Programmable Logic in the Space Radiation Environment

1. Programmable Logic in the Space Radiation Environment Presented by Kenneth A. LaBel Radiation Effects and Analysis Group Leader Electronics Radiation Characterization Project Manager Living with a Star Space Environment Testbed Experiments Manager ken.label@gsfc.nasa.gov

2. Acknowledgements • The entire Radiation Effects and Analysis Group at GSFC as well as Janet Barth (from whom I stole many charts!) • NASA HQ Code AE for supporting the NASA Electronic Parts and Packaging (NEPP) Program including the Program Manager, Chuck Barnes of JPL • Lew Cohn at Defense Threat Reduction Agency (DTRA) • The designers and systems engineers I’ve had the privilege to work with • Martha O’Bryan for graphics support

3. Abstract • Missions for the space environment differ from those of many terrestrial applications since they are presented with a radiation environment and long life requirements. Additionally, maintenance operations are extremely expensive if possible at all. The fundamentals of the radiation environment and radiation test techniques will be reviewed. Detailed specifications and failure modes will be analyzed, for each class of device and technology. Figures of merit will be given for specific devices in use. Design techniques to provide reliable operation in the radiation environment will be discussed as well as the analysis of device reliabilityissues such as single point failures and how to avoid them.

4. Outline • Introduction • Why we are here • Overview of Radiation Hazard • Overview of Radiation Effects • NASA and Mission Requirements • Radiation and Technology • A Radiation Hardness Assurance (RHA) Approach • Includes top-level discussion of mitigation • Radiation Effects on Programmable Technologies with Real-life Examples • Total Ionizing Dose (TID) • Single Event Effects (SEE) • Destructive • Non-Destructive • Design Techniques for Radiation Effects Mitigation • Summary

5. Introduction

6. SOHO/LASCO C3 CoronographJuly 14, 2000 Space Weather induces transients in a Charge-Coupled Device (CCD)

7. >$100B Circa 1998 100 90 80 1 8 Rad-Resist [<108 rad/s, <105 rad] 1 6 Rad-Hard [> 108 rad/s, >105 rad] 70 1 4 1 2 Number of Rad Tolerand Microelectronics Manufacturers 60 1 0 Billions of Dollars 8 6 50 4 2 2 A nalog/ 2 Digi tal 40 0 1 9 8 5 1 9 9 3 1 9 9 5 30 20 10 $1.4B (~1%) $0.4B (<0.25%) World NASA/Military RH/RT 0 Semiconductor Semiconductor Semiconductor Market Market Market The Space Semiconductor Market - Reduced Options for Risk Avoidance

8. Increased Radiation Awareness - Three Prime Technical Drivers • Commercial and emerging technology devices are more susceptible (and in some cases have new radiation effects) than their predecessors. • Limited radiation hardened device availability • There is much greater uncertainty about radiation hardness because of limited control and frequent process changes associated with commercial processes. • With a minimization of spacecraft size and the use of composite structures, • Amount of effective shielding against the radiation environment has been greatly reduced, increasing the internal environment at the device. • THESE THREE DRIVERS IMPLY THAT WE ARE USING MORE RADIATION SENSITIVE DEVICES WITH LESS PROTECTION.

9. The Space Radiation Environment

10. Space Radiation Environment Galactic Cosmic Rays (GCRs) Nikkei Science, Inc. of Japan, by K. Endo Solar Protons & Heavier Ions Trapped Particles Protons, Electrons, Heavy Ions Deep-space missions may also see: neutrons from background or radioisotope thermal generators (RTGs)

11. Components of the Natural Environment • Transient • Galactic Cosmic Rays (GCRs) • Hydrogen & Heavier Ions • Solar Particle Events • Protons & Heavier Ions • Trapped • Electrons, Protons, & Heavier Ions • Atmospheric & Terrestrial Secondaries • Neutrons

12. Source Modulator Protons Galactic Cosmic Rays Atmospheric Neutrons Heavier Ions Trapped Particles Trapped Particles Sun Dominates the Near-Earth Environment A True Dynamic System

13. 300 Cycle 19 Cycle 20 Cycle 21 Cycle 22 250 Cycle 18 200 150 Sunspot Numbers 100 50 0 1947 1997 Years Sunspot Cycle:An Indicator of the Solar Cycle after Lund Observatory Length Varies from 9 - 13 Years 7 Years Solar Maximum, 4 Years Solar Minimum

14. Solar Particle Events • Results in Increased Levels of Protons & Heavier Ions • Energies • Protons - 100s of MeV • Heavier Ions - 100s of GeV • Abundances Dependent on Radial Distance from Sun • Partially Ionized - Greater Ability to Penetrate Magnetosphere Than Galactic Cosmic Rays • Number & Intensity of Events Increases Dramatically During Solar Maximum • Models • Total Ionizing Dose & Displacement Damage Dose - SOLPRO, JPL, Xapsos/NASA • Single Event Effects - CREME96 (Protons & Heavier Ions)

15. Gradual Solar Events • Coronal Mass Ejections (CMEs) • Particles Accelerated by Shock Wave • Largest Proton Events • Decay of X-Ray Emission Occurs Over Several Hours • Large Distribution in Solar Longitude Holloman AFB/SOON

16. Impulsive Solar Events • Solar Flares • Particles Accelerated Directly • Heavy Ion Rich • Sharp Peak in X-Ray Emission • Concentrated Solar Longitude Distribution

17. Solar Proton Event - October 1989 Proton Fluxes - 99% Worst Case Event Counts/cm2/s/ster/MeV nT GOES Space Environment Monitor

18. Free-Space Particles: Galactic Cosmic Rays (GCRs) or Heavy Ions • Definition • A GCR ion is a charged particle (H, He, Fe, etc) • Typically found in free space (galactic cosmic rays or GCRs) • Energies range from MeV to GeVs for particles of concern for SEE • Origin is unknown • Important attribute for impact on electronics is how much energy is deposited by this particle as it passes through a semiconductor material. This is known as Linear Energy Transfer or LET (dE/dX).

19. 4 1 0 Z = 2 - 92 3 1 0 2 1 0 1 1 0 0 1 0 - 1 1 0 - 2 1 0 - 3 1 0 - 4 GEO 1 0 GTO - 5 1 0 MEO - 6 1 0 EOS - 7 LEO 1 0 - 8 1 0 - 1 0 1 2 1 0 1 0 1 0 1 0 GCR Abundance:Integral LET Spectra CREME 96, Solar Minimum, 100 mils (2.54 mm) Al LET Fluence (#/cm2/day) LET (MeV-cm2/mg)

20. 4 3 2 1 1 2 3 4 5 6 7 8 9 10 Trapped Particles in the Earth’s Magnetic Field: Proton & Electron Intensities AP-8 Model AE-8 Model Ep > 10 MeV Ee > 1 MeV #/cm2/sec #/cm2/sec A dip in the earth’s dipole moment causes an asymmetry in the picture above: The South Atlantic Anomaly (SAA) L-Shell NASA/GSFC

21. SAA and Trapped Protons:Effects of the Asymmetry in the Proton Belts on SRAM Upset Rate at Varying Altitudes on CRUX/APEX

22. Solar Cycle Effects:Modulator and Source • Solar Maximum • Trapped Proton Levels Lower, Electrons Higher • GCR Levels Lower • Neutron Levels in the Atmosphere Are Lower • Solar Events More Frequent & Greater Intensity • Magnetic Storms More Frequent --> Can Increase Particle Levels in Belts • Solar Minimum • Trapped Protons Higher, Electrons Lower • GCR Levels Higher • Neutron Levels in the Atmosphere Are Higher • Solar Events Are Rare

23. Magnetic Storm and the Electron Belts Space Weather Effect Courtesy: R. Ecofett/CNES

24. Secondary Particles May Also Effect Sensitive Technologies:Ionizing Particle Impacts to Focal Plane Arrays (FPAs) -Deltas are not spatially correlated primary Surrounding Material deltas FPA (latent emission) induced radioactivity secondaries natural radioactivity + Secondaries and delta electrons are time coincident with primary and have limited range Courtesy of Jim Pickel, SEE Symposium 2002

25. Basic Radiation Effects

26. Radiation Effects and Spacecraft • Critical areas for design in the natural space radiation environment • Long-term effects • Total ionizing dose (TID) • Displacement damage • Transient or single particle effects (Single event effects or SEE) • Soft or hard errors • Mission requirements and philosophies vary to ensure mission performance • What works for a shuttle mission may not apply to a deep-space mission

27. Total Ionizing Dose (TID) • Cumulative long term ionizing damage due to protons & electrons • Effects • Threshold Shifts • Leakage Current • Timing Changes • Functional Failures • Can partially mitigate with shielding • Low energy protons • Electrons

28. Displacement Damage • Cumulative long term non-ionizing damage due to protons, electrons, and neutrons • Effects • Production of defects which results in device degradation • May be similar to TID effects • Optocouplers, solar cells, CCDs, linear bipolar devices • Shielding has some effect - depends on location of device • Can eliminate electron damage • Reduce some proton damage Not particularly applicable to CMOS microelectronics

29. Single Event Effects (SEEs) • An SEE is caused by a single charged particle as it passes through a semiconductor material • Heavy ions • Direct ionization • Protons for sensitive devices • Nuclear reactions for standard devices • Effects on electronics • If the LET of the particle is greater than the amount of energy or critical charge required, an effect may be seen • Soft errors such as upsets (SEUs) or transients (SETs), or • Hard errors such as latchup (SEL), burnout (SEB), or gate rupture (SEGR) • Severity of effect is dependent on • type of effect • system criticality Destructive event in a COTS 120V DC-DC Converter

30. Types of Single Event Effects

31. Radiation Effects: The Root Cause in the Natural Radiation Environments • Total Ionizing Dose • Trapped Protons & Electrons • Solar Protons • Single Event Effects • Protons • Trapped • Solar • Heavier Ions • Galactic Cosmic Rays • Solar Events • Neutrons • Displacement Damage • Protons • Electrons • Spacecraft Charging • Surface • Plasma • Deep Dielectric • High Energy Electrons • Background Interference on Instruments

32. NASA and Radiation Requirements

33. Radiation Device Regimes for the Natural Space Environment • High • > 100 krads (Si) • May have • long mission duration • intense single event environment • intense displacement damage environment • Moderate • 10-100 krads (Si) • May have • medium mission duration • intense single event environment • moderate displacement damage environment • Low • < 10 krads (Si) • May have • short mission duration • moderate single event environment • low displacement damage environment Examples: Europa, GTO, MEO Type of device: Rad hard (RH) Examples: EOS, highLEO, L1, L2, ISSA Type of device needed: Rad tolerant (RT) Examples: HST, Shuttle, XTE Type of device needed: SOTA commercial with SEE mitigation Aeronautics must deal with neutron SEE environment

34. Mix of NASA Missions and Radiation Requirements~225 missions are currently in some stage of development • Informal study has been performed of percent of missions in each category

35. Implications of NASA Mission Mix • SEE tolerant is the major current need • “Radiation Tolerant” covers a large percentage of NASA needs • “Commercial” (non-hardened) devices or even boards and systems may be acceptable for some NASA missions (with the risks associated with commercial devices) • Even the low radiation requirement offers challenges for commercial devices • Example: Hubble Space Telescope has noted numerous anomalies on commercial microelectronics • Projects with rad hard needs struggle to meet requirements • Limited device availability or implications of adding mitigation • Two Further Notes: • Aero-Space (avionics/terrestrial) has issues with soft errors (typically induced by secondary neutrons) • NASA designs use all types of microelectronics from true rad-hard to Radio Shack COTS (Ex., shuttle experiment)

36. Ziatech ZT-6500 3U Compact PCI Pentium Board. International Space Station:Electronics Drivers • Radiation hazards (low earth, 57 deg inclination) • Primarily trapped protons, some GCR and solar particles • Radiation requirements • High amounts of effective shielding • Proton upset is prime driver; GCR is secondary • Non-radiation drivers • Large amounts of hardware • Serviceable • Philosophy • Use off COTS and COTS boards • Use proton ground tests to qualify hardware (controversial)

37. Space Shuttle: Electronics Drivers • Radiation hazards (Mostly ISS orbits) • Trapped particles, some GCR and solar particles • Radiation requirements • Shuttle upgrades require radiation tolerant • Experiments have none other than fail-safe • Non-radiation drivers • Serviceable • Short duration • Performance not a driver • Philosophy • “Radio Shack” for experiments

38. Europa: Electronics Drivers • Radiation hazards (Jovian Deep Space) • Trapped particles (electrons!), GCR, solar particles • Radiation requirements • High • Non-radiation drivers • 7 year storage of many instruments and systems • Temperature range • Philosophy • Radiation-hard where they can • Custom Radiation-hardened ASICs • Mitigation/shielding where needed

39. Radiation and Technology

40. Technology Triumvirate for Insertion Into Spaceflight Technology Development Ground Test, Protocols, and Models Reliable Technology for Space Systems On-orbit Experiments and Model Validation

41. NASA Needs for Microelectronics Technology • In general, NASA is tasked to • reduce time-to-launch (faster) • increase system performance (better), and • reduce spacecraft and instrumentsize and power as well as ground-based manpower (cheaper). • This implies that NASA microelectronics require • increased technical performance (bandwidth, power consumption, volume, etc.), and • increased programmatic performance (availability, cost, reliability). • Radiation tolerance is the “red-headed stepchild” of this process. • Current programs often “waive” or reduce reliability/radiation tolerance issues or design workarounds • “True” cost of commercial versus radiation hardened is often misunderstood

42. Procurement Screening Radiation Testing Availability Development Tools Sample Cost Factors for Selecting Commercial Versus Rad Hard Device • Prototypes • Manpower • Shielding • Circuit Mitigation • Development Path • Technical (re: need for Mflops) may be • the driver over cost • - Other factor to consider: risk

43. Radiation Issues for Newer Technologies • Proton induced single event upsets • Proton induced single event latchup • Neutron & Alpha induced upsets • Single events in Dynamic RAMs • Displacement damage in electronics • Single event functional interrupt • Stuck bits • Block errors in Dynamic RAMs • Single event transients • Neutron induced single event effects • Hard failures & latchup conditions • Multiple upsets from a single particle • Feature size versus particle track • Microdose • Enhanced low dose rate sensitivity (ELDRS) • Reduced shielding • Test methods for advanced packaged devices • Ultra-high speed & novel devices (e.g., photonics, InP, SiGe) • Design margins & mitigation • COTS variability • At-speed testing • Application-specific sensitivities In general, however, TID tolerance of deep submicron CMOS is improving

44. CMOS Microelectronics • Advantages: • Reduced power consumption with VCC <1V • Allows for enabling volume shrinkage for • space application • Sample Devices • COTS SDRAM, PowerPCs, Linears • Trend • Shrinking feature sizes • Reduced power supply voltages • SEE Knowledge: • SEL, SEU, SET sensitivities variable • Epi can help reduce SEE sensitivity in some, • but not all cases • Shrinking feature size devices • Lower critical charge required • Smaller target for ions • Noise reduction techniques to obtain • high-speed performance helps • reduce charge propagation • Hardened ultra-low power efforts at U of Idaho • (CULPRiT) • Comment: • Reliability issues of COTS (non-hardened) Data is flat for Intel PIII devices with three differing feature sizes and operating speeds

45. System Level Approach to Radiation Hardness Assurance (RHA)

46. Sensible Programmatics for Radiation Hardness Assurance (RHA): A Two-Pronged Approach • Assign a lead radiation engineer to each spaceflight project • Treat radiation like other engineering disciplines • Parts, thermal,... • Provides a single point of contact for all radiation issues • Environment, parts evaluation, testing,… • Each program follows a systematic approach to RHA • RHA active early in program reduces cost in the long run • Issues discovered late in programs can be expensive and stressful • What is the cost of reworking a flight board if a device has RHA issues?

47. Radiation and Systems Engineering: A Rational Approach for Space Systems • Define the Environment • External to the spacecraft • Evaluate the Environment • Internal to the spacecraft • Define the Requirements • Define criticality factors • Evaluate Design/Components • Existing data/Testing/Performance characteristics • “Engineer” with Designers • Parts replacement/Mitigation schemes • Iterate Process • Review parts list based on updated knowledge

48. Define the Hazard • The radiation environment external to the spacecraft • Trapped particles • Protons • Electrons • Galactic cosmic rays (heavy ions) • Solar particles (protons and heavy ions) • Based on • Time of launch and mission duration • Orbital parameters, … • Provides • Nominal and worst-case trapped particle fluxes • Peak “operate-through” fluxes (solar or trapped) • Dose-depth curve of total ionizing dose (TID) Note: We are currently using static models for a dynamic environment

49. Evaluate the Hazard • Utilize mission-specific geometry to determine particle fluxes and TID at locations inside the spacecraft • 3-D ray trace (geometric sectoring) • Typically multiple steps • Basic geometry (empty boxes,…) or single electronics box • Detailed geometry • Include printed circuit boards (PCBs), cables, integrated circuits (ICs), thermal louvers, etc… • Usually an iterative process • Initial spacecraft design • As spacecraft design changes • Mitigation by changing box location

50. Define Requirements • Environment usually based on hazard definition with “nominal shielding” or basic geometry • Using actual spacecraft geometry sometimes provides a “less harsh” radiation requirement • Performance requirements for “nominal shielding” such as 70 mils of Al or actual spacecraft configuration • TID • DDD (protons, neutrons) • SEE • Specification is more complex • Often requires SEE criticality analysis (SEECA) method be invoked • Must include radiation design margin (RDM) • At least a factor of 2 • Often required to be higher due to device issues and environment uncertainties