KR20010085809A - Flexible, modular, compact fiber optic switch - Google Patents

Abstract

According to the present invention, the optical fiber switch 400 has an optical fiber conversion module 100 for receiving and fixing the ends 104 of the optical fibers 106. The optical fiber conversion module 100 includes a plurality of reflective light beam deflectors 172 that can be selected as pairs for coupling the light beam 108 between the pair of optical fibers 106. The fiber optic switching module 100 also generates direction signals from each reflected light beam deflector 172 which indicates the direction of the fiber optic switching module. The port card 406 provided in the optical fiber switch 400 supplies drive signals to the optical fiber conversion module 100 for directing the at least one reflected light beam deflector 172. The port card 406 also receives direction signals generated by the reflective light beam deflector 172 with coordinates that specify a direction for the reflective light beam deflector 172. The port card 406 compares the received coordinates with the direction signals received from the reflected light beam deflector 172 and adjusts the drive signals supplied to the optical fiber conversion module 100 to receive the received coordinates and direction signals. Reduces any difference between The optical fiber switch 400 also employs light alignment to precisely orient the pairs of reflective light beam deflectors 172 that couple the light beam 108 between the optical fibers 106.

Description

Flexible Modular Compact Fiber Optic Switch {FLEXIBLE, MODULAR, COMPACT FIBER OPTIC SWITCH}

In recent years, the dramatic increase in telecommunications, which can have a significant impact on increasing Internet communications, requires the rapid introduction and commercial acceptance of technological innovations in fiber optic wired communication systems. For example, fiber optic communication systems have recently been introduced and installed to deliver digital telecommunications simultaneously with 4, 16, 32, 64 or 128 different light wavelengths propagating along a single fiber. While multi-wavelength fiber telecommunications have dramatically increased, while the bandwidth of a single fiber has increased, it is only available at both ends of the fiber, for example at two locations. When the light transmitted into one end of the optical fiber reaches the other end of the optical fiber, it is now possible to automatically propagate the light received at one end of the optical fiber onto a selected one of several different optical fibers, and also to receive the received light. There was no flexible, modular, compact N × N array of fiber optic switches to other destinations.

Historically, when telecommunications is communicated by electrical signals through pairs of copper wires, a telephone operator who is sitting on a manually operated telephone switchboard calls a pair of copper wires attached to a plug while simultaneously The incoming telephone call is physically connected to another pair of copper wires attached to the socket to complete the telephone circuit. The telephone operator's task of manually interconnecting pairs of lines from two telephones to achieve a telephone circuit is initially called a crossbar switch, which automates the operator's manual operation in response to telephone dialing signals. Replaced by an electromechanical device. For over 40 years, electromechanical crossbar switches for telecommunications have been replaced by electronic switching systems.

Presently, switches exist for fiber optic telephony and perform functions for optical fiber telephony similar or identical to electronic switching systems that perform crossbar switches and electrical telephony. However, currently available fiber optic switches are by no means ideal. In other words, existing optical fiber telecommunications technology is lacking as a switch and performs the same function for optical telecommunication as performed by an electronic switching system for a plurality of optical fibers.

One approach used in providing a 256 × 256 array of switches for fiber optic telecommunications first converts light received from an incoming fiber portion into an electrical signal and then transmits the electrical signal through an electronic switching network. The output signal from the electron switching network is used to generate a second beam of light and is passed into the output optical fiber portion. As will be appreciated by those skilled in the art of electronic and fiber optic communications, this approach to providing a 256 × 256 array of optical fiber switches is physically very large and requires electrical circuits that process electronic signals at extremely high speeds and are very expensive.

There are various limitations in complex electronic signals, and attempts to avoid switching between optical and electronic signals, for assembling optical fiber switches and coupling light beams directly from one optical fiber to another. An initial attempt to provide a fiber optic switch similar to an electric crossbar switch is to mimic the behavior of a telephone operator with machinery having only fiber optics rather than pairs of copper wires. US Patent No. 4,886,335, entitled "Optical Fiber Switch System", issued December 12, 1989, includes a conveyor and moves ferrules attached to both ends of the optical fiber. The conveyor moves the ferrule to the selected adapter and inserts the ferrule into the coupler / decoupler provided in the selected adapter. After the ferrule is inserted into the combiner / remover, light is transferred between the optical fiber extending in the ferrule and the optical fiber securely stored in the selected adapter.

The invention is entitled “Miniature 1 × N Electromechanical Optical Switch And Variable Attenuator” and US Patent No. 5,864,463 (hereinafter referred to as “463 patent”), issued January 26, 1999, refers to one optical fiber and multiple optical fibers. Another mechanical system for selectively coupling light between one optical fiber is disclosed. This US patent discloses a method of selectively coupling light between selected optical fibers by mechanically moving one end of one optical fiber along a linear array of both ends of one optical fiber and another optical fiber. A 1 × N array of switches uses a mechanical actuator to align one end of one optical fiber to a selected end of another optical fiber within 10 μm. The 1 × N array of switches uses light reflected back into the optical fiber moved from the immediately adjacent end of the selected optical fiber to more precisely align the ends of the input optical fiber with the input optical fiber. U.S. Patent No. 5,699,463, entitled "Mechanical Fiber Optic Switch," issued December 16, 1997, also aligns one end of one optical fiber with one of several other optical fibers assembled as a linear array. A method of superimposing a lens between both ends of two optical fibers is disclosed.

U.S. Patent No. 5,524,153 (hereinafter referred to as "153 Patent"), entitled "Optical Fiber Switching System And Method Of Using Same," issued June 4, 1996, describes two optical fiber switching devices adjacent to each other. A method of arranging two optically reversed groups is disclosed. Each switching device may align a predetermined one of the optical fibers of each switching device with a predetermined optical fiber of the optically inverted group of the switching devices. In the switching device, one end of each optical fiber is located adjacent to a beamforming lens and is received by a two-axis piezoelectric bender. The biaxial piezoelectric bender can bend the fiber so that light is emitted from one point of the fiber of a given optical fiber in the optically inverted group of the switching device. Pulsed light generated by a radiation emitting device ("RED") coupled to each optical fiber is transmitted from the fiber to the optical fiber selected from the group that is optically inverted. Pulsed light from the RED received by the optical fiber selected from the optically inverted group is processed to provide a signal and fed back to the piezoelectric bender for directing light from the optical fiber directly in the selected optical fiber.

Rather than mechanically affecting the alignment of light beams from one optical fiber to another by carrying out optical fiber or by bending one or both optical fibers, an optical switch has been proposed, which is an input. A micromachined moving mirror array is employed to selectively combine the light emitted from the optical fiber into the output optical fiber. 1 illustrates a device that can be used to fabricate a three stage, totally uninterrupted fiber optic switch disclosed in a paper presented to the OFC / IOOC of 1999, held February 21-26, 1999. The fiber optic switch employs a moving mirror array in which each polycrystalline silicon mirror can selectively reflect light at 90 °. In this proposed optical fiber switch, a relatively small 32 × 64 array of light switching arrays 52a _i (i = 1, 2,…) and 52b _k (k = 1, 2,…)] has 32 inputs or light to the output optical fibers (54a _n and 54b _n) receive and 32 input or output optical fibers (54a _n and 54b _n) from the light passes. 32 groups of 64 optical fibers 56a _{l, m} and 56b _{l, m transmit} light to each 32 × 64 array of light conversion arrays 52a _i and 52b _k and 32 × 32 array of light conversion arrays [58 _j (j = 1, 2, ... 64)].

Referring to FIG. 1, the complexity of the optical fiber switch will be readily apparent. For example, a 1024 × 1024 array of optical fiber switches assembled in accordance with the above proposal is mutually connected between the 32 × 64 array of light switching arrays 52a _i and 52b _k and the 32 × 32 array of optical switching arrays 58 _j . It requires 4096 individual optical fibers to connect. In addition, the 32 × 64 array of light conversion arrays 52a _i and 52b _k and the 32 × 32 array of light conversion array 58 _j require a total of 196,608 micromachined mirrors.

The polycrystalline silicon mirror proposed for the optical fiber switch shown in FIG. 1 is curved rather than optically smooth. In addition, these mirrors have the appropriate heat dissipation ability to convert a single 0.3 ㎽ wavelength of light and some of these wavelengths, while they cannot switch 10 or 20 such wavelengths. However, as noted above, fiber optic communication systems are already delivering more than 20 wavelengths over a single optical fiber, and if not so far, will soon deliver hundreds of wavelengths. If instead of a single wavelength of light, one optical fiber delivers 300 different wavelengths of light, each with 0.3 kW of power, then 100 kW of power reaches the polycrystalline silicon mirror proposed for this fiber switch. If the polycrystalline silicon mirror reflects 98.5% of the light, the polycrystalline silicon mirror should absorb substantially all of the rest, i.e. 1.5 kW of power. When absorbing 1.5 kW of power, the thermally nonconductive polycrystalline silicon is likely to be heated to an unacceptable temperature and also degrades the smoothness of the mirror.

FIELD OF THE INVENTION The present invention generally relates to the technical field of optical fibers, and more particularly to the free space of optical fiber switches in a reflective N × N arrangement.

1 is a block diagram illustrating a three stage, totally uninterrupted fiber optic switch of the proposed prior art;

2 is a ray tracing plan view showing a shape in which a light beam propagates through a reflection switching module of a converging light N × N array, which is a trapezoidal free space according to the present invention;

3 is a schematic plan or schematic elevation view showing a single light beam that can propagate between the A side and B side of the converging light N × N arrangement of the trapezoidal free space shown in FIG. 2 according to the present invention; .

FIG. 4A is a ray tracing perspective view showing a shape of propagating a light beam through a reflection switching module of a converging light N × N array in a rectangular free space as another embodiment according to the present invention; FIG.

FIG. 4B is a ray tracing top view showing the shape of propagating a converging light beam through the rectangular reflection conversion module shown in FIG. 4A in accordance with the present invention; FIG.

FIG. 5 is a ray tracing plan view showing a shape of propagating a light beam through a reflection switching module of a converging light N × N array which is a free space of a polygon as another embodiment according to the present invention;

FIG. 6 is a ray tracing plan view illustrating the shape of propagation of a light beam through a convergent light reflection conversion module, which is a trapezoidal free space according to the present invention, capable of coupling a light beam between any arbitrary selected pair of optical fibers;

FIG. 7 is a further embodiment of the present invention, which is more compact than the N × N array reflection switching module shown in FIG. 5, which propagates a light beam through a reflection switching module of the converging light N × N array, which is a trapezoidal free space. Ray tracing top view showing geometry.

FIG. 8A is an elevation view showing a preferred cylindrical microlens adapted for use in a reflection switching module in an N × N arrangement. FIG.

8B is an elevation view showing a microlens adapted for use in an N × N array reflection conversion module that may be located more densely between lenses and fibers;

FIG. 9 is a partial cross sectional view showing a block provided on both the A side and the B side of the N × N array reflection conversion module shown in FIG. 7 containing a tapered optical fiber collimator assembly; FIG.

FIG. 10 is a partial cross sectional view of the block shown in FIG. 9 containing a tapered optical fiber collimator assembly; FIG.

FIG. 11 is a partial cross sectional view showing a microlens adapted for use in an N × N array reflection switching module for simultaneously converting light transmitted by a double pair of optical fibers; FIG.

12 is an elevational view showing a preferred type of silicon wafer substrate used when making a torsional scanner.

FIG. 13 shows an electrostatically activated 2D torsional scanner specifically adapted for use with reflection conversion modules such as these reflection conversion modules shown in FIGS. 2, 4A and 4B, 5, 6 and 7. Floor plan.

14 is an enlarged plan view showing a torsionally flexible hinge used in a torsion scanner taken along line 14-14 of FIG.

FIG. 15A is a schematic cross sectional view showing a torsional scanner disposed on an insulating substrate having an electrode applied on the insulating substrate by a light beam so as not to reflect and vanish the mirror surface located at the rear of the device layer; FIG.

15B and 15C are another plan views showing the electrodes and a part of the insulating substrate taken along the lines 15b / 15c-15b / 15c.

FIG. 16A is an elevation view showing a strip of torsion scanner adapted for use in a reflection conversion module such as these reflection conversion modules shown in FIGS. 2, 4A and 4B, 5, 6 and 7;

16B is a cross-sectional view taken along line 16B-16B shown in FIG. 16A showing overlapping immediately adjacent strips of a torsion scanner to reduce the horizontal distance between immediately adjacent strips.

FIG. 16C is an elevation view showing a preferred strip of a torsion scanner adapted for use in a reflection conversion module such as these reflection conversion modules shown in FIGS. 2, 4A and 4B, 5, 6 and 7;

16D is a cross sectional view showing a preferred strip of a torsion scanner taken along line 16d-16d shown in FIG. 16C.

16E is a cross-sectional view taken along line 16d-16d shown in FIG. 16A showing a parallel arrangement for the strip of the torsion scanner shown in FIG. 16C.

FIG. 17A shows a vertical deviation strip of a torsion scanner capable of more densely arranging optical fibers in reflection switching modules such as these reflection switching modules shown in FIGS. 2, 4A and 4B, 5, 6 and 7. Floor plan.

FIG. 17B is a plan view illustrating a more dense mounting of deviation rows or columns of a torsion scanner that may be employed when all torsion scanners are manufactured as a single monolithic array rather than on a strip. FIG.

18A is a plan view illustrating another embodiment of a torsion scanner with an external torsionally flexible hinge facing diagonally with respect to the torsion scanner outer frame.

18B is a plan view showing an array of torsion scanners of the type shown in FIG. 18A.

FIG. 19A is a plan view showing another embodiment of a torsion scanner, wherein an internal torsionally flexible hinge is directed along the diagonal of the non-square mirror plate of the scanner;

FIG. 19B is a plan view illustrating another embodiment of the torsion scanner shown in FIG. 19A with two pairs of torsional flexible hinges properly oriented with respect to the crystallographic direction of silicon such that a torsion sensor in a torsion scanner having optimal properties can be fabricated.

FIG. 20A shows a dense arrangement of the torsion scanner shown in FIG. 18A adapted to enclose in a reflection diverting module such as these reflection diverting modules shown in FIGS. 2, 4A and 4B, 5, 6 and 7. Elevation too.

FIG. 20B shows a dense arrangement of the torsion scanner shown in FIG. 19A adapted to enclose in a reflection diverting module such as these reflection diverting modules shown in FIGS. 2, 4A and 4B, 5, 6 and 7. Elevation too.

FIG. 21 is a schematic cross-sectional view illustrating a strip of another embodiment of a torsion scanner secured to a substrate by moving the mirror strip to permit an arrangement in which the ends of the collimator lens and the optical fiber are positioned adjacent to the mirror surface on the torsion scanner.

22A is a front view showing a strip of a torsion scanner flip-chip bonded to a substrate.

FIG. 22B is a cross sectional view showing a strip of torsion scanner flip-chip bonded to a substrate taken along line 22b-22b shown in FIG. 22A.

FIG. 22C shows a strip of torsion scanner flip-chip bonded to a substrate taken along line 22c-22c shown in FIG. 22A.

FIG. 22D is a cross-sectional view illustrating a strip of torsion scanner flip-chip bonded to a silicon substrate having vias formed through the silicon substrate. FIG.

FIG. 23 is a ray tracing diagram illustrating scattering of light from portions of a torsion scanner surrounding a mirror surface of the torsion scanner; FIG.

FIG. 24 is a system level block diagram illustrating reflection switching modules such as these reflection switching modules shown in FIGS. 2, 4A and 4B, 5, 6 and 7;

25 is a perspective view illustrating a modular optical fiber switch in accordance with the present invention.

FIG. 26A is a block diagram illustrating the modular optical fiber switch shown in FIG. 25 with a port card and reflection switching module. FIG.

FIG. 26B illustrates one embodiment of a light detector that may be used in a light alignment servo that finely directs mirror pairs provided in the reflection conversion module. FIG.

FIG. 26C illustrates a composite light detector that can be used in a light alignment servo that finely directs mirror pairs provided in the reflection conversion module. FIG.

FIG. 27A illustrates a mirror provided in the reflection switching module included in the modular optical fiber switch shown in FIG. 25, such as one of the reflection switching modules shown in FIGS. 2, 4A and 4B, 5, 6, and 7. Block diagram showing a servo system that guarantees fine alignment.

FIG. 27B is a block diagram showing one channel (either x-axis or y-axis) of the dual-axis servo provided in the servo system shown in FIG. 27A; FIG.

FIG. 28A is a partial cross sectional view showing a double plate structure as another embodiment for receiving and fixing an optical fiber array; FIG.

FIG. 28B is an elevation view showing a profile for one type of hole that may be formed through one of the plates taken along line 28b-28b shown in FIG. 28A.

FIG. 28C is an elevation view showing an array of XY microstages formed through one of the plates taken along line 28c-28c shown in FIG. 28A.

FIG. 29A is an elevation view showing an XY microstage of the type having an array taken along lines 29a-29a shown in FIG. 28C.

29B and 29C are elevation views showing a portion of an XY microstage, which is another embodiment taken along the lines 29b / 29c-29b / 29c shown in FIG. 29A;

30A is a partial cross-sectional view showing a micromachined lens from a silicon substrate that can be electrostatically activated to move along the longitudinal axis of the lens.

30B is an elevational view showing a silicon micromachined lens taken along line 30b-30b shown in FIG. 30A.

FIG. 30C is a partial cross sectional view showing a micromachined lens from a silicon substrate similar to the lens shown in FIG. 30A, which may be electrostatically activated to move along the longitudinal axis of the lens; FIG.

FIG. 31 is an elevational view illustrating a method of directly coupling a light beam into one of the reflection conversion modules shown in FIGS. 2, 4A and 4B, 5, 6, and 7 from a routing wavelength demultiplexer.

The present invention provides an optical fiber switch capable of simultaneously coupling an incident beam of light transmitted on 1000 or more individual optical fibers to 1000 or more output optical fibers.

It is an object of the present invention to provide a simpler optical fiber switch which is switchable between a plurality of incident and exit beams of light transmitted on an optical fiber.

Another object of the present invention is to provide an efficient optical fiber switch that can switch between a plurality of incident and exit beams of light transmitted on an optical fiber.

It is still another object of the present invention to provide an optical fiber switch having low crosstalk between communication channels.

It is another object of the present invention to provide an optical fiber switch having low crosstalk between communication channels while the optical fiber switch is switched.

Another object of the present invention is to provide a high reliability optical fiber switch.

Another object of the present invention is to provide an optical fiber switch which does not exhibit dispersion.

It is another object of the present invention to provide an optical fiber switch that does not depend on polarization.

It is still another object of the present invention to provide an optically transparent fiber switch.

It is still another object of the present invention to provide an optical fiber switch which does not limit the bit rate for optical fiber telecommunications transmitted through the optical fiber switch.

Briefly, the present invention relates to an optical fiber switch, comprising an optical fiber conversion module for receiving and fixing both ends of an optical fiber. Further, in receiving and fixing both ends of the optical fiber, the optical fiber conversion module includes a plurality of reflective light beam deflectors, and is directed in response to a drive signal for coupling the light beam between pairs of optical fibers fixed to the optical fiber conversion module. May be selected as pairs. The fiber optic switching module also generates direction signals from each light beam deflector, indicating the direction of the light beam.

Further, in the optical fiber switching module, the optical fiber switch also has at least one postcard for supplying a drive signal to the optical fiber switching module directed to the at least one light beam deflector provided in the optical fiber switching module. The port card also receives direction signals generated by the light beam deflector together with coordinates specifying the direction for the light beam deflector. The port card compares the received coordinates with the direction signals received from the light beam deflector and adjusts the drive signal supplied to the optical fiber conversion module to reduce the predetermined difference between the received coordinates and the direction signal.

In a preferred embodiment, the optical fiber switching module of the optical fiber switch has first and second optical fiber storage groups separated from each other at opposite ends of the free space optical path. Each of these fiber optic storage groups is adapted to receive and secure both ends of the optical fiber. The optical fiber conversion module has lenses disposed in parallel at both ends of the optical fiber fixed to the first and second optical fiber storage groups, respectively, and along the optical path between the first and second optical fiber storage groups. Since each of these lenses is disposed with respect to one end of the combined optical fiber of the first and second optical fiber storage groups, respectively, the light beam can be transmitted through the immediately adjacent lenses and emitted from one end of the optical fiber, so that the second or first It propagates a quasi-balance beam through the optical path from the lenses on the optical fiber reservoir group side.

In a preferred embodiment, the optical fiber switch also has a set of first and second reflected light beam deflectors disposed on both sides along the optical path between the optical fiber reservoir groups. Each light beam deflector set is coupled to one of the optical fiber reservoir groups, and each light beam deflector set has the same plurality of light beam deflectors as the optical fibers in the combined optical fiber reservoir group. Each light beam deflector in the first or second set of reflective light beam deflectors is:

1. is combined with one of the optical fibers in the combined group of optical fiber reservoirs;

2. Follow a light path that can be emitted from a lens coupled with an optical fiber that strikes when the semi-balanced light beam is reflected from the light beam deflector through the light path;

3. Supply to a fiber optic conversion module oriented to reflect a semi-balanced light beam that can be emitted from the combined optical fiber so as not to be extinguished by being reflected by a selected one of the light beam deflectors in the second or first reflective deflector set. It can be activated by the drive signal.

The light beam deflector pair in this manner includes a light beam deflector belonging to one of the light beam deflector pairs belonging to the first set of reflective light beam deflectors and one light beam deflector belonging to the second set of reflected light beam deflectors. A pair of light beam deflectors activated into a selected one of the optical fiber reservoirs for entering the optical fiber which can be selected and oriented by the supplied drive signal and which can be fixed to the second optical fiber reservoir group or the first optical fiber reservoir group of the optical fiber reservoir. Combining the quasi-equivalent light beam propagating through the optical path from one end of one optical fiber fixed to the optical fiber reservoir so that either of the second optical fiber storage group or the first optical fiber storage group is sequentially reflected to the extinction.

In a preferred embodiment, the port card of the optical fiber switch has a driver circuit for supplying a drive signal to the optical fiber switching module for directing at least one light beam deflector provided in the optical fiber switching module. The portcard also has a dual axis servo that receives coordinates specifying a direction for the light beam deflector, and also receives a direction signal generated by the light beam deflector. The port card compares the received coordinates with the direction signal received by the light beam deflector and adjusts the drive signal supplied to the optical fiber conversion module to reduce the predetermined difference between the received coordinates and the direction signal.

These and other features, objects, and advantages will be understood or apparent to those skilled in the art from the following detailed description of the preferred embodiment, as illustrated with reference to several figures.

Best Mode for Carrying Out the Invention

Free Space,

Converging beam

Double bounce

Reflective switching module

FIG. 2 shows a trapezoidal, converging beam, reflection switching module in a double bound N × N arrangement according to the invention, denoted by the general reference numeral 10. The reflection switching module 100 of the N × N arrangement has sides 102a and 102b and is spaced apart from each other at opposite ends of the C-type free space optical path. Although described below, other geometric relationships to the sides 102a and 102b may occur for other configurations of the reflection switching module 100 in the N × N arrangement, having a C-type free space optical path. For the embodiment of the N × N array reflection switching module 100 shown in FIG. 2, the side surfaces 102a and 102b are preferably coplanar. Both sides 102a and 102b are adapted to the receiving and fixed ends 104 of N optical fibers 106, for example 1152 optical fibers 106. N optical fibers 106 are arranged in a rectangular array of 36 rows each containing 32 optical fibers 106. A lens 112 is disposed immediately adjacent the end 104 of each optical fiber 106 along the optical path between the sides 102a and 102b. Each lens 112 is disposed relative to the end 104 of the optical fiber 106, and is combined to produce light that can be emitted from the end 104 of the optical fiber 106 to which the optical fiber 106 is coupled, and is semi-equilibrium. Light propagates along the optical path between the sides 102a and 102b.

3 shows a single light beam 108 from a single optical fiber 106 that can propagate between sides 102a and 102b or vice versa. For wavelengths of light conventionally used for single mode optical fiber telecommunications, lens 112 is typically a microlens having a focal length of 2.0 to 12.0 mm. This lens 112 is preferably a semi-equilibrium beam having a diameter of approximately 1.5 mm propagating along a long path of 500-900 mm between the sides 102a and 102b. Since the reflection switching module 100 of the N × N arrangement preferably uses the maximum delay length of the lens 112, the ends 104 of each optical fiber 106 may be formed by the focal length of the lens 112 and the optical fiber 106. And add the Raleigh range of the light beam 108 emitted from As a result, if the end 104 of the optical fiber 106 moves by several microns along the lens 112, it is negligible in the direction along the quasi-equilibrium beam with the maximum delay length propagating between the sides 102a and 102b. Produces possible results. Typically the exit angle for a quasi-equilibrium beam having a maximum delay length from lens 112 is 1 in 1 millidian, or 0.001 radians. As described in more detail below, certain possible mismatches of the quasi-equilibrium beam having a maximum delay length due to a mismatch between the end 104 of the optical fiber 106 and the lens 112 may cause the plane to reflect the beam. It can be easily achieved by providing sufficiently large.

Mirror planes 116a and 116b, which are transmitted through the combined lens 112 and then the light beam 108 emitted from the ends of each optical fiber 106 are first combined with the vertical lens 112 and the optical fiber 106 pair. Indicated by the dotted line shown in Fig. 3] and then extinguished. Two-dimensional ("2D") torsion similar to those torsion scanners disclosed in U.S. Patent No. 5,629,790 (hereinafter referred to as "790 patent"), the mirror surface 116 of which is described in more detail below. Preferably provided by a scanner. The reflection switching module 100 of the N × N arrangement is responsible for the two sets 118a and 118b of the mirror surfaces 116 respectively disposed between the lenses 112 according to the light paths between the sides 102a and 102b. Equipped. Each set 118a and 118b has a plurality of individual and independent mirror planes 116, the mirror planes 116 in a pair of gimbals in which each mirror plane 116 can rotate about two non-parallel axes. Is supported by. The number of mirror surfaces 116 is equal to the number N of optical fibers 106 and lenses at the nearest sides 102a and 102b. After reflecting off the mirror surface 116a or 116b and extinguished, the light beam 108 propagating between the sets 118a and 118b shown in FIG. 2 is mirrored along the C-type light path between the sides 102a and 102b. Reflected at a selected one of the faces 116b or 116a and extinguished into one of the optical fibers 106 associated with the vertical lens 112 through one of the lenses 112 at the remote sides 102b or 102a.

4A and 4B show ray tracing for a light beam propagating through a reflection diverting module 100 in a rectangular converging N × N arrangement in another embodiment. The rectangular configuration of the N × N array reflection switching module 100 shown in Figs. 4A and 4B adopts a horizontally thin Z-shaped free space optical path. In this figure, the distance between the side surfaces 102a and the curved set 118a, the curved set 118a and the curved set 118b, and the curved set 118b and the side 102b are substantially the same, Those skilled in the art will appreciate that these distances need not be the same. Those skilled in the art will also recognize that the sets 118a and 118b can be bent to provide either one-dimensional (“1D”) or 2D convergence. Thus, for the configuration of the N × N array reflection conversion module 100 shown in Figs. 4A to 4B, the curved set 118a is moved closer to the side surface 102a, and the curved set 118b is moved. The advantage is that it can be moved closer to the side 102b. The shortening of the distance between these sides 102a and 102b and the curved sets 118a and 118b correspondingly extends the distance between the curved set 118a and the curved set 118b to form a parallelogram N × N arrangement. To generate the reflection conversion module 100. 5 shows ray tracing for a light beam propagating through a reflection switching module 100 of a polygonal N × N arrangement, which is another embodiment. The polygonal configuration of the N × N array reflection conversion module 100 shown in FIG. 5 also creates a Z-shaped free space optical path.

FIG. 6 comprises only one half of the N × N array reflection switching module 100 shown in FIG. 1, that is, the left half of the N × N array reflection switching module or the right half of the N × N array reflection switching module. A trapezoidal reflection conversion module 100 is shown. The reflection switching module 100 shown in FIG. 6 only includes a mirror 120 disposed in the center of the optical path between the sides 102a and 102b, and is essentially different from the reflection switching module shown in FIG. Do. With respect to the equivalent side 102a, the reflection switching module 100 shown in FIG. 6 is selectively selected between only half of the optical fibers 106 as the reflection switching module 100 of the N × N arrangement shown in FIG. While capable of combining light, the reflection conversion module 100 shown in FIG. 6 may combine light between any arbitrarily selected pairs for these optical fibers 106. FIG. 7 shows another trapezoidal N × N array reflection switching module 100 employing a mirror 120 to fold the optical path of the N × N array reflection switching module 100 shown in FIG. 5. . Folding the optical path into a W-type provides a more compact reflection switching module 100 than the N × N array reflection switching module 100 shown in FIG. 1.

Considering the light beam 108 schematically shown in FIG. 3 alone in a perspective view for an optical design, the reflection switching module (described above and shown in FIGS. 2, 4A, 4B, 5, 6 and 7) Some different embodiments for 100 differ primarily in the location of the mirror surfaces 116a and 116b and the folding of the light path along the light beam 108. For example, in the embodiment of the reflection switching module 100 of the N × N arrangement shown in FIGS. 4A and 4B, the mirror surfaces 116a and 116b are closer to the side surfaces 102a and 102b from the lens 112. It is located at approximately 1/3 of the path length between. Contrary to other configurations of the reflection conversion module 100 such as these reflection conversion modules shown in FIGS. 2, 5, 6 and 7, the mirror surfaces 116a and 116b have respective sides 102a and 102b. Adjacent to). However, one of ordinary skill in the art of optical design will appreciate the mirror faces 116a and 116b for the lens 112 and the end 104 of the optical fiber 106 that influence or influence various configurations, particularly other more detailed aspects of the optical design. It will be easy to understand the differences between the locating configurations.

Those skilled in the art of optical design will also appreciate the arrangement of the end 104 of the optical fiber 106 in one or more of the sides 102a and 102b, the combined lens 112 and the mirror surfaces 116a and 116b, respectively. It will be understood that there is a conceptual presence of an unlimited number of other possible geometric arrangements and lightpath shapes as well as the geometric arrangements and lightpath shapes shown in FIGS. 4A, 4B, 5, 6 and 7. For these other geometric arrangements of the free space optical path of the reflection diverting module 100, the priority for one arrangement over other possible arrangements is usually the suitability, size, ease of manufacture, and reflection for a particular light diverting application. Problems with relaxed mechanical resistance, reliability and cost to the assembly of the conversion module 100. Specifically, the reflection conversion module 100 of the trapezoidal converging beam N × N array having the W-type free space optical path shown in FIG. 7 is currently preferable for the following reasons.

1. The reflection conversion module 100 described above is secured within a standard 23 inch wide telecommunication rack.

2. Mechanical resistance is acceptable.

3. The effective delay length for the light beam 108 is long.

4. Operate on well separated electrical and optical cables.

As noted above, the light beam 108 produced by the lens 112 from the light emitted from the end 104 of the combined optical fiber 106 is first of one of the torsion scanners provided in the sets 118a and 118b. Abuts the combined mirror surface 116. As will be explained in more detail below, for the configuration of the reflection switching module 100 in the N × N arrangement shown in FIG. 7, the mirror surface 116 is provided by 36 linear strips of 32 torsional scanners. It is preferable. It is preferred that all 32 mirror surfaces 116 for each strip are substantially coplanar. As an example, within each strip a mirror surface 116 directly adjacent to each other may be disposed 3.2 mm apart, and a light beam in which immediately adjacent rows of the mirror surface 116 touches the mirror surface from immediately adjacent sides 102a and 102b. It is preferred to be spaced 3.2 mm apart with respect to 108.

In addition, for all of the various configurations for the N × N array of reflection switching module 100, the end 104 of the optical fiber 106, the lens 112, and the mirror surface 116 of the deactivated torsional scanner are at the side 102a. Located behind the set 118b of the mirror surfaces 116 with all of the light beams 108 generated by the light emitted from the optical fiber 106 having the ends 104 of the optical fiber 106 directed as described above. It is desirable to converge at point 122b. The light beam 108 emitted from the optical fiber 106 with the end 104 of the optical fiber 106 at the side 102b converges at a point 122a located behind the set 118a of the mirror surface 116. Similarly, the convergence point 122 is achieved by considering the mirror surface 116 at the opposite surface to the sets 118a and 118b. Convergence point 122 is a vertex of the angle at the faces extending from each mirror surface 116 through mirror surface 116 at opposite ends of these two mirror surfaces 116 and the other sets 118b or 118a. The angle is placed at the intersection of two lines that each divide two angles. Convergence point 122 is located at the height of the vertical half of the sets 118a and 118b. The geometric arrangement of the end 104 of the optical fiber 106, the lens 112, and the mirror surface 116 to create a prior convergence, with the same clockwise and counterclockwise rotation angles, in sets 118a or 118b. Sets 118a and 118b exhibiting the greatest movement when reflecting light beam 108 from one mirror surface 116 to a predetermined mirror surface 116 in another set 118b or 118a Provide a minimum angle of rotation with respect to the mirror surface 116. In the configuration for the reflection switching module 100 of the N × N arrangement shown in FIG. 7, when the mirror surface pairs 116a and 116b are separated by 650 mm along the light beam 108, the maximum angular rotation of the mirror surface 116 is achieved. Is approximately 3.9 ° clockwise and counterclockwise.

Although separate pairs of optical fibers 106 and lens 112 may be inserted into the grooves to assemble the sides 102a and 102b causing the convergence of the light beam 108 described above, the lens 112 and the optical fibers 106 For maximum density, it is preferable to use a monolithic block with appropriately drilled holes in the monolithic block. Each predrilled hole receives a conventional fiber optic ferrule securely stored against the end 104 of one of the lenses 112 and one optical fiber 106. The composite angle needed to align the lens 112 for the 2D convergence of the optical fiber 106 and the light beam 108 is provided by properly directing it to the hole drilled into the monolithic block.

FIG. 8A shows a preferred cylindrical microlens 112 that manufactures the focus of the cylindrical microlens as closely as possible or on the face 138 of the lens 112. As will be appreciated by those skilled in the optical fiber art, the optical fiber 106 is polished at an angle that can prevent the back end 104 of the optical fiber 106 from reflecting back from the end 104 of the optical fiber 106. Emits a light beam 108 at an angle about the centerline of the optical fiber 106. Since the end 104 of the optical fiber 106 is bent, the axis of the light beam 108 emitted from the end 104 of the optical fiber 106 converges on the longitudinal axis of the optical fiber 106. To align the longitudinal axis 144 of the lens 112 with the light beam 108, the face 138 of the lens 112 is bent to align the light beam 108 in the lens 112 to the center. In the focus of the lens 112 at the face 138 as described above, the end 104 of the optical fiber 106 is in one rolly range of the light beam 108 from the face 138, for example 50 to 60 μm. Are spaced apart by The diameter of the cylindrical surface 136 of the lens 112 is sufficiently large so that the light converging beam 108 is emitted from the lens 112 through the convex surface 142 through the convex surface 142 as a quasi-balance light beam 108. 108. This configuration for the lens 112 and the end 104 of the optical fiber 106 is described with respect to the longitudinal axis 144 of the lens 112 and the optical fiber collimator assembly 134 at the convex surface 142 of the lens 112. Center the quasi-balanced light beam 108 which is directed essentially parallel to the light beam 108 and the longitudinal axis 144. Conventional manufacturing latitude for the lens 112 described above produces an acceptable deviation in the exit angle and residual offset of the light beam 108 from the longitudinal axis 144 relative to the lens 112. do. For example, if the lens 112 is made of BK7 optical glass, and the end 104 of the optical fiber 106 is angled at 8 °, the angle of the light beam 108 in the lens 112 is 6.78 °, The lateral offset from the longitudinal axis 144 is 50 μm or less on both sides of the face 138 and 140 mm from the face 138. This well-centered light beam 108 can reduce the diameter of the cylindrical surface 136, allowing the lenses 112 to be positioned more closely together. This lens 112 is preferably made of Gradium material sold by LightPath Technologies, Inc.

FIG. 8B shows another embodiment of a “champagne cork” type microlens 112 with the advantage of spacing the lens 112 and the optical fiber 106 packed together on the sides 102a and 102b. do. Lens 112 has a smaller diameter face 132 that is housed in conical fiber collimator assembly 134 shown in FIG. 9. A larger diameter cylindrical surface 136 for the lens 112 protrudes out of the optical fiber collimator assembly 134. A champagne cork-like embodiment of the microlens 112 can be manufactured by grinding off a portion of the lens 112 shown in FIG. 8A.

As shown in FIG. 9, also when receiving either the cylindrical lens shown in FIG. 8A or the champagne cork microlens 112 shown in FIG. 8B, each optical fiber collimator assembly 134 also includes an optical fiber 106. Provided is a reservoir for receiving a conventional fiber optic ferrule 146 securely stored about its end 104. Converging blocks 152 disposed on both sides 102a and 102b of the reflection diverting module 100, respectively, in a number equal to the number N of optical fibers 106, as shown in FIG. Penetrated by As described above, the convergence of the light beam 108 is affected by the alignment of the optical fiber collimator assembly 134 and inserted into the conical hole 154. The optical fiber collimator assembly 134 and the conical hole 154 are preferably formed of the same material having identically shaped, mated, conical faces pointed at small angles. If configured in this manner, the optical fiber collimator assembly 134 is integrated into the converging block 152 when all of the optical fiber collimator assemblies 134 carrying the optical fiber 106 are installed into the mated apertures 154 of the optical fiber collimator assembly as a whole. It is fixed and sealed as if it sealed the inside of the reflection conversion module 100 in which the quasi-balanced light beam 108 propagates.

Converging block 152 may be simply mechanized by a single piece of metal or ceramic material, such as stainless steel. Alternatively, the converging block 152 may be made of Kovar, 42% nickel-iron alloy, titanium (Ti), tungsten (W) or molybdenum (Mo), which are suitably applied to prevent corrosion. All of these materials have an expansion coefficient that approximately matches lens 112 and minimize the birefringent effects that can occur when lens 112 is heated or cooled in an operating environment.

In addition to the above-described preferred manner of providing convergence by properly directing the optical fiber 106 and the lens 112 at the respective sides 102a and 102b, 1D convergence or 2D convergence can also be obtained in other ways. For example, the configuration of the optical fiber 106 and the lens 112 may provide some convergence where the arrangement of the mirror surfaces 116 to which the light beam 108 first strikes may provide the remainder of the convergence. For example, mirror surfaces 1116 of each column may be arranged along a cylindrical surface. Alternatively, the optical fiber 106 and the lens 112 do not provide any convergence to the mirror surface 116 arranged to provide all the convergences as shown in FIGS. 4A and 4B (ie, the light beam 108). Transmitted in parallel from the side surfaces 102a and 102b to the mirror surface 116. For example, mirror surfaces 116 may be arranged along a spherical surface in each column. In addition, the optical fiber 106, lens 112 and sets 118a and 118b provide either 1D convergence or 2D convergence behind sets 118a and 118b or to either of sets 118a and 118b. Can be arranged to For many other ways of adjusting the convergence of the light beam 108, choosing one approach over other possible approaches usually results in ease of manufacture, relaxed mechanical immunity to the assembly of the reflection conversion module 100, reliability and Problems related to price and the like.

The above convergence criteria affect the criteria for interacting with the reliability considerations as well as the optical design of the reflection conversion module 100. If each optical fiber 106 of the reflection diverting module 100 can be switched between 1152 optical fibers 106 carrying a light beam 108 having a total power of 100 kW, then all the transmissions through the reflection switching module 100 are carried out. The cumulative power of the light beam 108 exceeds 100 W at a given moment. However, assuming that on average, the same number of light beams 108 propagate in opposite directions between the side surfaces 102a and 102b, each set 118a or 118b of the mirror surface 116 averages at a given instant in time with a power of 50 Reflects a light beam 108 that delivers slightly higher power than W. In the worst case, in an analytical perspective, the light beam 108 delivering at least 50 W of power at a given moment touches either one or the other of the sets 118a or 118b of the mirror plane 116.

If the power supplied to the reflection conversion module 100 for directing the mirror surface 116 is cut off, at least 50 W and possibly 100 W or more of power within a short time, for example milliseconds, is managed at the convergence point. This amount of power destroys one or several mirror surfaces 116 provided in the sets 118a or 118b in which all the light beams 108 are concentrated. To prevent this false ending from occurring, both sets 118a and 118b are mirror plane 116 at the center of the mirror plane where the light beam 108 converges when the power to the reflection conversion module 100 is cut off. Omit. In order to detect this problem, the reflection conversion module 100 may have a photo detector behind this opening in the mirror surface 116.

In most telecommunications installations, the optical fiber is generally matched as a dual pair in which one fiber carries communications in one direction while another duplex pair carries communications in the reverse direction. Connectors are now available that are adapted to couple light between two double pairs of optical fibers that securely store two pairs of optical fibers in a single ferrule. The two pairs of optical fibers are switched at the same time, and the direction between the pair of optical fibers 106 and one of the other optical fibers located on the side 102b, one of which the reflection switching module 100 is respectively located on the side 102a. Since light can be coupled to either of these, proper adaptation of the lens 112 for use with a double pair of optical fibers 106 results in each being transmitted in opposite directions to the two optical fibers 106 for the double pair. A pair of mirror surfaces 116a and 116b can be used to divert light.

FIG. 11 shows a lens 112 adapted for use in the reflection conversion module 100 for simultaneously converting light transmitted by a double pair of optical fibers 106a and 106b. As shown in FIG. 11, a double fiber ferrule 146 carries a double pair of optical fibers 106a and 106b. The ends 104a and 104b of the optical fibers 106a and 106b and the faces 138a and 138b of the lens 112 are both obliquely polished. Angles of the faces 138a and 138b are formed to correct for the off-axis position of the optical fibers 106a and 106b and touch on the faces 138a and 138b from the optical fibers 106a and 106b. The light beams 108a and 108b are formed into the quasi-equilibrium beam exiting the convex surface 142 parallel to the longitudinal axis 144 but with a slight residual deviation from the longitudinal axis 144, and the reflection conversion module 100 Propagated in this manner. Both light beams 108a and 108b touch on the same pair of mirror surfaces 116a and 116b that are made large enough to reflect both light beams 108a and 108b simultaneously. Two semi-parallel beams of light beams 108a and 108b touch a double pair of optical fibers 106 on opposite sides 102a or 102b to another identically configured lens 112 and reflection conversion module 100. The lens 112 is arranged to couple the light beams 108a and 108b into each of the optical fiber pairs.

Torsion Mirror Configuration

As noted above, it is preferred that mirror surfaces 116a and 116b for sets 118a and 118b be provided by an electrostatically activated 2D torsional scanner of the type disclosed in the 790 patent. Both US Patent Application No. 08 / 885,883, filed May 12, 1997, and published Patent Cooperation Treaty (“PCT”) Patent Application International Publication No. WO 98/44571 are incorporated by reference and are preferred. Additionally provides more detailed information about the 2D torsional scanner. A hinge that mirror surface 116 can rotate about two non-parallel axes is attached for reference and has a torsion sensor of the type disclosed in US Pat. No. 5,648,618 (hereinafter referred to as "618 patent"). The torsion sensor provided in the hinge measures the rotation of the applied second frame or plate to provide the mirror surface 116 with respect to the first frame or the second frame, respectively.

As disclosed in the above patents and patent applications, it is preferred that the torsion sensor is manufactured by micromachining single crystal silicon using Simox, a silicon on insulator or a bonded silicon wafer substrate. Since these wafer substrates are very flat, stress free thin films, can be easily manufactured to as few micrometers as possible and can support the mirror surface 116, these wafer substrates are particularly a starting material for torsion scanner manufacture. It is preferable. As shown in FIG. 12, a silicon on insulator (“SOI”) wafer 162 includes an electrically insulated silicon dioxide layer 164 that separates single crystal silicon layers 166 and 168. Another portion of the torsion scanner is formed on the back side etched into the thick film silicon layer 168, while torsion bars and plates are formed in the thin film device silicon layer 166 that carry the mirror surface 116 of the torsion scanner. As is well known to those skilled in the micromachining art, the device silicon layer 166 has a front surface 169 farthest from the thick film silicon layer 168 and a back surface 170 of the silicon dioxide layer 164. The intermediate silicon dioxide layer 164 provides a complete etch stop for etching the wafer 162 from the backside of the wafer and yields torsion bars and plates with uniform thickness.

FIG. 13 shows a single electrostatically activated 2D torsional scanner 172 adapted to provide a mirror surface 116 for the reflection conversion module 100. Torsion scanner 172 has an outer reference frame 174 coupled to the opposite pair of hinges 176 that are flexible to outer torsion. A torsionally flexible hinge 176 supports an internal moving frame 178 that rotates about an axis achieved by the torsionally flexible hinge 176. An opposite pair of hinges 182 that are flexible to the inner torsion couples the midplane 184 to an inner moving frame 178 that rotates about an axis achieved by the torsionally flexible hinge 182. The axis of rotation achieved by the torsionally flexible hinge 176 and the torsionally flexible hinge 182 is non-equilibrium, preferably vertical.

It is important to note that the plate 184 of the torsion scanner 172 is a rectangle with a long side that is approximately 1.4 times wider than the height of the plate 184. Since the light beam 108 obliquely touches the mirror surface 116 transmitted by the plate 184 at a 45 ° angle, the plate 184 provided in the reflection conversion module 100 is rectangular. As a result, it is effective for the rectangular plate 184 to be square with respect to the reflection of the light beam 108 from the mirror surface 116. It is preferred that the plate 184 is 2.5 mm by 1.9 mm in size, with the inner moving frame 178 and torsionally flexible hinges 176 and 182 being between 5 μm and 15 μm thick. Torsionally flexible hinges 176 and 182 are between 200 μm and 400 μm in length and between 10 μm and 40 μm in width. The resonant frequency on both axes is on the order of 400 to 800 Hz, which can divert the light beam 108 to approximately 1 to 5 milliseconds between the two optical fibers 106. Both front 169 and back 170 of plate 184 are coated with the same metal adhesive layer and in complete stress balance, with titanium (Ti) or zirconium (500) below a metal reflective layer of 500-800 microns thick of gold (Au). Zr) is preferably coated with 10.0 to 100.0 kPa.

The torsionally flexible hinges 176 and 182 shown in more detail in FIG. 14 provide several advantages over conventional unfolded torsion bars. The invention is both entitled "Micromachined Members Coupled for Relative Rotation by Torsion Flexure Hinges", both by Slater Timothy G. and Neukermans Armand P. 1999 The U.S. Patent Application and Patent Cooperation Treaty ("PCT") International Patent Application, filed September 2, 2, is incorporated herein by reference and provides several advantages provided by torsionally flexible hinges 176 and 182. Initiate. Most striking for the reflection diverting module 100 is that the torsionally flexible hinges 176 and 182 are more compact than conventional unfolded torsion bars having an equivalent torsional spring constant. As a result, the use of torsionally flexible hinges 176 and 182 in place of conventional unfolded torsion bars allows the number of optical fibers 106 to be adapted to the sides 102a and 102b of the reflection conversion module 100. It is possible to make the torsion scanner 172 smaller, which can correspondingly increase and be mounted more closely together.

Each torsional scanner 172 provided in the reflection conversion module 100 has a pair of torsion sensors 192a and 192b of the type disclosed in the 618 patent. Torsion sensors 192a and 192b have a support member, ie plate 184 or internal movement frame 178, with respect to the support member, ie internal movement frame 178 or external reference frame 174, with a theoretical resolution of approximately 1.0 microradians. Measure the direction of. According to the detailed description of the 618 patent, when the torsion scanner 172 operates in the reflection diverting module 100, current is passed through two torsion sensors 192a and 192b between the pair of sensor current pads 194a and 194b. Flows in series. Thus, the torsion scanner 172 has a meandering metal conductor 196 bonded to the front surface 169 of the device silicon layer 166. The hinged metal conductor 196 starts at the sensor current pad 194a and is flexible to the lower torsion across the torsionally flexible hinge 176 immediately adjacent from the outer reference frame 174 onto the inner moving frame 178. A torsional sensor 192b of the X axis located at 182 is reached. The bracing metal conductor 196 extends from the torsion sensor 192b on the X axis onto a reflected stress balanced metal coating supplied to both sides of the plate 184 to provide a mirror surface 116, and the plate 184. And back onto the inner moving frame 178 across the upper torsionally flexible hinge 182. The torsional metal conductor 196 leads to a Y-axis torsion sensor 192a located in the hinge 176 which is flexible to the left torsion. The torsional metal conductor 196 travels around the external reference frame 174 from the torsion sensor 192a on the Y axis and into the sensor current pad 194b. A metal conductor across the hinge 176 that is flexible to the right torsion and disposed on the inner moving frame 178 on the opposite side of the fringing metal conductor 196 may be configured to connect a pair of inner hinge sensor pads 198a and 198b to the X axis torsion sensor. To 192b. Similarly, the metal conductor, which is one of the ones disposed along the side of the metallized conductor 196 on the outer reference frame 174, and bends around the opposite side of the torsion scanner 172 on the outer reference frame 174. The other metal conductor couples pairs of inner hinge sensor pads 202a and 202b to the torsion sensor 192a on the Y axis. The pair of grooves 204 are disconnected only through the device silicon layer 166 on opposite sides of the inner hinge sensor pads 198a and 198b, and with the sensor current pad 194a and the inner hinge sensor pads 198a and 198b. Increases electrical insulation between sensor current pad 194b and inner hinge sensor pads 202a and 202b.

The back side 170 of the plate 184 is shown for the electrode 214 and the sensor current pads 194a and 194b where the front side 169 is used to activate the rotation of the plate 184, as shown in FIG. 15A. It is desirable to provide a mirror surface 116 because it contacts the insulating substrate 212 leading to contact, both the inner hinge sensor pads 198a and 198b and the inner hinge sensor pads 202a and 202b, not shown in FIG. 15A. Do. The plate 184 for each torsional scanner 172 is also separated from the substrate 212 by a distance, for example 40 to 150 μm, by a spacer not shown in FIG. 15A. The separation between the plate 184 and the substrate 212 depends on how it moves relative to the edge of the plate 184 during rotation.

Note that for the reflection diverter module 100, a very thin plate 184, which is only a few micrometers thick, is preferred and can be fabricated using the device silicon layer 166 for the wafer 162. In many cases, plate 184 and torsionally flexible hinges 176 and 182 may be the same thickness as device silicon layer 166. Alternatively, as shown in Fig. 15A, the torsionally flexible hinge 176 can be thinned by etching. For example, the torsionally flexible hinge 176 may be 6 μm thick while the plate 184 may be 10 μm thick. Likewise, plate 184 may be thinned to reduce the moment of inertia of the plate by etching cavity 216 into plate 184 leaving reinforcing ribs 218 on thin plate 184.

Telecommunication system components, such as the reflection conversion module 100, exhibit high reliability. The plate 184 of the torsion scanner 172 that accidentally collides with the electrode 214 should not be attached to the electrode and should be rotated directly in the direction of the electrode. In addition, such accidental collisions should not damage torsion scanner 172 or any circuitry coupled to torsion scanner 172. To prevent the stiction, as shown in FIG. 13, rounded corners reduce the strength of the electrostatic field around the plate 184 and the inner moving frame 178. Rounding the periphery of the plate 184 also reduces the effective radius of rotation of the plate due to the compound rotation about the plate 184 about the axes achieved by the two twisting flexible hinges 176 and 182, respectively.

In addition, when the periphery of the plate 184 and the inner moving frame 178 is rounded, as shown in FIG. 15A, a position where the electrode 214 can be connected is covered by an electrically insulating material 219 such as polyimide. do. The plate 184 covers only portions of these electrodes 214 that can be connected to the electrically insulating material 219 to avoid the storage of charge on most of the electrodes 214. Likewise, during fabrication of torsion scanner 172, several silicon dioxide layers 164 may be left around plate 184, such that a metal reflective layer providing mirror surface 116 is never connected to electrode 214. It doesn't work. Alternatively, as shown in FIG. 16B, a hole 220 is formed through the metal of the electrode 214 in the possible connection area.

While operating the reflection diverting module 100, the torsional scanner 172 is at ground electrical potential while a drive voltage is supplied to the electrode 214. In order to reduce the electric charge current, when the plate 184 is connected to the electrode 214, a large resistance (eg, 1.0 kA) can be connected in series to the drive circuit for the electrode 214. Ideally, these resistors should be placed as close as possible to the electrodes 214, while the conductors connected between the electrodes 214 and the resistors can recover the off-field electric field that rotates the plate 184. Thus, in another embodiment, the electrode 214 is coated with a very high resistance, but in a selected region, such as these regions as shown in FIG. 16A, a weakly conductive material is removed from the electrode 214 for DC charge. To provide a bleed path. In addition, the inputs of all amplifiers connected with a torsion scanner 172, such as a torsion scanner that receives a direction signal from the torsion sensors 192a and 192b, are provided between the plate 184 and the electrode 214 under overvoltage conditions. Diode protection must be provided to prevent damage due to arcing or accidental connections.

There are several configurations in which the limiting factor for the density of the optical fiber 106 can be utilized to typically increase the mirror array at the sides 102a and 102b. For some reason, in particular, a number of contacts should be shown for each torsion scanner 172, and it is desirable for the torsion sensor 172 to be arranged into the strip 222 as shown in FIGS. 16A and 16B. By organizing the torsion scanner 172 into the strip 222, the density of the strip can be achieved when arranged as a two dimensional array of individual torsion scanners 172. Each strip 222 has a metal support frame 224 to which the substrate 212 is fixed.

As described in more detail below, since the strip 222 is flipchip bonded to the substrate 212, all electrical connections to the strip 222 are made between the strip 222 and the substrate 212. A flat polyimide supported multi-conductor ribbon cable 226 is connected to the substrate 212 to exchange electrical signals between the pads 194, 198 and 202 and the electrode 214. Since each support frame 224 can be an open frame that can have reinforcing ribs, the ribbon cable 226 can be flexed freely and guided from the substrate 212.

FIG. 16B shows that the substrate 212 and the strip 222 can be meandered along the stepped substrate 212 to the ribbon cable 226 without blurring the mirror surface 116. Arranging the strips 222 in this manner reduces the horizontal distance between the mirror surfaces 116 of the strips 222 immediately adjacent to the light beam 108. Since the light beam 108 touches approximately 45 ° on the mirror surface 116, the apparent distance between the immediately adjacent strips 222 is further subtracted by approximately factor 1,4, such that the plate 184 as described above. It is preferable that this rectangle is.

One disadvantage of constructing the strip 222 as shown in FIG. 16B is that the residual deviation between immediately adjacent strips 222 may be less than or equal to the thickness of the torsion scanner 172 and the substrate 212. . In addition, overlapping the immediately adjacent strip 222 and the substrate 212 inhibits the removal of a single bad strip 222 without disturbing the immediately adjacent strip 222.

16C and 16D illustrate strip 222 in an electrical lead 228 connected to a torsion scanner 172 that is plated or screened on one side, around one edge, and on another side of the substrate 212. And a preferred embodiment for the support frame 224. In this configuration for the electrical leads 228, the attachment of the ribbon cable 226 to the substrate 212 is not degraded. By plating or screening the electrical leads 228 on the substrate 212 and having several via holes through the substrate 212, the substrate 212 can be as narrow as the strip 222. Narrowing this range, the coupling strip 222, substrate 212 and support frame 224 can now be arranged as shown in FIG. 16E for both sets 118a and 118b. This may be achieved where the immediate adjacent strip 222 is needed by the optics of the reflective conversion module 100 rather than by mounting considerations. The optimum residual deviation between the immediately adjacent strips is approximately 0% to 10% of the distance between the plates 184 in the immediately adjacent strips 222. The configuration of the substrate 212 shown in FIG. 16D facilitates access to the substrate 212 and removal of the strip 222 without disturbing the adjacent support frame 224. Note that if necessary, electrical leads 228 may appear around both corners of the substrate 212. This capability is utilized to separate the lead 228 carrying the high voltage drive signal supplied between the plate 184 and the electrode 214 from the lead 228 carrying the signal from the torsion sensors 192a and 192b. The advantage is that it can be.

Without reducing the size of the plate 184, as shown in FIG. 17A, the density of the optical fiber 106 at the sides 102a and 102b determines the vertical distance between the torsion scanners 172 in the strip 222. By halving vertically, it can be increased by replenishing the torsion scanner 172 of the immediately adjacent strip 22. Due to the convergence criteria for arranging the light beam 108 within the reflection diverting module 100, supplementing the torsion scanner 172 in the immediately adjacent strip 222, the optical collimator is turned into a semi-hexagonal dense array with a semi-rectangular array. Affects the reorganization of the holes 154 that receive the assembly 134. While the torsion scanner 172 is replenished, the torsion scanner 172 does not increase the density of the torsion scanner 172 in the immediately adjacent strip 222, so that the arrangement of the torsion scanner 172 is not limited to the lens 112 or the like. In order to extend the diameter of either of the optical fiber collimator assemblies 134, the density of the optical fiber 106 is increased at the sides 102a and 102b and the space between the immediately adjacent optical fibers 106 is limited.

The density of the torsion scanner 172 can be further increased by manufacturing the torsion scanner 172 when it is a complete monolithic two dimensional array than when it is a strip 222. As shown in FIG. 17B, when the torsion scanner 172 is replenished in the immediately adjacent column, the torsion of the torsion scanner 172 into the empty space generated between the torsion scanners 172 in the immediately adjacent row or column of the array. The flexible hinge 176 can intermesh with each other. Interdigitation of this torsionally flexible hinge 176 provides a short distance between the centers of the plates 184 of the torsion scanner 172 in adjacent rows or columns, and is a hexagonal torsional scanner 172. ) And corresponding sides 102a and 102b more closely approach the optical fiber 106.

Another embodiment of the strip 222 is to face the flexible hinges 176 and 182 torsionally at 45 ° with respect to the vertical and horizontal axes of the support frame 224. 18A and 18B show torsionally flexible hinges 176 and 182 that effectively use the area for the strip 222 more than the configuration where the torsionally flexible hinges 176 and 182 are oriented in parallel and perpendicular to the strip 222. Shows the diagonal configuration for. When using the diagonal configuration for the flexible hinges 176 and 182 for torsion towards 45 ° relative to the outer reference frame 174, the torsionally flexible hinge does not increase the area occupied by the torsion scanner 172. It can be made longer. The plate 184 is elongated in one direction to correct the 45 ° contact angle of the light beam 108. Due to the elliptical light beam 108 touching the plate 184, the light beam 108 may be edged off, conventionally providing octagonal plate 184 a room for external reference frame 174. Sides of the outer reference frame 174 are oriented in the <110> crystal direction of silicon for ease of manufacture. The configuration for this torsion scanner 174 is directed towards torsion sensors 192a and 192b along the <100> crystal direction of silicon. Thus, the p-type device silicon layer 166 or wafer 162 with p-type implantation should be used to manufacture the torsion sensors 192a and 192b. The <110> and <100> crystal orientations of the silicon can be exchanged as the process changes appropriately.

Using the arrangement of the torsion scanner 172 shown in FIG. 18B, the plate 184 of size 1.5 × 2 mm can be spaced by 2.5 mm to effectively increase the density of the mirror surface 116 by factor 1.4. . When viewed at an angle of incidence of approximately 45 ° of the light beam 108, the strip is tilted 54 °. In this configuration, the strip 222 is directed at 45 ° relative to the support frame 224. The orientation to this strip 222 requires that the mirror surface 116 block the light beam 108 entirely. The support frame 224 faces 45 °, and all strips 222 can be the same length, more effectively using the area on the wafer 162.

19A illustrates another embodiment of a torsion scanner 172 that further shortens the distance between immediately adjacent mirror surfaces 116 in the reflection conversion module 100 by further reducing the size of the torsion scanner 172. In the above detailed description, it is evident that positioning the torsionally flexible hinges 176 and 182 at the corners rather than the sides of the plate 184 has the advantage of reducing the size of the torsion scanner 172. In FIG. 19A, elliptic curve 232 shows the outline of light beam 108 striking on mirror surface 116 of plate 184. Since the light beam 108 does not touch the edge of the plate 184, the hinge 182 flexible to internal torsion may rotate relative to the plate 184 taking up unused edge space. As with the configuration of the torsion scanner 172 shown in FIG. 18A, the hinges 176 that are flexible to external torsion are continuous to occupy the corners of the external reference frame 174.

Positioning the torsionally flexible hinge 182 at the corners of the plate 814 as shown in FIG. 19A not only reduces the size of the torsion scanner 172 but also reduces the angle when the plate 814 rotates about both axes simultaneously. Reduces the synthesis of When the plate 184 rotates about the axis achieved by the hinges 176 and 182 that are flexible at both twists, combining the angles increases the distance by moving the edges of the plate 184. Combining the angles, there is a necessary separation between the plate 184 and the substrate 212 which increases the voltage to be supplied between the plate 184 and the electrode 214 equivalently to rotate the plate 184. Is increased. However, if the plate 184 has a non-square aspect ratio as typically occurs for the plate 184 provided in the reflection diverting module 100, the torsionally flexible hinges 176 and 182 shown in FIG. 19A The torsion sensors 192a and 192b are no longer directed along the orthogonal crystal direction, i.e., either the <100> or <110> direction of silicon. This is undesirable as the torsion sensors 192a and 192b in the torsionally flexible hinges 176 and 182 respond to both the bending and torsion of the torsionally flexible hinges 176 and 182.

Since the plate 184 shown in FIG. 19A has an aspect ratio of approximately 1.4: 1, the axes of rotation 236a and 236b are achieved by hinges 176 and 182 that are flexible to cross at approximately 70.5 °. However, if the axes of rotation 236a and 236b are delicately directed in the new direction until these axes of rotation intersect at 90 ° as shown in Fig. 19B, then the outer reference frame 174 is directed to the <110> crystal orientation of silicon. When aligned along, the torsionally flexible hinges 176 and 182 may be oriented in a direction along the single crystal direction of silicon, for example the <100> crystal direction. Configured as shown in FIG. 19B, the torsion scanner 172 provides a certain amount of space for the flexible hinge 182 to the internal torsion at the corners of the plate 184 reducing the size of the torsion scanner 172. do. Further, while the configuration of the torsion scanner 172 shown in Fig. 19B preserves the crystallographic directions of the torsion sensors 192a and 192b, the synthesis effect is significantly reduced because it is not completely removed. However, in the configuration of the torsion scanner 172 shown in FIG. 19A, the perpendicular axis of rotation achieved by the torsionally flexible hinges 176 and 182 is directed in a direction oblique to the length and width of the plate 184. Nevertheless, since only a small angular rotation of the plate 184 occurs during the operation of the reflection conversion module 100, the area of the plate 184 to which the light beam 108 touches is slightly changed when the plate 184 rotates. do.

Incorporating the torsion scanner 172 shown in FIG. 18A or 19A into one of the sets 118a or 118b of the mirror surface 116, these torsional scanners needing to rearrange the shape of the set 118a or 118b. Minimize each of its benefits. A preferred arrangement for the strip 222 'of the torsion scanner 172 shown in FIG. 18A is shown in FIG. 20A. As described above and shown in FIG. 20A, a strip 222 ′ is mounted at a 45 ° angle with respect to the horizontal base 242 of the reflection diverting module 100. In FIG. 20A, a support frame 224 ′ carrying the strip 222 ′ is also mounted at a 45 ° angle with respect to the horizontal base 242. The two axes achieved by the torsionally flexible hinges 176 and 182 relative to the plate 184 are indicated by the X and Y axes 244 shown in FIG. 20A. The maximum angle of rotation with respect to the plate 184 of the axis is achieved by the torsionally flexible hinges 176 and 182 where the mirror surface 116 permits the same torsional scanner 172 in further sets 118b or 118a. A serrated rectangular field 246 of the addressable torsional scanner 172 is achieved in the possible sets 118a or 118b.

This optimal rectangular field 246 is a flat cut edge and has a side that is approximately diagonal to the strip 222 '. For the arrangement shown in FIG. 20A, the longest strip 222 ′ is at least 1.4 than the torsion scanner 172 required for the rectangular array of torsion scanners 172 assembled into the strip 222 shown in FIG. 16A. You should have more boats. However, the torsion scanner 172 may be excluded from the position of the sets 118a or 118b that are not addressable from the other sets 118b or 118a. Only a few strips 222 ′ shown in FIG. 20A need to be total length. These strips 222 'having several torsional scanners 172 may be entirely removed. For example, by using 40 strips 222 'including up to 44 torsional scanners 172, 1152 torsional scanners in sets 118a or 118b with very small scan angles and relatively small mirror sizes. (172) arrays are possible. Different arrangements provide 1132 torsional scanners 172 that measure only 1.59 x 2.2 mm and require deflection angles of 3.69 ° and 3.3 °. The strip 222 ′ of the torsion scanner 172 is directed in an average 55 ° direction relative to the optical fiber collimator assembly 134. Although the arrangement shown in FIG. 20A substantially increases the density of the torsion scanner 172 and the corresponding optical fiber collimator assembly 134 slightly more complex, many torsion scanners may be subjected to a particular rotation angle specified for the plate 184. Can be addressed.

FIG. 20B shows similar rearrangements in sets 118a and 118b of the torsion scanner 172 of the type shown in FIG. 19B. In the arrangement for the torsion scanner 172 shown in Figure 19B, the strip 222 'and support frame 224 "are oriented vertically similar to Figure 16A. However, the X axis with respect to the plate 184 rotation And Y axis 244 is directed at 45 [deg.] With respect to strips 222 " and the support frame 224 " of the strips. X and Y axes (with respect to strip 222 " and support frame 224 " The inclined direction of 224 is again set by another set 118b or 118a of the mirror surface 116 to which the maximum angle of rotation relative to the plate 184 of the corresponding torsion scanner 172 can be addressed and Means a serrated octagonal or truncated rectangular field 256 of the torsion scanner 172 addressable at 118b. A rectangular field 256 when these torsion scanners 172 are in a p × q array. Is achieved, the optimal field coverage for the strip is determined by the diagonal X and Y axes 224. A square field or rectangular field with 0.7 to 1.2 pq regions symmetrically arranged, thus the aspect ratio for rectangular field 256 is slightly thinner and longer relative to strip 222 " direction (e.g., 1.0: 1.3). . When the set 118a or 118b faces the strip 222 " and the support frame 224 " in the horizontal direction, the elongated portion of the rectangular field 256 becomes horizontally longer than vertical. For ease of manufacture, all strips 222 ″ are of the same length. Similar to the arrangement of torsion scanner 172 shown in FIG. 20A, again the rectangular field 256 can be excluded from the torsion scanner 172. There is an area, again, with the advantage of slightly tapering the other strip 222 ", except for the short strip 222" along the sides of the rectangular field 256 with several torsional scanners 172. In the embodiment shown in FIG. 20B, for an angle of rotation of 1.8 × 2.4 mm plate 184 and the plate 184 about the X and Y axes of 5.6 ° and 3.7 °, the arrangement is twisted to approximately 1,500 pieces. The number of scanners 172 is significantly increased.

In the configuration of the reflection conversion module 100, which is further described below, the converging block 152 in which the optical fiber collimator assembly 134 is positioned some distance from at least some of the sets 118a and 118b of the mirror surface 116. Is fixed to. This configuration for the reflection diverter module 100 requires very good alignment of the collimator to the mirror plane 116. FIG. 21 shows an arrangement in which the collimating lens 112, the optical fiber 116 and the torsion scanner 172 bring the strips 222 closer together, thereby mitigating the tolerance for their alignment. In this figure, the substrate 212 widens the beam-folding and deflection assembly 264 by widening the strip 222 and the mirror strip 262 attached to the surface of the substrate 212 opposite the strip 222. To achieve. The beam-folding and deflection assembly 264 comprises a mirror strip of one beam-folding and deflection assembly 264 in which the quasi-balanced light beam 108 touches the mirror surface 116 provided by the immediately adjacent torsional scanner 172. 262 is re-arranged into a regular structure that is reflected and destroyed. In the arrangement shown in FIG. 21, since all the lenses 112 are located at the same short distance from the mirror surface 116 coupled to the lens, the alignment of the light beam 108 with respect to each mirror surface 116 of the lens is not possible. The limit for this is reduced. Convergence of the light beam 108 may be provided in one dimension by arranging the immediately adjacent beam-folding and deflection assembly 264 at slightly different angles. Two-dimensional convergence can be obtained by properly positioning the optical fiber 106 and the lens 112 with respect to the mirror surface 116 respectively coupled to the lens. In the arrangement shown in FIG. 21, since the substrate 212 is adjacent to the mirror surface 116 coupled to the lens, an almost full length of 500-900 mm long path between the sides 102a and 102b is achieved by the sets ( Being between pairs of mirror surfaces 116 of 118a and 118b reduces the angle that plate 184 should rotate.

As shown in FIG. 13, all electrical connections to the torsion scanner 172 occur at the front surface 169 of the device silicon layer 166, and as shown in FIG. 15A, the light beam 108 is connected to the device silicon layer. Reflected and extinguished in the metal layer coated on the back surface 170 of 166. In order to form an electrical connection between the substrate 212 and the torsion scanner 172 in the strip 222, the strip 222 is preferably flip chip connected to the substrate 212. The substrate 212 can be modified more than one strip 222 by using the substrate 212 larger than the strip 222. Substrate 212 can be manufactured in a number of different ways.

Substrate 212 may be made of 100 silicon wafers. If the substrate 212 is made of a silicon wafer, the cavity 272 can be anisotropically etched into the substrate 212 to provide space for rotation of the plate 184 and to the plate 184 and the cavity 272. The space between the located electrodes 214 is precisely controlled. Electrical insulation between the leads 228 and between the electrodes 214 can be obtained by forming an electrically insulating oxide on the surface of the silicon substrate 212. Electrode 214 may be integrated into silicon substrate 212 or applied onto a silicon surface within each cavity 272.

If the substrate 212 is made of a silicon wafer, the electronic circuit also has the advantage that it can be integrated into the substrate. The circuitry provided in the silicon substrate 212 provides a current source for providing current to the torsion sensors 192a and 192b of the torsion scanner 172 and a torsion indicative of the direction of the internal moving frame 178 and plate 184. The differential amplifier receives a signal from the sensors 192a and 192b and the amplifier supplies a high voltage signal to the electrode 214 which activates the rotation of the plate 184. Incorporating these different types of electronic circuitry into the substrate 212 significantly reduces the number of leads that must be provided in the ribbon cable 226. The number of leads in ribbon cable 226 can be further reduced by having one or more multiplexer circuits in silicon substrate 212.

A light detector, which is responsive to the wavelength of light present in the light beam 108 and disposed on the surface of the substrate 212 adjacent to the strip 222 outside the shadow created by the mirror surface 116, is part of the light beam 108. Has the advantage that it can be provided on the substrate 212 to detect the case that is outside the mirror surface 116. With respect to the wavelengths of light used for optical fiber telecommunications, such a photo detector may be used to transmit light at a wavelength at which silicon is used for optical fiber telecommunications, even when a portion of the light beam 108 is off the mirror surface 116. Therefore, the case where it is covered by the parts of the strip 222 other than the mirror surface 116 is detected.

Strip 222 is coupled to substrate 212 by solder bumps 276 or other bonds formed by solder reflow. Solder bumps 276 strongly interconnect pads 194, 198, and 202 of torsion scanner 172 to pads and strips 222 on substrate 212. Flip-chip bonding of the strip 222 and the substrate 212, which is a similar material, keeps the strip 222 flat because no stress is introduced as the temperature coefficient between the strip and the substrate is perfectly consistent.

If the substrate 212 is made of silicon or polycrystalline silicon, as shown in FIG. 22D, a number of very small electrically conductive vias 282 are described by Calmes described on page 1500 of the 99 transducers. It may be formed through the silicon wafer during fabrication of the substrate 212 using a process similar to that disclosed in the paper submitted by et al. Holes for vias 282 are first formed through the wafer using a standard Bosch deep reactive ion etch ("RIE") process. The holes can be 50 μm wide and 500 μm deep. The wafer is oxidized to achieve an electrically insulating oxide layer 284 that isolates the pores from the peripheral wafer. Highly doped polycrystalline silicon layer 286 is grown over oxide layer 284 by providing conductive paths to the pores along the wafer surface. Obtaining sufficient conductive polycrystalline silicon layer may also require vapor phase doping of the polycrystalline silicon layer 286 comprising phosphorus. Conductive polycrystalline silicon 286 formed in this manner is electrically connected to both sides of the wafer. Ring portions 288 may be etched through the polycrystalline silicon layer 286 around each via 282, so that vias from each other ( It is desirable to electrically isolate 282. In order to increase the electrical conductivity of the substrate 212 and to facilitate the formation of electrical connections to the vias 282, one or more additional metal layers may be coated in a suitable pattern on either side of the substrate 212. .

22D illustrates a method of mounting strip 222 to substrate 212 having vias 282. Electrical connections between the strip 222 and the vias 282 of the substrate 212 are again formed by the solder bumps 276. Elastomeric layer 292 secures polyimide and copper foil 294 forming ribbon cable 226 on the side of substrate 212 furthest from strip 222 to non-ferrim scanner 172. Ballgrid or TAB bumps 298 connect with conductive vias 282 to achieve electrical connection with the polyimide and copper foil 294. In this manner, a very large amount of connection is made through the substrate 212 with vias 282 which are relatively low electrical resistance.

If the substrate 212 is made of polycrystalline silicon or pyrex glass, the cavity 272 may be etched into the substrate. However, if substrate 212 is made of Pyrex glass, electrode 214 must be applied onto the surface of cavity 272. Substrate 212 may also be made of a suitable ceramic, such as aluminum oxide, and is preferably aluminum nitride having a coefficient of thermal expansion that more closely matches the silicon forming strip 222. If strip 222 is made of ceramic material, spacers should be screened onto substrate 212 to provide space for rotation of plate 184 and precise control between plate 184 and electrodes 214. Achieve the desired separation. However, forming spacers on the surface of the ceramic substrate 212 typically requires repeated coating to achieve sufficient spacing between the electrode 214 and the plate 184.

It is demonstrated that the inclined side surface 302 formed by the 111 side exposed by anisotropic etching of the handling silicon layer 168 of the wafer 162 shown in FIG. 15A has proved very beneficial for flip chip bonding. Notably. The sides 302 substantially protect the mirror surface 116 on the back 170 of the plate 184 from damage during manufacture while simultaneously mechanically reinforcing the strip 222, while the inclination angle of the sides is approximately The light beam 108 reaching the mirror surface 116 at a 45 ° angle is hardly blurred. In addition, the mirror surface 116 protects from contamination by expanding the very thin coating 304 similar to those coatings used for masking the integrated circuit (“IC”) across the backside of the handling silicon layer 168. Can be.

Due to the presence of the handling silicon layer 168 around the mirror surface 116, the flip chip configuration for mounting the torsional scanner 172 also has the advantage of reducing light scattering as shown in FIG. 23. The inclined sides 302 and the periphery of the backside of the handling silicon layer 168 effectively absorb stray light that strikes the antireflective layer when the light beam 108 is switched between the mirror surfaces 116. It may be covered with an antireflection layer 312. The inclined side surface 302 also allows the side surfaces 102a and 102b of the side from which the light beam 108 propagates to receive stray light when the light beam 108 is switched between the mirror surfaces 116 from the light beam 108. Scatter it at a very large angle to prevent it.

24 is shown in FIGS. 2, 4A and 4B, 5, 6, and 7 embedded in an environmental housing 352 that completely encloses the light path through which the light beam 108 propagates, as described below. A reflection switching module 100, such as these reflection switching modules, is shown schematically. As noted above, the reflection diverting module 100 mechanically interconnects the sides 102a and 102b and the sets 118a and 118b and maintains them in a strongly aligned manner. An environmentally sealed environmental housing 352 that protects the reflection diverting module 100 may provide temperature regulation to maintain an appropriate operating environment for the reflection diverting module 100. Controlled dry gas, such as nitrogen, can flow through the environmental housing 352 to prevent moisture from condensing in the reflection diverting module 100. The environmental housing 352 may also be slightly pressurized to exhaust ambient atmosphere from the reflection diverting module 100. The environmental housing 352 has a desaturable microdryer 353 as disclosed in US Pat. No. 4,528,078 to control moisture in the atmosphere within the reflection diverting module 100. The wall 354 of the environmental housing 352 is penetrated by the electrical supply 356 for the ribbon cable 226. The optical fiber collimator assembly 134 safely stores the end 104 of the optical fiber 106 plugged directly into the converging block 152 which is fed through the environmental housing 352. In this manner, the environmental housing 352 almost hermetically seals the reflection diverting module 100. In the environmental housing 352, in order to reduce the possibility of optical mismatch, the ribbon cable 226 is carefully routed to apply stress to the reflection conversion module 100, in particular the support frame 224 and the substrate 212. Should be prevented.

Fiber optic switch

25 is generally referenced 400 at which a modular fiber optic switch in accordance with the present invention is shown. The optical fiber switch 400 has a standard 23 inch wide area telecommunication rack 402 at the base on which the environmental housing 352 containing the reflection diverting module 100 is located. The environmental housing 352, which includes all the torsional scanner 172, is on a particular bearing on the road surface just below the leg 402 and is very flexiblely connected to the leg 402. Supporting the environmental housing 352 on a particular bearing rest which minimizes vibrations and the like and thermally couples the environmental housing 352 to the road surface portion strengthens the thermal regulation of the environmental housing.

Port card

Mounted in the rack 402 on the environmental housing 352 are a number of double sockets 404 provided in the port card 406 adapted to receive a double pair of optical fibers 106. One double pair optical fiber 106 sends one light beam 108 to the optical fiber switch 400, and another double pair optical fiber 106 receives one light beam from the optical fiber switch 400. The port card 406 is arranged either horizontally or vertically in the rack 402 and can be individually removed or installed without the immediate adjacent port card 406 intervening. As in the common example, in the telecommunications industry, the port card 406 is hot-swappable. The reflection diverting module 100 may include an extra mirror surface 116 such that if several mirror surfaces 116 are broken, the ribbon optical switch 400 may maintain the operational capability of the entire ribbon optical switch. In principle, it will be readily apparent that all or some fewer optical fibers 106 connected to the port card 406 can receive the light beam 108 from the port card. Likewise, all or some fewer optical fibers 106 connected to the port card 406 can deliver the light beam 108 to the port card 406. The optical fibers 106 may be organized in double pairs as shown in FIG. 26A, but need not be so organized.

In the block diagram shown in FIG. 26A, all items for the left side of the dotted line 412 are provided in the port card 406, and all the items for the right side of the dotted line 414 are provided in the reflection switching module 100. . The area between the dashed lines 412 and 414 shows the back side of the rack 402. Each port card 406 includes the electronics, alignment optics, and electro-opticals needed to control the operation of a portion of the reflection conversion module 100. Therefore, all the optical fibers 106 are provided in the reflection switching module 100 connected to the port card 406. Similarly, all the torsion scanners 172 having a mirror surface 116 to which a given light beam 108 can reach are connected to the port card 406 through the substrate 212 and the ribbon cable 226 of the torsion scanner. Each port card 406 is connected to 16 or 32 optical fibers 106, half of which may receive light beam 108 from the port card 406, and half of which may receive the light beam 108. 406 is preferably, but need not be. In FIG. 26A, the odd subscript optical fibers 106 ₁ , 106 ₃ ,..., 106 _2n-1 transmit the light beam 108 to the reflection conversion module 100, while the even subscript optical fibers 106 ₂ , 106 ₄ ,. 106 _2n ) receives the light beam 108 from the reflection conversion module 100.

The port card 406 has a light source 422 and a tap or directional coupler 424 for supplying and coupling light into the optical fiber 106 for use in the servo alignment of the reflection conversion module 100. The directional coupler 424 also supplies the light received from the reflection conversion module 100 through the optical fiber 106 to the photo detector 426. The port card 406 also includes a drive, detection and control electronics 432, such as a digital signal processor ("DSP"), with circuitry coupled to the port card, to provide an electrode (e.g. 214 and the torsion sensors 192a and 192b included in each torsion scanner mounted on the substrate 212 to exchange electrical signals through the ribbon cable 226. The drive, detection, and control electronics 432 controls the orientation of the mirror surface 116 with an implemented servo loop that ensures proper orientation of the mirror surfaces, and also controls the supervisor processor (through the RS232 data communication link 438). 436). The backside between the dashed lines 412 and 414 also allows all ribbon cables 226 to be connected to the connector 442, the data communication link 438, and other power such as the power required to operate the drive, detection and control electronics 432. It is provided with a heterogeneous mixed electrical connection.

In orienting a pair of mirror surfaces 116, one of each of sets 118a and 118b directs one light beam 108 to one optical fiber 106 and side 102b to side 102a. Coupled between one optical fiber 106 for the other, and two mirror planes 116 initially use pre-determined angular coordinates that specify rotation about two axes for each mirror plane 116 in the pair. Is appropriately directed. Thus, ignoring the extra redundant mirror surface 116 provided in the reflection switching module 100 with respect to the reflection switching module 100 in the N × N arrangement, the optical fiber switch 400 is applied to each torsion scanner 172. The 4 × N ² values for the direction signals generated by the provided torsion sensors 192a and 192b should be stored. Accordingly, the reflection switching module 100 is maintained in the management processor 436, has a lookup table 452 shown in Fig. 27A, and uses direction signals during a valid period of operation of the optical fiber switch 400. Store ² values for 4 × N.

The 4 × N ² values for the direction signals generated by the torsion sensors 192a and 192b included in each torsion scanner 172 may be initially determined analytically. During assembly of the fiber optic switch 400, the analytically determined coordinates and direction signals are finely adjusted to correct manufacturing tolerances. It is also desirable that these coordinates and direction signals can be updated throughout the operational lifetime of the optical fiber switch 400. Thus, because the torsion sensors 192a and 192b are good at characterization, the lookup table 452 stores correction data for initial values of coordinates and direction signals, for example sensor residual deviations and temperature corrections.

In a preferred embodiment of the optical fiber switch 400, the high frequency servo system uses direction signals generated by the torsion sensors 192a and 192b in the direction control of each mirror surface 116. The frequency response of this high frequency servo system allows for the correct orientation for the pairs of mirror planes 116 when switching from one pair of optical fibers 106 to another pair of optical fibers. The high frequency servo system also maintains the orientation of all mirror surfaces 116 despite mechanical shocks and vibrations. During operation of the optical fiber switch 400, in order to ensure the precise orientation of the pairs of mirror surfaces 116, the optical fiber switch 400 also employs a low frequency optical feedback servo, which is described in more detail below.

In preferentially facing the pair of mirror surfaces 116, one of each of the sets 118a and 118b directs one light beam 108 of one optical fiber 106 and side 102b of the side 102a. Coupled between the other optical fiber and transmitted to each of the two dual-axis servos 454 provided on the port card for each torsion scanner 172 exchanging signals with the port card 406 from the lookup table 452. Stores the value for the direction signal.

Each biaxial servo 454 transmits a drive signal to the electrode 214 provided on the substrate 212 via the ribbon cable 226 to rotate the mirror surface 116 in a predetermined direction. The two torsion sensors 192a and 192b included in each torsion scanner 172 return the respective direction signals of the torsion scanner to the respective biaxial servos 192a and 192b through the ribbon cable 226. Each dual axis servos 454 compares the value for the direction signal received from the lookup table 452 with the direction signal received from the dual axis servos coupled to the torsion sensors 192a and 192b. When there is a predetermined difference between the stored value for the direction signal received from the lookup table 452 and the direction signal received from each of the torsion sensors 192a and 192b of the dual axis servos, the dual axis servos 454 This predetermined difference is reduced by appropriately correcting the drive signals delivered to the electrodes.

FIG. 27B shows either the X-axis or the Y-axis of the dual axis servos 454, one of two co-channels. As shown in the figure and described above, the current source 462 provided in the port card 406 supplies current to the torsion sensors 192a and 192b connected in series to the torsion scanner 172. Differentiating the output signals from one or the other of the torsion sensors 192a and 192b, in Fig. 27, an instrument in which the X axis torsion sensor 192b is provided on the port card 406 via the ribbon cable 226 It is supplied in parallel to the input of the amplifier 463. Instrument amplifier 463 delivers an output signal to error amplifier 464 in proportion to the signal produced by torsion sensor 192b on the X axis.

As described above, the drive, detection, and control electronics 432 of the port card 406 includes a DSP 465 that executes a computer program stored in a random access memory 466 (" RAM "). In addition, when the value for the direction signal specifying the direction with respect to the mirror surface 116 is stored in the RAM, it is supplied to the lookup table 452 held by the management processor 436. The computer program executed by the DSP 465 retrieves each coordinate, i.e., either the appropriate X or Y axis, and passes the search results to the digital-to-analog converter [467: DAC]. The DAC 467 converts each coordinate retrieved from the DSP 465 in digital data form into an analog signal which the DAC 467 transfers to the input of the error amplifier 464.

The output of the error amplifier 464 integrates the signal in proportion to the difference between the analog signal representing each coordinate and the signal from the instrumentation amplifier 463 proportional to the signal generated by the torsion sensor 192b on the X-axis ( To the input of 472). An integrated circuit 472 consisting of an amplifier 473, a network of resistors 474 and a capacitor 475 directs the output signals to the input of the sum amplifier 476a and the inverting amplifier 477 to the input. The inverting amplifier 477 delivers the output signal to the input of the second sum amplifier 476b. In addition, signals are respectively received directly from the integrated circuit 472 and indirectly from the integrated circuit 472 via the inverting amplifier 477, and the inputs of the sum amplifiers 476a and 476b also receive a fixed bias voltage. Receive. Sum amplifiers 476a and 476b deliver an output signal to the input of the pair of high voltage amplifiers 478, each proportional to the sum of each input signal of the sum amplifiers. The high voltage amplifiers 478 respectively transmit drive signals to either the X-axis or Y-axis electrode 214 of the torsion sensor 172 through the ribbon cable 226.

In this manner, the dual axis servos 454 transmit the differential drive signal to the electrodes of the torsion scanner 172, each symmetrically larger and smaller than the voltage determined by the bias voltage supplied to the sum amplifiers 476a and 476b. Supply. In addition, the drive signal supplied by the dual-axis servos 454 to the electrode 214 is properly corrected to output values from the torsion sensors 192a and 192b and values for the direction signal specified in the lookup table 452. Reduces any difference that may exist between.

At room temperature single crystal silicon is composed of a material whose mechanical properties of torsionally flexible hinges 176 and 182 remain stable for many years because they are not plastically deformed, have no dislocations, have no losses, and do not fracture with fatigue. As a result, the combination of long-term stability for torsionally flexible hinges 176 and 182 and torsion sensors 192a and 192b depends on the direction signal that the lookup table 452 supplies to the pair of dual axis servos 454. The value for affects near precise alignment of the pairs of mirror faces 16.

However, as disclosed in the 463 and 153 patents, the inclusion of an optical servo loop in the optical fiber switch ensures precise alignment. In order to implement such an optical servo loop, as shown in FIG. 26A, each port card 406 provided in the optical fiber switch 400 has one optical detector 426 and one for each optical fiber 106. It has a direction coupler. Each directional coupler 424 couples approximately 5% to 10% of light propagating through one optical fiber provided in the directional coupler 424 to an optical fiber having 95% to 90% of light remaining in the original optical fiber. . As a result, when the light source 422 turns on 5% to 10% of the light emitted by the light source 422 into the directional coupler 424, the optical fiber on the reflection conversion module 100 side ( Some other 95% to 90% of the light propagated along 106 is transmitted into the input optical fiber 106, for example optical fiber 106 ₁ , for delivery onto the reflection conversion module 100. The reflection conversion module 100 couples this combined light from an input optical fiber 106, for example an optical fiber 106 ₁ , to an output optical fiber 106, for example an optical fiber 106 ₂ . When the light arrives, the directional coupler 424 is coupled to the output optical fiber 106, for example optical fiber 106 ₂ , wherein 5% to 10% of the light received from the reflection conversion module 100 is received from the optical fiber 106. From photo detector 426 connected to directional coupler 424 via directional coupler 424.

The optical fiber switch 400 utilizes the capability to introduce light into the optical fiber 106 for delivery through the reflection conversion module 100 and to recover some of the transmitted light to operate certain pairs of mirror surfaces 116. It is desirable to analyze and adjust the condition and to ensure precise alignment of the pairs of mirror surfaces 116 while operating the optical fiber switch 400.

In operation of this optical servo portion with respect to the optical fiber switch 400, the optical servo is directed in a direction in which alignment light propagates through the pair of mirror surfaces 116 (i.e., from the optical fiber 106 to the output optical fiber 106 or Vice versa] aligns the pair of mirror surfaces 116. As a result, in principle, the port card 406 receives only half of the optical fibers 106 provided in the optical fiber switch 400 (e.g., all the input optical fibers 106 or all the output optical fibers 106). It needs to be attached to. However, in telecommunications, in order to facilitate flexible and reliable operation of the optical fiber switch 400, in fact, all the directional couplers 424 connected to the input optical fiber 106 and the output optical fiber 106 must be a light source ( 422 may be mounted.

Referring now to FIG. 26B, outputs from all the directional couplers of port card 406 supply light to a telecom-signal-strength photo-detector [482]. All telecommunication signal strength light detectors 482 are some of the light propagating along the optical fiber 106 into the reflection conversion module 100 regardless of whether the optical fiber 106 is an input optical fiber 106 or an output optical fiber 106. Receive and respond. Thus, before the pair of mirror surfaces 116 are optically precisely arranged, the output signals from the two telecommunication signal intensity light detectors 482 may cause the port card 406 to have a light source 422 for that purpose. It is indicated whether light must be supplied from or whether the input optical fiber 106 can transmit a telecommunication signal of sufficient intensity to arrange the light. If the signal from the pair of telecommunication signal light detector 482 indicates that the two optical fibers 106 are not delivering enough light to arrange the light, then the port card 406 is a light source 422. Is turned on to obtain the light required for the light arrangement, while the light present on the input optical fiber 106 is used for that purpose.

One method for using light introduced into the optical fiber 106 from the light source 422 shown in FIG. 26B is to envision a method of using 850 nm light from a relatively inexpensive laser diode for the light source 422. It became. In this method, the alignment photodetector 484 sensitive to the red wavelength of light may be an inexpensive silicon photodetector. However, the light is also generated at 850 nm by the light source 422 and the input optical fiber 106 also emits light that is probably larger than the optical telecommunication wavelength, e.g. 1310 GHz, or greater than the power generated by the light source 422. Simultaneously transmits at 1550 Hz. In order to ensure separation of the 850 nm aligned light generated by the light source 422 _2j -1 and supplied to the reflection conversion module 100 through the optical fiber 106 _2j -1 from the light at the optical telecommunication wavelength, reflection conversion The output of the directional coupler 424, which emits some of the light received by the photocard 406 from the module 100, directs this light onto the dichroic mirror 486 _2j . The dichroic mirror 4486j reflects the 850 nm alignment light and reflects it to the alignment light detector 484, while light of the optical telecommunication wavelength can be transmitted onto the telecommunication signal monitoring light detector 488. If the reflection conversion module 100 is entirely bidirectional so that a given optical fiber 106 can be an input or output optical fiber 106 at a given moment, the dichroic mirror 4486 _ja is a telecommunication signal monitoring photo detector 488 _{2j-. 1} ) should be used by the directional coupler 424 _2j-1 to separate the light from the light source 422 _2j-1 from the light at the optical telecommunication wavelength it receives.

For various reasons, a pair of mirror surfaces 116 are preferentially optically precisely arranged to achieve a connection through the reflection switching module 100 between the input optical fiber 106 and the output optical fiber 106, and the light source 422 ) And use the light input to the optical fiber switch 400 of the optical telecommunication wavelength in the periodic alignment check. The configuration of light source 422 and light detector 426 is maintained as shown in FIG. 26B. Operating in this manner, the telecommunication signal intensity light detector 482, which first receives light at an optical telecommunication wavelength input into the optical fiber switch 400 through the double socket 404, may provide an optical loss or Detect modulation loss. While operating such an optical fiber switch 400, the telecommunication signal monitoring light detector 488 periodically monitors and maintains the quality of the light transmitted through the reflection conversion module 100. Used in conjunction with The test demonstrates that the direction signal from the torsion sensors 192a and 192b is supplied to the biaxial servo 454 which maintains proper alignment of the mirror surface 116 for an extended period of time, such as several hours. As a result, after the pair of mirror surfaces 116 are optically precisely aligned, the relatively irregular adjustment of the mirror direction is provided with the deflection, temperature change, support frame 224 in the torsion sensors 192a and 192b. To compensate for the mechanical creep of the reflection conversion module 100 and possibly the substrate 212 or the like.

In another method for detecting light alignment supplied at 850 nm from light source 422, dichroic mirror 4482j and light detectors 484 and 488 coupled to dichroic mirror are shown in FIG. 26C. It is substituted with the composite sandwich type photodetector shown. In the composite sandwich detector shown in the figure, the silicon photodetector 492 is mounted over a long wavelength photodetector 494, such as a germanium (Ge) or indium gallium arsenide (InGaAs) photodetector. The composite sandwich type photodetector absorbs shorter alignment wavelengths in the silicon photo detector 492. However, longer wavelengths of optical telecommunication light pass through the silicon photo detector 492 absorbed by the long wavelength photo detector 494 without virtually weakening. Using a complex sandwich photodetector separates the two signals as a whole. The InGaAs photodetector is replaced with a second Ge photodetector to detect long wavelength light, but with less sensitivity than the InGaAs photodetector. However, because the directional coupler 424 has difficulty using 850 nm of light for alignment that results in a multi-mode device, some of the alignment light is coupled into and out of the optical fiber 106 over time.

In order to eliminate the difficulty of using 850 nm light while optically precisely aligning the pair of mirror surfaces 116, optical telecommunication wavelengths, such as 1310 GHz or 1550 GHz from the light source 422, may also be used. It is possible and preferable to supply light. Light at these wavelengths can be provided by an expensive vertical resonant surface emitting laser (vcsel). Vertical resonant surface emitting lasers do not have precise wavelengths or the stability of such inexpensive laser sources, but the precision and stability provided by the laser sources do not require optical alignment of the pair of mirror surfaces 116. Using light at the optical telecommunication wavelength, the alignment light detector 484 can be eliminated, with the advantage that the coupling coefficient for the directional coupler 424 is higher and more stable than the light at 850 nm. Thus, vertical resonant surface emitting lasers need to supply power for less light or light alignment than laser diodes producing 850 nm of light.

If the initial light alignment for the pairs of mirror planes 116 needs to use an inexpensive laser that generates light at the optical telecommunication wavelength for the light source 422, then the cost of that source would be to use a 1 × N array of optical switches. Can be shared between the directional couplers 424. This 1 × N array of optical switches is so large that it can provide light to all port cards 406. Alternatively, to enhance reliability, several such optical telecommunication lasers, each with one smaller 1 × N array of optical switches that provide light only to the directional coupler 424 included in a single port card 406, may be used. It may be provided.

Light beam alignment

In a telecommunications network, an optical fiber switch 400 is provided and achieves the most important reliability and availability. Thus, the mirror surface 116 always forms a connection that preferentially couples light from one optical fiber 106 to another optical fiber 106 through the precise reflection switching module 100, while the connection is continued. It is very important to be under dual-axis servo 454 control where quality is maintained. As described above, in the connection shown in Figs. 26B and 26C, all port cards 406 are mirror surfaces by light generated by one of the light sources or light sources 422 input to the optical fiber switch 400. Provide the ability to monitor precise alignment of the pairs of 116.

The optical fiber switch 400 uses the capability of the port card 406 to align the optical alignment to the pairs of mirror planes 116 by monitoring the quality of coupling between the pairs of optical fibers 106 connected to the reflection conversion module 100. To facilitate. In monitoring the coupling quality, the mineral oil switch 400 tilts each mirror surface 116 in the pair slightly from the direction specified by the value for the direction signal stored in the lookup table 452, ie both mirrors. While dithering the face 116, it simultaneously monitors the intensity of the light beam 108 coupled between the two optical fibers 106. In general, monitoring the intensity of the light beam 108 coupled between two optical fibers 106 requires coordinates between two of at least 36 port cards 406 included in the optical fiber switch 400. The process must be managed at least by the management processor 436 shown in FIG. 26A. Thus, if it is necessary or useful to optically align the pair of mirror surfaces 116, the management processor 436 may issue appropriate commands to the RS232 port 502 provided on the data communication link 438 and each port card 406. Is transmitted to the DSP 465 provided in each related port card 406 shown in FIG. 27B. The command sent by the management processor 436 includes two DACs (e.g., provided by the dual axis servo 454 which the DSP 465 slightly tilts the mirror surface 116 in the direction controlled by the dual axis servo 454). The coordinate data is sent to 467). Since this change in direction changes the light beam 108's contact on the lenses 112 associated with the output optical fiber 106, the amount of light coupled into the combined optical fiber 106 is changed. This change in the light coupled into the optical fiber 106 causes the output light to be delivered to the photo detector 426 provided at the port card 406 and coupled through the directional coupler 424. In order to be able to detect such a change in light, a computer program executed by the DSP 465 from an analog-to-digital converter [504: "ADC") coupled to the light detector 426 as shown in FIG. 27B. Obtain optical density data. In either or both of the DSP 465 or the management processor 436 on the port card 406, the optical fiber switch 400 allows two optical fibers 108 to couple the light beam 108 between the two optical fibers 106. This light density data is analyzed to precisely align the mirror surface 116.

After the mirror surface 116 is optically precisely aligned, the optical fiber switch 400 which checks light from the input optical fiber 106 dithers only the mirror surface 116 to which the input light beam 108 first touches the reflection switching module. Through 100 is coupled to the appropriate output optical fiber 106. When the reflection diverting module 100 is properly aligned and couples light between a particular pair of optical fibers 106, it provides for light from the input light beam 108 caused by dithering this particular mirror surface 116. Density modulation should only appear in the corrected output fiber 106 and not in the other fiber 106.

After the pair of mirror surfaces 116 are optically aligned as described above and after checking the input light coupled into the appropriate optical fiber 106 through the reflection conversion module 100, the optical fiber switch 400 periodically mirrors. The quality of the dithering of the face 116 is used to monitor the connection quality. The computer program executed by the management processor 436 makes use of the alignment data obtained in this manner for updating each coordinate data stored in the lookup table 452 in a timely manner, and also keeps a log of such data. By doing so, the long-term reliability of the optical fiber switch 400 can be analyzed.

FIG. 28A illustrates another embodiment structure for receiving and securing an optical fiber 106 that may be used at the sides 102a and 102b instead of the converging block 152 and the optical fiber collimator assembly 134. In the structure shown in FIG. 28A, the clamp plate 602 micromachining silicon securely stores the optical fiber 106. The throttle plate 604 also has micromachining of silicon and can adjust the ends of the optical fiber 106 that protrude through both from side-to-side and up and down. And secures the end 104 to the adjusted position of the end. Clamp plate 602 is penetrated by an array of holes 606 etched through a 1.0-2.0 mm thick silicon substrate using a Bosch deep RIE process. Holes 606 having a diameter several orders of magnitude larger than the optical fiber 106 typically have a diameter of 100 to 125 μm that matches the outer diameter of a typical optical fiber 106. If the clamp plate 602 is to be thicker than 1.0 to 2.0 mm, two or more plates can be placed and registered in kinematic parallel to each other using the V groove and rod. After being registered, two or more parallelly placed clamp plates 602 may be glued together.

The holes 606 position the optical fibers 106 precisely within a few microns with respect to each other. Holes 606 having a high depth-to-diameter ratio, for example 10: 1 or more, easily fix the optical fiber 106 in the longitudinal direction. In order to easily insert the optical fiber 106 into the hole 606, a pyramidal inlet 608 for the hole 606 shown in FIG. 28A is formed on one side of the clamp plate 602 using anisotropic etching. Can be.

While hole 606 may be formed like a right cylinder, hole 606 may also have a more complex cylindrical profile, as shown in FIG. 28B. Hole 606 may be RIE or wet etched to provide a profile for cantilever 612 that protrudes into hole 606. Since the cantilever 612 is positioned relative to the residual hole 606, inserting the optical fiber 106 into the hole causes the cantilever 612 to bend slightly. In this manner, the cantilever 612 holds the optical fiber 106 firmly against the wall of the hole 606, while allowing the optical fiber 106 to slide along the length of the hole 606. The hole 606 may couple other more complex structures for securing the optical fiber 106 with respect to the hole 606. For example, a portion of each hole 606 may be formed by the profile shown in FIG. 28B, while the remainder etched against the indication from opposite sides of the clamp plate 602 may be shaped like a right cylinder. .

After the clamp plate 602 is manufactured, the optical fiber 106 is inserted through all the holes 606 until all the optical fibers 106 protrude equally several millimeters out of the clamp plate 602, for example, 0.5 to 3.0 mm. do. When the optical fiber 106 is protruded, the optical fiber can be easily bent behind the clamp plate 602. Protruding all the optical fibers 106 equally can be ensured for the stop device during compression and assembly of the ends of the optical fibers 106. The optical fiber 106 may be secured to the clamp plate 602 by simply holding it by gluing, soldering, or frictionally engaging with the cantilever 612.

The control plate 604 shown in FIG. 28C has an XY microstage array that is also etched through a 1.0-2.0 mm thick silicon substrate using a Bosch deep RIE process. Each XY microstage 622 has a hole 624 adapted to receive an end of the optical fiber 106 that projects through the clamp plate 602. The distance between the holes 624 through the adjustment plate 604 is the same as those holes through the clamp plate 602 and can be formed by the profile shown in FIG. 28B. Each optical fiber 106 is securely fastened in the hole 624.

29 shows in more detail one of the XY microstages 622 provided in the adjusting plate 604. A similar monolithic silicon XY stage is disclosed in US Pat. No. 5,861,549 (hereinafter referred to as the "549 patent"), issued January 19, 1999. 29A shows the entirety of an XY microstage 622 formed monolithically from a silicon substrate using RIE etching. The outer base 632 surrounding the XY microstage 622 is described in 1988, Rev. SCI. It is coupled to the intermediate Y axis stage 634 by four bends of the type disclosed in the paper published by Teague et al., Instum., 59, pg.67. Four similar curves 642 couple the Y axis stage 634 to the X axis stage 644. Since the curved portions 636 and 642 are paraflex type, they are appropriately expanded for the XY movement motion envisioned with respect to the hole 624. The XY microstage 622 needs to be able to move and position the ends of the optical fiber 106 over a small distance on the bends 636 and 642 to avoid excessive stress. Other configurations for the curved portions 636 and 642 can also be used, as are these curved portions disclosed in the 549 patent.

The XY microstage 622 easily omits certain actuators, but the Y axis stage 634 can be secured against a metal ribbon, such as gold, kovar, tungsten, molybdenum, aluminum or wire connections 652. Similarly, X axis stage 644 may also be secured relative to Y axis stage 634 with metal ribbon or wire connections 654. The material selected for connections 652 and 654 preferably has an expansion coefficient equal to or close to that of silicon. However, if the connections 652 and 654 are short, for example 100 μm, for a 20 PPM differential expansion coefficient between silicon and metal (eg aluminum), the movement of the X axis stage relative to the outer base 632 is It is approximately 20Å per degree Celsius. Metals other than aluminum provide greater thermal stability.

In adjusting the XY microstage 622, the connections 652 and 654 are first bonded to the Y axis stage 634 and the X axis stage 644, respectively. By looking at the end of the optical fiber 106 through the microscope and simultaneously pulling the metal connections 652 and 654, the X axis stage 644 moves along both the X and Y axes to position the end 104 at a particular location. You can. After the X axis stage 644 moves the end 104 to an appropriate location, the connections 652 and 654 are joined or spot welded to that location. The XY microstage 622 has a lever 622 shown in FIG. 29C to reduce the movement of the X axis stage 644 compared to the movement of the distal end 664 of the XY microstage 622. For the XY microstage 622 shown in the figure, etching to form the stages 634 and 644 yields a cantilevered lever 622 from the Y axis stage 634. The connection 654 is preferentially bonded to both the X-axis stage 644 and the lever 622. Similarly, the end of the lever 622 at the distal end from the lever in which the connection 666 is engaged with the Y-axis stage 634. Is fixed to. The X axis stage 644 properly positions the end 104 before the connection 666 is joined or spot welded to the Y axis stage 634. Alternatively, as shown in FIG. 29C, the connection 654 may be omitted from the XY microstage 622, such that between the X-axis stage 644 and the cantilevered lever 622 from the Y-axis stage 634. It is substituted with a flexible pushpin [672: well known in the art to combine the. Opposite ends of the flexible pushpin 672 are coupled to the X-axis stage 644 and the lever 622 by the curved portion 674, respectively. In the embodiment of the XY microstage 622 shown in FIG. 29C, one connection 666 for securing the X axis stage 644 when the end 104 of the optical fiber 106 is at a particular position of the optical fiber. Need only. Also, the movement of the X axis stage 644 is bidirectional because the flexible pushpin 672 can push and pull on the X axis stage 644.

Although the previous description of the lever 622 described only the movement of the X axis with respect to the X axis stage 644, the similar lever is the movement of the Y axis relative to the Y axis stage 634 with respect to the outer base 632 and the X axis stage. It can be integrated into the outer base 632 to achieve movement of the Y axis relative to 644.

As described above, the XY microstage 622 may fix and adjust the ends of the optical fiber 106 along the X and Y axes of the XY microstage. However, properly focusing the lens with respect to the end 104 of the optical fiber 106 may require associated movement of either the end 104 or the lens 112 along the longitudinal axis 144. . The separation between the end 104 of the optical fiber 106 and the lens 112 can be adjusted in several different ways. SPIE Proc., Vol. A paper published by Bright et al. In 2687, pg. 34, discloses a polycrystalline silicon mirror that behaves like a piston that can be placed electrostatically perpendicularly onto a substrate from which it is manufactured.

FIG. 30A shows that the monolithic one-sided surface can be disposed electrostatically along the longitudinal axis 144 perpendicular to the substrate on which the one-sided convex lens is manufactured, and micromachined the SOI wafer 162 using RIE etching. The convex lens 112 is shown. In order to be able to statically displace the lens 122 along the longitudinal axis 144, as shown in FIG. 30B, the lens 112 is connected to the wafer 162 by three V-shaped curved portions 682. Is supported around the device silicon layer 166. One end of each curved portion 682 extending to a portion of the periphery of the lens 112 is coupled to the peripheral silicon layer 166, while the other end is coupled to the lens 112. Except for the deflection electrode 684 disposed on the right side of the lens 112 shown in FIG. 30A and electrically insulated from the wafer 162, the entire assembly is constructed as one monolithic silicon structure. Electrostatic attraction between the curved portion 682 and the lens 112 coupled with the electrode 684 produced by supplying an electrical potential between the electrode 684 and the device silicon layer 166 is along the longitudinal axis 144. Pull the lens to the (684) side.

Silicon lenses suitable for IR fiber transmission are commercially available and can be adapted for use in the present invention. Thus, small individual commercially available microlenses can be placed in a cavity that is etched into a flat membrane supported by curved portion 682. Alternatively, lens 112 may be formed using RIE, while curved portion 682 is formed. Another embodiment is also to first rotate the lens 112 in diamond shape while protecting the lens from etching while forming the bend 682 using RIE. Further, another embodiment first uses the RIE to form the curved portion 682 while protecting the area where the lens 112 is formed, and rotating the lens 112 in a diamond shape. After the lens 112 and the bend 682 are formed in any such manner, the wafer 162 under the lens and the bend is removed by anisotropic etching to expose the silicon dioxide ring 164. The backside of the lens 112 manufactured in this manner is optically flat.

Instead of the electrostatic actuation action, the lens 112 can move electromagnetically along the longitudinal axis 144. As shown in FIG. 30C, the electrode 684 disposed adjacent to the lens 112 shown in FIG. 30A is permanently directed by the magnetic field of the electrode parallel to the longitudinal axis 144 of the lens 112. Replaced by a magnet 692. Coil 694 also surrounds lens 112. It is preferred that electrical leads from coil 694 appear symmetrically to device silicon layer 166 through bend 682 to ensure linear displacement of lens 112. According to the direction of the current supplied to the coil 694, the lens 112 moves to or from the end 104 side of the optical fiber 106.

In many telecommunication applications for the optical fiber switch 400, the light reaching the optical fiber switch 400 may be delivered through a routing wavelength demultiplexer, which may typically be in the form of an integrated chip in advance. The effective cost in manufacturing the routed wavelength demultiplexer is the cost of connecting from the planar circuit of the routed wavelength demultiplexer to the output optical fiber. When the reflection switching module 100 of the optical fiber switch 400 is properly configured, there is no need to configure a connection between the routing wavelength demultiplexer and the optical fiber. In addition, the output light beam from the routing wavelength demultiplexer is coupled to the free space for the lens 112 of the reflection conversion module 100, which may be provided with an antireflective coating that simply reduces reflection.

FIG. 31 shows an arrangement of routing wavelength demultiplexer 702 with several demultiplexed planar waveguides 704. The demultiplexer planar waveguide 704 emits the light beam 108 directly to the lens 112 side that is in contact with the demultiplexer planar waveguide, thereby eliminating any need to couple the routing wavelength demultiplexer 702 to the optical fiber. A substrate 706 of the routing wavelength demultiplexer 702 leading to the demultiplexer type planar waveguide 704 may be positioned adjacent the lens 112 to supply the input light beam 108 to the reflection conversion module 100. Similarly, if the output light beam 108 remains in the reflection conversion module 100, the lens 112 directly couples the light beam 108 to a demultiplexer type planar waveguide 704 in which the light beam can be multiplexed into one or several output optical fibers. can do. By providing and preserving some redundant output and input holes 154 in a converging block 152 for use by the wavelength converter, the optical fiber switch 400 is received from a predetermined optical fiber coupled to the optical fiber switch 400. It can provide wavelength conversion for light.

Although the present invention has been described with respect to the presently preferred embodiments, it is to be understood that this detailed description is purely illustrative and not restrictive. As a result, without departing from the spirit and scope of the present invention, there is no doubt that various changes, modifications and / or applications other than the present invention can be suggested to those skilled in the art after reading the above detailed description. Accordingly, the following claims are intended to embrace all such alterations, modifications or other applications that fall within the true spirit and scope of this invention.

Claims (20)

An optical fiber conversion module, First and second optical fiber storage groups separated from one another at opposite ends of the free space optical path; Lenses disposed in parallel at optical fiber ends fixed to the first and second optical fiber storage groups, respectively, and arranged along an optical path between the optical fiber storage groups; First and second reflected light beam deflector sets disposed on both sides along an optical path between the optical fiber reservoirs; Each said optical fiber storage group is adapted to receive and fix ends of optical fibers, respectively; Since each of the lenses is disposed with respect to an end of the optical fiber coupled to the first and second optical fiber storage groups, respectively, a light beam that can be emitted from the end of the optical fiber is transmitted through the immediately adjacent lenses so that the second or second 1 propagates as a semi-balanced beam through the light path from the lenses on the optical fiber reservoir group side; The respective light beam deflector sets are coupled to one of the optical fiber reservoirs and have the same number of light beam deflectors as the optical fiber in the optical fiber reservoir group to which one optical fiber reservoir is coupled, The first or second light beam deflector set, Is combined with one of the optical fibers in the combined optical fiber storage group; A quasi-equilibrium light beam disposed along the light path and capable of being emitted from lenses coupled to the optical fiber is reflected from the lenses through the light path to reach the light beam deflector; In the second or first light beam deflector set may be activated by drive signals supplied to the optical fiber conversion module directed to reflect the quasi-balanced light beam that may be emitted from the combined optical fiber. Reflects off and disappears from a selected one of the light beam deflectors of A pair of light beam deflectors, one light beam deflector for the pair belonging to the first light beam deflector set and one belonging to the second light beam deflector set fixed to the second or the first optical fiber storage group One optical fiber that can be secured to the optical fiber reservoir of either the first or the second optical fiber reservoir group to be sequentially reflected and extinguished in the pair of activated light beam deflectors to retract the optical fiber that can be made. An optical fiber switching module that can be selected and oriented by the drive signal supplied to the light beam deflectors to couple the quasi-balanced light beam propagating through the light path from an end into a selected one of the optical fiber reservoirs. The optical fiber switching group of claim 1, wherein the first optical fiber storage group is located at side A of the optical fiber switching module, the second optical fiber storage group is located at side B of the optical fiber switching module, and the side A is spaced from side B. And the first and second light beam polarizer sets are also spaced apart from each other. The optical fiber conversion module according to claim 2, wherein the optical path between side A and side B is C type. The optical fiber conversion module according to claim 2, wherein the optical path between side A and side B is Z-shaped. The optical fiber conversion module according to claim 2, wherein the optical path between side A and side B is W type. 3. The optical fiber conversion module of claim 2, wherein a mirror is disposed between the light beam polarizer sets to fold the light path between the light beam polarizer sets. The optical fiber conversion module of claim 1, wherein a mirror is disposed between the light beam polarizer sets to fold the light path between the light beam polarizer sets. The optical fiber of claim 1, wherein the first optical fiber storage group has only one optical fiber storage and the second optical fiber storage group has the remaining optical fiber storage, so that the optical fiber conversion module can be fixed to one optical fiber storage. Coupling a light beam from an end of the optical fiber to an end of a predetermined optical fiber that can be secured to the second optical fiber storage group. The optical fiber storage module of claim 1, wherein the lenses included in the optical fiber conversion module have facing surfaces oriented at an angle inclined with respect to the longitudinal axis of each lens, so that they are stored in the optical fiber reservoirs emitted at an angle to the centerline of the optical fiber. Wherein the light beams from the ends of the optical fiber may be aligned with the longitudinal axis of the lenses. The lens of claim 1, wherein the lenses provided in the optical fiber conversion module are formed with a smaller diameter outer surface disposed closer to an end of the optical fiber that can be fixed to one of the optical fiber reservoirs. And a larger diameter outer surface disposed further from an end of the optical fiber that can be secured to one of the optical fiber reservoirs than the smaller diameter outer surface of the lenses. The optical fiber collimator of claim 1, wherein the individual optical fiber reservoirs are conical and adapted to receive a separate, paired, conical optical fiber collimator assembly that carries lenses coupled to an optical fiber that can be secured to the optical fiber reservoir. The optical fiber collimator assembly is adapted to receive and fix an end of the optical fiber collimator assembly. 12. The optical fiber conversion module of claim 11, further comprising an environmental housing surrounding the light path through which the light beam propagates. 13. The optical fiber conversion module of claim 12, wherein the environmental housing provides a temperature regulation to maintain a stable operating environment for the optical fiber conversion module. 13. The optical fiber conversion module of claim 12, wherein the drying gas passing through the environmental housing prevents moisture from condensing in the optical fiber conversion module. 13. The optical fiber conversion module of claim 12, wherein the environmental housing is pressurized to exhaust air around the environmental housing that houses the optical fiber conversion module. 13. The optical fiber conversion module of claim 12, wherein the environmental housing includes a microdryer that is unsaturated to prevent moisture from condensing in the optical fiber conversion module. 13. The optical fiber conversion module of claim 12, wherein the wall of the environmental housing is penetrated by an electrical supply through which the drive signals are transmitted. As an optical fiber switch, A plurality that can be selected as pairs oriented in response to drive signals supplied to the optical fiber conversion module for receiving and fixing ends of the optical fibers and for coupling a light beam between a pair of optical fibers fixed to the optical fiber conversion module. An optical fiber conversion module having a reflected light beam deflector; At least one port card for supplying the driving signals to the optical fiber switching module for directing at least one light beam deflector provided in the port card, and receiving the direction signals generated by the light beam deflector, The optical fiber conversion module also generates direction signals from each light beam deflector indicating the direction of the optical fiber conversion module, The port card also receives data specifying the direction for the light beam deflector, includes these received data with the direction signals received from the light beam deflector, and adjusts the drive signal supplied to the optical fiber conversion module. Thereby reducing a predetermined difference between the received data and the direction signals. An optical fiber switching module for receiving and fixing ends of the optical fibers, the optical fiber switching module being directed in response to drive signals supplied to the optical fiber switching module for coupling the light beam between a pair of optical fibers fixed to the optical fiber switching module. A port card adapted for use in an optical fiber switch having a plurality of reflected light beams that can be selected as pairs, wherein the fiber optic switching module is also adapted for use in generating optical signals from each light beam deflector indicating the direction of the light beam deflector, The port card, A driver circuit for supplying the drive signals to the optical fiber switching module for directing at least one light beam deflector provided in the driver circuit; A dual axis servo that specifies a direction to the light beam deflector and also receives direction signals generated by the light beam deflector, And comparing the data with the direction signals received from the light beam deflector and adjusting the drive signals supplied to the optical fiber conversion module to reduce a predetermined difference between the received data and the direction signals. 20. The port card of claim 19, wherein said driver circuit supplies electrostatic drive signals to said fiber optic switching module.