WO2003089971A2

WO2003089971A2 - Optical device with alignment compensation

Info

Publication number: WO2003089971A2
Application number: PCT/US2003/012252
Authority: WO
Inventors: Steven K. Case; Timothy A. Skunes; Michael Knipfer; Gregory S. Mowry
Original assignee: Cyberoptics Corporation
Priority date: 2002-04-22
Filing date: 2003-04-22
Publication date: 2003-10-30
Also published as: WO2003089971A3; GB0424352D0; AU2003239153A1; CN1646960A; GB2403817A; GB2403817B

Abstract

An optical device (120) is provided which includes a plurality of optical modules (122, 126) and alignment compensation modules (152). The alignment compensation modules (152) remove residual alignment errors of the optical device.

Description

OPTICA DEVICE WITH ALIGNMENT COMPENSATION

BACKGROUND OF THE INVENTION The present invention relates to optical components and optical devices fabricated from such components. More specifically, the present invention relates to alignment compensation of such devices.

Optical devices are being increasingly used in various industries and technologies in order to provide high speed data transfer such as a fiber optic communication equipment. In many applications there is a transition or an incorporation of optical devices where previously only electrical devices were employed. An optical device typically consists of a number of components which must be precisely assembled and aligned for the device to operate and function efficiently. Example components include fibers, waveguides, lasers, modulators, detectors, gratings, optical amplifiers, lenses, mirrors, prisms, windows, etc. Historically, optical devices such as those used in fiber optic telecommunications, data storage and retrieval, optical inspection, etc. have had little commonality in packaging and assembly methods. This limits the applicability of automation equipment for automating the manufacture of these devices since there is such a disparity in the device designs. To affect high volume automated manufacturing of such devices, parts of each individual manufacturing line have to be custom-designed. In contrast, industries such as printed circuit board manufacturing and semiconductor manufacturing have both evolved to have common design rules and packaging methods. This allows the same piece of automation equipment to be applied to a multitude of designs. Using printed circuits as an example, diverse applications ranging from computer motherboards to cellular telephones may be designed from relatively the same set of fundamental building blocks. These building blocks include printed circuit boards, integrated circuit chips, discrete capacitors, and so forth. Furthermore, the same automation equipment, such as a pick and place machine, is adaptable to the assembly of each of these designs because they use common components and design rules.

Further complications arise in automated assembly of optical devices. Such assembly is complicated because of the precise mechanical alignment requirements of optical components. This adds to problems which arise due to design variations. The problem arises from the fact that many characteristics of optical components cannot be economically controlled to exacting tolerances. Examples of these properties include the fiber core concentricity with respect to the cladding, the location of the optical axis of a lens with respect to its outside mechanical dimensions, the back focal position of a lens, the spectral characteristics of a thin-film interference filter, etc. Even if the mechanical mounting of each optical element were such that each element was located in its exact theoretical design position, due to the tolerances listed above, the performance specifications of the optical device may not be met.

To appreciate the exacting alignment requirements of high performance optical devices, consider the simple example of aligning two single mode optical fibers. In this example, the following mechanical alignments are required to ensure adequate light coupling from one fiber to the other: the angle of the fibers with respect to each other, the fiber face angle, the transverse alignment (perpendicular to the light propagation direction) and the longitudinal spacing (parallel to the light propagation direction) .

Typical single mode optical fibers used in telecommunications for the 1.3 μ to 1.6 μm wavelength range have an effective core diameter of about 9 microns and an outside cladding dimension of 125 microns. The typical tolerance for the concentricity of the core to the outside diameter of the cladding is 1 micron. If the outside claddings of the two fibers were perfectly aligned and there is no angular misalignment or longitudinal spacing, the cores may still be transversely misaligned by as much as 2 microns. This misalignment would give a theoretical coupling loss of about 14 percent or 0.65 dB. This loss is unacceptable in many applications. It would be desirable to provide an optical device which addresses some of the deficiencies of the prior art . SUMMARY OF THE INVENTION

In one example aspect, an optical device is provided which comprises a plurality of optical modules and an alignment compensation module. Each optical module includes an optical component to operably couple to a relative reference mount. The relative reference is configured to couple to a fixed reference mount. A plurality of optical modules mounted on the fixed reference mount form the optical device. The alignment compensation module removes residual alignment errors of the optical device.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a perspective view of a fiber optic demultiplexer device. '

Figure 2 is an optical schematic representation of the fiber optic demultiplexer device of Figure 1.

Figure 3 is an exploded perspective view of the fiber optic demultiplexer device of Figure 1.

Figure 4 is a side sectional view showing a fiber optic collimator module. Figure 5 is a bottom view of a relative reference showing optical module registration features . Figure 6 is an enlarged top view of the fixed reference from Figure 3 that shows fixed reference registration features.

Figure 7 is a side sectional view of an optical filter module.

Figure 8 is an optical schematic representation of a fiber optic demultiplexer with alignment compensation modules.

Figure 9 is an optical schematic representation showing a residual alignment error and an alignment compensation module with an afocal lens pair.

Figure 10 is an optical schematic representation showing the removal of an alignment error by linearly displacing one of the lenses of Figure 9. Figure 11 is a perspective view of an alignment compensation module with a lever arm alignment structure .

Figure 12 is a front plan view of the alignment compensation module of Figure 11. Figure 13 is an optical schematic representation of an alignment compensation module that uses a rotated, afocal lens pair to remove an alignment error .

Figure 14A is an optical schematic representation showing a residual alignment error and an alignment compensation module with an afocal lens.

Figure 14B is an optical schematic representation showing the removal of an alignment error by rotating the lens shown in Figure 14A. Figure 15 is a side sectional view of an alignment compensation module with a single lens in an adjustable, rotating mount.

Figure 16 is a perspective view of a fiber optic demultiplexer device with alignment compensation modules .

Figure 17 is an optical schematic representation of an alignment compensation module with Risley wedge prisms . Figure 18 is an optical schematic representation of a laser transmitter with a single, tilted, optical window for removing alignment errors.

Figure 19 is an optical representation of a laser transmitter with two tilted, optical windows for removing alignment errors.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention includes various aspects that reduce or eliminate many of the problems associated with the prior art. The present invention offers an optical device fabricated from optical modules which are prealigned in standardized optical modules. Each optical module can be aligned with sub- micron precision with respect to registration features. Registration features on a module can be aligned with matching features on a substrate. This is similar to mounting an electrical component in or on a printed circuit board. Optical devices can be easily fabricated by mounting prealigned optical modules in the optical "circuit board" and using align ent compensation modules. The prealign ent of the optical module can compensate for variations between components to substantially reduce component variability. The use of alignment compensation modules greatly relaxes optical module and substrate tolerances at the small additional complexity of performing a final alignment with an alignment compensation module. The prealigned optical modules are well suited for automated fabrication of devices. The modules can be fabricated in silicon using techniques which are well known in the art of silicon processing. However, any appropriate material can be used. Preferable materials are those which are used with existing electrical or optical components. Further, the invention can be used with active devices such as lasers, modulators, detectors, etc.

Electrical conductors can be fabricated on the various layers for coupling to active optical components. Electrical circuitry including analog and digital circuitry can also be fabricated directly on the modules or on the fixed reference mount.

In one aspect, the present invention provides an optical device formed from at least two optical modules in which optical components are mounted to optical component mounts. The optical component mount is fixed to a relative reference mount such as a base mounting plate at a desired position and orientation. The relative reference mount is coupled to a fixed reference mount such as a substrate such that the optical component is maintained at a desired position and orientation relative to the fixed reference mount. In this general configuration, the optical component can be pre-aligned to a desired spacial reference and orientation by adjusting the optical component mount relative to the reference mount prior to fixing their relative positions. This can be used to provide general component pre-alignment as well as compensate for the variations which can arise between optical components. The following description sets forth a number of specific examples, however, in various aspects,, the present invention is not limited to the specific configurations, components or techniques set forth herein. In general, an optical device is fabricated from two optical modules which include respective optical components. The components are mounted to the optical component mounts which are positioned and oriented to achieve a desired osition and orientation of the optical components relative to base mounting plates

(also referred to as relative reference mounts) . The base mounting plates mount to a reference substrate such that the optical components are in substantial alignment. A substrate is one example of a fixed reference mount and any appropriate fixed reference mount with an appropriate shape and configuration can be used.

The optical component modules of the present invention can be pre-assembled and pre-aligned to an appropriate reference such that a final optical device is fabricated by simply mounting the assembled optical modules on the reference substrate. An alignment compensation module (discussed below) can also be employed to assist in aligning the optical device. For example, the substrate can be in a planar substrate which can be thought of as an optical "circuit board" which receives optical modules to form an optical, opto-electrical or opto-mechanical device.

The present invention provides prealigned optical modules which can reduce or eliminate the effects of component variability. This is achieved by adjusting a component mount (holder) relative to a registration feature on a base mounting plate. The relationship between the component and the registration feature is then fixed. Precise registration features are provided on the base mounting plate such that it can be inserted into an optical "circuit board" to fabricate devices which comprise multiple optical component modules. The optical modules are well suited for automated assembly of optical devices because they are in standardized packages, prealigned and can be easily mounted on a reference substrate. Optical modules can be manually placed into the optical "circuit board" or the process can be automated. The particular optical modules are typically standardized to facilitate such automation. Further, this configuration allows assembly of devices in a "top downward" fashion in which optical modules are moved downward into an optical "circuit board" which facilitates process automation. The present invention provides an optical device comprising a plurality of optical modules in which optical variations due to component variability are eliminated or significantly reduced. This provides uniformity across multiple optical modules which is particularly desirable for automated assembly. In one aspect, the invention can be viewed as providing several stages of alignment of an optical device. A first stage of alignment is provided between the component mount (holder) and the optical component, for example using a V-groove registration feature as shown or other technique. A second stage of alignment is between the optical component mount and registration features of the relative reference mount. This also eliminates or reduces alignment variations due to component variability. A third stage of alignment occurs between the optical module and the reference substrate. A final alignment stage occurs using an alignment compensation to remove residual alignment errors. In another example aspect, the optical element has an optical characteristic which varies in space relative to at least one dimension. The optical component is aligned with reference features on the relative reference mount by fixing the position of the component mount relative to the registration features of the relative reference mount to thereby align the optical characteristic. In one aspect, the first stage of alignment is eliminated and the optical element is directly aligned with the registration features of the relative reference mount and no mount/holder is used.

The use of alignment compensation modules greatly relaxes optical module and fixed reference tolerances at the small additional complexity of performing a final alignment with an alignment compensation module. In one aspect, optical modules are prealigned such that errors in the alignment of an optical characteristic of the module due to component variability are greatly reduced. Active prealignment can be performed in which light interacts with the optical component and is used to adjustment the alignment of the component to compensate for errors introduced due to variability of the component. Such active prealignment refers to alignment in which light is actively passed through, reflected by, received by or generated by an optical component and that component is aligned with respect to registration features of a relative reference mount which holds the component. An example fiber optic demultiplexer and laser transmitter will now be presented for illustrating alignment compensation modules. However, alignment compensation modules may be used with other optical devices and other types of optical components.

Figure 1 is a perspective view of fiber optic demultiplexer 120. The operation of demultiplexer 120 will be briefly described by referring to the optical schematic of Figure 2. Input fiber 124A typically carries several signals that are encoded onto different optical wavelengths. The output of fiber 124A is collimated by lens 122A. Optical filters 126 are typically coated with multiple dielectric layers that transmit certain portions of the optical spectrum and reflect the remaining portions. Optical filter 126A transmits a certain portion of the optical spectrum in order to extract or demultiplex one or more signals from the signals encoded onto input fiber 124A. Light transmitted by optical filter 126A is focused by lens 122B onto output fiber 124B. The combination of filter 126A and lens 122B demultiplex the desired signal (s) onto output fiber 124B. Additional signals are sequentially demultiplexed by optical filters 126B, 126C, and 126D and lenses 122C, 122D, and 122E respectively. Any remaining signals are then focused by lens 122F onto fiber 124F. The exploded perspective view of demultiplexer 120 in Figure 3 shows prealigned fiber optic collimator modules 131 and prealigned optical filter modules 133. Figure 4 is a side sectional view of fiber optic collimator module 131. Lens 122, strain relief material 123, and fiber 124 form fiber optic collimator 125 such as those manufactured by Lightpath Technologies of Orlando, FL. Collimator pointing error tolerances of 0.1 - 1.0 degrees are common. To a first order approximation, the pointing error of a fiber optic collimator is the angular propagation direction of the collimated beam that maximizes the amount of light coupled into the fiber as measured with respect to the angle def±ned by the outside mechanical dimensions of lens 122 or another mechanical feature such as a cylindrical metal tube that houses lens 122. Lens 122 is attached to spherically shaped lens mount 130. Features on top of lens mount 130, such as a v-groove or channel, aid in securing and aligning lens 122 to mount 130. Receptacle 142 in relative reference mount 140 receives mounts 130. Receptacle 142 is shown as conical shaped depression in relative reference 140.^' Mount 130 swivels in receptacle 142 in the θ_x direction and θ_γ direction to affect angular alignment of collimator 125. Collimator module registration features 144, shown as protrusions, are provided on relative reference 140. Collimator module registration features 144 mate with matching registration features on fixed reference 128. In one embodiment, collimator module registration features 144 substantially constrain movement in six degrees of freedom when physically coupled to the mating registration features (shown in Figure 3) in fixed reference 128. The pointing axis of collimator 125 is actively prealigned in the θ_x direction and θ_γ direction relative to optical module registration features 144. Upon proper angular alignment of collimator 125 with respect to registration features

144, mount 130 may be secured to relative reference

140 by appropriate means such as adhesive, solder, welding, or other appropriate attachment technique. A substantial percentage of the aforementioned collimator pointing error is removed prior to inserting fiber optic collimator module 131 into fixed reference 128. Receptacle 142 may have other shapes such as spherical sockets, holes, or holes with chamfers that allow mount 130 to swivel on the receptacle. Collimator 125 may also be secured to a mount that contains a receptacle and the relative reference may contain a spherical feature to permit angular adjustment of the collimator 125.

Figure 5 is a bottom view of relative reference 140, from Figure 4, showing optical module registration features 144. Figure 6 is an enlarged top view of fixed reference 128 from Figure 5 that shows fixed reference registration features 145 that mate with registration features 144 of optical module 131. Fixed reference registration features 145 are shown as v-groove depressions in fixed reference 128. This configuration provides an example of a kinematic-type registration or alignment technique. V-grooves 145 and optical module registration features 144 may be made with high precision, for example, by anisotropic etching of silicon using well-known techniques. One example kinematic technique is described in U.S. Patent No. 5,748,827, entitled "TWO-STAGE KINEMATIC MOUNT". In another embodiment, optical module registration features are spherically shaped protrusions that mate with v- groove registration features aligned at 120 degree intervals with respect to each other on fixed reference 128. The optical module registration features typically constrain movement in six degrees of freedom when physically coupled to the mating fixed reference registration features.

Figure 7 is a side sectional view of prealigned optical filter module 133. Optical filter 126 is attached to spherically shaped filter mount 132. Receptacle 148 in relative reference mount 146 receives mount 132. Mount 132 swivels in receptacle 148 in the θ_x direction and the θ_γ direction to affect angular alignment of optical filter 126. Optical module registration features 150 are provided on relative reference 146. The angle of optical filter is actively prealigned in the θ_x direction and θ_γ direction relative to optical module registration features 150. Upon proper angular alignment of optical filter 126 with respect to optical module registration features 150, mount 132 may be secured to relative reference 146 prior to inserting relative reference 146 into fixed reference 128 by appropriate means described above such as adhesive, solder, welding, or other appropriate attachment technique. Fixed reference 128 has registration features as shown in Figure 3 that mate with optical module registration features 150. The optical module registration features 150 typically constrain movement in six degrees of freedom when physically coupled to the mating registration features in fixed reference 128. Lens 122 and filter 126 may be secured to their respective mounts by appropriate means such as adhesive, solder, welding, or other appropriate attachment technique. Mounts 130 and 132 may be transparent to allow appropriate radiation to secure mounts 130 and 132 to receptacles 142 and 148, respectively, such as with adhesive or by laser soldering. Relative references 140 and 146 may also be transparent to facilitate adhesive curing or la.ser soldering. The mounts and relative references may be of appropriate materials, or a combination of materials, such as metal, glass, ceramic, semiconductor, or plastic and have coatings to facilitate bonding of the mounts to the relative references. Lens mounts 130 and filter mounts 132 may also contain additional mechanical features to aid in gripping, manipulating and aligning the mounts. The various mounts and fixed references may also be made by molding. Many sources of alignment error may accumulate during the manufacture of an optical device such as that of demultiplexer 120. These error sources include slight errors associated with actively prealigning and securing optical components and optical component mounts into optical modules, slight mismatches between the optical module registration features and the fixed reference registration features, and small errors in the position, size, and orientation of the fixed reference registration

•features. For example, degradation of the light coupling efficiency into fiber 124F shown in Figures

1 and 2 may occur due to the error sources just mentioned for the various collimator modules 131, optical filter modules 133, and fixed reference 128.

The statistical accumulation of these errors may leave a residual alignment error. Use of an alignment compensation module substantially removes residual alignment errors as will be explained with reference to Figures 8-12.

Figure 8 is an optical schematic representation of demultiplexer 121. Demultiplexer 121 is similar to demultiplexer 120 shown in Figure 2 with addition of alignment compensation modules 152, shown in block diagram form, between optical filters 126 and lenses 122.

An optical schematic of an alignment compensation module 152 is shown in Figure 9. Substantially collimated light beam 141, shown propagating along a Z axis, passes through alignment compensation module 152 where it is refracted by lenses 154 and 156 to produce substantially collimated light beam 143. Light beam 143 is then focused by lens 122. A residual alignment error is also illustrated in Figure 9, albeit greatly exaggerated. That is, there is a residual alignment error between the position of the light focused by lens 122 and the core of fiber 124. Lens 154 is a relatively weak lens with slightly positive focal length and lens 156 is a relatively weak lens with slightly negative focal length. Together, lenses 154 and 156 form an afocal pair and do not appreciably change the state of collimation of a light beam. In Figure 10, lens 154 has been displaced in the Y direction, compared to its original position in Figure 9. The angular propagation direction of light beam 143 is now changed, relative to the original angular propagation direction of light beam 141, so that light focused by lens 122 is displaced in the Y direction and coupled into the core of fiber 124 with high efficiency. Lens 154 may also be displaced in the X direction to compensate for alignment errors in the X position of the focused light beam. Displacing, or decentering, lens 156 has a similar effect on alignment compensation, although it must be displaced in the opposite direction of lens 154 since it has a negative focal length. The angular propagation direction of light beam 143 entering lens 122 may be adjusted by appropriately displacing lenses 154 and 156 in the X-Y plane in order to compensate for residual alignment errors. In one embodiment, lenses 154 and 156, of alignment compensation module 152, are displaced independently and in substantially orthogonal directions to improve convergence of the alignment compensation.

Figure 11 shows a perspective view of an alignment compensation module 152 on a fixed reference 129. Fixed reference 129 is similar to fixed reference 128 shown in Figure 1 with the addition of alignment compensation modules in front of lenses 122B, 122C, 122D, 122E, and 122F. Lenses 154 and 156 are mounted to tilt plates 159A and 159B, respectively. This is an example of a lever arm alignment structure for adjusting and securing lenses 154 and 156. A front view of alignment compensation module 152 is shown in Figure 12. Pivots 161A and 161B rest in v-grooves 163 of fixed reference 129. V- grooves 167A and 167B mate with pivots 161A and 161B and allow tilt plates 159A and 159B, respectively, to rotate in the θ_z direction. As tilt plate 159A rotates in the θ_z direction, the center of lens 154 is displaced along a nearly linear path labeled A-A. Likewise, as tilt plate 159B rotates in the θ_z direction, the center of lens 156 is displaced along a nearly linear path labeled B-B that is substantially perpendicular to path A-A. Residual alignment errors may be substantially eliminated by tilting plates 159A and 159B to remove residual alignment errors and then secured in place with solder 165A and 165B, respectively, or other appropriate means such as with adhesive or by welding. Other mechanical configurations are possible to allow tilt plates 159A and 159B to rotate, such as ball bearings in receptacles.

The focal lengths of lenses 154 and 156 can be chosen such that displacements of lenses 154 and 156 result in much smaller displacements of focused light in the X-Y plane of fiber 124 as can be seen from Figures 9 and 10. Choosing the focal lengths of lenses 154 and 156 in this fashion eases the accuracy and resolutions requirements for manipulators that tilt plates 159A and 159B. Also, any displacement errors of lenses 154 and 156 that occur during the securing of plates 159A and 159B result in negligible displacement errors of focused light in the X-Y plane of fiber 124. Alignment compensation module 152 shown in Figures 11 and 12 allows "top downward" assembly onto fixed reference 129 as well as manipulation of tilt plates 159A and 159B from above which is advantageous for an automated assembly system.

In another aspect, alignment compensation module 152 includes afocal lens pair 154 and 156 that may be rotated together to compensate for residual alignment errors as shown in Figure 13. This rotation, or tilt, has the effect of displacing lenses 154 and 156 in opposite directions in the X-Y plane. As shown in Figure 13, afocal lens pair 154 and 156 has been rotated in the θ_x direction to remove an alignment error in the Y direction at the core of fiber 124. Similarly, a rotation of afocal lens pair 154 and 156 in the θ_γ direction will remove residual alignment errors in the X direction at the core of fiber 124. Rotating lens pair 154 and 156 in the θ_x direction and the θ_γ direction changes the angular propagation direction of light beam 143 and allows residual alignment errors to be removed.

In one example embodiment, alignment compensation module 152 includes a single lens 158 that may be rotated in the θ_x direction and θ_γ direction to compensate for residual alignment errors as shown in Figures 14A and 14B. An optical schematic of an alignment compensation module 152 is shown in Figure 14A. Substantially collimated light beam 141, shown propagating along a Z axis, passes through alignment compensation module 152 where it is refracted by lens 158 to produce substantially collimated light beam 143. Light, beam 143 is then focused by lens 122. A residual alignment error is also illustrated in Figure 14A, albeit greatly exaggerated. That is, there is a residual alignment error between the position of the light focused by lens 122 and the core of fiber 124. Lens 158 is afocal and does not appreciably change the state of collimation of light beam 141. As shown in Figure 14B, lens 158 has been rotated in the θ_x direction to remove an alignment error of the focused light in the Y direction at the core of fiber 124. Similarly, a rotation of lens 158 in the θ_γ direction will remove residual alignment errors in the X direction at the core of fiber 124. Rotating lens 158 in the θ_x direction and θ_γ direction changes the angular propagation direction of light beam 143 in the θ_x direction and the θ_γ direction, respectively. Residual alignment errors may be removed by appropriately rotating lens 158 in the θ_x direction and the θ_γ direction.

Lens 158 has curved surfaces 174 and 176. Surfaces 174 and 176 may have either spherical or aspherical curvatures. Curvatures 174 and 176 are selected so that relatively large rotations of lens 158 produce small angular deviations in the propagation direction of light beam 143. Lens 158, designed to be afocal, does not substantially change the degree of collimation of incoming light beam 141. Rotations of alignment compensation module 152 in the θ_x direction and the θ_γ direction are used to change the angular propagation direction of light beam 143 and slightly displace the focused position of the light to compensate the optical beam alignment with respect to the core of optical fiber 124.

Alignment compensation module 152 may use an afocal optical system that is rotated in the θ_x direction and θ_γ direction when used in a substantially collimated light beam. In this aspect, the optical elements of alignment compensation module 152 are not limited to one or two refractive elements, but may also be a combination of one or more reflective, refractive, and diffractive elements to form an afocal system that deviates the angular propagation direction of substantially collimated light beam when this combination is rotated in the θ_x direction and θ_γ direction. This is in contrast to a plane parallel plate, that when rotated in the θ_x direction and θ_γ direction, laterally shifts the location of an optical beam, but does not change its angular propagation direction. Alignment compensation module 152 can be designed such that shifts of mounts 166 that inevitably occur after alignment and securing affect final alignment much less than shifts of the optical components themselves such as mirrors 126 and collimators 125. For example, applicants have found it advantageous that the optical elements in alignment compensation module 152 need to be rotated about five to ten times further in angular measure to remove residual alignment errors than if collimator 125 were rotated directly to remove a residual alignment error. This reduces the sensitivity of the final alignment, making it easier to align, and reduces affects due to any mount shifts within alignment compensation module 152.

Figure 16 is a perspective view of demultiplexer 121 from Figure 8 showing alignment compensation modules 152. A side sectional view of alignment compensation module 152 is shown in Figure 15. Lens 158 is attached to spherically shaped lens mount 166. Receptacle 162 in alignment compensation mount 160 receives lens mount 166. Lens mount 166 swivels in receptacle 162 in the θ_x direction and θ_γ direction in order to compensate for residual alignment errors as discussed with reference to Figures 14A and 14B. Registration features 164 are provided on alignment compensation mount 160. Registration features 164 mate with corresponding features (not shown) on fixed reference 127 of Figure 16 to position alignment compensation modules 152 into fixed reference 127. Alignment compensation mount 160 may be secured to fixed reference 127 appropriate means such as adhesive, solder, welding, or other appropriate attachment technique. Upon removing residual alignment error by tilting or rotating lens 158 in the θ_x direction and θ_γ direction, mount 166 may be secured to alignment compensation mount 160 by appropriate means such as adhesive, solder, welding, or other appropriate attachment technique. Lens 158 may be secured to mount 166 appropriate means such as adhesive, solder, welding, or other appropriate attachment technique. Mounts 166 may be transparent to allow appropriate radiation to secure mount 166 to receptacles 162 such as with adhesive or by laser soldering. Relative references 160 may also be transparent to facilitate adhesive curing or laser soldering. Mount 166 and relative reference 160 may be of appropriate materials, or a combination of materials, such as metal, glass, ceramic, semiconductor, or plastic and have coatings to facilitate bonding of mount 166 to relative reference 160. Mount 160 may also contain additional mechanical features to aid in gripping, manipulating and aligning. Receptacle 162 may have other shapes such as a spherical socket, hole, hole with a chamfer, or simply a planar surface, that allow mount 166 to swivel on the receptacle in the θ_x direction and the θ_γ direction. Lens 158 may also be secured to a mount that contains a receptacle and the relative reference may contain a spherical feature to permit angular adjustment of lens 158. Fixed reference 127 may also contain a receptacle and mount 166 may be placed directly into this receptacle without further need of relative reference 160. The various mounts, relative references, and fixed references may also be made by molding. Alignment compensation module 152 is shown in Figures 8 and 16 to be between optical filters 126 and lenses 122. The present invention is not limited to this configuration. For example, alignment compensation module 152 could also be placed between mirrors 126A and 126B, to compensate for residual alignment errors at fiber 124C. Also, with proper design, alignment compensation module 152 may be inserted into diverging or converging beams to change the beam position and compensate for residual alignment errors. It is also not necessary to populate fixed reference 127 with all alignment compensation modules 152 if the alignment at a particular fiber 124 is within specification without alignment compensation module 152.

Typically, many prealigned optical modules may be manufactured and stored for later insertion into fixed references to rapidly build complex optical devices such as demultiplexer 121. In one aspect the alignment compensation module is used to perform a final alignment of one or more prealigned optical components. In such an embodiment, the tolerance requirements for a prealigned optical module can be relaxed because the final compensation performed by the alignment compensation module is used to remove any residual alignment errors. This reduces the cost of manufacturing the prealigned optical module.

Figure 17 is an optical schematic representation of another aspect of alignment compensation module 152. Substantially collimated light beam 141 passes through alignment compensation module 152 where it is refracted by two optical wedges 170 and 172, commonly referred to as Risley prisms. Wedges 170 and 172 may be rotated independently in the θ₂ direction to change the angular propagation direction of light beam 143 and remove residual alignment errors.

Figure 18 shows another aspect of alignment compensation module 152 for removing residual alignment errors in laser transmitter 182. Figure 18 is an optical schematic representation of laser transmitter 182. Diverging light from laser source 180 is collected by lens 186 where it is converted to converging light beam 151. Optical window 184 has two planar surfaces 187 and 189. Optical window 184 may be made of be made of optical glass, plastic, or other materials that are transparent at the wavelengths of interest. The optical window may be made of silicon, for example, at telecommunication wavelengths between 1.3 and 1.6 um since silicon is transparent at these wavelengths. Optical window 184 may be rotated in the θ_x direction and the θ_γ direction to displace the light beam 153 in the Y direction and X direction, respectively, to remove residual alignment errors with respect to the core of optical fiber 188. Optical window 184 may also be rotated in the θ_z direction with another rotation in either the θ_x direction or the θ_γ direction to remove residual alignment errors. Another aspect of alignment compensation module 152 for removing residual alignment errors is shown in Figure 19. Figure 19 is an optical schematic representation of laser transmitter 182. Alignment compensation module consists of optical windows 184 and 190. Optical windows 184 and 190 may be independently rotated in the θ_z direction to shift the position of light beam 153 in the X direction and Y direction to compensate for residual alignment errors with respect to the core of optical fiber 188. Optical windows 184 and 190 may be rotated in v- grooves that are integral to their respective mounts and secured with solder, adhesive, or welded in place . Although the present invention has been described with reference to preferred embodiments, workers skilled in the art will recognize that changes may be made in form and detail without departing from the spirit and scope of the invention. For example, entire multimodule assemblies can be arranged on an optical "circuit board" to fabricate a complex opto-electronic assembly. The optical component can be any type of active or passive optical, opto-electrical or opto-mechanical component and not limited to the specific examples set forth herein. The optical component can be aligned and its orientation fixed using any suitable or desirable means. The specific components and examples set forth herein are provided to demonstrate various aspects of the invention and do not limit the scope of the invention. Other elements, shapes, components, configurations, etc. are within the scope of the invention. Any appropriate registration technique that provides the desired amount of constraint may be used. Typically, the registration technique is highly repeatable and accurate in all six degrees of freedom. Any appropriate material can be used for the various components. In one specific aspect, the relative reference mount and other components are formed from a single crystal material such as silicon. In another aspect, these components can be fabricated from any electrical material including semiconductors or ceramics. Other materials include machinable materials 5. such as steel, aluminum, metal alloys, etc. depending on requirements of a particular implementation. An assembled optical module can be used to fabricate an optical device using a "pick and place" machine or any suitable or desirable means. In such an embodiment, the 0 chamfers or bevels on the edges of the component mount can facilitate mechanical gripping of the mount. Similarly, the various components of the invention can be fabricated using any desired technique. Solders are known in the art and any appropriate solder can be selected to obtain the desired characteristics. The optical component can be coupled directly to the relative reference mount without a separate component mount. As used herein, "light" is not necessarily visible light. Further, the optical component can be any active or passive optical, opto-electrical or opto-mechanical element.

The alignment compensation modules of the present invention are not limited to compensating the final alignment of an optical beam with respect to optical fibers and may be used for compensating, for example, the final optical beam alignment with respect to the active area of a photodetector or may be used to angularly align the collimated beam of light exiting a laser collimator. In one aspect, alignment compensation modules can be comprised of a single lens or optical element or multiple lenses or optical elements. When multiple lenses or elements are used, the lenses and elements can rotate together and may be held in a single mount, or moved separately. Not all of the lenses and elements must rotate and may be stationary or exhibit some other form of movement such as a translation or rotation.

Claims

WHAT IS CLAIMED IS:

1. An optical device, comprising: a substrate having a first registration feature and a second registration feature; a first optical module comprising: an optical component; a relative reference mount including a registration feature configured to align with the first registration feature of the substrate, the optical component held at a prealigned spacial orientation relative to the registration feature; a second optical module, comprising: an optical component; and a relative reference mount including a registration feature configured to align with the second registration feature of the substrate, the optical component held at a prealigned spacial orientation relative to the registration feature; and the first and second optical modules arranged to transfer light therebetween, an alignment compensation module configured to interact with the light to provide a final alignment to the light to thereby compensate for residual alignment errors .

2. An optical alignment compensation module to align a substantially collimated light beam with an initial angular propagation direction along a Z axis comprising: an afocal optical system configured to deviate the angular propagation direction of the light beam when rotated about a Y axis perpendicular to the initial angular propagation (Z axis) direction, the amount of deviation less than the amount of rotation of the afocal optical system; and a mount configured to rotate the afocal optical system about the Y axis perpendicular to the initial angular propagation direction (Z axis) .

3. A method of making an optical device comprising: obtaining prealigned optical modules having registration features; obtaining a fixed reference with registration features configured to mate with the registration features of the prealigned optical modules; mounting the optical modules onto the fixed reference and aligning the registration features; obtaining an alignment compensation module; and adjusting a path of light through the optical device with the alignment compensation module.

4. A method of compensating the alignment of an optical system to align a light beam with an initial angular propagation direction along a Z axis, the method comprising: obtaining an alignment compensation module which includes an afocal optical system configured to deviate the angular propagation direction of the light beam when rotated about a Y axis perpendicular to the initial angular propagation (Z axis) direction, the amount of deviation less than the amount of rotation of the afocal optical system; and rotating the afocal optical system to adjust the path of the light beam through the optical system.

5. The invention of claims 1, 3 or 4 wherein the light beam is collimated.

6. The invention of claims 1, 3 or 4 wherein the light is substantially converging.

7. The invention of claims 1, 3 or 4 wherein the light is substantially diverging.

8. The invention of claims 1, 2, 3 or 4 wherein the alignment compensation module includes a lens element .

9. The invention of claims 1 or 3 wherein the alignment compensation module includes a lens element which is translated substantially perpendicular to a Z axis to adjust light beam alignment.

10. The invention of claim 9 including a lever arm alignment structure to translate the lens element.

11. The invention of claims 1 or 3 wherein the alignment compensation module includes an optical element with substantially two planar surfaces.

12. The invention of claim 11 wherein the optical element is rotated to provide the final alignment.

13. The invention of claims 1 or 3 wherein the alignment compensation module includes an afocal optical system configured to alter an angular propagation direction of the light.

14. The invention of claim 13 including an adjustable mount configured to rotate the afocal optical system.

15. The invention of claims 2 or 14 wherein the mount includes a spherical surface and a receptacle.

16. The invention of claims 2 or 14 wherein the mount is fixedly secured with an attachment material.

17. The invention of claims 1 or 3 wherein at least one optical element has an optical characteristic which varies relative to at least one dimension and wherein the optical characteristic is aligned with a reference defined relative to a registration feature of the respective relative reference mount.

18. The invention of claim 1 wherein the optical device is a fiber optical demultiplexer.