During the 1980's and 1990's the methods used to manufacture inertial MEMS devices could be divided into two groups; bulk and surface micromachining. Institutions which developed high precision inertial MEMS devices usually employed bulk micromachining processes. This was done to fabricate devices with large proof masses and stiff beams which result in a high scale factor, as well as high drive, and sense frequencies. New processes have been developed which are based on silicon on insulator (SOI) wafers. These processes combine the advantages of bulk and surface micromachining while enabling the etching of thick proof masses. This paper illustrates the manufacturing and performance advantages of an SOI inertial MEMS process.

MEMS gyros are often approximated as coupled, dual mode, resonant devices that transfer energy between two mechanical modes when subjected to rotational rate. These approximations are two degree-of-freedom (DOF) analyses with a mode-to-mode coupling described by Coriolis, elastic, and damping cross terms. In practice, MEMS gyros often exhibit unused, collateral modes with resonant frequencies close to the gyro's two operational modes. These additional modes provide mechanisms for the transfer of energy that are independent of rotational rate and, as a result, can generate errors in the gyro output. The purpose of this paper is to show the effects that collateral modes have on the bias of coupled-mode rate gyros. In this paper, the conventional 2 DOF analysis is extended into a three DOF model that includes one collateral mode. It is shown that the 3 DOF model is able to predict bias errors where the 2 DOF approach does not.

In this paper an excitation scheme employing simultaneous harmonic forcing and parametric excitation is applied to an electrostatically actuated MEMS gyroscope in order to improve the rate resolution performance to near inertial grade. A multiples scales perturbation method is used to investigate the dynamics of the gyroscope and facilitate in the design of a control methodology that enables the parametric pumping phenomena to the realized practically. The analysis shows that the quality factor of the primary mode of the gyroscope may be increased arbitrarily through parametric excitation. This allows forcing levels for the primary mode to be reduced by several orders of magnitude whilst sustaining the primary mode amplitude. Simulation of the oscillator scheme, which is highly non-linear, is achieved using MATLAB Simulink and is applied to a micro-ring gyroscope. The simulation demonstrates the Q-factor of the primary mode is increased by two orders of magnitude whilst the harmonic forcing amplitude is reduced by the same order, when the control scheme is operating. Agreement between the perturbation analysis and MATLAB Simulink models is within 8%. The increase in the Q-factor by two orders of magnitude results in a decrease in the electrical noise due to feedthrough by two orders of magnitude. This will enable a significant improvement of resonant gyroscope performance.

The ability to meet demanding specifications and stringent price points continues to drive market insertion for MEMS inertial sensors in cost-sensitive automotive applications and more recently, consumer markets. This paper examines major markets and drivers for gyroscopes and accelerometers, which will grow from a total of $835 million in 2004 to almost $1350 in 2009, a CAGR of over 10%.

New advances in Micro Electro-mechanical Systems (MEMS) deformable mirrors with high actuator count, high precision, and low cost have greatly influenced the design thinking for high resolution instruments on the next generation of large aperture telescopes for ground based astronomy. The use of MEMS as general purpose active optical components will open up a variety of capabilities for both sensing and controlling of astronomical light. In this talk we will describe the current thinking for instrument concepts on the Thirty Meter Telescope project and discuss some of our preliminary results from MEMS and AO system concept testing at the UCO/Lick Laboratory for Adaptive Optics.

The next major step in the study of extrasolar planets will be the direct detection, resolved from their parent star, of a significant sample of Jupiter-like extrasolar giant planets. Such detection will open up new parts of the extrasolar planet distribution and allow spectroscopic characterization of the planets themselves. Detecting Jovian planets at 5-50 AU scale orbiting nearby stars requires adaptive optics systems and coronagraphs an order of magnitude more powerful than those available today - the realm of "Extreme" adaptive optics. We present the basic requirements and design for such a system, the Gemini Planet Imager (GPI.) GPI will require a MEMS-based deformable mirror with good surface quality, 2-4 micron stroke (operated in tandem with a conventional low-order "woofer" mirror), and a fully-functional 48-actuator-diameter aperture.

MEMS is one of several emerging technologies for fabricating wavefront correctors for use in adaptive optics systems. Each technology has its own advantages and disadvantages. In order to compare devices, it is useful to define a task and make a comparison based upon the effectiveness of each device for this task. Such an approach implies, of course, that device A might be better suited for task X whereas device B is better suited for task Y. In adaptive optics, this situation is already known: deformable mirrors that are relatively effective at compensating for atmospheric turbulence are not necessarily the mirrors that one would choose for correction of the aberrations of the eye. This is essentially because the statistical modal distribution of the aberrated wavefronts in each case are different. In this talk, we shall present a method for systematically evaluating the effectiveness of different mirror (or transmissive) technologies in adaptive optics in the eye. It uses a model for the aberrations of the eye (such as that developed by Thibos et al¹) and a least squares fitting procedure. Results will be presented for at least 4 mirrors, including a 12x12 MEMS device. The key point is that it is the effectiveness of each actuator signal that is important, not the raw number of actuators.

A MEMS deformable mirror has recently been employed in the AO system of an adaptive optics scanning laser ophthalmoscope (AOSLO). MEMS allows for a more compact, efficient and effective system. The AO system in the AOSLO operates with a modal closed loop. Aberrations after AO reduce the wave aberration to less than 0.1 microns RMS in most eyes. Results show improved resolution, brightness and contrast. Images of patches of retina show a well resolved cone photoreceptor mosaic as they change in size with eccentricities ranging from 0.6 degrees to 4.23 degrees from the fovea.

The development of adaptive optics for the human eye to correct aberrations, to restore accommodation after lens extraction due to cataract and to correct age-related presbyopia have interest of academia and industry. We report on optics for a new accommodative intraocular lens which uses a two-element varifocal Alvarez lens. This lens has two refractive elements with cubic surfaces which, in combination, form a varifocal lens when the elements are shifted relatively to each other perpendicular to the optical path. The accommodative function of the lens will be driven by the ocular ciliary muscle. The refractive elements of the dual-optic intraocular lens are designed to provide a near emmetropic on-axis vision with a >4 dioptre accommodation range. The anterior element has a spherical lens to correct for the overall refraction of the eye, aspheric terms to correct the corneal asphericity and a cubic term as accommodative component; the posterior element has a cubic shaped surface only. The modular transfer function shows that the image on the retina reaches a diffraction limited performance for the on-axis vision in combination with the aspheric correction for aberrations of the cornea. We conclude that the varifocal lens is uniquely suitable for application as an intraocular accommodative lens because of its optical quality and ample accommodative power.

Adaptive laser resonators with deformable MOEMS mirrors under closed-loop control are discussed and experimental results are presented. The requirements for deformable mirrors and for closed-loop control systems of these mirrors are analyzed. Several deformable mirrors have been characterized and the results are presented. Currently available membrane mirrors deform under laser load and need further development before they can be used for aberration correction of solid state lasers above some tens of Watts. Nevertheless, the results are encouraging and the requirements are within reach of currently available technology. Finally, we demonstrate an Nd.YVO₄-laser with a closed-loop adaptive resonator and more than 6 W of output power. The closed-loop system was able to compensate artificially introduced aberrations from a phase plate.

Formation of the given laser beam intensity and phase is an important practical and scientific problem. Semipassive bimorph flexible mirror is one of the most widely used devices for this purpose. We present a novel approach of multilayer bimorph (multimorph) mirrors and a numerical model to simulate them, based on a variation approach of the finite elements method. The multilayer bimorph mirror consists of a substrate and a number of piezoceramic layers. The electrode grid of each layer is determined separately to reproduce low order aberrations.

The use of adaptive and active optics (AO) is enabling the construction and test of flexible optical systems with performances unprecedented. This flourishing of technical advances is also due to the availability of new technologies that are much lower in cost, much easier to implement and use. Among these new technologies the use of Micro-Electro-Machined (MEM) mirrors is one of the primary sources of innovation. Several groups are actively working in bringing to fruition AO systems based on MEMs technologies and at the same time several groups are working to improve the MEMs technology and tailor it more and more towards various aspects of the AO problems. In this paper we will presents an overview of MEMs adaptive optics problems. We will especially focus on our experience in this field and discuss results from our AO system. We will discuss pros and cons on the use of MEM for adaptive optics and elaborate on our experience on field-testing of these devices. This paper will also briefly discuss the broader use of adaptive optics in fields other than atmospheric compensation.

In most adaptive optics systems, there are two elements that control wavefront correction, a fast steering mirror that corrects tip and tilt and a deformable element that corrects higher order aberrations. By mounting the deformable element onto the tip/tilt platform, complete wavefront compensation is now possible at one location in an optical system. The advantage of mounting a lightweight Micro Electro-Machined (MEM) deformable mirror on a tip/tilt stage is both fewer optical components and a simpler alignment process. The impact on the frequency stability of the of the MEM device on the tip/tilt platform is approximately 5% of the driving frequency.

The original proposal of wavefront-sensorless aberration correction was suggested by Muller and Buffington[1]. In this technique we attempt to correct for wavefront aberrations without the use of conventional wavefront sensing. We apply commands to a corrective element with adjustable segments in an attempt to maximise a metric which correlates to image quality. We employ search algorithms to find the optimal combination of actuator voltages on a DM to maximise a certain sharpness metric. The "sharpness" is based on intensity measurements taken with a CCD camera. It has been shown that sharpness maximisation, using the simplex algorithm, can minimise the aberrations and restore the Airy rings of an imaged point source. This paper demonstrates that the technique can be applied to extended objects which have been aberrated using a Hamamatsu SLM to induce aberrations. The correction achieved using various search algorithms are evaluated and presented.

An important consideration in the design of an adaptive optics controller is the range of physical shapes required by the DM to compensate the existing aberrations. Conversely, if the range of surface shapes achievable with a DM is known, its suitability for a particular AO application can be determined. In this paper, we characterize one MEMS DM that was recently developed for vision science applications. The device has 140 actuators supporting a continuous face sheet deformable mirror having 4mm square aperture. The total range of actuation is about 4μm, achieved using electrostatic actuation in an architecture that has been described previously. We incorporated the MEMS mirror into an adaptive optics (AO) testbed to measure its capacity to transform an initially planar wavefront into a wavefront having one of thirty-six orthogonal shapes corresponding to the first seven orders of Zernike polynomials. The testbed included a superluminescent diode source emitting light with a wavelength 630nm, a MEMS DM, and a Shack Hartmann wavefront sensor (SHWS). The DM was positioned in a plane conjugate to the SHWS lenslets, using a pair of relay lenses. Wavefront slope measurements provided by the SHWS were used in an integral controller to regulate DM shape. The control software used the difference between the the wavefront measured by the SHWS and the desired (reference) wavefront as feedback for the DM. The DM is able to produce all 36 terms with a wavefront height root mean square (RMS) from 1.35μm for the lower order Zernike shapes to 0.2μm for the 7th order.

We have demonstrated that a microelectrical mechanical systems (MEMS) deformable mirror can be flattened to < 1 nm RMS within controllable spatial frequencies over a 9.2-mm aperture making it a viable option for high-contrast adaptive optics systems (also known as Extreme Adaptive Optics). The Extreme Adaptive Optics Testbed at UC Santa Cruz is being used to investigate and develop technologies for high-contrast imaging, especially wavefront control. A phase shifting diffraction interferometer (PSDI) measures wavefront errors with sub-nm precision and accuracy for metrology and wavefront control. Consistent flattening, required testing and characterization of the individual actuator response, including the effects of dead and low-response actuators. Stability and repeatability of the MEMS devices was also tested. An error budget for MEMS closed loop performance will summarize MEMS characterization.

Various applications in modern optics are demanding for Spatial Light Modulators (SLM) with a true analog light processing capability, e.g. the generation of arbitrary analog phase patterns for an adaptive optical phase control. For that purpose the Fraunhofer IPMS has developed a high-resolution MEMS Micro Mirror Array (MMA) with an integrated active-matrix CMOS address circuitry. The device provides 240 x 200 piston-type mirror elements with 40 μm pixel size, where each of them can be addressed and deflected independently at an 8bit height resolution with a vertical analog deflection range of up to 400 nm suitable for a 2pi phase modulation in the visible. Full user programmability and control is provided by a newly developed comfortable driver software for Windows XP based PCs supporting both a Graphical User Interface (GUI) for stand-alone operation with pre-defined data patterns as well as an open ActiveX programming interface for a direct data feed-through within a closed-loop environment. High-speed data communication is established by an IEEE1394a FireWire interface together with an electronic driving board performing the actual MMA programming and control at a maximum frame rate of up to 500 Hz. Successful application demonstrations have been given in eye aberration correction, coupling efficiency optimization into a monomode fiber, ultra-short laser pulse modulation and diffractive beam shaping. Besides a presentation of the basic device concept the paper will give an overview of the obtained results from these applications.

This paper presents a MEMS DM that is a hybrid of surface micromachining and bulk micromachining. The combination of fabrication techniques resulted in a DM that has demonstrated 7.6 μm stroke at 125 V, 98.6% fill factor, and excellent optical quality of better than 16 nm rms after packaging. Preliminary cyclic testing over 110 hours and 10⁷ cycles showed no noticeable changes to the actuator positions after cycling.

MicroAssembly Technologies has developed an optical MEMS process based on the batch transfer of microstructures. This "plug and play" process enables integration of ultra-flat SOI reflectors with high-aspect MEMS actuators and high-voltage CMOS driver circuits. Interferometer measurements confirm that assembled mirror devices are extremely flat.

The paper proposed a novel curb structure to elevate the micro optical devices by the driving force of micro array thermal actuator, MATA. The effects of spring structure and curb structure on the maximum displacements and the variation of surface flatness of the elevated micro mirror varied with operation voltage are investigated. The motion behaviors of the elevated micro mirror are stimulated and analyzed to get the maximum displacement and inclined angle of the device. The results demonstrate the wider width, longer pitch and more pitch numbers of spring structure are; the maximum displacement of the elevated micro mirror is larger. Compared the effects of spring structure and curb structure on the maximum displacement of the elevated micro mirror, there are more influence on the variation of maximum displacement due to the varied spring structure than the varied curb structure. On the other hand, the variation of surface flatness of the elevated micro mirror is more significant by the varied pitch number of spring structure and the varied width of curb structure. The maximum displacement and inclined angle of proposed micro optical device are 58.6μm and 17.04°C, respectively.

We report on the development of a new MEMS deformable mirror (DM) system for the hyper-contrast visible nulling coronagraph architecture designed by the Jet Propulsion Laboratory for NASA's Terrestrial Planet Finding (TPF) mission. The new DM is based largely upon existing lightweight, low power MEMS DM technology at Boston University (BU), tailored to the rigorous optical and mechanical requirements of the nulling coronagraph. It consists of 329-hexagonal segments on a 600μm pitch, each with tip/tilt and piston degrees of freedom. The mirror segments have 1μm of stroke, a tip/tilt range of 600 arc-seconds, and maintain their figure to within 2nm RMS under actuation. The polished polycrystalline silicon mirror segments have a surface roughness of 5nm RMS and an average curvature of 270mm. Designing a mirror segment that maintains its figure during actuation was a very significant challenge faced during DM development. Two design concepts were pursued in parallel to address this challenge. The first design uses a thick, epitaxial grown polysilicon mirror layer to add rigidity to the mirror segment. The second design reduces mirror surface bending by decoupling actuator diaphragm motion from the mirror surface motion. This is done using flexure cuts around the mirror post in the actuator diaphragm. Both DM architectures and their polysilicon microfabrication process are presented. Recent optical and electromechanical characterization results will also be discussed, in addition to plans for further improvement of DM figure to satisfy nulling coronagraph optical requirements.

This ongoing work concerns the creation of a deformable mirror by the integration of MEMS actuators with Nanolaminate foils through metal compression boning. These mirrors will use the advantages of these disparate technologies to achieve dense actuation of a high-quality, continuous mirror surface. They will enable advanced adaptive optics systems in large terrestrial telescopes. While MEMS actuators provide very dense actuation with high precision they can not provide large forces typically necessary to deform conventional mirror surfaces. Nanolaminate foils can be fabricated with very high surface quality while their extraordinary mechanical properties enable very thin, flexible foils to survive the rigors of fabrication. Precise metal compression bonding allows the attachment of the fragile MEMS actuators to the thin nanolaminate foils without creating distortions at the bond sites. This paper will describe work in four major areas: 1) modeling and design, 2) bonding development, 3) nanolaminate foil development, 4) producing a prototype. A first-principles analytical model was created and used to determine the design parameters. A method of bonding was determined that is both strong, and minimizes the localized deformation or print through. Work has also been done to produce nanolaminate foils that are sufficiently thin, flexible and flat to be deformed by the MEMS actuators. Finally a prototype was produced by bonding thin, flexible nanolaminate foils to commercially available MEMS actuators.

This work concerns the development of a technology that uses Nanolaminate foils to form light-weight, deformable mirrors that are scalable over a wide range of mirror sizes. While MEMS-based deformable mirrors and spatial light modulators have considerably reduced the cost and increased the capabilities of adaptive optic systems, there has not been a way to utilize the advantages of lithography and batch-fabrication to produce large-scale deformable mirrors. This technology is made scalable by using fabrication techniques and lithography that are not limited to the sizes of conventional MEMS devices. Like many MEMS devices, these mirrors use parallel plate electrostatic actuators. This technology replicates that functionality by suspending a horizontal piece of nanolaminate foil over an electrode by electroplated nickel posts. This actuator is attached, with another post, to another nanolaminate foil that acts as the mirror surface. Most MEMS devices are produced with integrated circuit lithography techniques that are capable of very small line widths, but are not scalable to large sizes. This technology is very tolerant of lithography errors and can use coarser, printed circuit board lithography techniques that can be scaled to very large sizes. These mirrors use small, lithographically defined actuators and thin nanolaminate foils allowing them to produce deformations over a large area while minimizing weight. This paper will describe a staged program to develop this technology. First-principles models were developed to determine design parameters. Three stages of fabrication will be described starting with a 3x3 device using conventional metal foils and epoxy to a 10-across all-metal device with nanolaminate mirror surfaces.

Highly performing adaptive optical (AO) systems are mandatory for next generation giant telescopes as well as next generation instrumentation for 10m-class telescopes, for studying new fields like circumstellar disks and extra-solar planets. These systems require deformable mirrors with very challenging parameters, including number of actuators up to 250 000 and inter-actuator spacing around 500μm. MOEMS-based devices are promising for future deformable mirrors. We are currently developing a micro-deformable mirror (MDM) based on an array of electrostatic actuators with attachment posts to a continuous mirror on top. In order to reach large stroke for low driving voltage, the originality of our approach lies in the elaboration of a sacrificial layer and of a structural layer made of polymer materials. We have developed the first polymer piston-motion actuator: a 10μm thick mobile plate with four springs attached to the substrate, and with an air gap of 10μm exhibits a piston motion of 2μm for 30V, and measured resonance frequency of 6.5kHz is well suited for AO systems. The electrostatic force provides a non-linear actuation, while AO systems are based on linear matrices operations. We have successfully developed a dedicated 14-bit electronics in order to "linearize" the actuation. Actual location of the actuator versus expected location of the actuator is obtained with a standard deviation of 21 nm. Comparison with FEM models shows very good agreement, and design of a complete polymer-based MDM has been done.

Many adaptive optics (AO) applications require mirror arrays with hundreds to thousands of segments, necessitating a CMOS-compatible MEMS process to integrate the mirrors with their driving electronics. This paper proposes a MEMS actuator that is fabricated using low-temperature polycrystalline silicon-germanium (poly-SiGe) surface-micromaching technology (total thermal budget is 6 hours at or below 425°C). The MEMS actuator consists of three flexures and a hexagonal platform, on which a micromirror is to be assembled. The flexures are made of single-layer poly-SiGe with stress gradient across thickness of the film, making them bend out-of-plane after sacrificial-layer release to create a large nominal gap. The platform, on the other hand, has an additional stress-balancing SiGe layer deposited on top, making the dual-layer stack stay flat after release. Using this process, we have successfully fabricated the MEMS actuator which is lifted 14.6 μm out-of-plane by 290-μm-long flexures. The 2-μm-thick hexagonal mirror-platform exhibits a strain gradient of -5.5×10^-5 μm^-1 (equivalent to 18 mm radius-of-curvature), which would be further reduced once the micromirror is assembled.

A high-stroke micro-actuator array was designed, modeled, fabricated and tested. Each pixel in the 4x4 array consists of a self-aligned vertical comb drive actuator. This micro-actuator array was designed to become the foundation of a micro-mirror array that will be used as a deformable mirror for adaptive optics applications. Analytical models combined with CoventorWare^(R) simulations were used to design actuators that would move up to 10 microns in piston motion with 100V applied. Devices were fabricated according to this design and testing of these devices demonstrated an actuator displacement of 1.4 microns with 200V applied. Further investigation revealed that fabrication process inaccuracy led to significantly stiffer mechanical springs in the fabricated devices. The increased stiffness of the springs was shown to account for the reduced displacement of the actuators relative to the design.

In this paper we review the use of a 3-dimensional MEMS fabrication process to prototype long stroke (>10 μm) actuators as are required for use in future adaptive optics systems in astronomy and vision science. The Electrochemical Fabrication (EFAB^TM) process that was used creates metal micro-structures by electroplating multiple, independently patterned layers. The process has the design freedom of rapid prototyping where multiple patterned layers are stacked to build structures with virtually any desired geometry, but in contrast has much greater precision, the capability for batch fabrication and provides parts in engineering materials such as nickel. The design freedom enabled by this process has been used to make both parallel plate and comb drive actuator deformable mirror designs that can have large vertical heights of up to 1 mm. As the thickness of the sacrificial layers used to release the actuator is specified by the designer, rather than by constraints of the fabrication process, the design of large-stroke actuators is straightforward and does not require any new process development. Since the number of material layers in the EFAB^TM process is also specified by the designer it has been possible to gang multiple parallel plate actuators together to decrease the voltage required for long-stroke actuators.

Electrostatic Membrane Deformable Mirror (MDM) technology developed using silicon bulk micro-machining techniques offers the potential of providing low-cost, compact wavefront control systems for diverse optical system applications. Electrostatic mirror construction using bulk micro-machining allows for custom designs to satisfy wavefront control requirements for most optical systems. An electrostatic MDM consists of a thin membrane, generally with a thin metal or multi-layer high-reflectivity coating, suspended over an actuator pad array that is connected to a high-voltage driver. Voltages applied to the array elements deflect the membrane to provide an optical surface capable of correcting for measured optical aberrations in a given system. Electrostatic membrane DM designs are derived from well-known principles of membrane mechanics and electrostatics, the desired optical wavefront control requirements, and the current limitations of mirror fabrication and actuator drive electronics. MDM performance is strongly dependent on mirror diameter and air damping in meeting desired spatial and temporal frequency requirements. In this paper, we present wavefront control results from an embedded wavefront control system developed around a commercially available high-speed camera and an AgilOptics Unifi MDM driver using USB 2.0 communications and the Linux development environment. This new product, ClariFast^TM, combines our previous Clarifi^TM product offering into a faster more streamlined version dedicated strictly to Hartmann Wavefront sensing.

The use of Micro-Electro-Machined (MEM) devices as deformable mirrors (DM) for active and adaptive optics is increasing dramatically. Such increases are due to both the cost and simplicity of use of these devices. Our experience with MEM DMs has been positive, however the controlling protocols of these devices presents some issues. Based on our experience and needs we decided to design a generic controller based on a fast communication protocol. These requirements have pushed us to design a system around a USB 2.0 protocol. In this paper we present our architectural design for such controller. We present also experimental data and analysis on the performance of the controller. We describe the pros and cons of such approach versus other techniques. We will address how general such architecture is and how portable is to other systems.

A CMOS electronics driver chip to control a deformable MEMS mirror has been developed. With the advances in CMOS technology, it has become possible to design and fabricate electronics operable at higher voltages than those in traditional integrated circuits. Since MEMS structures require relatively high operating voltages to achieve electrostatic force, these high voltage CMOS processes offer promise for miniaturization of the corresponding drivers. Using the capability of low voltage logic together with high voltage output stages, a compact driver chip has been designed and fabricated. The chip was developed and fabricated though a high voltage CMOS process. The driver is digitally controlled through address and data input bits, and through a smart low-voltage to high-voltage transition output stage, voltages of up to 300V are output to each mirror electrode. A compact design allows the control of 144 channels through a single chip with 8-bit resolution at 100Hz refresh rate. The low-voltage stage consists of address logic together with latch stages to store the data, which in turn is converted to a high voltage signal through a current mode, binary weighted scheme. This technique combines the digital-to-analogue conversion stage and a high-voltage amplifier stage, thus saving on substrate area. Using this method, the 144 channel high-voltage driver was fabricated on a single chip less than 3.5cm² in area. In this paper, design, fabrication and testing of these drivers are reported.

This paper describes the design, microfabrication and testing of a pre-aligned array of fiber couplers with integrated out-of-plane microlenses using direct UV lithography of SU-8. The fiber bundle coupler includes a refractive microlens array and two fiberport collimator arrays. With the optical axis of the pixels parallel to the substrate, each pixel of the microlens array can be pre-aligned with the corresponding pixels of the fiberport collimator array as defined by the lithography mask design. This out-of-plane polymer microlens array was pre-aligned with the fiber collimator arrays with no additional adjustment and assembly required. This novel approach helps to dramatically reduce the running cost and improve the alignment quality and coupling efficiency.