Presentation Applications for Presentation Dr. A. Fenner Milton - - PDF document

Presentation Title Presentation Title

Presented to Name By Name Presenter’s Title Presentation Date

• Dr. A. Fenner Milton

Director Night Vision & Electronic Sensors Directorate

Presented by Dr. Don Reago Principal Deputy Technology and Countermine Night Vision and Electronic Sensors Directorate 19 March 2008

Future Army Applications for IR Focal Plane Arrays Future Army Applications for IR Focal Plane Arrays

Approved for Public Release, distribution unlimited.

• To discuss potential applications of IR

FPAs to future Army needs

– Cooled and Uncooled Purpose

MCT on Si/GaAs ?

MWIR MCT on Si? InSb

LWIR QWIP ? Staring MWIR InSb for Smaller Optics

Common Modules

Scanning LWIR MCT

SADA I/II

PtSi

TE Cooled MCT

InSb

25 µm Uncooled Manportable/Low Cost 1990 1995 2000

Long Range Airborne (at altitude)

2005 2010 17 µm Uncooled

Multiband MBE MCT

• n CdZn Te

Small Pixel MCT SLS ? Large Multiband Arrays Very Large Arrays InSb Compact Systems

Helo and Ground to Ground

1 2 3

Imaging IR FPA Technology Roadmap

Goal: Affordable Day/Night/IR pilotage and situational awareness for two pilots and crew UH-60, CH47 Approach: Distributed Aperture Sensor using stitching software (no gimbal) feeding multiple head mounted displays, multiple cameras covering wide field of regard with high resolution forward and lower resolution side/rear coverage

High Resolution Forward Baseline (0.8 – 0.9 mR IFOV) For forward sector: Need two low cost cameras with large (1.5K x 1.5K) arrays (MWIR or LWIR) with short integration time (t = 2-3 msec) For side cameras: Need low-cost night imaging with short integration time (2-3 msec). Possibilities include fast uncooled, I2TV and/or SWIR

H 90° - 180° V 60° – 90°

+

Low Cost Side/Rear Night Imaging (IFOV TBD)

Future Application #1: Pilotage for Utility and Lift Rotocraft

Current 3rd Gen: 640x480 MCT on CZT MBE Grown LW/MW Current 2nd Gen:

480x4 Scanning MCT 480x1153 Effective (1.5 S/D)

Future 3rd Gen:

720x1280 MCT

• n Si

Goal: Achieve current 2nd Gen FOVs with new 3rd Gen dual band staring focal plane arrays (improve range and maintain FOV)

Reduction in cost required for larger dual band arrays

Gen 3 B-Kit

Future Application #2: Large Format Gen 3 FLIR (Dual Band MW/LW)

55o FOV IR I2 SWIR

• IR 20o/27o X 15o/20o FOV

320 X 240 VOx Microbolometer

• I2 Circular 40o/46o FOV –

18mm 3rd GEN Image Tube - Direct View

• SWIR 55o X 41o – Miniature

Camera

20 20o

• FOV

FOV

Goal: Increase soldier’s passive situation awareness using SWIR imagery combined with existing image intensifier goggle and uncooled IR

• IR for long range targeting
• I2 for high resolution exterior mobility and low

light levels

• Passive SWIR for WFOV situational awareness

Inexpensive, large format SWIR camera needed for wide FOV application

‘Urban’ ENVG (UENVG)

SWIR Camera

ENVG

40o FOV

Future Application #3: SWIR Overlay on the Head

• Persistent Surveillance: 24-hour

airborne imagery of entire urban battlespace

• Potential Capabilities:

– Very Large Areas: > 25 km2–64 km2 – Moderate Resolution:

• 1 – 0.5m for vehicle tracking
• 0.5 – 0.25m for dismounts

– Moderate Update Rate: 1 – 2 Hz for low-speed urban targets – Altitude: Aircraft must remain above 15,000 ft for survivability

• Potential Solutions:

– Very large staring FPAs – Large scanning or step/stare FPAs

MWIR Night image 0.5m GSD

Operational Concept

New application with many possible detectors (HgCdTe, InSb, QW/SLS) Image size is the dominant characteristic (moderate

resolution/f# demands)

Future Application #4: IR Sensors for Airborne Persistent Surveillance Future Application #5: Distributed Aperture Sensor (DAS)

Current 73º-11 cameras 115 º - 6 cameras 105 º / 70 º - 7 cameras

Goal: Provide vehicle occupants a simultaneous, real time, 360o x 90o dome of

situational awareness coverage of the immediate surroundings, enabling visibility, day or night while on the move.

IR FPA 640 x 480 1280 x 1024 1280 x 1024 Slices / IR cameras 7 + 4

• verhead

4 + 2 OH 5 + 2 OH

• Horz. FOV

per camera 73º 115º 70º (Front) 105º (Side/Rear) Azimuth Resolution 2 mrad 1.6 mrad 0.95 mrad (Front) 1.4 mrad (Side/Rear)

Sensor Design Trade Study

Challenges:

• Affordability for ground vehicle
• High optical flow rates
• Resolution

Inexpensive, large format uncooled IR camera with short time constants critical for full performance in moving ground vehicle

Potential Solutions:

• Fast uncooled IR (low tc)
• Near IR/SWIR TV
• Image Fusion (FLIR/I2)

Potential advantages of Si or GaAs over CdZnTe

• Availability
• Lower cost/cm2
• Scaleable technology
• To 1st order, processing cost is not dependent on

size

• Better thermal match to readout circuit
• Mature processing capability
• Improved durability and toughness

Biggest Disadvantage

• Higher dark current than CdZnTe

6 x 6 cm CdZnTe 6” (15 cm) diameter silicon

The Army needs low cost, high performance LW/MW/SW IR FPAs irrespective of material system (II-VI or III-V) or detector type (p/n or SLS)

6 inch Si: 120 640 x 480 or 37 1280 x 720 with 20 micron pitch

Number of die/wafer drives end cost

Low Cost Substrate Alternatives to CdZnTe

In situ doped p-type cap layer n-type HgCdTe base layer Buffer layer lattice-matched to HgCdTe ZnTe layer - controls CdZnTe orientation Si substrate

• automatic thermal expansion

match to Si ROIC

Si Readout Circuit HgCdTe CdTe Buffer Layer n-type Si Substrate Vs CdZnTe

In situ doped p-type cap layer n-type HgCdTe base layer Buffer layer lattice-matched to HgCdTe ZnTe layer - controls CdZnTe orientation match to Si ROIC

Si Readout Circuit HgCdTe CdTe Buffer Layer n-type Si Substrate Vs CdZnTe Si Readout Circuit HgCdTe CdTe Buffer Layer n-type Si Substrate Vs CdZnTe

Need to mitigate lattice and thermal mismatch which cause dislocations in MCT

HgCdTe/Si Hybrid FPAs Technical Approach

Comparison: MCT vs. MCT LWIR (Si vs. CdZnTe)

LWIR Parameter MCT/Si MCT/CZT QE (%)

80-90 85-95

RoA (ohm-cm2)

100-200 200-500

Waveband (µm)

10.5 10.5

Operability (%)

95-99 >99

Dark Current (A/cm2)

3.4-6.7e-5 1.3-3.4e-5

Operating Temp (K)

80 80

Substrate & Max Size

6-in 7x7 cm2 CZT

Approx cost pixel (20002)

0.3X for 20002 (Enables > 30002) 1x

MCT/Si enables very large FPAs, allows lower cost and increased reliability -- Need to improve MCT/Si operability

Comparison: MCT and InSb MWIR

MWIR Parameter MCT/Si 80K MCT/Si 110K InSb

QE (%) 80-90 80-90 99 (w/o CO2 notch) ~97 w / notch RoA (ohm-cm2) 1e7 >1e5 1e6 Waveband (µm) 5.3 5.3 5.3 Operability (%) >99-99.6 >99 640x480 99.9 2k2 99.7 Dark Current (A/cm2) <7e-9 <9.5e-8 7e-9 Operating Temp (K) 80 110 80 Substrate & Max Size 6-in Si 6-in Si 4” in production 6” available from foundry, estimate 2 yrs to production Approx cost pixel (20002) 0.6x 0.6x 1x

MCT/Si allows multiple bands, higher operating temperatures and lower cost -- InSb currently in production with large FPAs

• Remove scanner
• Remove cooler
• Reduce size/weight
• 4.5 – 2.7 lbs

BST (TFFE) 320 X 240, 2 mil MCT Monolithic VOx 320X240, 2 mil Monolithic VOx 320X240, 1 mil Monolithic VOx 640X480, 1 mil

• Improve

performance

• Remove chopper
• Smaller
• ptics
• 1.9 lbs
• 18° FOV

Use Uncooled IR (LWIR) to Maintain Range Performance with Lower Weight Improved Performance with Continued Technology Developments

1998 2003 2004 2005

• 5.5 – 4.5 lbs
• TE Cooled MWIR
• Longer range enables

M/HTWS

• 2.5/3.9 lbs
• 18°/9° FOVs

Progression in TWS Technology Uncooled Applications: Current and Future

Current Applications:

• Rifle Sights
• Helmet Mounted Imagers
• Driving Sensors
• UGS

25 um Future Applications:

• Far Target Locator
• Distributed Aperture
• Aviation Pilotage
• Missile Seekers
• Threat Warning

VOx – 640x480

Application Higher Resolution Shorter Time Constant Higher Frame Rate Multi-band / wavelength

Far Target Locator Distributed Aperture Aviation Pilotage Missile Seekers Threat Warning

Improvements Required For Future Applications:

Keys to future application expansion are FPAs with smaller pixels and shorter time constants without reducing S/N performance

Improvements in Uncooled FPAs

Segment of 640x480 shows pixilation Segment of 1024x768 shows Improved resolution

Benefit of 17 micron pixel pitch

1024x768, 17 micron pixel provides about the same size active area for drop-in replacement of 640x480 (25 micron pixels) without major changes and 40% increase in range

Camera #1 has a thermal time constant of 11.4ms Camera #2 has a thermal time constant of 3.8ms

Benefit of shorter time constant

3.8 msec time constant provided more contrast with less blur than the 11.4 msec time constant

New materials and processes create short time constant and smaller pixels

.... 2 1

/ 1 2 2 2 2

+ + ∆ + ∆ = ∆

f sub bias sub TF sNeTD

B kT C T T b T T α χ α γ

Thermal Johnson 1/f

Reduced C decreases SNR Ο(C-1/2) Reduced G increases SNR Ο(G)

Uncooled Detector Parameters

(slide show version)

[ ]

      ∆       +       + + − = ) ( 1 4 1 1 1 1 ) ( ) (

2

ω η ωτ ω α ω

s s s D r e B

T dT dW A F i g G H I Y ( )

G A Y

d TCR •

≈ α

r

g G C + = τ

Time Constant: Signal: Signal/Noise Reduce τ: Increase G but decrease SNR linearly Reduce C but decrease SNR Ο(C1/2) Decrease pixel size: Reduces signal Reduces C with reduced Ad Ο(Ad-1)

Smaller pixels require decreased G and reduction in other noise factors to maintain SNR. Increased TCR can significantly improve SNR, provided other noise factors stay constrained

(...) ) / 1 ( 1 / ) / (

2 2

• +

⋅ ⋅ ≈ ∆

TCR sub d NETD

C kT A G T α

.... 2 1

/ 1 2 2 2 2

+ + ∆ + ∆ = ∆

f sub bias sub TF sNeTD

B kT C T T b T T α χ α γ

Thermal Johnson 1/f

Uncooled Detector Parameters

(print version p.1 of 2)

[ ]

      ∆       +       + + − = ) ( 1 4 1 1 1 1 ) ( ) (

2

ω η ωτ ω α ω

s s s D r e B

T dT dW A F i g G H I Y

r

g G C + = τ

Time Constant: Signal: Signal/Noise

Smaller pixels require decreased G and reduction in other noise factors to maintain SNR. Increased TCR can significantly improve SNR, provided other noise factors stay constrained

Uncooled Detector Parameters

(print version p.2 of 2)

( )

G A Y

d TCR •

≈ α

r

g G C + = τ

Time Constant: Signal: Signal/Noise

Smaller pixels require decreased G and reduction in other noise factors to maintain SNR. Increased TCR can significantly improve SNR, provided other noise factors stay constrained

(...) ) / 1 ( 1 / ) / (

2 2

• +

⋅ ⋅ ≈ ∆

TCR sub d NETD

C kT A G T α

Reduce τ: Increase G but decrease SNR linearly Reduce C but decrease SNR Ο(C1/2) Decrease pixel size: Reduces signal Reduces C with reduced Ad Ο(Ad-1) Reduced C decreases SNR Ο(C-1/2) Reduced G increases SNR Ο(G)

Future Uncooled Considerations

• Present devices order of magnitude or more away from fundamental

limits

• 1/f noise is increasingly the dominant noise term. Becomes more

severe as the detector becomes smaller and thinner. Must reduce prefactor: EBM and SiGe show promise, increase TCR without inc. 1/f prefactor, materials study and development.

• Johnson noise comparable to temperature fluctuation. Reduce with

higher TCR, more bias drive. Complicates ROIC design with bias drive subtraction circuits.

• Drive toward lower NeTD, better performance: reduce heat capacity

and better thermal isolation: Holes in the detector, better lithography, continuous bias may allow electronic control of time constant. 17 µm imagery from BAE Advanced Pixel Structures Higher TCR and Lower Noise Will Enable Future Applications

• MCT still forms backbone of US Army FLIR programs; however,

InSb and uncooled now preferred for certain applications

• Emerging requirements for new FPAs in LWIR, MWIR and SWIR

– Many applications will require very large FPAs (> 1Kx1K) – Cost/producibility will be a major issue—uncooled very attractive

• InSb (MWIR) and InGaAs (SWIR) are the preferred approaches to

large format IR FPAs today

– Substrate size and wavelength/band limitations – No good cooled LWIR alternative today

• Progress in MCT on low cost substrates (LCS) has been slower than

expected but recently encouraging

– Good RoA’s and low average dark current – Operability still a concern—defects or manufacturing issues? – Some good looking arrays in all three wavebands

• Large format uncooled sensors on the horizon with new 17um pixel

structures

– Continued progress depends on new approached to raise TCR and lower noise factors (especially 1/f)

Large format, lower cost arrays critical to many future applications

Summary & Conclusion