Telecentric 1xN optical fiber switches

A telecentric 1×N optical switch according to the present invention switches between output fibers without the need for active alignment by utilizing a telecentric lens group translated in directions perpendicular to the axis of the input fiber. The telecentric lens group directs the input beam to the selected output fiber by translating to a location associated with the selected fiber.

Skip to: Description  ·  Claims  · Patent History  ·  Patent History
Description
BACKGROUND OF THE INVENTION

[0001] 1. Field of the Invention

[0002] The present invention relates to apparatus and methods for coupling an input optical fiber selectively to one of a plurality of output fibers. In particular, the present invention is a 1×N fiber switch.

[0003] 2. Description of the Prior Art

[0004] Currently, there are a number of ways to implement fiber-to-fiber switches, where an input optical fiber is coupled selectively to one of a plurality of output fibers. A first method involves bringing the cut and polished surface of the input fiber into close proximity to the similarly cut and polished end of the desired output fiber. If the fibers' cores (where the light is guided) are positioned closely and accurately enough, most of the light from the input fiber will enter the core of the output fiber. This kind of switch requires accurate positioning of the fibers to a fraction of a micron, if low losses and repeatability are to be accomplished.

[0005] A second switching method involves collimating the light from the input fiber using a lens. The collimated beam is then reflected into a collimator and hence directed into the desired output fiber using a movable mirror. Each output fiber has its own collimator. This type of switch requires each output fiber-collimator to be aligned to a very small fraction of a degree (a few arc seconds) in order to maintain sufficiently low-loss coupling. In addition, the mirrors must accurately reproduce the same output beam angle for each output fiber.

[0006] A third type of switch involves passing the light from the input fiber through an interferometer with two possible outputs, such as a Mach-Zender interferometer. By manipulating the path length of one arm of the interferometer, the input light is directed to either of the two possible outputs. Free-space or fiber interferometers are expensive and must remain stable to a small fraction of a wavelength. Waveguide interferometers require very accurately positioned couplers in order to efficiently couple light from fibers to the waveguide switch and back to the fiber.

[0007] To summarize, all of the known 1×N switching methods require high precision alignment of a number of their optical components. When such switches are to be used with single mode fibers, as are used in optical networking, the required precision of the switch components exceed the accuracy achieved by normal manufacturing processes. Therefore, expensive and time consuming active alignment processes, whereby a component is adjusted while a metric of the output is monitored, are required for each output fiber, often in several stages. Hence, at least N active alignments for each 1×N switch are required for all currently available switches.

[0008] Many uses of 1×N switches, including optical networks, require that the loss be substantially the same for each of the N outputs, and further that this loss be accurately repeated each time the switch is set to each output. For use in optical networks, it is further required that the switch retain this uniformity and repeatability over a substantial environmental temperature range.

[0009] A need remains in the art for a 1×N optical fiber switch which does not require active alignment steps for each output fiber, and which has uniform and predictable losses.

SUMMARY OF THE INVENTION

[0010] An object of the invention is to provide 1×N optical fiber switches which do not require active alignment steps for each output fiber, and which have uniform and predictable losses. A 1×N optical switch according to the present invention switches between output fibers by translating a telecentric optical element in a plane perpendicular to the optical axis of the input fiber. The telecentric optical element is preferably a compound lens group. It collects the light emitted from the input fiber and focuses it on a selected output fiber in the output fiber bundle. The telecentric optical element is telecentric on both the input and the output sides.

[0011] This invention describes a 1×N optical fiber switch whose construction does not require placement accuracies greater than what is common in ordinary manufacturing. Thus, this switch may be built economically using normal pick and place or fixturing techniques common in automated manufacturing. The number of outputs, N, of the switch may be a very large number such as 1000 or even more. In addition, the switch described by this invention can achieve coupling efficiencies and repeatability equaling or exceeding the state of the art in optical switches.

[0012] A telecentric 1×N optical switch for coupling an input beam from an input fiber to a selected one of N output fibers includes a telecentric optical element and means for translating the telecentric optical element to a specific one of a group of predetermined locations to direct the beam from the input fiber to a selected output fiber associated with the specific location. The means for translating translates the telecentric optical element in directions perpendicular to the axis of the input fiber. The means for translating includes a computer for controlling translation and at least one driver for inducing translation responsive to the computer. In general the means for translating comprises a horizontal driver and a vertical driver.

[0013] In the preferred embodiment, the computer further comprises means for determining the predetermined locations in a test configuration and means for storing the predetermined locations.

[0014] Generally, the telecentric optical element comprises a lens group. For example, the lens group might comprise six lenses.

BRIEF DESCRIPTION OF THE DRAWINGS

[0015] FIG. 1 is an isometric cutaway view showing the preferred embodiment of a 1×N telecentric switch according to the present invention.

[0016] FIG. 2 is a side cutaway view showing the telecentric lens group of FIG. 1.

[0017] FIG. 3 is a flow diagram illustrating an automatic process for determining locations for the telecentric lens group of FIGS. 1 and 2 to select ouput fibers.

[0018] FIG. 4 is a block diagram illustrating the set up for executing the process of FIG. 3.

[0019] FIG. 5 is a table of specifications of an example telecentric lens group of FIG. 1.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0020] The present invention relates to apparatus and methods for coupling an input optical fiber 102 selectively to one of a plurality of output fibers 104. A 1×N optical switch 114 according to the present invention switches between output fibers using a telecentric optical element, such as a telecentric lens group 106.

[0021] As shown in FIG. 1, coupling between an input fiber 102 and a bundle of output fibers 104 is accomplished by translating a compound lens group 106 in a plane perpendicular to the optical axis 112. Lens group 106 collects the light emitted from input fiber 102 and focuses it on a selected output fiber 104 in the output fiber bundle.

[0022] In order for lens group 106 to efficiently couple the light from input fiber 102 to the selected output fiber 104 (especially if the fibers are single-mode fibers) lens group 106 must be telecentric on both the input and output sides. What this means, as shown in FIG. 2, is that a ray input that is parallel to the optical axis of the lens will always result in an output ray that is also parallel to the optical axis, regardless of whether the input ray is translated away from the axis or not.

[0023] In order to provide coupling to a bundle of fibers, both horizontal and vertical displacement are required. Hence, horizontal motion of lens group 106 is accomplished by applying horizontal stage driver 118 to horizontal translation stage element 108. Vertical motion of lens group 106 is accomplished by applying vertical stage driver 120 to vertical translation stage element 110. See FIG. 5 for one possible configuration of lens group 106.

[0024] In FIG. 2, a cone of rays parallel to the optical axis but input off-axis at input position 3, results in an output cone of rays that is also parallel to the optical axis at output position 3.

[0025] Thus, for a single input fiber 102, the cone of light that is output remains parallel to optical axis 112, even as lens group 106 is translated perpendicular to optical axis 112. Furthermore, the output cone is translated perpendicular at a rate twice as fast as the lens group. By means of appropriate translations of lens group 106, the light from input fiber 102 can be coupled efficiently into any selected output fiber 104.

[0026] As an example, telecentric lens group 106 might comprise six lenses 202, 204, 206, 208, 210, and 212.

[0027] In the preferred embodiment, switch 114 is computer controlled (for example by means of an on-board microprocessor). The lens group positions need not be specified in advance—in the preferred embodiment, the lens is calibrated by an automatic process after construction. FIG. 3 is a flow diagram illustrating an example of such a process. Refer also to FIG. 4, which illustrates the alignment configuration. The steps of the process are:

[0028] In step 302, the sharp focus of lens group 106 is temporarily spoiled, for example by placing a plane-parallel glass plate 408 between lens group 106 and either output bundle 104 or input fiber 102. This spreads the focal point out so that a coarse search pattern can find the approximate position for coupling to each output. This step is optional. If it is used, steps 308 and 310 may be repeated with the spoiling removed, as indicated by arrow 309, for very precise results.

[0029] In step 304, a source 402 is connected to input fiber 102. In step 306 detectors 404 are placed adjacent to output fibers 104 to detect the amount of light appearing at each output during step 308. (Alternatively, a single detector can be switched to each of the outputs in turn.)

[0030] In step 308, computer 406 scans lens group 106 (via control signals 412 and 414, to translation stage elements 110 and 108) in a search pattern to find the best coupling position for each of the outputs. Computer 406 monitors detectors 404 via signals 410 to determine how much light is appearing at each output fiber 104 at the various search pattern locations. As the search pattern is executed, computer 406 stores the locations of best coupling to each of the output fibers in memory in step 310.

[0031] Step 312 indicates that the positioning may be detuned for some output fibers 104 in order to end up with equivalent loss at each output fiber. These detuned positions would replace the corresponding locations previously stored in step 310.

[0032] In use, test source 402 is removed and the actual input signal is coupled to input fiber 102. Detector array 404 is removed and replaced with the appropriate output coupling. Detuning plate 408 has been removed. Computer 406 still controls horizontal and vertical movement of lens group 106, according to the stored location for each output fiber. When switch 114 is commanded to select a particular output, it goes to the stored location for that output. If desired, a feedback signal may may used to maintain accuracy.

[0033] FIG. 5 is a table of specifications of an example telecentric lens group of FIGS. 1 and 2. Surface numbers and lens numbers are as labeled in FIG. 2. Thus, for example, lens 1 is lens 202 in FIG. 2, and has first surface 1 and second surface 2. Telecentric lens groups are known in the art, and the group specified in FIG. 5 is only one example.

Claims

1. A telecentric 1×N optical switch for coupling an input beam from an input fiber to a selected one of N output fibers comprising:

a telecentric optical element; and
means for translating the telecentric optical element to a specific one of a group of predetermined locations to direct the beam from the input fiber to a selected output fiber associated with the specific location;
wherein the means for translating translates the telecentric optical element in directions perpendicular to the axis of the input fiber.

2. The switch of claim 1 wherein the means for translating includes a computer for controlling translation and at least one driver for inducing translation responsive to the computer.

3. The switch of claim 2 wherein the means for translating comprises a horizontal driver and a vertical driver.

4. The switch of claim 2, wherein the computer further comprises means for determining the predetermined locations in a test configuration and means for storing the predetermined locations.

5. The switch of claim 1 wherein the telecentric optical element comprises a lens group.

6. The switch of claim 5 wherein the lens group comprises six lenses.

7. The method for switching an input beam from an input fiber to a selected one of N output fibers comprising the steps of:

coupling the input beam to a telecentric optical element; and
translating the telecentric optical element to a specific one of a group of predetermined locations to direct the input beam from the input fiber to a selected output fiber associated with the specific location;
wherein the translating step translates the telecentric optical element in directions perpendicular to the axis of the input fiber.

8. The method of claim 7 wherein the translating step includes the step of generating control signals and the step of inducing translation responsive to the control signals.

9. The method of claim 8 wherein the translating step translates horizontally and vertically.

10. The method of claim 8, further including the step of determining the predetermined locations in a test configuration and the step of storing the predetermined locations.

11. The method of claim 10, wherein the step of determining further includes the steps of:

spoiling the focus of the optical element prior to determining the locations a first time;
unspoiling the focus of the optical element after determining the locations a first time; and
determining the locations a second time after unspoiling the focus of the optical element.

12. The method of claim 10, further including the step of modifying some of the locations to detune some of the couplings between the input fiber and the output fibers.

13. The switch of claim 7 wherein the telecentric optical element comprises a lens group.

14. A telecentric 1×N optical switch for coupling an input beam from an input fiber to a selected one of N output fibers comprising:

a telecentric lens group;
a driver element for translating the telecentric lens group to a specific one of a group of predetermined locations to direct the beam from the input fiber to a selected output fiber associated with the specific location; and
a computer for selecting the specific location;
wherein the driver element translates the telecentric lens group in directions perpendicular to the axis of the input fiber, responsive to the computer.

15. The switch of claim 14 wherein the driver element comprises a horizontal driver and a vertical driver.

16. The switch of claim 14, wherein the computer further comprises means for determining the predetermined locations in a test configuration and means for storing the predetermined locations.

17. The switch of claim 14 wherein the lens group comprises six lenses.

Patent History
Publication number: 20030210856
Type: Application
Filed: May 10, 2002
Publication Date: Nov 13, 2003
Inventor: Robert H. Cormack (Boulder, CO)
Application Number: 10144209
Classifications
Current U.S. Class: Single Pole Multiple Throw (relay Switch) (385/22)
International Classification: G02B006/35;