The invention relates to the field of fiber optics, and more particularly to birefringence in single-mode fibers.
Linear birefringence in a single-mode fiber optic sensing coil employing the Faraday effect has the effect of reducing the sensitivity of the coil to magnetic fields. The sensitivity decreases with increasing length and birefringence of the coil. Birefringence occurs even if the sensing fiber is straight, due to the presence of residual linear birefringence resulting from core ellipticity. Even nominally circular core single mode fibers possess some core ellipticity due to imperfections in the production process. Additionally, bending or other anisotropic stresses on the core contribute to birefringence, such as when the fiber is formed into a coil.
Linear birefringence tends to suppress the response of the fiber to magnetic fields. As this effect is temperature dependent, and each fiber has different characteristics, it is difficult to produce a sensing coil whose characteristics are known over the operating temperature range. Moreover, measuring the coil temperature may be impracticable and expensive.
In the short section of fiber that is used for a sensing coil (a few meters to 50 meters), the ellipticity is relatively constant. This arises from the method of manufacture of optical fiber, where the fiber is drawn down from a preform. A preform is a version of the fiber that has a larger diameter than the fiber, but is shorter in length. Exemplary length ratios (fiber/preform, for fibers manufactured by KVH Industries) are approximately 4000:1. In other words, 0.25 mm of preform length yield approximately 1 meter of fiber. Over this short preform length the ellipticity will not change significantly, so that the ellipticity can be considered as being constant in that section of fiber.
Several approaches have been proposed to overcome the deleterious effects caused by linear birefringence when the fiber is used for current sensing. For example, circular birefringence can be introduced by torsion-type twists to “swamp” the linear birefringence and achieve coupling of the linear modes. Alternatively or in addition, the preform or the fiber can be spun or rotated in an oscillating manner during drawing to reduce polarization mode dispersion (PMD). These remedies tend to complicate the fiber or coil fabrication process, may require annealing, and are expensive.
It would therefore be desirable to provide an optical fiber for magnetic field and current sensing applications and a method of producing such an optical fiber that has increased sensitivity to a magnetic field and reduced temperature dependence by reducing the deleterious effects caused by linear birefringence and that can be produced at less cost.
The invention relates to reducing the effect of linear birefringence in an optical fiber, in particular a circular-cored single-mode fiber. According to one aspect of the invention, a method is disclosed of reducing linear birefringence in an optical fiber, which may include subdividing a length of the optical fiber into a plurality of sections, and introducing between the sections a twist having a predetermined sense of rotation and a twist angle. Twist, as used herein, may be a permanent rotational displacement of the fiber from its initial position over a distance which is a small fraction of a beat length. In one method, the twist occurs over a length of less than 1 mm. According to another aspect of the invention, an optical fiber with reduced linear birefringence may include a plurality of fiber sections, and a twist located between adjacent sections. The sense of rotation of the twist between sections along the optical fiber may be reversed after an accumulated twist between previous fiber sections is substantially equal to or greater than 90°. The distance over which the accumulated angle is 90° may preferably be less than half of the beat length. As is known in the art, a beat length is the distance along the fiber required for two propagating waves having differing velocities to move out of and back into synchronization.
Embodiments of the invention may include one or more of the following features. The twist angle may be less than or equal to 90°, with the twists produced by heating the fiber to a softening point of the fiber material, without affecting the shape of the fiber core. (Twists greater than 90° may be used, although they offer no benefit over twists of less than 90°.) More particularly, the orientation of the fiber is substantially maintained between twists, i.e., the fiber sections themselves are not twisted. The twists may be produced either with a consistent sense of rotation of the twists along the fiber or by reversing the sense of rotation, as described above.
Further features and advantages of the present invention will be apparent from the following description of preferred embodiments and from the claims.
The following figures depict certain illustrative embodiments of the invention in which like reference numerals refer to like elements. These depicted embodiments are to be understood as illustrative of the invention and not as limiting in any way.
By way of background information, consider a linearly polarized wave launched along one of the two residual birefringent axes of an optical fiber. It will have a different propagation velocity than a wave launched in the other orthogonal axis. Although this difference is typically not great, it is sufficient to adversely effect the Faraday effect sensitivity. The sensitivity, S, of a fiber optic coil to an electric current flowing through a wire enclosed by the coil can be shown to be:
S=ρ sin δ/δ,
wherein
From this expression it may be seen that the sensitivity of the coil, all other things held constant, is a sin(x)/x function of the product of the linear birefringence and the length of the coil, multiplied by a factor that is proportional to both the current and the length of the coil. Birefringence itself causes the sensitivity of the coil to vary from the ideal by an unknown amount, and since the birefringence has a stress based component, it has an unknown temperature dependence.
Referring now to
At first, the introduction of discrete rotations at the splice 16 may appear to be a piecewise approximation to the aforedescribed technique of spinning the fiber as it is drawn. However, with the technique of the present invention, any well-made single mode fiber may be utilized, since discrete rotations may be introduced by either an approximately 90° splice or a melted twist as described in the co-pending U.S. patent application Ser. No. 09/337,223 to Dyott, having a filing date of Jun. 22, 1999, which is incorporated herein by reference. Accordingly, the method of the invention may be considerably less costly than making special fiber. It is noted that optical fiber 10 may typically include a coating (not shown) which may be removed from fiber 10 in the area of splice 16, such that the twist can be made. Subsequently, the coating may need to be replaced. In the context of the present invention, the word “twist” refers to a rotation of the fiber in the longitudinal direction about its core by either a splice or a melted twist.
Successive twists generally of approximately 90° may continue in the same sense of rotation or reverse rotation, since the fiber generally has 180° rotational symmetry, so that an approximately 90° rotation has the same effect as an approximately −90° rotation. In other words, the fiber in this embodiment of the present invention may have two identifiable orthogonal axes of linear polarization arising out of the residual ovality of the core or residual stresses from the drawing process, and over the very short distance involved, these axes may not rotate with respect to the fiber itself. An approximately 90° twist is effective as it introduces the maximum angular change per twist.
If this occurs often enough, it may minimize the effects of linear birefringence and may stabilize the fiber sensitivity to the Faraday effect. In the embodiments of the twist described here, it is clear that the relatively “abrupt” junction between the fiber sections results in interchange of energy between the modes. The junction may preferably be long enough so as not to cause reflection and scattering of the propagating light, but also preferably not be so long that it can be considered adiabatic and slowly rotates the plane of polarization.
It may not be sufficient to use only one twist at the center. This depends on the differential retardance (birefringence) between the modes. It is desirable to restrict the cumulative retardance along the fiber to a fraction of a beat length; that is, the cumulative retardance may preferably be less than 180° in any section of fiber length. The smaller the total retardance per fiber section, the better the expected performance may be. Fibers may have from about 10° to several thousand degrees of retardance per meter. More likely, however, retardance may be much less than 500°/m for typical modem single-mode communication fibers. By using three twists, i.e., by creating four sections, the cumulative retardance may be limited to that of one quarter of the length of the entire fiber. The number of twists per meter may likely depend on the characteristic properties of the fiber.
It has been found that, in using torsion-type twists, two approximately 360° turns for a fiber having a measured retardance of about 100°/m (considered a large value), may be sufficient to render the fiber insensitive to polarization effects. In stress torsion induced circular birefringence, the magnitude of the coupling effect depends on the relative values of the circular and linear birefringence. While the effect of the torsion-type twists and the twist method of the present invention may not be exactly equal, it is unlikely that the circular birefringence due to practical torsion twist rates will have an effect greater than the twist technique of the present invention. Using the twist technique of the present invention, the two whole (approximately 360°) turns could be accomplished with 8 twists, creating 9 fiber sections.
Modern communications fibers may have a typical beat length of 6 meters, implying a residual retardance of approximately 60°/m. As such, only about 4 sections may be needed. The sections need be only approximately the same length, and the twist angles only approximate to achieve the desired result.
To increase the effectiveness of the method even further, the twist technique may be combined with other techniques, such as annealing or a toroidal coil (U.S. Pat. No. 6,023,331 to Blake et al.) or a Berry's phase coil.
While the invention has been disclosed in connection with the preferred embodiments shown and described in detail, various modifications and improvements thereon will become readily apparent to those skilled in the art. Accordingly, the spirit and scope of the present invention is to be limited only by the following claims.
This application claims priority to, and incorporates by reference, the entire disclosure of U.S. Provisional Patent Application No. 60/222,389, filed on Aug. 2, 2000.
Number | Name | Date | Kind |
---|---|---|---|
4571650 | Ojima et al. | Feb 1986 | A |
4603931 | Ruffman | Aug 1986 | A |
4615582 | Lefevre et al. | Oct 1986 | A |
4630229 | D'Hondt | Dec 1986 | A |
4630890 | Ashkin et al. | Dec 1986 | A |
4637722 | Kim | Jan 1987 | A |
4668264 | Dyott | May 1987 | A |
4669814 | Dyott | Jun 1987 | A |
4697876 | Dyott | Oct 1987 | A |
4712866 | Dyott | Dec 1987 | A |
4733938 | Lefevre et al. | Mar 1988 | A |
4740085 | Lim | Apr 1988 | A |
4755021 | Dyott | Jul 1988 | A |
4756589 | Bricheno et al. | Jul 1988 | A |
4765739 | Koizumi et al. | Aug 1988 | A |
4776700 | Frigo | Oct 1988 | A |
4796993 | Sonobe et al. | Jan 1989 | A |
4798438 | Moore et al. | Jan 1989 | A |
4815817 | Levinson | Mar 1989 | A |
4842409 | Arditty et al. | Jun 1989 | A |
4848910 | Dupraz | Jul 1989 | A |
4883358 | Okada | Nov 1989 | A |
4887900 | Hall | Dec 1989 | A |
4926040 | Turpin et al. | May 1990 | A |
4943132 | Huang | Jul 1990 | A |
5033854 | Matthews et al. | Jul 1991 | A |
5048908 | Blonder et al. | Sep 1991 | A |
5048962 | Kurokawa et al. | Sep 1991 | A |
5074665 | Huang et al. | Dec 1991 | A |
5080489 | Nishikawa et al. | Jan 1992 | A |
5106193 | Fesler et al. | Apr 1992 | A |
5133600 | Schröder | Jul 1992 | A |
5135555 | Coyle, Jr. et al. | Aug 1992 | A |
5289257 | Kurokawa et al. | Feb 1994 | A |
5289258 | Szafraniec et al. | Feb 1994 | A |
5298047 | Hart, Jr. et al. | Mar 1994 | A |
5331404 | Moeller et al. | Jul 1994 | A |
5351123 | Spahlinger | Sep 1994 | A |
5359413 | Chang et al. | Oct 1994 | A |
5365338 | Bramson | Nov 1994 | A |
5412471 | Tada et al. | May 1995 | A |
5418881 | Hart et al. | May 1995 | A |
5459575 | Malvern | Oct 1995 | A |
5469257 | Blake et al. | Nov 1995 | A |
5469267 | Wang | Nov 1995 | A |
5471301 | Kumagai et al. | Nov 1995 | A |
5493396 | Sewell | Feb 1996 | A |
5500909 | Meier | Mar 1996 | A |
5504684 | Lau et al. | Apr 1996 | A |
5552887 | Dyott | Sep 1996 | A |
5559908 | August et al. | Sep 1996 | A |
5654906 | Youngquist | Aug 1997 | A |
5655035 | Burmenko | Aug 1997 | A |
5664037 | Weidman | Sep 1997 | A |
5682241 | Mark et al. | Oct 1997 | A |
5694512 | Gonthier et al. | Dec 1997 | A |
5701177 | Kumagai et al. | Dec 1997 | A |
5701376 | Shirasaki | Dec 1997 | A |
5767509 | Cardova et al. | Jun 1998 | A |
5781675 | Tseng et al. | Jul 1998 | A |
5822487 | Evans et al. | Oct 1998 | A |
5854864 | Knoesen et al. | Dec 1998 | A |
5898496 | Huang et al. | Apr 1999 | A |
5943466 | Henderson et al. | Aug 1999 | A |
5946097 | Sanders et al. | Aug 1999 | A |
5987195 | Blake | Nov 1999 | A |
6023331 | Blake et al. | Feb 2000 | A |
6025915 | Michal et al. | Feb 2000 | A |
6047095 | Knoesen et al. | Apr 2000 | A |
6075915 | Koops et al. | Jun 2000 | A |
6148131 | Geertman | Nov 2000 | A |
6163632 | Rickman et al. | Dec 2000 | A |
6185033 | Bosc et al. | Feb 2001 | B1 |
6233371 | Kim et al. | May 2001 | B1 |
6301400 | Sanders | Oct 2001 | B1 |
6341503 | Miller et al. | Jan 2002 | B1 |
6351310 | Emge et al. | Feb 2002 | B1 |
6370289 | Bennett | Apr 2002 | B1 |
6389185 | Meise et al. | May 2002 | B1 |
6396965 | Anderson | May 2002 | B1 |
6429939 | Bennett et al. | Aug 2002 | B1 |
6519404 | Yoshida et al. | Feb 2003 | B1 |
6535657 | Dyott | Mar 2003 | B1 |
6539134 | Dyott | Mar 2003 | B1 |
6542651 | Bennett | Apr 2003 | B1 |
6563589 | Bennett et al. | May 2003 | B1 |
6594020 | Dyott | Jul 2003 | B1 |
6703821 | Dyott | Mar 2004 | B1 |
6707558 | Bennett | Mar 2004 | B1 |
20020018212 | Bennett et al. | Feb 2002 | A1 |
20020025098 | Dyott | Feb 2002 | A1 |
20020027659 | Dyott | Mar 2002 | A1 |
20020180416 | Dyott | Dec 2002 | A1 |
Number | Date | Country |
---|---|---|
33 05 104 | Aug 1984 | DE |
36 15 305 | Nov 1987 | DE |
36 15305 | Nov 1987 | DE |
37 42 201 | Jun 1989 | DE |
37 42202 | Jun 1989 | DE |
0 551 874 | Jul 1993 | EP |
0 586 242 | Mar 1994 | EP |
0 686 867 | Dec 1995 | EP |
0 722 081 | Jul 1996 | EP |
856 737 | Aug 1998 | EP |
0 871 009 | Oct 1998 | EP |
0 872 756 | Oct 1998 | EP |
10300797 | Nov 1998 | EP |
2 535 463 | May 1984 | FR |
2 535 463 | May 1984 | FR |
07209398 | Aug 1995 | JP |
WO 8300232 | Jan 1983 | WO |
WO 9853352 | Nov 1998 | WO |
WO9858268 | Dec 1998 | WO |
WO0036425 | Jun 2000 | WO |
Number | Date | Country | |
---|---|---|---|
20020031292 A1 | Mar 2002 | US |
Number | Date | Country | |
---|---|---|---|
60222389 | Aug 2000 | US |