Single Event Transients in Linear Integrated Circuits - NASA/GSFC ...

33 downloads 0 Views 6MB Size Report
3. Computer Simulation. - Device simulation. - Circuit simulation. 4. SET testing. –. Broad Beam of Protons and Heavy ions. –. Pulsed laser light. 5. Case Studies.
Single Event Transients in Linear Integrated Circuits Stephen Buchner QSS Group Inc./NASA GSFC

Dale McMorrow Naval Research Laboratory

L. Cohn A. Clark

K. LaBel S.Buchner, C.Poivey, J.Howard, R.Ladbury

A.Sternberg, Y.Boulghassoul, T.Holman L.Massengill J.Rowe

D.McMorrow, W.Lotshaw

R.Pease

M.Savage, T.Turflinger, J.Titus D. Platteter

1

Course Outline 1. Introduction 2. Fundamental Mechanisms – – –

Charge generation Charge collection Device Response

3. Computer Simulation -

Device simulation Circuit simulation

4. SET testing – –

Broad Beam of Protons and Heavy ions Pulsed laser light

5. Case Studies 6. Mitigation 7. Conclusions

1. Introduction

2

Introduction • Definition of a SET An SET is a temporary disruption to the output of a device (transistor) or circuit caused by an ionizing particle passing through the device.

• SETs occur in Digital and Analog circuits. DSETs and ASETs

• SETs occur in bipolar, CMOS, BiCMOS and compound semiconductor (III-V) technologies.

Introduction • This Short Course deals with ASETs in Linear Integrated Circuits. – – – – – – – –

Op-amps Comparators Voltage references Pulse width modulators Voltage controlled oscillators DC/DC Converters Phase Lock Loops Digital-to-Analog Converters

3

Introduction • SETs appear as voltage “glitches” at the outputs of: R1 – – – –

Op-amps, Comparators, Voltage refs, DACs

R2

Vin

Vout

+

10

Amplitude (V)

9 8 7 6 5 4 0

2

4

6

8

10

12

14

Time (µ µs)

Introduction • SETs, or voltage glitches, are just another form of noise. Effect on system determined by amplitude and width (∆Vmin,∆tmin). ∆Vout R1

R2

∆t

Vin

Follow-on Circuit

+ CLK

4

Introduction 6.5

Q2 Q4 Q5

6.0 5.5

ASETs in Op-amps (LM124)

5.0 6.0

• ASETs assume a wide variety of shapes. In the LM124 there are at least five different types. • Each ASET type also has a variety of amplitudes.

4.0

2.0

Q9 Q 19

Vout (V)

0.0 7.0

Q 18

6.5 6.0 5.5 5.0 10

Q 20

9

• Not all ASETs pose a threat. • Transient widths often related to circuit bandwidth.

8 7 6 5 8 7 6 5 4 3 2 1

Q 16

Buchner et al IEEE TNS 2004 0

5

10

15

T im e (m icrose co n ds)

Introduction ASETs in Voltage Compators (LM119) 5

• ASETs are all either positivegoing or negative-going depending on whether output is high or low.

4

Q6

3 2 1 0 5 4

Q3

3

• The shape stays the same as ∆V and ∆t change.

2 1 0 5 4

Q2

3 2 1 0 0

0.2

0.4

0.6

0.8

1

Buchner et al HEART 2001

µs) Time (µ

5

Introduction SETs appear as disruptions to the outputs of: Voltage Controlled Oscillators

Pulse Width Modulators

Chen et al IEEE TNS 2003

Howard et al Data Workshop 2003

Introduction Satellites Exhibiting ASETs (Partial List) • • • • • •

Topex/Poseidon (OP-15)………….. 1993 SOHO (LM/PM139)………………... 1996 Cassini (LM139)……………………. 1998-2002 TDRS (LM139)………………………1997 MPTB (LM124 and LM139) ………. 1997 MAP (LM139)………………………..2001

6

Introduction ASETs on MAP • Launched June 30, 2001. • A reset of the spacecraft processor occurred on November 5, 2001. • Caused the spacecraft to go into “safehold.” • Solar event on Nov. 5, 2001.

Poivey RADECS 2002

Introduction Processor Reset Circuit contains Comparators +5V (Min=4.3V)

+5V R=1k

R=3.6k 2.9V

EEP_RESET (∆ ∆=2.5V) LM139

R=5.1k

Vref=2.5V HW_RESET (∆ ∆=2.5V)

+5V R=100k

2.9V LM139 Vref=2.5V

5V

R=1.5k C=10 µF

R=10k

LM139 Vref=2.5V Poivey RADECS 2002

7

Introduction ASETs for the LM139 Comparator

Output Voltage (V)

Vt = 2.5 V

4

2

0

-2

-4 1.2 10

2

LM 139, V =+/-5, δ V =800m V, LET=18.7 M eV-cm /m g cc

-5

1.4 10

-5

1.6 10

-5

1.8 10

-5

i

2 10

-5

2.2 10

-5

2.4 10

-5

2.6 10

-5

2.8 10

-5

Tim e (s) Poivey RADECS 2002

Introduction Summary: • ASETs occur in a wide variety of linear devices. • The nature of the ASET depends on the device. • ASETs have caused problems in a number of spacecraft.

8

2. Fundamental Mechanisms

Fundamental Mechanisms Interaction of ions with matter Ionization 1. Coulomb interaction between nucleus and bound electrons. 2. It requires, on average, 3.6 eV to create 1 electron-hole pair in Si. 3. Energetic electrons (delta rays) create additional e-h pairs as they move through lattice. 4. Result is a track of charge with a diameter < 1 µm.

SILICON

9

Fundamental Mechanisms Interaction of ions with matter Ionization 1. Coulomb interaction between nucleus and bound electrons. 2. It requires, on average, 3.6 eV to create 1 electron-hole pair in Si. 3. Energetic electrons (delta rays) create additional e-h pairs as they move through lattice. 4. Result is a track of charge with a diameter < 1 µm. 5. Charge Track expands rapidly via diffusion.

SILICON

Fundamental Mechanisms Interaction of ions with matter Ionization 1. In the presence of a p-n junction charge is collected via drift in the junction electric field.

N-type

P-type

SILICON

10

Fundamental Mechanisms Unbiased p-n Junction. n-Type

E

p-Type

Depletion Layer

EC

EF

EV

Fundamental Mechanisms Forward-biased p-n junction. n-Type

E

-

p-Type

+ Depletion Layer

EC EV

11

Fundamental Mechanisms Reverse-biased p-n junction. n-Type

+

E

p-Type

-

Depletion Layer

EC EV

Fundamental Mechanisms Charge Collection via Drift Large electric field collects charge via drift, an efficient process. E +

-

n-Type

+

+

p-Type

-

-

n-Type EC

+

EV +

p-Type

12

Fundamental Mechanisms Charge Collection by Diffusion Electrons and holes generated far from the junction will diffuse slowly to the junction. +

100 µm

+

EC EV

-

E

-

n-Type p-Type

+

Fundamental Mechanisms ASETs Shapes

Electric Potential (V)

Fast (drift) and slow (diffusion) components. (Shape depends on applied bias).

Drift Component

Diffusion Component

Time (ps)

13

Fundamental Mechanisms Bipolar Transistor Two back-to-back p-n junctions. Emitter

Base

n

p

Collector n

EC

EF

EV

Emitter

Base

Collector

Fundamental Mechanisms Forward-active mode Holes injected into C/B junction drift into base and turn the transistor more on. Emitter

Base

n

p

Collector n

V2

V1 -

-

+ +

Emitter

Base

Collector

14

Fundamental Mechanisms Charge Collection in Vertical n-p-n Transistor. E

C n+

B p

n+

p+

p+

n-epi n+

p-substrate

Fundamental Mechanisms Differential pair: Used as inputs for op-amps and voltage comparators. V+

V2 V1

+

I2

I1 V1

If: V1 = V2 then I1 = I2 and Vout(1)=Vout(2)

R

R Vout(1)

Tr1

Tr2

Vout(2) V2

IE

15

Fundamental Mechanisms Differential pair: If: V1 > V2 then I1 > I2 and Vout(2) is high Vout(1) is low

V+

R

R Vout(1)

I2

I1 V1

Tr1

Vout(2) V2

Tr2

IE

Fundamental Mechanisms Differential pair: Assume V1 > V2 V+

R

R I2

I1 V1

As V1-V2 increases, more charge is needed to produce an ASET.

Tr1

Tr2

V2

IE

16

Fundamental Mechanisms LM111

• Input transistors are most sensitive to SETs. VIN

VOUT

• SETs propagate from input transistor to output.

Fundamental Mechanisms Summary • Particles passing through matter produce electronhole pairs. • Electric fields associated with p-n junctions separate the electrons and holes. • This causes a change in the potential at the node. • In a BJT the charge can make a transistor more conducting and that alters the internal potentials. • The potential disturbance travels through the circuit to the output where it appears as an ASET.

17

3. Computer Simulation

Computer Simulation • Device simulator – E.g. SILVACO, PISCES, etc (Only two published reports)

• Circuit simulator – E.g. SPICE.

18

Computer Simulation 1. Device Simulation Simulation of the device behavior by the solution of the following equations describing charge flow in the device. • Maxwell’s Equations • Poisson’s Equations • Continuity Equations • Carrier Transport Equations Non-linear differential equations cannot be solved analytically. A computer is used to solve the equations numerically in space and time.

Computer Simulation 1. Device Simulation B

E p

n+ n-epi

p+

400 µm

p+

p-substrate

Substrate p-n-p

VBE = -0.42 V – weakly on. Johnston IEEE TNS 2000

19

Computer Simulation 1. Device Simulation

B

E p

n+ n-epi

p+

400 µm

p+

Johnston IEEE TNS 2000

p-substrate

Substrate p-n-p

Emitter Base

Collector

Computer Simulation 1. Device Simulation Issues: • Device simulation does not include the circuit response. A circuit simulator (SPICE) would have to be included to model the propagation of the ASET through the circuit (Mixed Mode). • It is also impractical for doing an entire circuit that might have 50 transistors or more.

20

Computer Simulation 2. Circuit Simulation (SPICE): Require: • Circuit interconnects • Identities of devices (transistors, resistors, etc) • Gummel-Poon values for transistors

Validate SPICE Model for Normal Operation: • Calculate the circuit response (output) to various input signals to validate the model

Validate SPICE Model for ASETs: • Comparison of calculated ASETs with experimental ASETs using a feedback process until the transient shapes match.

Computer Simulation Identify Circuit Elements and Interconnects Q19

Q16

Q15

LM124 Q21

Q20 R1

Q5

Q17 Q2

Q4

Q3

C

Q9

21

Computer Simulation Identify Circuit Elements and Interconnects V+ Q16

Q15 Q14

Q19 Vin(+) C1

Q12

Q18 Q20

LM124

Vin(-)

Q13

R2 Q7

R1

Vout

Q6 Q11

Q17 Q21

Q9 Q5

Q2 Q3

Q8

Q10

Q4

V-

Computer Simulation Need SPICE Transistor Parameters LM124 • Obtain SPICE transistor parameters from data sheets or from manufacturer. • Otherwise, must isolate transistor with focused ion beam (FIB) milling and measure I-V curves to obtain Gummel-Poon parameters.

22

Computer Simulation SPICE Validation – Small and Large Input Signals

Input: 100 mV 10 µs +

-

Output

LM124 Sternberg IEEE TNS 2002

Computer Simulation Spice Model (LM124) Validation – Shape of I(t) does not affect ASET V+

I(t)

Q16

Q15 Q14

Q19 Vin(+)

C/B

C1

Q12

Q18 Q20 Vin(-)

R1

Vout

Q6 Q11

Q17

I(t)

Q21

E/B

Q13

R2 Q7

Q9 Q5

Q2 Q3

Q8

Q10

Q4

I(t) E/C

V-

23

Computer Simulation SPICE Simulation – ASET Validation via Iteration Compare SETs generated by pulsed laser or ion microprobe with those calculated using SPICE. Agreement? Poor

Good

Modify circuit and device parameters.

Stop.

Computer Simulation SPICE Simulation – ASET Validation

24

Computer Simulation ASET Dependence on Compensating Capacitance

Sternberg IEEE TNS 2002

Computer Simulation Constant Gain but Different Values of R1/R2 R1 Vin

R2

+

Gain =

R1 = 10 R2

Vin = -60 mV Vout = 600 mV Sternberg IEEE TNS 2002

25

Computer Simulation Summary • Device simulation provides detailed information about mechanisms responsible for ASETs. • Circuit simulation is suitable for studying ASETs with response times longer than the charge collection time. • The validity of the circuit model is established by comparing actual and calculated responses to input signals, but that is not sufficient……….. • The model must also be validated by comparing calculated ASET shapes with those obtained experimentally. • Circuit simulation can be used to study effects of circuit parameters on ASET shapes.

4. ASET Testing

26

ASET Testing Testing Approaches: 1. Broad beam of heavy ions or protons 2. Pulsed Laser 3. Ion micro-probe 4. Radioactive source (252Cf)

ASET Testing

1. Broad Beam of Heavy Ions

27

ASET Testing (Broad Beam) Standard Approach for SEU Testing

2. Irradiate Device with Ions having selected LET

1. Program Device

3. Read Number of Upsets

# of Upsets

σ = Fluence

4. Calculate Cross Section

σ(LET) 5. Irradiate Device with Ions having different LET

ASET Testing (Broad Beam) Standard Approach for SEU Testing

Cross-Section (cm2)

The goal is to predict the error rate in space

• Fit the data with a Weibull curve. • Use the curve as input in a program such as CREME96 to calculate SEU rate.

LET (MeV.cm2/mg)

28

ASET Testing (Broad Beam) Modifications for ASET Testing

DUT

Computer

Oscilloscope

Probe

1. Carefully establish device configuration (voltages, trigger levels etc). 2. Establish a method for counting the ASETs 3. Use active probe or test in actual application 4. Capture and store all transients for later analysis of ∆V vs ∆t.

ASET Testing (Broad Beam) Store all Transients 6 9

5 4

8 7 6

3 2

5 4 3

1 0

2 1

-5 0

1.6 1.4 1.2

8 1.4 7 1.2

1

6

0.8

5

1 0.6 0.4 0.2 0 -5 -0.2

0.8 4 0.6

Width

3 0.4 2

0 1

0.2 0

5

0

-5 0

10 5

10

0 -5 -0.2

15

5 0

-5

Amplitude 15 20 10 5

0

5

25 20

30 25

15 10

20

10

15

30 25

15

30 20

20

25 25

30

Adell et al. IEEE TNS 2000.

29

ASET Testing (Broad Beam) Analysis of Test Data – Pulse Height vs Pulse Width LM124 - LET = 2.8 MeV·cm2/mg 15

Output Signal, V

Pulse Amplitude (V)

10

5

5 0 -5 -10

-5

-15 0

10

20

30

Time, µs

-15

0

5

10

15

Pulse Width (µs)

ASET Testing (Broad Beam) Analysis of Test Data – Pulse Height vs Pulse Width

SET Amplitude, V

10 5

LM124

0

LET = 53.9 MeV·cm2/mg; Voltage Follower.

-5 -10 -15

0

5

10

15

SET Pulse Width, µs

30

ASET Testing (Broad Beam) Effect of Application on σ(LET) 2

Cross-Section (cm )

1.E-02

σsat

1.E-03

1.E-04

Rate ~ 1.E-05

σsat (LETth)2

1.E-06 0

20

40

60

80 2

LET (MeV.cm /mg)

100

120 Poivey GSFC Report

ASET Testing (Broad Beam) Factors that affect ASET Cross-Section 1. 2. 3. 4. 5.

Oscilloscope trigger level Input voltage Supply voltage Gain Output loading

31

ASET Testing (Broad Beam) Factors that affect ASET Cross-Section

Oscilloscope trigger level ∆VT = 50 mV o ∆VT = 300 mV ∆ ∆VT = 700 mV

LM111

Koga et al. IEEE TNS 1993

ASET Testing (Broad Beam) Factors affecting ASET Cross-Section: (V2-V1)

V2 LM139 V1

Koga et al. IEEE TNS 1997

32

ASET Testing (Broad Beam)

Cross Section (cm2/device)

Factors affecting ASET Cross-Section: Vs V+

LM111

V-

Vs = +/-15V, ∆Vin=50 mV

1.0E-03 Vs = +/-5V, ∆Vin=50 mV

1.0E-05

1.0E-07 0

10

20

30

40

50

60

70

LET (MeV.cm2/mg) Koga et al. IEEE TNS 1997

ASET Testing (Broad Beam) Factors affecting ASET Cross-Section: Output Loading 6

5.0 V V+

Amplitude (V)

5

R

4

R=0.17 kΩ Ω

3 2 1 0 -1

LM119

6

0

2

4

6

8

10

µs) Time (µ

V-

Amplitude (V)

5 4 3

R=1.7 kΩ Ω

2 1 0 0

2

4

6

8

10

µs) Time (µ

33

ASET Testing (Broad Beam) Effects of Ion Range

TAMU: • 15 MeV/amu has range ~ 150 µms • 25 MeV/amu has range ~ 400 µms • 40 MeV/amu has range ~ 1,000 µms BNL: • 2 - 8 MeV/amu has range ~ 50 µms

Poivey GSFC Report

ASET Testing

2. Pulsed Laser

34

ASET Testing (Heavy Ions) Pulsed Laser ASET Testing Technique •

Use light instead of particles to generate free carriers (electrons and holes). Particles

Absorption

Coulomb interaction between nucleus of incident particle and bound electrons of Si

Absorption of photons by bound electrons of Si

ASET Testing (Pulsed Laser) Comparison of Ion and Laser-light Induced Charge Tracks λ=800 nm

0

30 MeV Ar

2 µ

Depth in Material, µm

2



λ=590 nm

0

1/e Contour

4

6

w(z)

1/e Contour

4

6

w(z)

8

8

10

10 -4

-2

0 Distance, µm

2

4

-4

-2

0

2

4

Distance, µm

35

ASET Testing (Heavy Ions) Pulsed Laser ASET Testing Technique • • •

Can focus light to a diameter of ~ 1 µm to obtain spatial information – origins of ASETs. Light source must be a pulsed laser with pulse width shorter than the response time of the circuit ~ 1 ps. Particularly well suited for studying ASETs in linear devices because sensitive areas are large compared to size of beam and relatively little metal on surfaces to block beam.

ASET Testing (Pulsed Laser) Two Approaches: 10

5

10

4

10

3

10

2

10

1

10

0

One Photon:

Absorption Coefficient, cm

-1

600 nm

10

600 nm, 800nm, 1.06 µm

800 nm

Above band gap Single-photon absorption

1.06 µm

Two Photon: λ > 1.15 µm 1.26 µm

-1

400

600

800 1000 Wavelength, nm

1200

Sub-bandgap Two-photon absorption

1400

36

ASET Testing (Pulsed Laser) Two Approaches:

1-Photon

2-Photon -

Ec

Ec

600 nm

1260 nm

Eg=1.1eV

Ev

Ev

- - +-

-

-

+-

ASET Testing (Pulsed Laser) One-Photon Absorption SEE Experiment Carrier generation equation:

dN ( r , z ) αI ( r , z ) = dt ω 0

2



1 /e C o n to u r

w (z )

4



Carrier generation is proportional to the intensity of the incident laser pulse



Because the loss is linear in the incident pulse intensity, the pulse experiences exponential attenuation from the surface of the material:

6

I ( r , z ) = I o e −αz

8

10 -4

-2

0

2

4

P o s it io n , µ m

37

ASET Testing (Pulsed Laser) Two-Photon Absorption SEE Experiment Carrier generation equation:



0

dN ( r , z ) αI ( r , z ) β 2 I 2 ( r , z ) = + ω 2 ω dt

w (z )

20

10

2

1 /e C o n to u r 30

Depth in Material, µm

N α I

-4

-2

0 P o s itio n , µ m

2

Carriers are generated by nonlinear absorption at high pulse irradiances by the simultaneous absorption of two photons • Carriers are highly concentrated in the high irradiance region near the focus of the beam • Because of the lack of exponential attenuation, carriers can be injected at any depth in the semiconductor material

4

ASET Testing (Pulsed Laser) Carrier Density Distribution: 1-Photon vs. 2-Photon Absorption 0

0

800 nm

1/e Contour

4

6

w(z)

2

 µ

w(z)

1/e Contour

30

Depth in Material, µm

2

Depth in Material, µm 20 10

NαI

8

10 -4

-2

0 Distance, µm

2

4

-4

-2

0 2 Position, µm

4

38

ASET Testing (Pulsed Laser) Pulsed Laser SEE Experimental Apparatus ps or fs laser source laser pulse

PD1

λ/2 Polarizer

DUT

xyz    

 

ccd PD2 camera

ASET Testing (Pulsed Laser)

Validate Technique

39

ASET Testing (Pulsed Laser) Devices Sensitive to SETs (LM124) V+ Q16

Q15 Q14

Q19 Vin(+)

Q13

R2 Q7

C1

Q12

Q18 Q20

R1

Vout

Q6

Vin(-)

Q11

Q17 Q21

Q9 Q5

Q2 Q3

Q8

Q10

Q4

V-

ASET Testing (Pulsed Laser) SETs: Comparison of Pulsed Laser Light and Low LET Heavy Ions 1.5

Q18

1.2 1.4

Vout (V)

1.1

LM124 Inverting Configuration:

1.3

1.0

1.2

Q2

Vdd = +/-6 V Vin = 60 mV

1.2

1.2

1.1

1.1

1.0

1.0 40 MeV Cl 590 nm Laser

Q6 0

5 10 Time ( µ s)

0.9

15

40 MeV Cl 590 nm Laser

Q20 0

5

10

15

20

Time (µs) Buchner et al. IEEE TNS 2003

40

ASET Testing (Pulsed Laser) SETs: Comparison of Pulsed Laser Light and High LET Heavy Ions 6.5

5.5 Vout (V)

7.0

Q2 Q4 Q5

6.0

6.0

LM124 Voltage Follower:

5.5

5.0

5.0

6.0

10 9 8 7 6 5

4.0 2.0 0.0

Q18

6.5

Q9 Q19 0

5 Time, µs

10

15

Q20

0

5

10

Vdd = +/-15 V Vin = 5 V

15

Time, ms Pease et al. IEEE TNS 2002

ASET Testing (Pulsed Laser)

How do we handle these transients?

41

ASET Testing (Pulsed Laser) Generate Plots of Amplitude vs Width Q19

Q20, C1

Q16

Q15

Q21 R1

Q5

Q17 Q2

Q4

Q3

Q9

C

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q20 (C1) Q20 V∆t Plot 10

5

Output Signal, V

SET Amplitude, V

10

0 -5 -10

0 -5 -10

-15 -20

5

-15

0

5

10 Pulse Width, µs

15

20

0

10

20

30

Time, µs

42

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q20 (C1) Q20 V∆t Plot

10

5

Output Signal, V

SET Amplitude, V

10

0 -5 -10

0 -5 -10

-15 -20

5

-15 0

5

10

15

0

20

10

20

30

Time, µs

Pulse Width, µs

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q20 (C1) Q20 V∆t Plot

10

5

Output Signal, V

SET Amplitude, V

10

0 -5 -10

0 -5 -10

-15 -20

5

-15 0

5

10 Pulse Width, µs

15

20

0

10

20

30

Time, µs

43

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q20 (C1) Q20 V∆t Plot

10

5

Output Signal, V

SET Amplitude, V

10

0 -5 -10

0 -5 -10

-15 -20

5

-15 0

5

10

15

0

20

10

20

30

Time, µs

Pulse Width, µs

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q20 (C1) Q20 V∆t Plot

10

5

Output Signal, V

SET Amplitude, V

10

0 -5 -10

0 -5 -10

-15 -20

5

-15 0

5

10 Pulse Width, µs

15

20

0

10

20

30

Time, µs

44

ASET Testing (Pulsed Laser) Generate Plots of Amplitude vs Width Q19

Q20, C2

Q16

Q15

Q21 R1

Q5

Q17 Q2

Q4

Q3

Q9

C

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q20 (C2) Q20 V∆t Plot

SET Amplitude, V

10 5 0 -5 -10 -15 -20

0

5

10

15

20

Pulse Width, µs

45

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q20 (C2) Q20 V∆t Plot

10

5

Output Signal, V

SET Amplitude, V

10

0 -5 -10

0 -5 -10

-15 -20

5

-15 0

5

10

15

0

20

10

20

30

Time, µs

Pulse Width, µs

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q20 Q20 V∆t Plot

10

5

Output Signal, V

SET Amplitude, V

10

0 -5 -10

0 -5 -10

-15 -20

5

-15 0

5

10 Pulse Width, µs

15

20

0

10

20

30

Time, µs

46

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q20 Q20 V∆t Plot

10

5

Output Signal, V

SET Amplitude, V

10

0 -5 -10

0 -5 -10

-15 -20

5

-15 0

5

10

15

0

20

10

20

30

Time, µs

Pulse Width, µs

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q20 Q20 V∆t Plot

10

5

Output Signal, V

SET Amplitude, V

10

0 -5 -10

0 -5 -10

-15 -20

5

-15 0

5

10 Pulse Width, µs

15

20

0

10

20

30

Time, µs

47

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q20 Q20 V∆t Plot

10

5

Output Signal, V

SET Amplitude, V

10

0 -5 -10

0 -5 -10

-15 -20

5

-15

0

5

10

15

0

20

10

20

30

Time, µs

Pulse Width, µs

ASET Testing (Pulsed Laser) Generate Plots of Amplitude vs Width Q19

Q16

Q15

R1

Q21 Q5 Q17 Q2

Q4

Q3

C

Q9

48

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: R1 Q20 V∆t Plot

10

Output Signal, V

SET Amplitude, V

10

5

0

-5

5

0

-10 0.0

0.5

1.0

1.5

2.0

2.5

3.0

0

Pulse Width, µs

5

10

15

Time, µs

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: R1 Q20 V∆t Plot

10

Output Signal, V

SET Amplitude, V

10

5

0

-5

5

0

-10 0.0

0.5

1.0

1.5

2.0

Pulse Width, µs

2.5

3.0

0

5

10

15

Time, µs

49

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: R1 Q20 V∆t Plot 10 Output Signal, V

SET Amplitude, V

10

5

0

-5

5

0

-10 0.0

0.5

1.0

1.5

2.0

2.5

0

3.0

5

10

15

Time, µs

Pulse Width, µs

ASET Testing (Pulsed Laser) Generate Plots of Amplitude vs Width Q15 Q19 Q16 Q21 Q5 Q17 Q2

Q4

Q3

C

Q9

50

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q19 Q20 V∆t Plot

0 Output Signal, V

SET Amplitude, V

0 -5 -10 -15 -20

-5 -10 -15 -20

0

5

10

15

20

25

30

0

10

Pulse Width, µs

20

30

Time, µs

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: Q16 10

10

5

5

Output Signal, V

SET Amplitude, V

Q20 V∆t Plot

0 -5 -10

0 -5 -10 -15

-15 -20

-20 0.0

5.0

10.0

15.0

Pulse Width, µs

20.0

25.0

-5

0

5

10

15

20

25

30

35

Time, µs

51

ASET Testing (Pulsed Laser) 590 nm Pulsed Laser ASET Data: All Nodes Q5

R1,Q6,Q16

10

Q18

SET Pulse Amplitude, V

Q4

5

Q2

0 -5

Q20

-10 -15

Q9,Q16,Q19 Q20

-20 0

5

10

15

20

25

30

SET Pulse Width, µs

ASET Testing (Pulsed Laser) Comparison of Heavy Ion and Laser Data

Q5

R1,Q6,Q16

10

10

Q18

SET Pulse Amplitude, V

Q4

SET Amplitude, V

5 0 -5 -10 -15

5

Q2

0 -5

Q20

-10 -15

Q9,Q16,Q19 Q20

-20 0

5

10

SET Pulse Width, µs

15

0

5

10

15

20

25

30

SET Pulse Width, µs

52

ASET Testing (Pulsed Laser)

Sub-bandgap 2-photon absorption induced SEE: • Deposit charge at different depths • Backside irradiation

ASET Testing (Pulsed Laser) Comparison of 1-Photon and 2-Photon SET

Output Signal, V

5 4 3 2 1

LM119 Q6 1260 nm 590 nm

0 -0.1

0.0

0.1

0.2

0.3

0.4

0.5

Time, µs

53

ASET Testing (Pulsed Laser) LM124 Q20: General Characteristics

P (C2)

P (E)

N+ (B)

P (C1)

Overlayers

P

C1

C2

N (base)

E

12 µm

N+ (buried layer)

B

P+ (iso)

P+ (iso) P (substrate)

McMorrow et al. IEEE TNS 2003

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

P+ (iso) P (substrate)

6

P

Output Signal, V

P (C2)

4 2 0

Z = -15 µm -2 0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

54

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

6

P

P+ (iso)

Output Signal, V

P (C2)

4 2 0

Z = -6 µm -2

P (substrate)

0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

P+ (iso) P (substrate)

6

P

Output Signal, V

P (C2)

4 2 0

Z = -2 µm -2 0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

55

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

6

P

P+ (iso)

Output Signal, V

P (C2)

4 2

Z = 0 µm

0 -2

P (substrate)

0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

P+ (iso) P (substrate)

6

P

Output Signal, V

P (C2)

4 2

Z = 4 µm

0 -2 0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

56

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

6

P

P+ (iso)

Output Signal, V

P (C2)

4 2

Z = 7 µm

0 -2

P (substrate)

0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

P+ (iso) P (substrate)

6

P

Output Signal, V

P (C2)

4 2

Z = 13 µm

0 -2 0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

57

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

6

P

P+ (iso)

Output Signal, V

P (C2)

4 2

Z = 20 µm

0 -2

P (substrate)

0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

P+ (iso) P (substrate)

6

P

Output Signal, V

P (C2)

4 2

Z = 25 µm

0 -2 0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

58

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

6

P

P+ (iso)

Output Signal, V

P (C2)

4 2

Z = 38 µm

0 -2

P (substrate)

0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

P+ (iso) P (substrate)

6

P

Output Signal, V

P (C2)

4 2

Z = 44 µm

0 -2 0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

59

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

6

P

P+ (iso)

Output Signal, V

P (C2)

4 2 0

Z = 47 µm -2

P (substrate)

0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

ASET Testing (Pulsed Laser) Z” Dependence: LM124 Q20 C1-epi Junction (Inverting Configuration; gain of 20) Overlayers P (C1)

N (base)

12 µm

N+ (buried layer) P+ (iso)

P+ (iso) P (substrate)

6

P

Output Signal, V

P (C2)

4 2 0

Z = 53 µm -2 0

10

20

30

Time, µs McMorrow et al. IEEE TNS 2003

60

ASET Testing (Pulsed Laser) Overlayers P (C1)

P (C2)

P

0

-1

N (base) -2

N+ (buried layer)

“C1” Transient

-3

P+ (iso)

0

P+ (iso)

10

20

30

P (substrate) 6

C1-Sub “Shunt”

5

0.0

4 3

-0.5

2 1 0

-1.0

“C2” Transient -1.5

0

10

20

-1

0

10

20

30

30

ASET Testing (Pulsed Laser)

Sub-bandgap two-photon absorption - backside irradiation

61

ASET Testing (Pulsed Laser) Backside “Through-Wafer” TPA Illumination

Circuit Layer(s)

Substrate transparent to single photon sub-bandgap excitation

Tightly focused twophoton excitation source

Surface elements opaque to optical excitation

Region of 2 Photon Carrier Generation McMorrow et al. IEEE TNS 2004

ASET Testing (Pulsed Laser) Photomicrograph of Q20 in LM124 Captured with NIR InGaAs Camera

Back Side

Front Side

62

ASET Testing (Pulsed Laser) Backside “Through-Wafer” TPA Illumination LM124 Operational Amplifier 1.5

Q18

Front Side Back Side

Output Signal, V

1.0 0.5 0.0 Q18 2.0 1.0 0.0 0

5

10

15

20

McMorrow IEEE TNS 2004

Time, µs

ASET Testing (Pulsed Laser)

X-Y Scanning……..

63

ASET Testing (Pulsed Laser) 2D Scan of LM6144 Op-Amp

Area 1

V. Pouget, IXL, France

ASET Testing (Pulsed Laser) 2D Scan of LM6144 Op-Amp

10 9

Amplitude (V)

8 7 6 5 4 3 2 1 0 0.00

10.00

20.00

30.00

40.00

50.00

µs) Time (µ

V. Pouget, IXL, France

64

ASET Testing (Pulsed Laser) 2D Scan of LM6144 Op-Amp

V. Pouget, IXL, France

ASET Testing Summary • ASET testing has unique aspects. – Sensitivity to configuration – Need to capture transients – Analysis to determine which transients are of concern

• The approaches discussed in this section are in large part a result of the DTRA program. • The best approach is to use a number of different experimental methods including: – Broad beam of heavy ions – Focused beam of heavy ions – Pulsed laser (1-photon and 2-photon)

• ASET sensitivity depends critically on configuration.

65

5. Case Studies

ASETs Originating in Resistors in LM119 Voltage Comparator

66

Case Studies – LM119 LM119 Circuit

Sternberg, IEEE TNS 2002

Case Studies – LM119 Resistor R11 shows ASET sensitivity Laser

Simulation

Laser

Sternberg, IEEE TNS 2002

67

Case Studies – LM119 Comparison of experimental and simulated results

Sternberg, IEEE TNS 2002

Spreading Resistance for Input Transistors in LM111 Voltage Comparator

68

Case Studies – LM111 Validating the SPICE Model for the LM111 I(t) C/B

I(t) E/B

I(t) E/C

Pease et al. IEEE TNS 2002

CASE STUDIES LM111 Validating the SPICE Model for the LM111

Photomicrograph Of LM111 input

Location of ASETs in LM111 Induced by ion microprobe Pease et al. IEEE TNS 2002

69

Case Studies – LM111 Validating the SPICE Model for the LM111

Collector

Base

Emitter

Sternberg, IEEE TNS 2002

Case Studies – LM111 Validating the SPICE Model for the LM111

Pease, IEEE TNS 2002

70

ASETs originating in Resistor in LM124

Case Studies – LM124 Validating the SPICE Model for the LM124

Heavy Ion

71

Case Studies – LM124 Validating the SPICE Model for the LM124

“R1”

Case Studies – LM124 Validating the SPICE Model for the LM124 4

Amplitude (V)

3

2

1

0

-1 -10

-8

-6

-4

-2

0

2

4

6

8

10

Time (µ µs)

Heavy Ion

Pulsed Laser

72

Case Studies – LM124 Validating the SPICE Model for the LM124 V+ Q16

Q15 Q14

Q19 Vin(+) C1

Q12

Q18 Q20 Vin(-)

Q13

R2 Q7

R1

Vout

Q6 Q11

Q17 Q21

Q9 Q5

Q2 Q3

Q8

Q10

Q4

V-

Case Studies – LM124 Validating the SPICE Model for the LM124 SPICE

R1 has a transistor structure but the base is left floating and so it acts as a resistor.

Pease, SEE Symposium 2004

73

Long-Duration Pulses (LDPs)

Case Studies ASET width determined by device bandwidth. OP293

LMH6628 4.5 4

300 us

Ampltiude (V)

3.5 3

30 ns

2.5 2 1.5 1 0.5 0 0

50

100

150

200

250

300

350

400

Time (ns)

Unity gain bandwidth = 35 kHz

Unity gain bandwidth = 300 MHz

74

Case Studies – LM6144 Long-Duration Pulses observed in Heavy-Ion Testing of LM6144 Op-Amp.

Unity-gain bandwidth - 17 MHz (∆t < 1µs)

Only observed for LETs > 50 MeV.cm2/mg

Boughassoul et al. IEEE TNS 2004

Case Studies – LM6144 Laser Scan identified the origins of the LDPs

-9 V to –9.25 V

9.2 V to 9.8 V

Boughassoul et al. IEEE TNS 2004

75

Case Studies – LM6144 Laser Scan identified the origins of the LDPs Effect of light: • Illuminator • Scattered laser light • Room lights Strikes to other transistors • Stops the transient LDPs (in dark) depend on supply voltage: Vdd = ± 10V, ∆t=25 ms Vdd = ± 7.5V, ∆t=45 ms Vdd = ± 5.0V, ∆t=100 ms Boughassoul et al. IEEE TNS 2004

Case Studies – LM6144 Used SPICE to model effects • Able to simulate all effects • Pulse Length vs supply voltage

Boughassoul et al. IEEE TNS 2004

76

Case Studies – LM6144 Used SPICE to model effects

• Suggested hardening approach – adding capacitors across C/B of two transistors in the bias/startup circuit.

Boughassoul et al. IEEE TNS 2004

6. ASET Mitigation

77

ASET Mitigation Approaches • Repeat the Measurement  Housekeeping measurement can be repeated three times

•Device Level  Epitaxial layers (SOI) to limit charge collection

• Circuit Level  Low bandpass filter to remove ASETs

• Circuit/Subsystem Level  Triple Modular Redundancy (TMR)

ASET Mitigation Triple Modular Redundancy Rin

Cin

Van Vonno, IEEE TNS 2001

Rin Cin

78

ASET Mitigation Minimum Input Overdrive for ASET-free Operation - Depends on Input Resistance and Capacitance 160

Cin = 0 pF

Input Overdrive (mV)

140 120

9V 15 V 30 V

100 80

9V Cin = 100 pF 15 V 30 V

60 40 20 0 0

20

40

60

80

Input Resistance (Kohms)

100

120 Van Vonno, IEEE TNS 2001

7. Summary & Conclusions

79

Summary & Conclusions Which LM139 Caused Reset on MAP? +5V

+5V R=1k

R=3.6k 2.9V

EEP_RESET (∆ ∆=2.5V) LM139

R=5.1k

Vref=2.5V HW_RESET (∆ ∆=2.5V)

+5V R=100k

2.9V LM139 Vref=2.5V

5V

R=1.5k C=10 µF

R=10k

LM139 Vref=2.5V Poivey RADECS 2002

Summary & Conclusions • ASETs have caused anomalies in spacecraft. • They occur in linear devices when particle radiation passes through a sensitive node. • A powerful approach for studying ASETs is to use a combination of simulation, broad-beam and focusedbeam of heavy ions, and pulsed laser light. • Linear devices are unique in that their ASET sensitivity depends on configuration. Testing in one condition does not automatically mean the data is valid for another condition. • There is many different approaches to reducing the ASET sensitivity of linear devices.

80