A new short wavelength infrared (SWIR) focal plane measurement facility at the Naval Ocean Systems Center is described. The following aspects of the system are discussed: spectral response measurement, noise spectral density measurement, frequency response measurement, and cryogenic Dewars.
A computer-controlled system has been developed for collecting output data from a multiplexer, an individual detector, or a multiplexed array of detectors, and generating pulse height distribution statistics in both tabular and graphical form. Array data from each pixel can be kept separate to produce statistics for each pixel or can be combined to produce statistics for the array as a whole. Both time and voltage thresholds can be established such that only those events above threshold are included in the statistics. The effect of time-dependent drift in the mean value of the outputs can also be minimized by computing a "running mean" which is subtracted from each data point as it is collected. Multiplexed or detector outputs are digitized with 15 bits of resolution at a rate up to 500,000 samples per second. There is no limit to the number of samples that can be collected to generate a given distribution.
An expert system has been developed to assist in analyzing IR detector test data. This data is taken in the production of Common Module detector-dewar assemblies. These assemblies contain on the order of 1500 internal electrical connections, and test data may consist of more than 1500 data items. Testing is done at several stages in the manufacturing sequence to determine if any rework is needed prior to the next assembly operation. Rework of a defect should be done at the earliest possible stage to minimize cost and jeopardy to the unit. Thus it is important to identify defects from their signature in the test data with consistency and accuracy. Coldfinger is the name of the expert system designed to help in this task. It was written in PROLOG and implemented on a Macintosh PC. The user friendly interface of the Macintosh made factory training easier and faster. Coldfinger's software code contains 165 facts, 348 tools and 138 rules. The facts describe the electric wiring topology and adjacency of connections, including probabilities for shorts between any two wires. The tools are used to compare test data results locally and globally. The rules have been distilled from the experience of real production personnel familiar with the product. Coldfinger operates in real time, taking about two minutes to analyze a set of test data with no defects, and only seven minutes with a hypothetical data set containing a large number of defects. Unlike real experts, Coldfinger does not eat or sleep so that assistance in decision making is available around the clock.
A unique and versatile vacuum chamber has been designed and built by R.G. Hansen and Associates for the Jet Propulsion Laboratory's Infrared Focal Plane Technology Group. This chamber is equipped with multiple ports for cryogen and electrical vacuum feedthroughs, pumping units, vacuum gauges, sources, and detector camera heads. The design incorporates a liquid-nitrogen-cooled optical table and radiation shield for low-background infrared detector testing. Focal planes can be tested in this chamber at temperatures ranging from 300 K to that of liquid helium. The test chamber design emphasizes versatility so that the system can accommodate various detectors and testing conditions. This paper describes the design and construction of this low-background infrared focal plane test chamber and discusses some of its distinctive features. An analysis of the test chamber's performance is also be presented.
A scanning electron microscope (SEM) has been modified and conjoined with a cryogenic chamber to enable application of a focused electron beam to probe long-wavelength infrared (LWIR) detectors in a low IR photon environment. A beam blanker installed in the SEM allows the temporal as well as the spatial response of a detector to be measured. The apparatus has been applied to study the effects of ionizing radiation in Si impurity band detectors and HgCdTe detectors. Examples of these applications are presented.
Schottky barrier diodes made from platinum silicide are outstanding candidates for staring imagery in the MWIR band. This use was first suggested by Shepherd et. al. in 1974 who reported results on small arrays. The technology has advanced rapidly and large, two dimensional arrays of staring Schottky diodes are presently available. The detectors are fabricated on high quality, VLSI grade silicon substrates. Two dimensional arrays of these diodes have extremely high uniformity of optical responsivity. This fact allows for simple application of non-uniformity correction algorithms which give MRT's of less than 0.05K. The rapid advancement of PtSi as an infrared imaging technology is in a large part due to the advanced state of silicon substrate technology.
The improvements in the development of focal plane array technology are greatly impacting the ability to test the arrays. To ease this impact, the Air Force Space Technology Center is coordinating the Focal Plane Test Program. This program consists of 1) a Focal Plane Test Working Group, 2) a Round Robin Test Schedule, and 3) interaction with the users of the focal plane arrays.
Infrared absorption cross sections of As in Si near zero Kelvin have recently been measured in two different investigations. The average of the integrals of the cross section over photon wavenumber was 8.64 x 10-13 cm-1. This is nearly equal to the value predicted by the oscillator-strength sum rule. Between 500 and 1000 cm-1, the absorption cross sections reported here agree very well with 0.7 times the currently accepted formula for the photoionization cross section of As in Si. Calibration errors in spreading resistance measurements on epitaxial layers seem to be the cause of the 0.7 multiplicative error in the photoionization formula. Above 1000 cm-1, 0.7 times the value from the formula predicts a larger photoionization cross section than the absorption cross sections reported here. This is apparently caused by the impact ionization of donor electrons from impurity atoms by energetic photoionized electrons.
The test and evaluation of long wave infrared (LWIR) hybrids is not an insignificant task, especially when large numbers of devices need to be examined. Device temperatures are typically below 20 degrees Kelvin, and low optical backgrounds are imperative. Noise equivalent irradiance (NEI) measurements require exceedingly low station noise floors, while the desire to reduce integration times calls for faster acquisition systems. Reporting data on the large number of pixels in a typical array demands a streamlined way to relate the performance of the part to the test engineer. A hybrid, as used in this paper, is a detector array that is indium bumped to a readout chip. It is also commonly referred to as a sensor chip assembly (SCA). This paper describes some of the hybrid testing developed on the Precursor Above the Horizon Sensor (PATHS) program as well as a few techniques that proved invaluable for the characterization of the low background LWIR hybrids.
A long wavelength infrared relative spectral response measurement system has been developed. Measurements from this system are combined with low background blackbody-based broadband measurements to produce accurate absolute spectral response measurements. The requirement for this capability was driven by the development of LWIR detectors intended for operation at backgrounds of 109 to 1014 photons/cm2 sec and the subsequent need for accurate determination of the figures of merit of the detectors at these backgrounds. The heart of the spectral response system is an Optronics Laboratories model 735D triple grating, subtractive mode, double monochromator with a spectral range of 2 to 30 microns. Details of the monochromator design are presented and the complete spectral response system is described. Sample data from the relative spectral response test is shown and the repeatability of these tests is demonstrated. The monochromator uses pyroelectric reference detectors to measure the relative spectral irradiance on the detector under test. The relative response curves of the pyroelectric reference detectors have been measured by comparison to a blackened thermocouple detector. Samples of these measurements are presented. Test instrumentation, test dewar optical components and test methods for the broadband radiometric measurements are also discussed. Sample data from several broadband tests is shown and the accuracy of these measurements and potential sources of error are analyzed. Absolute detector response is obtained from the relative response and broadband measurements by using the relative response curve to convert the total broadband photon flux to equivalent photon flux at the peak photon response wavelength. This procedure and the required numerical integration software are also discussed. Comparative radiometric measurements of front illuminated Si:As IBC detectors are discussed.
An extremely low background research dewar has been designed, and several have been built and are currently being used for characterization of visible and infrared photon detectors. The dewar has a background in the 106 photons/cm2-s range over our detection capability range of 0.4 to 28 micrometers wavelength. This low background environment facilitates visible and infrared detector experiments involving photon counting, long charge integration times or devices with large surface areas. The temperature of the device under test can be regulated between 4.3 and 20 K. The dewars have been in constant use for several years and have proved to be both flexible and trouble-free. The dewar design incorporates many features that are useful for infrared detector testing and characterization studies in general. Flood illumination of the detector under test can be provided by an internal bandpass-filtered hot resistor source, pinhole shutter, light-emitting diode, optical fiber, or any combination of these sources. The entire test stage is mounted as a plug-in module to accommodate packages with different pin configurations. Calibrated reference detectors are attached to the side of the test detector socket. The leads between the detector test socket and the vacuum feedthroughs are less than 15 centimeters long, to keep corresponding stray capacitances low. The combination of extremely low photon background, flexibility and trouble-free operation results in a dewar that is well-suited for the measurement of sensitive infrared detectors in a research environment.
With the development of large high-performance, self-scanned infrared and visible focal plane arrays (FPAs) for military and scientific applications, fairly sophisticated optical apparatus is required and novel computer-aided data acquisition and processing techniques are needed for complete testing and characterization of these devices. For many advanced tactical and strategic applications, the data generated by a single focal plane sensor chip ca n exceed several tens of megapixels per second and the very low noise can result in a dynamic range in excess of 105 at the lower pixel rates. Also, the number of detectors in a single device can range from a few hundred in the case of a simple scanning sensor to a few million in the case of large staring arrays for scientific applications. This paper describes the capabilities developed over the past several years at the Lockheed Palo Alto Research Laboratory for the detailed test and characterization of a broad range of advanced technology focal plane arrays. Hardware and software techniques for the automated measurement and characterization of key FPA performance parameters such as photoresponse, spectral response, noise, detector RA product, frequency response, optical and electrical crosstalk, and spatial response and MTF will be discussed. Specific examples of data will be shown with special emphasis on automated measurement of frequency response, spatial response and MTF, and spectral response.
An automatic test station for infrared focal plane arrays and multiplexers has been developed by Honeywell EOD. The Computer Aided Mosaic Array Test station (CAMAT) has characterized arrays of 64x128 and is capable of handling arrays of 256x256 and larger. Data rates up to 5 MHz and dynamic ranges of up to 16 bits can be measured. A modular approach is used in both hardware and software. Data acquisition and focal plane drive equipment are custom built while signal processing and display is done by commercial hardware. The software is menu driven and includes data acquisition, statistics, FFT, and graphics output modules in addition to standard sensitivity calculations. A test procedure editor allows for simple development of customized test procedures.
A 64-channel readout device with a 64:1 multiplexer output, designed for use in cryogenic, infrared focal plane applications, is being tested extensively in a production environment. For the test system, an existing product was modified and expanded to provide the accuracy, flexibility, and throughput needed to meet test requirements. Unique test approaches, coupled with innovative throughput enhancement techniques, streamlined operations to allow automated testing of 500 devices per day on one test set with minimum operator intervention. The system executes more than 20 sophisticated tests at wafer level, in order to grade each device as faulted or specification-acceptable. Confidence in the quality and reliability of the deliverable product has also been increased. Creative software modules were designed and integrated with a standard software package (developed and refined over 5 years of device testing), forming a test program which fulfills demanding test requirements. These new modules facilitate viewing of raw data, display of "quick-look" reduced data, failure analysis, and re-sorting of device grades after testing. Additional benefits derived from this development effort include automated testing of packaged parts at both ambient and cryogenic temperatures, and reduced characterization test time for similar devices. Anticipated future improvements include greater utilization of the pipeline processor already present in the system architecture, and continued development to generalize the software package to allow swift reconfiguration for testing other readouts and hybrid arrays of varying size and performance.
An advanced test station has been developed to test integrated focal plane array (FPA) hardware. This test system is intended to test readouts, sensor chip assemblies (SCAs), modules, and complete FPAs. The test station was developed in response to existing and anticipated project requirements that exceed the performance capabilities of existing test stations. A new generation test station will require enhanced performance, such as the ability to accommodate increased data rates, generate lower noise drive electronics, and provide improved diagnostic capability. The goal was to develop a test vehicle that was configurable enough to accommodate the projected FPA test requirements for the next several years, thereby eliminating the need for a new design test station for each individual program. Two types of programs were considered when specifying the new test station: development programs where a substantial amount of diagnostic and analysis capability are required, and production programs where test time and ease of use are critical. If a flexible test station design could be achieved, the nonrecurring costs to implement a new program could be greatly reduced, both in test equipment design and in training test personnel. Since the goal of the CDATS program was to develop a test station that could be reproduced for many projects, the recurring cost of this test station should also be reasonable. The CDATS system is, we believe, a low cost system capable of being configured or easily modified to test most current and future FPA components.
The CECOM Center for Night Vision and Electro-Optics (C2NVEO) infrared nfrared Focal Plane Array test facility is being upgraded to improve speed and test capability. Prior to the upgrade,the test station was capable of testing FPA's, but it did not meet present and future testing needs. Deficiencies such as; high system noise, inflexible dewar configuration, Limited levels of automation, in addition to the inability to perform focussed spot and FPA spectral responsivity measurements were reasons to improve the test station. It was important during this development that the ability to characterize a diverse variety of Infrared (IR) Focal Plane Array designs, sizes, formats, materials, and applications be maintained or increased. Considerations relating to this will be discussed. In the future, the test station will be able to test HgCdTe and PtSi arrays on a much larger scale than at present and test other new technologies as well. The goal is to be able to fully characterize one Focal Plane Array per day.
An extensive material characterization facility has been developed to support a modern epitaxial silicon growth laboratory. The growth laboratory provides material for producing impurity band conduction (IBC) IR detector arrays. The laboratory consists of two sections. Material from the research section is used for producing advanced IBC detectors. Material from the manufacturing section is used for producing detector arrays for producibility studies The characterization facility currently includes a computer-controlled variable temperature Hall effect system, a spreading resistance system, and an electrochemical capacitance-voltage dopant profiling device. A low frequency capacitance-voltage system has been assembled from components for specific use with heavily doped silicon exhibiting impurity band conduction. The variable temperature Hall effect measurement allows determination of majority dopant species and concentration, carrier mobility, and total compensation concentration. Spreading resistance yields rapid doping profile measurement of .1 μm to 100 μm thick epitaxial layers. The electrochemical C-V system allows accurate measurement of sharp doping transitions with the capability of 10 Å resolution. The low frequency C-V system can perform measurements on test samples or completed devices to determine compensation concentration for material doped so heavily that impurity band conduction interferes with the ability of the Hall measurement to provide this information. Representative data from each system will be shown and discussed. Application of each measurement to implement material quality control as well as material development will be examined.
An automatic detector test system to characterize short and medium wavelength photovoltaic (PV) detectors is described. The system is designed to measure impedances from more than 1012 Ω to less than 100 Ω and to measure noise from source impedances between 104 Ω and 1010 Ω at frequencies from 1 Hz to 100 kHz. The system includes multiplexing to test up to 200 elements mounted in a test dewar. Multiplexing is performed with specialized stepping switches for low series resistance, high isolation resistance and low stray capacitance. State of the art high bandwidth, low noise, current sensitive preamplifiers with seven gain ranges cover the wide range of source resistance. The test dewar temperature can be controlled from 80 K to 300 K with a resolution of 0.01 K. Dual signal measurement channels are included for speed and flexibility. The custom built electronics is in two different types of modules and is controlled by a small computer. Optical baffling is used to reduce stray reflections and increase test accuracy.
Since 1987, SBRC has been under contract to the US Air Force (Aeronautical Systems Division, Wright-Patterson Air Force Base) for the Photovoltaic Mercury Cadmium Telluride Infrared Focal Plane Array Manufacturing Technology program. One task in this program is the development of automated visual inspection of arrays to increase effective testing, improve throughput and reduce costs. Each array produced by this program, while only a quarter inch square, has 16,384 diodes. Thorough visual inspection, electrical and physical wafer screening, and final performance testing of arrays is both difficult and time consuming. Physical defects can range from excess debris between diodes to missing or smeared indium bumps. These conditions adversely effect electrical performance and must be identified. The automated visual inspection system will be used to screen out unacceptable arrays, and to reduce costly downstream steps of hybridization (i.e. mating of an array to a readout) and final testing. This will result in reduced labor and better early attrition which will significantly lower the overall product cost, as well as increase the throughput capacity. The automated inspection system, currently under development at SBRC, is built around an Applied Intelligent Systems AIS-5000TM parallel processor. Other modules of the system include a microscope with quadocular head (four viewing ports), computer controlled X-Y stage, and variable lighting schemes. This paper will report on system hardware and software development, algorithm development and verification, and the manner in which automated inspection complements current testing methods.
This paper describes the development, operation, and output of a fully automated, low photon flux background, longwave infrared (LWIR) detector radiometric test capability. The facility includes three low background, broadband, LWIR absolute radiometric response test stations and two medium background relative spectral response test stations. The reasons for developing such a capability are discussed. An overview of the two station types is given with block diagrams, photographs, and descriptions of the electronic, mechanical, cryogenic, and optical components. A description of the available automated test routines is presented with a discussion of the parameter space over which the tests may be run. Sample data plots for several tests of a front illuminated Si:As IBC detector are included. The automation of the system is discussed, including software design, system configuration, operator interface, test control, data archiving, and data reporting.
The architecture of a versatile, reliable, and expandable automatic test system is described. The system incorporates a low noise computer-like control and address bus into which a number of boards can be inserted. Individual boards have been designed to date to multiplex 16 detector inputs and bias one of them, to bias and multiplex 8 detectors, and to power internal dewar bias networks with a programmable low noise power supply. Test systems built with this architecture have been configured to test arrays of up to 180 elements at measurement frequencies from DC to 5 MHz. Noise levels well under 1 nV per root Hertz, with negligible stray signal contamination, are routinely obtained over the entire frequency range. Design details of individual modules will be described in addition to description of the overall system design and EMI control philosophy.
This paper describes a method to determine absorbed radiation dose inside a dewar at cryogenic temperatures. Results of using thermoluminescent dosimeters (TLDs) placed externally on the dewar and then extrapolated to obtain the inside dose are presented. TLDs measure the cumulative dose and a Si pin diode monitors real time dose during each Linac pulse. Dosimetry methods are taken from results from electron tests at a Linac using 4, 12 and 18 MeV electrons and a neutron test using Californium-252 (252Cf) fission neutrons.
A low infrared background helium dewar has been built to evaluate the response and recovery of infrared focal plane arrays to transient electron bursts with mean source energy of 1.4 MeV. The dewar has thin beryllium windows in the vacuum and helium shields, open areas in the nitrogen shield, and a mount that allows the device under test to be perpendicular to the incident electron beam. Beryllium provides shield support and optical background reduction while minimizing electron energy absorption. Dose rates as high as 1012 Rad(Si)/s can be achieved at the device under test. A Febetron 705 flash x-ray (FXR) generator at Rockwell International is used to produce the electron beam. The dewar was used with the FXR electron beam generator at various dose rates. The dosimetry performed in the dewar/electron beam environment will be discussed, including an analysis of the electron beam energy spectrum at the device position.
A facility capable of producing very high gamma flux levels and rapid total dose accumulation for testing infrared detectors and multiplexed hybrid arrays has been developed. Effective gamma fluxes from 108 to 1014 gamma/cm2-sec equivalent are routinely achieved. Total dose testing has been performed on various devices at rates between 0.5 rads(Si)/second and 50,000 rads(Si)/second. Testing has been performed at temperatures between 5K and 77K and optical backgrounds as low as 1 • 1011 photons/cm2-sec. In addition to the electron beam, a low level gamma source facility and complete electronics test suite are described. Data will be presented for several types of detectors and hybrids to demonstrate test capabilities and device response.
A low-temperature cryostat to be used in high fluence neutron work, which uses titanium as the primary construction material, is described. Results of the material studies used to make the choice are presented and machining and joining techniques used in the fabrication are described. A heatsink design that provides rapid cooldown is described; it can be used in device studies in the temperature range up to 40° Kelvin using liquid helium as the refrigerant.
Experience shows that there is no only-way to accomplish radiation testing in the field. Many good tests have been done with sketchy plans and written procedures but by a test crew that has the knack to respond quickly to implications of each succeeding test result and to things that go wrong. Other good tests have been conducted with detailed written plans and procedures, though usually with some on-the-spot modifications. Failure most often results from either poor planning or rigid procedures, carelessness in preparation or test conduct, non-response to the unforeseen, and occasionally from facility failure.
A low dose rate (3 rad(Si) per sec) radiation test facility of novel design has been constructed. The apparatus is used for radiation testing of infrared detectors and associated readout circuits. Utilizing a one Curie cobalt 60 source, an ionizing event flux approaching 1 x1010cm-2s-1 is achieved. System design, capabilities, and dosimetry are presented. Also, examples of actual test data are given. The uniqueness of the system arises from the fact that the movement of the radioactive source from the lead pig into the test dewar is controlled from outside the lead experiment enclosure ("hut") via a tungsten-alloy control rod which passes through a small hole in the wall of the hut, virtually eliminating operator exposure. Additionally, the cryostat is a continuous-flow type, and the design of the test dewar allows for the cobalt 60 source to remain at room temperature at all times. Radiation testing using this facility is very convenient, and event rates are highly repeatable due to the indexed, lockable source control rod.
The Mosaic Array Test System (MATS) has been developed by the Air Force Weapons Laboratory (AFWL) to perform radiation effects characterization of infrared detectors, readout devices, and infrared hybrid arrays. This paper describes the key features of MATS, the testing methods used to perform radiation effects measurements, and experimental results which illustrate the current system capabilities.
The design and construction of a system to test common module detector dewar assemblies with a simulated cooler vibration is described. Pseudo-random vibration levels up to 10 g RMS at frequencies up to 20 kHz can be applied to the unit under test, with control accuracy better than 1 dB. A test time of about 3 seconds per element is routinely obtained for arrays up to 180 elements. The system is software configurable to address a variety of testing requirements; vibration levels and frequencies and noise analysis frequencies and bandwidths are variable. Unique features of this system include a true three axis vibration control system which is integrated into the main system controller, and a custom built 180 channel preamplifier/multiplexer array, and a small size and quiet operation for production area compatibility. The system is compatible with a family of plug in preamplifier cards, further enhancing its adaptability. The system has been in continual use for several years and has proven to be versatile, adaptable, and reliable.
Nonuniformity correction is a necessary image processing step for achieving background limited performance in staring IR focal plane arrays. In developing successful correction techniques, drift in the array output must be considered in order to ensure that the correction coefficients remain valid over a reasonable length of time. We discuss two generic sources of drift, namely system drift and drift originating within the individual unit cells. The latter source of drift is synonymous with 1/f noise and can be shown to have a direct effect on the spatial noise of the array as a function of time. Examples of these phenomena in two current state-of-the-art MWIR staring arrays are given.
Neutron fluence testing of large dimension electrical components at cryogenic temperatures has traditionally been performed at fast burst reactors. A new facility, using Californium-252 (252Cf) as a neutron radiation source, was designed and built at the IRT Corporate Headquarters in San Diego. This paper discusses the physical properties of the californium isotope, describes the facility and its test capabilities, and presents the results of the 1 MeV calibration as well as IR and visible sensor testing.
Santa Barbara Research Center has developed nuclear radiation testing methodology to measure the performance of IR. detectors in simulated total ionizing dose, transient dose rate (and flux), and neutron environments. The total dose test for IR detectors required that temperature, IR background, bias, and vacuum conditions be maintained and controlled during in-situ measurements of performance parameters; these conditions were accomplished with a specially fabricated Helitrans dewar adapted to a Cobalt-60 internal irradiator. This irradiator and dewar configuration with the appropriate portable measurement equipment allowed I-V and C-V measurements before, during and after irradiation under the same test conditions. This same portable test set was also used to measure detector performance parameters following irradiations with flash X-ray machines. The dose rate test methodology produced consistent detector response and recovery profiles from different X-ray machines at different test facilities. The transient flux test utilized portable equipment including a multichannel analyzer, pulse-shaping amplifier, and a low-noise dewar with built-in amplifier to collect the pulses caused by gamma-flux-induced charge in the detectors. This test was performed at a Cobalt-60 source range which produced a good simulation of the actual environment. The neutron fluence test used portable radiation test equipment and dewars specifically designed to minimize the induced activity expected during irradiation of the detectors for example, in a TRIGA fast-neutron reactor. In this case, the TRIGA reactor core was positioned and shielded in the pool to produce predominantly fast neutrons with an acceptably low level of reactor gammas. The general test methodology, consistent with MIL-STD-883C, provides a capability to test detectors in any ionizing or displacement nuclear radiation type.
The Advanced Components Evaluation System at NOSC is a computer controlled test system for collecting and processing radiation effects data from multiplexed infrared detector arrays. The system is designed to support testing at a variety of radiation sources such as neutron, gamma, and x-rays. The various components of the test system will be discussed, and sample data formats from neutron tests performed at IRT in LaJolla California will be presented.
Infrared detector focal plane arrays made from platinum silicide (PtSi) are currently being used in applications where good reliability, low cost, and high producibility are required. Future use of these sensors in extreme operating environments has also been suggested. We have tested small arrays of detectors in constant acceleration environments up to 300,000 times that of gravity, and in dynamically changing acceleration stress up to 5000 g/sec. Measurements of both the optical and electrical performance of the detectors were taken after acceleration. Our measurements show that there are no changes in the infrared spectral characteristics of the samples subject to either transient or steady state acceleration up to 300,000 g. The only changes noted in the samples were found from electrical measurements where excess dark current was noted at static accelerations greater than 200,000 g. Our results show that platinum silicide should be an excellent candidate for applications which anticipate large static and dynamic acceleration stresses. Detailed experimental data is presented.