Bolometers Running Backward: The Synergy Between

Uncooled IR Sensors & Dynamic IR Scene Projectors

Paul Bryant*a, Steve Solomonb, Jay Jamesc
a

Left Coast Consulting, 1222 Manitou Lane, Santa Barbara, CA 93101

b
Acumen Scientific, P.O. Box 6084, Santa Barbara, CA 93160
c
Santa Barbara Infrared, Inc., 30 S. Calle Cesar Chavez, Suite D, Santa Barbara, CA 93103
ABSTRACT
The leading IR scene projection (IRSP) device technology, resistive emitter arrays, has grown from its early roots in the
uncooled microbolometer community into a separate and highly specialized field of its own. IRSP systems
incorporating microbolometers running backwards are critical tools now ubiquitous in laboratory testing and
evaluation of high performance IR sensors and their embedded algorithms. Adoption of IRSPs has reduced the scope of
flight/field testing, producing dramatic resource savings and strong system development advantages.
Modern IRSP systems provide the capability to project high-resolution (1024 x 1024), high-temperature (750 K)
dynamic MWIR-LWIR imagery at frame rates up to 200 Hz, with 16-bit input resolution. Novel IRSP systems are now
being developed to test advanced FPAs and sensors requiring wide-format (768 x 1536), cryogenic background (50-80
K), fast-framing (400 Hz), and/or very high-temperature (2500 K) dynamic IR simulation in order to be properly
evaluated.
The ongoing cycle of sensor improvement and test system evolution is perfectly illustrated by the parallel development
of IRSP and emerging FPA/sensor technologies. The cross-pollination of technology between the sensor and projector
domains continues to bring innovation to both communities. Technological trends related to semiconductor and microelectrical-mechanical system (MEMS) device fabrication, real-time digital video processing, and EO system design are
being exploited by both sensor and projector developers alike with advantages realized by both.
This paper presents a lighthearted overview of the technical evolution of IRSP from its early microbolometer roots,
discusses current and emerging IRSP capabilities, illustrates the device-level to system-level synergy between sensors
and projectors, and offers a peek into the advanced EO simulation capabilities and technologies which will be required to
address emerging FPA and sensor trends.
Keywords: Emitter Array, IR Scene Projection, Micro-Bolometer, MEMS Fabrication.

1. INTRODUCTION
Uncooled micro-bolometer technology is now ubiquitous in the infrared sensor world. Ever-improving radiometric
performance and the emergence of multiple component- and system-level producers has turned this niche thermal
imaging market segment into a growing and competitive marketplace. Utilizing 3-D bridge structures to suspend
individual resistor elements above a CMOS readout array, modern microbolometers typically require a combination of
licensed and proprietary MEMS materials, fabrication processes, and design techniques. To enable the advancement of
next-generation thermal imaging devices, microbolometer R&D continues at a rapid pace, with smaller and higherperformance MEMS pixel technologies at the cutting-edge.
IR scene projection (IRSP) development began in the 1970s, with early filaments and Bly cells eventually replaced with
*

[email protected]; (805) 708-3029; http://www.leftcoastconsulting.net

resistive arrays based on the same core microbolometer MEMS pixel technology originally developed for IR sensing
applications. Resistive arrays provide a unique IR simulation capability, due to their combination of wide radiometric
dynamic range and high thermal resolution. Now commonly used in a variety of hardware-in-the-loop (HWIL) test
applications, resistive array based IR scene projection (IRSP) systems have benefited greatly from the development of
microbolometer technology. Interestingly, IRSP technology has now reached the point at which unique innovations in
MEMS pixel design and fabrication are flowing back into the IR world, with applications and potential benefits in both
sensor and projector communities.

2. FROM BOLOMETERS TO EMITTERS

(or, Hey, I wonder if we could run these bolos in reverse)
During the 1970s, it was recognized that a dynamic IRSP capability was required in order to effectively evaluate the
latest and greatest IR imaging systems of the day. Static infrared targets were not adequate for true characterization of
the emerging sensors and algorithms, and field tests were too time consuming, risky, and costly. This T&E gap
prompted a series of studies to identify and evaluate potential dynamic IR simulation technologies. As we will discuss
in section 2.4, resistive arrays emerged as the leading technology for all but a few IR scene projection applications.
Before we review the history of IRSP device research and the chronology of resistive array development, we turn our
attention to the MEMS or more accurately, the micro-optical-electrical-mechanical system (MOEMS) structure of
microbolometers and resistive emitters, a review of pixel thermal physics, and a comparison of FPA device and system
drivers for both classes of device.
2.1. Pixel Operation & Key Interfaces
For both microbolometer and resistive emitter pixel structures, a resistor element is suspended above the IC substrate by
two support legs. An optical etalon is formed between the absorptive layer and a reflective layer typically deposited
onto the top surface of the substrate/IC. In the case of microbolometer pixels, incident radiation causes a temperature
(and hence resistance) increase within the pixel, thereby producing an irradiance-dependent current integrated and
processed by the readout IC (ROIC). In the case of resistive emitter pixels, the thermally sensitive resistor film used in
the microbolometer device is replaced with a unique resistive film capable of handling the extreme thermal excursions
required by IR simulation applications.
During scene projection, a unique drive current is applied to each emitter pixel via a read-in IC (RIIC). This applied
current causes power to be dissipated in each pixel body (P = I2R), and the pixel temperature to rise dependent upon
dissipated power and thermal losses (neglecting radiative losses). Depending on the physical temperature of the emitter
and the in-band emissivity of the resonant pixel structure, IR radiation is emitted (Lout Tphys, ) and subsequently
projected to the UUT by a collimator.
A typical resistive emitter pixel structure and RIIC drive topology is illustrated in Figure 1.

Figure 1 Typical Resistive Emitter Pixel Architecture (left) & Electronic Drive Scheme (right)

Microbolometer and resistive emitter arrays usually require thermal stabilization, although for different reasons. Selfheating and/or thermal drift in microbolometer arrays results in image artifacts, and thereby limits radiometric
performance. As we will discuss in more detail later, the large drive currents employed in running resistive arrays to
useful simulated IR temperatures give rise to very high levels of array power dissipation and spatial power density. For
reasons associated with both elimination of image artifacts and prevention of overheating/damage to the scene projector,
resistive emitter arrays are generally liquid-cooled.
The optical interfaces to bolometer and emitter arrays are similar in that they both require specific optics with sufficient
transmission, low aberration and distortion, high MTF, etc. Microbolometer optical systems are typically designed to
strike a balance between performance, mass, cost, and other constraints associated with development of highperformance, compact EO systems. IR scene projection optics, on the other hand, are usually driven by the need for the
IRSP system to introduce minimal degradation into the imagery, with mass and cost often considered secondary to
performance. As such, the worlds of sensor and projector optics share much of the same basic technical emphasis, with
a dramatically different emphasis on size, weight, and cost. As we will discuss later, novel designs incorporating
modularity and zoom functions are facilitating the development of sensors and projectors alike. Figure 2a shows a
typical emitter pixel and projected image, while Figure 2b shows a typical mircobolometer pixel and core imagery.

Figure 2a Resistive Emitter Pixel (left) & Projected MWIR Image (right)
(courtesy of Santa Barbara Infrared)

Figure 2b Uncooled Micro-Bolometer Pixel (left) & LWIR Image (right)

(courtesy of Raytheon Vision Systems)

While most microbolometer arrays convert photons into analog video signals, emitter arrays convert 16-bit digital scene
data into an array of unique analog current levels applied to each emitter. Although the sensor community is
increasingly incorporating A/D conversion into ROIC designs, IRSP arrays have included on-chip, 14- or 15-bit DACs
since the 1990s. This early shift to an all-digital data interface was driven by the difficulty of optimizing the analog
interface between a high frame rate emitter array and a remotely located DAC (or even multi-DAC) subsystem.

2.2. Hardware-in-the-Loop (HWIL) Testing

IR scene projectors are used in an increasing number of HWIL test environments. These applications include the test
and evaluation of FLIRS, imaging IR seekers, IR countermeasures, infrared search and track (IRST) systems, IR threat
warning sensors, and space/strategic sensors.
The IRSP is typically employed within a dynamic, feedback-controlled architecture, as illustrated in Figure 3. In this
type of HWIL installation, the IRSP receives digital video from a scene-generation computer typically an SGI
machine, but increasingly based on desktop PC technology. The IRSP mounts on the outer arm of a flight motion table
and projects imagery to the UUT located at the tables center. UUT image/tracking data and guidance commands are
sent to a simulation processor, which performs real-time coordination of UUT line-of-sight (LOS) and scene generator
imagery.

Figure 3 Typical Hardware-in-the-Loop (HWIL) Simulation Architecture

Latency compensation is often implemented as a set of real-time azimuth/elevation/rotation parameters sent from the
simulation processor to a point near the output of the IRSP data path, enabling x, y, and rotation to updated based on
current UUT attitude.
2.3. IRSP Performance Drivers
(or, Scene projector performance: How much is enough?)
For modern IRSP systems, array format is often required to be at least 10242 in order to support the more stressing test
applications. Pixel size has held constant at approximately 50 m, due largely to concerns associated with LWIR
output/diffraction and pixel fill factor. Frame update rates are typically supported in the range of 20-400 Hz, thereby
allowing simulation at integer multiple of the UUT frame rate and minimizing spatial-temporal sampling artifacts. Both
snapshot and raster update modes allow optimum synchronization with both scanning and staring sensors under test.
The maximum apparent (simulated) temperature of production IRSP systems is currently in the range of 700-800 K
(MWIR) and 500-600 K (LWIR). Development is already underway to extend the upper end of the simulated
temperature range to at least 2500 K the effective temperature of a late M-class star! While maximum apparent
temperatures are rising, some HWIL applications such as space-based interceptor T&E require minimum simulated
temperatures in the cryogenic regime. This of course requires an array specifically designed to operate with very low
substrate temperature, yet with top-end radiant output comparable to traditional systems.
Typically, each pixels radiance rise/fall time must be less than approximately 5 ms, and scene-dependent cross talk must
be less than 1-3%. Operability must be greater than 99%, and cluster defects maintained to an acceptable level. Pixel
substitution and interpolation functions available to sensors offer nothing to improve imagery projected by IRSP systems
no photons at a given emitter location means no radiant output, period. IRSP operability is thus a tremendous
challenge, exacerbated by increasing array formats and the inherent difficulty of fabricating MEMS devices
monolithically onto CMOS wafers.

2.4. History of IRSP Technologies

(or, Theres more than one way to make Max Planck happy)
As mentioned previously, the quest for suitable IRSP technologies began in the 1970s. Initial candidate technologies for
IRSP use included the Bly cell, the liquid crystal light valve, resistor arrays, Texas Instruments digital micromirror
device, deformable mirror devices, scanned laser devices, the infrared CRT, a vanadium dioxide spatial light modulator
and later, laser diode arrays. Most of these devices had one or more limitations precluding their use in the most stressing
HWIL environments, and resistive arrays emerged as the leading technology for all but a few IR scene projection
applications. Figure 4 shows a historical overview of IRSP device R&D, from the early days of the Bly cell, to the
current generation of systems relying predominantly on resistive arrays, to the ongoing research in the areas of VCSEL
arrays, plasmas, and quantum/photonic crystal devices.

Figure 4 Historical Overview & Time Line of IRSP Technologies

Figure 5 Chronology of Resistive Emitter Array Development for HWIL Applications

Figure 5 illustrates the chronology of resistive array development for IR scene projection. The history of resistive arrays
began with the early 128 x 128 Electro-Optek device, and evolved to the 1282, 5122, and 10242 offerings from BAe and
the 1282, 5122, and 672 x 544 (DIRSP) configurations produced by Honeywell. After purchasing an exclusive license to
the Honeywell MEMS/emitter technology in 2001, Santa Barbara Infrared is now the industry leader in IRSP, with 5122
cryogenic 5122, and 10242 systems in production, and a wide-format 768 x 1536 system in development.

3. PIXEL-LEVEL DEVICE PHYSICS

(or, Foiled by Conservation of Energy, Once Again)
Although microbolometers and emitters rely on different methods of modulating resistor temperature (radiative
absorption vs. electrical heating), the fundamental thermal performance of both types of device relies upon the same type
of heat (or power) balance equation.
3.1. Pixel Performance
For an emitter pixel suspended above a substrate with temperature Tsub and driven by a constant current, the relationship
between pixel power, conductive and radiative losses, heat capacity, and the thermal derivative of pixel physical
temperature with respect to time is given by:

C(T )

dTphys
4
4
= I 2 R(T )" G(T )[Tphys " Tsub ]" g[Tphys " Tbg ]
dt

[1]

where key parameters are defined as follows:

Tapp " Apparent Temperature

R(T) " Emitter Resistance

C(T) " Heat Capacity
G(T) " Thermal Conductance
g " Geometric/Radiative Factor

Tphys " Physical Temperature

Tsub " Substrate Temperature
Tbg " Background Temperature

As (a) the thermal derivative goes to zero and (b) the device is operated in a temperature regime limited by conductive
losses (i.e. real world temperatures for current microbolometers
and emitters), we are left with the fundamental steady!
!
state equation relating pixel power to conductive losses:

I 2 R(T ) = G(T )[Tphys " Tsub ]

[2]

For an emitter driven to a given physical temperature per equation [1] or [2], the pixels simulated (apparent)
temperature may be computed by solving for the variable Tapp in the following equation:

!" 2

"2

# B (T
"

"1

app

)d" = ff # $ opt ( " )%& ( ")% B" (Tphys )d"

[3]

"1

where 1 and 2 are the lower and upper spectral band edges, B(T) is the spectral radiant emittance (Planck function), ff
is the pixels optical fill factor, OPT is the transmission of the projection optics and package window(s), and () is the
pixels spectral emissivity.
! The design of emitters as with microbolometers is complicated by the fact that critical
material parameters such as heat capacity and thermal conductance are sensitive to process variations. Emitter design is
further complicated by the fact that (a) multiple parameters depend upon instantaneous physical temperature and (b)
each pixel must operate over a temperature range of 700-800 K.

3.2. Mo Betta Emitters

(or, Its gettin hot in here)
As indicated previously, each emitter must be able to slew 700-800 K in 3-5 ms to be useful in modern IRSP
applications. This obviously places extreme demands upon the emitter material set (i.e. bridge dielectric, leg and
resistor metals, etc.) and operating environment (i.e. temperature and vacuum). The latest generation of 1024 x 1024
emitter arrays provides a maximum apparent temperature of approximately 700-750 K, draws 100 A peak current, and
dissipates over 1 kW, peak! A new high-temperature emitter development program is striving for apparent temperatures
of ~ 2500 K, and hence physical temperatures of approximately 3000 K! Figure 6 illustrates a high-temperature BAe
emitter array driven to incandescence.

courtesy of BAe

Figure 6 Resistive Array Driven to Incandescence

(courtesy of BAe)

In addition to handling the extreme temperature environment required for IRSP, emitters must also provide a high degree
of physical robustness, good radiometric stability, and high array operability. As with bolometers, a successful emitter
design and material set must provide high IR emissivity (absorption), suitable thermal conductance to substrate, and low
pixel heat capacity. Who said it would be easy?

4. IRSP SYSTEM CHALLENGES

(or, Honey, Where Can I Dump this Kilowatt?)
With the current generation of 1024 x 1024 emitter arrays drawing 200 A and dissipating over 1000 W, cooling and
thermal stabilization are crucial to system survival and simulated performance. Power density for the current and
emerging emitter arrays is approximately 40 W/cm2, thus requiring novel thermal design solution to remove heat (and
return current) from the package. To achieve this, each emitter array is reflow soldered to a specialized heat spreader
within the device package. The heat spreader is soldered to a meso-channel fluid plenum/cooler, through which a
special coolant (typically HFE 7100) is controlled and re-circulated by a high-capacity chiller. Figure 7 illustrates a
1024 x 1024 projector core, and shows the typical 2-D distribution of heat across the emitter substrate during active
projection.
The 200 A maximum emitter current is sourced to the array through hundreds of parallel wire bonds on the RIIC surface,
and returned through the backside-metalized RIIC substrate, through the heat spreader and meso-channel cooler, to a
copper heat sink. From the heat sink, low inductance buss bars route the emitter supply and return nodes to switching
power converters located in the projector head. Power converter inputs are connected to a main power supply in the
remote command and control electronics (C&CE) rack.

Figure 7 1024 x 1024 Scene Projection Core (left) & Thermal Distribution Across Array (right)
(courtesy of Santa Barbara Infrared)

As temperature requirements become more stressing and emerging sensors demand higher-resolution emitter array
formats it is clear that these thermal and electrical design solutions must continue to improve for future IRSP systems.

5. SENSOR/PROJECTOR SYNERGY
(or, Two Niche Communities Are Better than One)
The continuing arms race between sensors and test systems has historically resulted in the cross-pollination of key
MEMS, electronics, and optics technologies. Next-generation FLIRs, trackers, visible imagers, and laser range finders
are contributing advancements to the scene projector realm and vice versa, as illustrated in Figure 8.

Figure 8 The Cyclical Nature of IR Sensor and T&E Technology Advancement

Pixel uniformity, operability, and device architecture are evolving for both microbolometers and emitters. New highspeed digital video processing technologies are enabling new sensor and projector capabilities. Improvements in optical
system design are facilitating the efficient development of new sensors, and improving the flexibility of IR scene
projectors. Together, these complementary developments will continue to benefit both sensors and projectors.
5.1. Pixel Uniformity & Operability
(or, Why is nobody else trying to make 2 x 2-inch bolo arrays?)
The monolithic MEMS fabrication processes used in microbolometer and emitter production are complex, sensitive, and
costly. Fabrication time is generally on the order of 3-6 months, with many adjustment and tuning steps involved.
Although many MEMS technology licensees work to improve and optimize the core process(es), numerous variability

and defect mechanisms give rise to uniformity and operability limitations which in turn drive yield, cost, and lead time.
Non-uniformity correction (NUC) is applied in both sensors and projectors to improve spatial uniformity at the expense
of dynamic range. However, the sophisticated pixel substitution and interpolation functions routinely implemented in
modern sensors are simply not possible in the projector domain no output, no projected pixel!

Figure 9 The Importance of Array Operability!

5.2. Emitter & Micro-Bolometer MEMS Innovations

(or, How much $#!% can we fit in this five-microgram sack?)
Figures 10a through 10d show examples of MEMS structure innovations created to improve bolometer and/or emitter
performance, increase operational flexibility, and offer novel solutions to optical switching problems at the array level.
These MEMS
developments
will surely
to Eyelid
improve
IR sensors
and projectors.
Artificial
Actuator
with High-Frequency
Emitter
Pixels with Reduced
Masscontinue,
& Low driven by the ongoing need
Optical Switching Capability (max frequency ~ 5-10 kHz)

Due to Gossamer MEMS Architecture (39 m, 51 m)

courtesy of Santa Barbara Infrared, Inc.

Figure 10a Emitter Pixels with Reduced Mass & Low

Due to Gossamer MEMS Architecture
(courtesy
of Santa Barbara
Advanced
Micro-Bolometer
ArrayInfrared)
with Optimized
Fill Factor and Conductance (640 x 480, 20 m)

courtesy of Raytheon Vision Systems

Figure 10c Advanced Microbolometer Array

with Optimized Fill Factor and Conductance
(courtesy of Raytheon Vision Systems)

courtesy of RTI International

Figure 10b Artificial Eyelid Actuator with

5-10 kHz Optical Switching Capability
(courtesy
of RTI International)
Experimental Micro-Bolometer Structure
with Variable Responsivity (tunable thermal conductance)

courtesy of University of Minnesota

Figure 10d Experimental Micro-Bolometer Structure with

Variable Responsivity/Tunable Thermal Conductance
(courtesy of University of Minnesota)

5.3. Signal Processing & Data Throughput

(or, Digital video data rates: Too much of a good thing)
In the sensor domain, data throughput and storage performance are typically driven by high-performance hyperspectral
imaging applications. High-rate data streams are in the range of 50-100 MB, with resulting data cubes for typical
collection scenarios in excess of 100 GB. Lossless data compression schemes, optical interconnection, high-bandwidth
downlinks, and other technologies are striving to maximize sensor performance and mission utility. Real-time functions
such as adaptive NUC and smart FPA processing promise to further improve sensor performance in the future.
In the projector domain, technical evolution is being driven by high-resolution IRSPs such as the Large- and WideFormat Resistive Array (LFRA/WFRA) projects. To support 200 Hz simulations, these systems require data throughput
of over 400 MB/s, with real-time implementation of NUC, translation/rotation, convolution, and other functions. Fiber
optic links between C&CE and projector core improve performance and reduce electronic integration risk. Minimizing
latency is important in IRSP applications, and projector systems now include real-time horizontal, vertical, and rotational
(x, y, ) scene adjustment to compensate for changes in UUT attitude during scene data transfer. As with many other
key performance parameters, the effectiveness of latency compensation depends upon the electronic architecture, data
path, and firmware implementation. As of the last several years, the custom (and somewhat arcane) computing and
video electronics subsystems utilized in IR scene projection are being replaced by COTS-based design-solutions.
Figure 11 illustrates the complexity of hyperspectral data cubes in the sensor world, and the importance of highresolution, high-throughput video data paths to IR scene projector performance.

Figure 11 Hyperspectral Data Cube (left) & High-Resolution IRSP Imagery (right)

5.4. Optical System Design

(or, Ive looked at photons from both sides now)
Ongoing improvements in optical system design, modeling, fabrication, and coating improve the evaluation of new
thermal imaging systems, and increasingly allow IR scene projection systems to be adapted to a variety of T&E
applications and UUT types. Custom solutions such as modular and zoom architectures allow the improved in-process
screening of thermal imaging cores, thereby minimizing failures at higher levels of assembly due to image artifacts. The
same types of modular and zoom architectures are applied to IRSPs, enabling HWIL facilities to re-configure for testing
of different UUTs and simulated scenarios without the need for unique refractive collimators.
Modular optics typically offer discrete adjustment of focal length to optimize projected IFOV for the UUT and/or
simulation at hand. As IRSP systems are usually operated with multiple emitter pixels optically mapped to a single UUT
pixel (to minimize spatial aliasing artifacts), the ability to tune projected IFOV is extremely useful. Zoom optics, while
more costly to develop and manufacture, offer continuous variation of IFOV and FOV within the design range of the

collimator. Full zoom capability provides maximum flexibility to match the projected image to the UUT with ease of
re-configuration and minimal re-alignment time. Figures 12a and 12b illustrate modular and zoom architectures

Figure 12a Novel Approaches to Sensor & Projector Optics:

Modular (left) and Zoom (right) Architectures
(courtesy of StingRay Optics)

Figure 12b Adjustable IR Lens for IRSP NUC

and/or Camera Core Image-Based Screening
(courtesy of Santa Barbara Infrared)

6. SENSOR SIMULATION TRENDS

(or, What Have You Done for Me Lately?)
As sensors incorporate increasingly advanced architectures and offer higher levels of performance and operational
flexibility, EO simulation systems must necessarily keep pace in terms of FOV, spatial resolution, dynamic range, frame
update rate and mode, temporal characteristics, and uniformity/operability. Table 1 lists a variety of emerging sensor
trends and operational parameters, with their corresponding simulation/projection challenges.

Table 1 Summary of Emerging IR Sensor & Scene Projection Trends

High-resolution UUTs are being addressed by emitter arrays of 768 x 1536 pixels, and beyond. 3rd-gen FLIRs and
broadband imaging suites are currently supported by optically combined IRSPs, with photonic crystals, laser diode
arrays, and other simulation technologies in development. Hyperspectral sensors have spawned a series of novel
arbitrary spectrum generators, capable of projecting distinct tuned spectra for each pixel location in a line image, with
area projection under development. Extended dynamic range sensors are currently addressed by resistive arrays
combined with discrete, scanned laser spots. For the future, technologies such as laser diode/VCSEL arrays and UVVIS-IR plasmas are in development. Laser-driven projection of sub-image pixel clusters formed and steered by 2-D

analog mirror-based MEMS arrays will enable high frame rate simulation for emerging UUTs. 3-D imaging/LADAR
simulation poses one of the most difficult scene projection challenges, although a variety of approaches are being
explored including laser diode and VCSEL arrays with vertically integrated electronic control, leveraging ongoing
developments in the sensor realm.

SUMMARY
Resistive arrays for IR scene projection evolved from the early days of microbolometer development, and currently offer
a unique IR simulation capability. While bolometers and emitters are limited by the same thermal and physical
constraints, MEMS advancements are pushing both bolometer and emitter technologies forward. The cross-pollination
of electronic and optical technologies and architectures further increases the pace of development and the feasibility of
the most challenging sensor and projector solutions. This trend will surely continue, to the benefit of sensor and
simulation system developers worldwide.

ACKNOWLEDGMENTS
The authors gratefully acknowledge the consultation and assistance of Santa Barbara Infrared, RTI International,
Raytheon Vision Systems, StingRay Optics, the Tri-Services IRSS P3I Program, Air Force Research Lab, Missile
Defense Agency, Dr. Owen Williams, Mr. Tom Joyner, and Sophie the cat.

REFERENCES
1.

O. Williams, G. Goldsmith, & R. Stockbridge, History of Resistor Array Infrared Projectors: Hindsight is Always
100% Operability, Technologies for Synthetic Environments: Hardware-in-the-Loop Testing X, Proceedings SPIE
Vol. 5785 (2005).

2.

T. Joyner, IR Scene Projector History, IRSP Working Group Session, 2005.

3.

P. Bryant, J. Oleson, J. James, S. McHugh, J. Lannon, D. Vellenga, S. Goodwin, A. Huffman, S. Solomon, G.

Goldsmith, MIRAGE: Developments in IRSP Systems, RIIC Design, Emitter Fabrication, and Performance,
Technologies for Synthetic Environments: Hardware-in-the-Loop Testing X, Proceedings SPIE Vol. 5785 (2005).

4.

S. Solomon, R. Ginn, S. Campbell, M. Jalali, G. Goldsmith II, High Temperature Materials for Resistive IRSPs,
Technologies for Synthetic Environments: Hardware-in-the-Loop Testing XI, Proceedings SPIE (2006).

5.

D. Murphy, M. Ray, A. Kennedy, J. Wyles, C. Hewitt, R. Wyles, E. Gordon, T. Sessler, S. Baur, D. Van Lue, S.
Anderson, R. Chin, H. Gonzalez, C. Le Pere, S. Ton, and T. Kostrzewa, Expanded Applications for High-Performance
VOx Microbolometer FPAs, Infrared Technology & Applications XXXI, Proceedings SPIE (2005).