarization, it is reflected by polarizing beam splitter 34, instead of transmitted through it. In this manner, beams c and d are recombined by the beam splitter 34 as they exit from opposite ends of the fiber optic loop to form a single recombined beam 42. Alternately, instead of twisting the cable 90.degree. from end to end, a half-wave plate having a 45.degree. axis can be placed at one end of the fiber.

As explained above, if the fiber optic loop was rotating while light beams c and d were passing through it in opposite directions, there will be a relative phase difference between the light comprising the two beams. Accordingly, recombined beam 42 would comprise light of two different phases.

Recombined beam 42 is separated from the incoming beam 32 by a nonreciprocal 68 element (for example an optical switch or Faraday circulator) and is directed to a birefringent quarter-wave plate 44. The plate has different speeds of propagation through it for light of different polarizations. Accordingly, the relative phase difference between the two beam components which comprise recombined beam 42 is altered. In particular, a quarter-wave plate shifts phase by .pi./2, i.e., a quarter-wavelength. Accordingly, a quarter-wave plate would convert light that is linearly polarized to light which is circularly polarized.

If there has been no relative phase shift between beams c and d, then recombined beam 42 will be like incident beam 32, i.e., linearly polarized at 45.degree. as shown in FIG. 3A. A birefringent quarter-wave plate such as quarter-wave plate 44 would convert a 45.degree. linearly polarized light beam such as shown in FIG. 3A to a circularly polarized beam such as shown in FIG. 3C. However, if there has been a relative phase shift of .DELTA..sub.p due to rotation of the fiber optic loop, then recombined light beam 42 will no longer be perfectly linearly polarized but will be elliptically polarized prior to passing through quarter wave plate 44. FIG. 3D illustrates an exemplary polarization vector graph of recombined light beam 42 if there had been a small phase shift in the loop. Quarter wave plate 44 would convert the elliptical polarization of FIG. 3D to the elliptical polarization shown in FIG. 3E.

Had the phase shift been in the opposite direction, i.e., -.DELTA..sub.p, then the polarization of recombined beam 42 would have been as shown in FIG. 3F, which would be converted by quarter-wave plate 44 to the form shown in FIG. 3G. As illustrated by FIGS. 3F and 3G, the tilt of the graphical representation of the polarization vector indicates the direction in which the fiber optic loop was rotating whereas the degree to which the polarization vector is elliptically stretched from a pure circle before passing through plate 44, or from a straight line, after passing through plate 44, indicates the magnitude of the relative phase difference .DELTA..sub.p.

The beam 46 exiting from quarter-wave plate 44 is applied to a second 45.degree. linear polarizing beam splitter 48 which splits beam 46 into two separate beams 50 and 52. If there has been a relative phase shift, the magnitude of the split beams 50 and 52 will not be equal. The relative intensities of the beams is a direct function of the phase difference of the two distinct beam components of the incident beam 46, which, in turn, is a function of the rotation direction and velocity of the optical fiber loop. Beams 50 and 52 are immediately introduced into a balanced detector 56. The balanced detector 56 determines the relative intensities of beams 50 and 52. The output of the balanced detector is directly indicative of the direction and magnitude of the phase shift and thus the rotation of the device. The output of detector 56 can, for instance, be forwarded to a microprocessor 58 for further analysis and/or recording. If the interferometer is used as a gyro, microprocessor 58 calculates the rotation direction and speed of the fiber optic loop based upon the relative intensities of beams 50 and 52 using well known techniques.

It should be understood by those skilled in the art that the two beams introduced to the optical fiber need not be linearly polarized but can be any two orthogonally polarized beams which are time reversed from one another and for which the fiber is able to maintain those polarizations. If circular polarization maintaining fiber was employed, the fiber would not need to be twisted 90.degree. or need a half-wave plate at one end.

The optical fiber interferometer disclosed herein can be used as a fiber optic gyro for detecting rotation, for example, in planes or boats. The entire apparatus simply could be mounted to a boat. It requires no moving parts and is extremely sensitive.

The invention would also be useful in applications relating to "squeezing" of optical pulses as discussed in K. Bergman and H. A. Haus, Squeezing in Fibers With Optical Pulses, Optic Letters (1991) and M. Shirasaki and H. A. Haus, Squeezing of Pulses in a Non-Linear Interferometer, Journal of the Optical Society of America (1990), which references are incorporated herein by reference. In optical squeezing applications, the optical fiber could not be a linear polarization maintaining fiber, but could be an elliptical or circular polarization maintaining fiber.

A portion of the optical fiber cable 36 may be a Kerr medium as shown at 60 in FIG. 2. A Kerr medium is a medium in which birefringence is induced in a light beam by means of passing it through an electric field such as between the plates 62 and 64 of a charged capacitor 66. A Kerr medium is a medium having non-linear refractive properties. Particularly, in a Kerr medium, the index of refraction depends on the optical power. A Kerr medium can be used in optical "squeezing" as described in the above referenced articles.

Having thus described a few particular embodiments of the invention, various alterations, modifications and improvements will readily occur to those skilled in the art. For instance, it should be understood by persons of skill in the art that the two light beams which originally enter the optical fiber need not be of equal phase or intensity as long as the relative phases and intensities are known. Further, if and when circular polarization maintaining fiber is available, the loop may be constructed of such fiber and the two beams introduced to the loop may be circularly polarized instead of linearly polarized and quarter wave plate 46 could be eliminated. Also, neither of the polarizing beams splitters need be 45.degree. polarizing splitters. Such alterations, modifications and improvements as are made obvious by this disclosure are intended to be part of this description though not expressly stated herein, and are intended to be within the spirit and scope of the invention. Accordingly, the foregoing description is by way of example only, and not limiting. The invention is limited only as defined in the following claims and equivalents thereto.