We review our previous work to develop an uncooled subMMW detector capable of achieving an NEΔT
~0.5K at 30Hz frame rate and describe our approach to develop a staring subMMW camera based on a 2D
array of such detectors. Both predicted and measured results of performance metrics (responsivity, NEP,
response time, spectral bandwidth, NEΔT) are presented. The measured performance agrees reasonably well
with predictions and is consistent with attaining our NEΔT goal. Thus far, 1×4 detector arrays have been
fabricated, and 1×8 focal plane arrays have been developed and tested. We briefly discuss our vision to
achieve a128×128 detector array needed for a practical staring subMMW imager and describe the technology
challenges needed to realize it.
We review our previous work to develop an uncooled THz detector capable of achieving an NEΔT ~0.5K at
30Hz frame rate and describe our approach to develop a staring THz camera based on a 2D array of such
detectors. Both predicted and measured results of performance metrics (responsivity, NEP, response time,
spectral bandwidth, NEΔT) are presented. The measured performance agrees reasonably well with predictions
and is consistent with attaining our NEΔT goal. Thus far, 1x4 detector arrays have been fabricated, and 1x8
focal plane arrays have been developed and tested. We briefly discuss our vision to achieve a128x128
detector array needed for a practical staring THz imager and describe the technology challenges needed to
We report experimental results on recently developed MEMS-based, uncooled THz detectors and imaging
applications for linear focal plane arrays constructed from them. The detector incorporates a broadband
micro-antenna coupled to an impedance-matched microbridge. Micro-antennas were fabricated having cut-on
frequencies of 500GHz, 650GHz, and 1.5THz, each with bandwidth of several hundred GHz. Sensitivity and
frequency response of the detectors is predicted to be ~6pW/√Hz (with backplane) and 7kHz, respectively,
and supporting measurements of the first devices will be presented. Fully integrated 1x8 linear focal plane
arrays have been assembled and will be used in on-going imaging demonstrations.
Born of necessity of application, the Vertical Cavity Surface Emitting Laser (VCSEL) is now found in nearly all optical networking systems based on standards such as the IEEE 802.3z and ANSI X3.t11. Reliability continues to be the hallmark of the technology, and the volume manufacturing aspects are now realized. While VCSEls satisfying optical networking standards continue to provide the highest volume applications, the advantages of the technology are beginning to enable novel optical equipment. This paper explores development of VCSELs at wavelengths from 650 to 850nm, and the commercial applications of these devices in both the data communications and optical sensing arenas. VCSELs operating at longer wavelengths are also being developed, but are not at a stage of commercialization to be discussed in this forum.
The system architecture and the first prototype demonstrator system for the VCSEL-based Interconnects in VLSI Architectures for Computational Enhancement (VIVACE) program is described. The main goal of the VIVACE program is to build a high bi-section bandwidth free-space optically interconnected switch and to demonstrate it in a system of multiple compute nodes running a distributed algorithm. The prototype demonstrator system developed is a stand alone first-generation VIVACE Optical Network Interface Card (VONIC) communicating to another VONIC through a parallel- data fiber link. This system was developed to test the signal integrity and Bit Error Rate between two VONICs.
A brief summary of both VCSEL technology and guided-mode grating resonant filters (GMGRFs) is presented. We then discuss benefits and issues of integrating the two technologies, emphasizing control of wavelength, polarization, and laser cavity modes. We present a GMGRF design suitable for a 980 nm InGaAs VCSEL and show that a significant loss (-4%) in reflectivity results from the slight loss associated with the minimum mirror conductivity required to inject current through the mirror. Experimental data are presented at 850 mm for gratings designed for and fabricated on fused silica substrates and illustrate that GMGRFs are also very sensitive to other forms of loss such as scatter caused by roughness in the grating lines. We suggest a hybrid approach of a GMGRF on a reduced distributed Bragg reflector stack as a means to circumvent the high sensitivity to loss in the GMGRF.
We have designed, fabricated, and tested hybrid refractive/diffractive optical elements in acrylic and cyclic olefin copolymer polymers. The elements were tested for optical performance before and after various environmental conditions.
An optical device is presented that uses the highly wavelength dispersive nature of diffractive optics to provide a means of either removing a small waveband of incident radiation from a scene, to function as a tunable notch filter, or passing only a small waveband of incident radiation from the scene, to function as a tunable bandpass filter. A prototype design example is presented along with preliminary performance analysis.
The DARPA-funded Consortium for Optical and Optoelectronic Technologies for Computing (CO-OP) recently completed the first DOE Foundry run delivering ten samples to each of nineteen users, each with a unique design. The binary optics process was used to provide a maximum of eight phase levels at a design wavelength of 850 nm. Averaged over all users and all samples, an etch depth error of one percent and alignment accuracy within 0.25 micron were achieved. This paper summarizes the details of the process results.
Using LIGA techniques, a transmissive diffraction grating device in permalloy with variable, and controllable, grating period has been designed and fabricated for use as a tunable infrared spectral filer. Typical device parameters exhibit parameters exhibit periods approximately 8-30 microns, permalloy grating lines approximately 3 microns wide by 10- 50 microns high. In Ell polarization, the device acts as low pass spectral filter with cutoff wavelength determined by the variable grating period. Recent developments in fabrication of the gratin structure and actuator are reviewed briefly. Emphasis is given to results of recent performance modeling both for thinner and shorter grating walls suitable for shorter wavelength applications and to demonstrate the effects of an incident spherical wave.
An overview is presented of diffractive optics and micro-optics technologies and their application to infrared imaging systems. The technology overview describes several fabrication methods, emphasizing the importance of diamond machining for infrared systems and materials, and discusses various optical functions commonly implemented with diffractive elements. Applications are discussed both as a general survey by citing an extensive list of references and as a more detailed presentation of three examples.
LIGA technology has been used to fabricate linear gratings having free-standing nickel walls a few micrometers wide and as much as 50 micrometers high and period on the order of 10 micrometers . With additional MEMS processing steps, such as devices are intended for use in a tunable infrared filter. Prediction of optical performance is a particularly challenging problem for gratings with these parameters and materials and requires a robust Maxwell solver. We have applied our own code, described elsewhere, in the form of a finite element implementation of equivalent variational problem to examine the optical properties of this class of gratings. Here, we describe our predicted results for transmittance as a function of wavelength and polarization for various grating parameters and incident conditions. Measurements of fabricated gratings were also carried out, and the predictions are shown to agree well with the measured data. The filter cutoff is shown to be sensitive to cone angle of the incident radiation, and the implications of this effect on system performance are discussed.
An optical device is presented that uses the highly wavelength dispersive nature of diffractive optics to provide a means of either removing a small waveband of incident radiation from a scene, to function as a tunable notch filter, or passing only a small waveband of incident radiation from the scene, to function as a tunable bandpass filter. Design examples are given along with system performance analysis. Experimental verification is also presented.
We present an overview of diffractive optics technology and the advantages this technology offers when applied to head-mounted displays (HMD). We show especially the impact on weight reduction when diffractive elements are used to correct chromatic aberrations in full-color HMDs. We discuss the effect of higher diffractive orders on image quality and show how to model these effects. Finally, we present the results of a demonstration of a diffractive element in a conventional monochromatic HMD, compare the performance of the hybrid and conventional systems, and demonstrate the validity of our model.
Mathematical models for diffractive optics have been developed, and implemented as numerical codes, both for the “direct” problem and for the “inverse” problem. In problems of the “direct” class, the diffractive optic is specified, and the full set of Maxwell’s equations is cast in a variational form and solved numerically by a finite element approach. This approach is well-posed in the sense that existence and uniqueness of the solution can be proved and specific convergence conditions can be derived. As an example, we consider a low order metallic grating, where other approaches are known to have convergence problems, and show the variational method yields exceptionally good convergence. In problems of the “inverse” class, some information about the diffracted field (e.g., the far-field intensity) is given, and the problem is to find the periodic structure in some optimal sense. A new approach is described that applies relaxed optimal design methods to give entirely new grating structures; wave propagation is based on the Helmholtz equation. An example of an angle-optimized antireflective structure and of an ideal array generator are presented.
Honeywell has developed a quantitative image quality model for the Helmet Mounted Display (HMD) electro-optical systems that will predict the optical performance and image quality of a given system configuration. The linear systems model includes modules for the image intensifier objective, image intensifier tube, fiber optic faceplates and tapers, charge coupled device (CCD) camera, liquid crystal display (LCD) or CRT image source, relay optics, electronic filtering and preprocessing, and a perception model for the eye. Sine wave and square wave system response are predicted via modulation transform function (MTF) calculations as well as the maximum resolution and a measurement of just noticeable differences (jnd's) as perceived by the human eye. The model will allow the system designer to quickly and inexpensively evaluate complex systems tradeoffs and modifications to advanced HMD systems.
We present results characterizing the effects of processing errors on the performance of staircase kinoforms, commonly known as `binary optical devices.' Diffraction efficiency and modulation transfer function data are given for various types of processing errors present in staircase kinoforms of a f/10 Fresnel phase lens having two, four, and eight phase levels. Processing errors include etch depth, linewidth, and mask alignment. Processing errors, especially mask alignment, are shown to have the greatest impact on diffraction efficiency and very little effect on image quality.
Two recent demonstrations of diffractive optics for uncooled staring 8 - 12 micrometers infrared imagers are reviewed and critically examined. Although each demonstration by itself yielded impressive results, when viewed from a broader system standpoint and considering current state-of-the-art in uncooled detector arrays, it is concluded that there is not a compelling reason to use diffractive elements in current systems. Alternatives are suggested.
This paper describes the fabrication and testing of 64 X 96 arrays of microlenses used for fill-factor enhancement of uncooled infrared detector arrays. Each lenslet represents a f/0.9 Fresnel phase lens at 10 micrometers wavelength. All arrays were etched into silicon wafers as either 8-level or 16-level staircase kinoforms. When integrated with a detector array having 20 fill factor, these microlens arrays were capable of increasing the magnitude of the measured signal with f/2.2 fore-optics by 2.5-fold.
We describe an integral approach to the rigorous solution of Maxwell's equations for diffraction from a biperiodic grating having arbitrary profile and periodicity. Emphasis is placed on the implementation of solutions and is addressed to the engineering community. The method described is amenable to mathematical convergence analysis. A numerical example is given which indicates convergence of solutions.
Results are presented of an on-going experimental program to characterize the effects of processing errors on kinoform performance. Diffraction efficiency and modulation transfer function data are given for various types of processing errors present in staircase kinoforms of a f/10 Fresnel phase lens having two and four levels. Processing errors include etch depth, linewidth, and mask alignment. Processing errors, especially mask alignment, are shown to have the greatest impact on diffraction efficiency and very little effect on image quality.
We examine the role of processing errors on diffraction efficiency of binary optical elements and the validity of the Fourier
model to predict diffraction efficiency. We show that mask alignment error can significantly degrade efficiency. Models based
on the Fourier theory can adequately predict both the magnitude of diffraction efficiency and its sensitivity to processing errors
for optically slow elements (f/b). For optically fast elements (<f/2), a rigorous model is required for good predictive
accuracy. We use our model to indicate the sensitivity of diffraction efficiency to processing errors.