Single Event Transients in Linear Integrated Circuits€¢ Each ASET type also has a variety of amplitudes. • Not all ASETs pose a threat. • Transient widths often related to circuit

1

Single Event Transients in Linear Integrated Circuits

Stephen BuchnerQSS Group Inc./NASA GSFC

Dale McMorrowNaval Research Laboratory

K. LaBelS.Buchner, C.Poivey, J.Howard, R.Ladbury

A.Sternberg, Y.Boulghassoul,

T.Holman L.Massengill

J.Rowe

D.McMorrow, W.Lotshaw

M.Savage, T.Turflinger,

J.TitusD. Platteter

L. CohnA. Clark

R.Pease

2

Course Outline1. Introduction2. 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 Studies6. Mitigation7. Conclusions

1. Introduction

3

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

4

Introduction

• SETs appear as voltage “glitches” at the outputs of:– Op-amps, – Comparators, – Voltage refs, – DACs

4

5

6

7

8

9

10

0 2 4 6 8 10 12 14Time (µµµµs)

Am

plitu

de (V

)

R1

Vout

-

+Vin

R2

Introduction

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

-

+Vin

R1

R2

CLK

Follow-onCircuit

∆∆∆∆t

∆∆∆∆Vout

5

Introduction

0 5 10 15123456785

6

7

8

9

10

5.0

5.5

6.0

6.5

7.00.0

2.0

4.0

6.0

5.0

5.5

6.0

6.5

Q16

Time (m icroseconds)

Q20

Vou

t (V)

Q18

Q9Q19

Q2Q4Q5

• ASETs assume a wide variety of shapes. In the LM124 there are at least five different types.

ASETs in Op-amps (LM124)

Buchner et al IEEE TNS 2004

• Each ASET type also has avariety of amplitudes.

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

to circuit bandwidth.

Introduction

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

ASETs in Voltage Compators (LM119)

Buchner et al HEART 2001

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

Q6

Q3

Q2

0

1

2

3

4

5

0 0.2 0.4 0.6 0.8 1

Time (µµµµs)

0

1

3

4

5

2

0

1

3

4

5

2

6

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

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

Satellites Exhibiting ASETs (Partial List)

7

Introduction

• 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.

ASETs on MAP

Poivey RADECS 2002

Introduction

R=3.6k

LM139

LM139LM139

R=5.1k

R=10kR=1.5k

R=1k

R=100k

C=10 µµµµFVref=2.5V Vref=2.5V

Vref=2.5V

+5V (Min=4.3V)

+5V

EEP_RESET (∆∆∆∆=2.5V)

+5V

Poivey RADECS 2002

Processor Reset Circuit contains Comparators

2.9V

2.9V 5V

HW_RESET (∆∆∆∆=2.5V)

8

IntroductionASETs for the LM139 Comparator

-4

-2

0

2

4

1.2 10-5 1.4 10-5 1.6 10-5 1.8 10-5 2 10-5 2.2 10-5 2.4 10-5 2.6 10-5 2.8 10-5

Out

put V

olta

ge (

V)

T ime (s)

LM139, Vcc

=+/-5, δVi=800mV, LET=18.7 MeV-cm2/mg

Vt = 2.5 V

Poivey RADECS 2002

Introduction

• 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.

Summary:

9

2. Fundamental Mechanisms

Fundamental MechanismsInteraction of ions with matter

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

Ionization

10


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.

Ionization

SILICON


1. In the presence of a p-n junction charge is collected via drift in the junction electric field.

Ionization

SILICON

N-type

P-type

11

Fundamental MechanismsUnbiased p-n Junction.

EC

EV

EF

p-Typen-Type

Depletion Layer

E

Fundamental MechanismsForward-biased p-n junction.

EC

EV

p-Typen-Type

Depletion Layer

- +

E

12

Fundamental MechanismsReverse-biased p-n junction.

EC

EV

p-Typen-Type

Depletion Layer

+ -

E

Fundamental MechanismsCharge Collection via Drift

n-Type

p-Type

EC

EV

+ -

n-Type+

+

+

-

-

-

p-Type+-

Large electric field collects charge via drift, an efficient process.

E

13

Fundamental MechanismsCharge Collection by Diffusion

n-Type

p-TypeEC

EV

+-

-

+

-+

100 µµµµm

Electrons and holes generated far from the junction will diffuse slowly to the junction.

E

Fundamental MechanismsASETs Shapes

Time (ps)

Ele

ctri c

Pot

entia

l (V

)

Drift Component

Diffusion Component

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

14

Fundamental MechanismsBipolar Transistor

Emitter Base Collector

Emitter CollectorBase

n p n

EC

EV

Two back-to-back p-n junctions.

EF

Fundamental Mechanisms

Holes injected into C/B junction drift into base and turn the transistor more on.

Emitter CollectorBase

n p n

V1

V2


+

--

+

Forward-active mode

15


C E B

p-substrate

n+

n+ n+

n-epi p+p+p

Charge Collection in Vertical n-p-n Transistor.

Fundamental MechanismsDifferential pair:Used as inputs for op-amps and voltage comparators.

If:V1 = V2thenI1 = I2andVout(1)=Vout(2)

-

+V1

V2

V+

V1

R

I1

R

I2

V2Tr1 Tr2

IE

Vout(2)Vout(1)

16


V+

V1

R

I1

R

I2

V2Tr1 Tr2

IE

Vout(2)

If:V1 > V2thenI1 > I2andVout(2) is highVout(1) is low

Differential pair:

Vout(1)


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

Differential pair:

V+

V1

R

I1

R

I2

V2Tr1 Tr2

IE

Assume V1 > V2

17

Fundamental MechanismsLM111

• Input transistors are most sensitive to SETs.

• SETs propagate from input transistor to output.

VIN

VOUT


• Particles passing through matter produce electron-hole 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.

Summary

18

3. Computer Simulation

• Device simulator– E.g. SILVACO, PISCES, etc (Only two

published reports)• Circuit simulator

– E.g. SPICE.

Computer Simulation

19

1. Device SimulationSimulation of the device behavior by the solution of the following equations describing charge flow in the device.

Computer Simulation

• Maxwell’s Equations• Poisson’s Equations• Continuity Equations• Carrier Transport Equations

Non-linear differential equations cannot be solvedanalytically. A computer is used to solve the equationsnumerically in space and time.

Computer Simulation

Johnston IEEE TNS 2000

1. Device Simulation

VBE = -0.42 V – weakly on.Substrate p-n-p

p-substrate

n+

n-epi

pp+p+

E B

400

µm

20

Computer SimulationJohnston IEEE TNS 20001. Device Simulation

Substrate p-n-p

p-substrate

n+

n-epi

pp+p+

E B

400

µm


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.

21

Computer Simulation2. 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

LM124

Q17

Q21

Q19

C

Q20

Q16 Q15

R1

Q9

Q2

Q3

Q4

Q5

22

Q2Q3 Q4

Q5

Q19Q15Q16

Q14

Q13

Q11

Q10Q8

Q7Q12

Q9

Q6

Q21Q17

Q18 Q20 R1

R2

C1

Vin(-)

Vin(+)

V+

V-

Vout

Q2Q3 Q4

Q5

Q19Q15Q16

Q14

Q13

Q11

Q10Q8

Q7Q12

Q9

Q6

Q21Q17

Q18 Q20 R1

R2

C1

Vin(-)

Vin(+)

V+

V-

Vout

Computer Simulation

Identify Circuit Elements and Interconnects

LM124

Computer SimulationNeed 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-Poonparameters.

23

Computer Simulation

SPICE Validation – Small and Large Input Signals

Sternberg IEEE TNS 2002

+

-

Input:100 mV10 µµµµs

Output

LM124

Q2Q3 Q4

Q5

Q19Q15Q16

Q14

Q13

Q11

Q10Q8

Q7Q12

Q9

Q6

Q21Q17

Q18 Q20 R1

R2

C1

Vin(-)

Vin(+)

V+

V-

Vout

Q2Q3 Q4

Q5

Q19Q15Q16

Q14

Q13

Q11

Q10Q8

Q7Q12

Q9

Q6

Q21Q17

Q18 Q20 R1

R2

C1

Vin(-)

Vin(+)

V+

V-

Vout

Computer SimulationSpice Model (LM124) Validation –Shape of I(t) does not affect ASET

I(t)

I(t)

I(t)

C/B

E/B

E/C

24

Computer Simulation

SPICE Simulation – ASET Validation via Iteration

Compare SETs generated by pulsed laser or ion microprobe with those calculated using SPICE.

Agreement?Poor

Modify circuit anddevice parameters.

Good

Stop.

Computer Simulation

SPICE Simulation – ASET Validation

25

Computer Simulation

ASET Dependence on Compensating Capacitance


Computer Simulation

Constant Gain but Different Values of R1/R2


+

-R2

R1

Gain = = 10R1R2

Vin

Vout = 600 mV

Vin = -60 mV

26

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

27

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

28

ASET Testing (Broad Beam)Standard Approach for SEU Testing

1. Program Device 2. Irradiate Device withIons having selected LET

3. Read Number of Upsets

σ =# of UpsetsFluence

4. Calculate Cross Section 5. Irradiate Device withIons having different LET

σ(LET)

ASET Testing (Broad Beam)Standard Approach for SEU Testing

LET (MeV.cm2/mg)

Cro

ss-S

e ctio

n (c

m2 )

• Fit the data with a Weibull curve.• Use the curve as input in a

program such as CREME96to calculate SEU rate.

The goal is to predict the error rate in space

29

ASET Testing (Broad Beam)Modifications for ASET Testing

DUT Oscilloscope ComputerProbe

1. Carefully establish device configuration(voltages, trigger levels etc).

4. Capture and store all transients for later analysis of ∆∆∆∆V vs ∆∆∆∆t.

2. Establish a method for counting the ASETs3. Use active probe or test in actual application

ASET Testing (Broad Beam)Store all Transients

0

1

2

3

4

5

6

-5 0 5 10 15 20 25 300

1

2

3

4

5

6

7

8

9

-5 0 5 10 15 20 25 30-0.2

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

-5 0 5 10 15 20 25 300

1

2

3

4

5

6

7

8

-5 0 5 10 15 20 25-0.2

0

0.2

0.4

0.6

0.8

1

1.2

1.4

-5 0 5 10 15 20 25 30

Width

Amplitude

Adell et al. IEEE TNS 2000.

30

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

LM124 - LET = 2.8 MeV·cm2/mg

-15

-5

5

15

0 5 10 15

Pulse Width (µs)

Pul

se A

mpl

itude

(V)

0 10 20 30

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs

ASET Testing (Broad Beam)

Analysis of Test Data – Pulse Height vs Pulse Width

0 5 10 15-15

-10

-5

0

5

10

SE

T A

mpl

itude

, V

SET Pulse Width, µs

LM124LET = 53.9 MeV·cm2/mg;

Voltage Follower.

31

1.E-06

1.E-05

1.E-04

1.E-03

1.E-02

0 20 40 60 80 100 120

LET (MeV.cm2/mg)

Cro

ss-S

ectio

n (c

m2 )


Poivey GSFC Report

Effect of Application on σσσσ(LET)

Rate ~ σsat

(LETth)2

σsat


Factors that affect ASET Cross-Section

1. Oscilloscope trigger level2. Input voltage3. Supply voltage4. Gain5. Output loading

32


Koga et al. IEEE TNS 1993

Factors that affect ASET Cross-Section

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

LM111



Factors affecting ASET Cross-Section: (V2-V1)

LM139

V2

V1

33



Factors affecting ASET Cross-Section: Vs

LM111

V+

V-

1.0E-07

1.0E-05

1.0E-03

0 10 20 30 40 50 60 70

LET (MeV.cm2/mg)

Cro

ss S

ectio

n (c

m2 /d

evic

e) Vs = +/-15V, ∆Vin=50 mV

Vs = +/-5V, ∆Vin=50 mV

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

LM119

V+

V-

-1

0

1

2

3

4

5

6

0 2 4 6 8 10

Time (µµµµs)

Am

plitu

de (V

)

R=0.17 kΩΩΩΩ

R=1.7 kΩΩΩΩ

R

5.0 V

0

1

2

3

4

5

6

0 2 4 6 8 10

Time (µµµµs)

Am

plitu

de (V

)

34


Poivey GSFC Report

Effects of Ion Range

TAMU:• 15 MeV/amu has range ~ 150 µµµµms• 25 MeV/amu has range ~ 400 µµµµms• 40 MeV/amu has range ~ 1,000 µµµµms

BNL:• 2 - 8 MeV/amu has range ~ 50 µµµµms

2. Pulsed Laser

ASET Testing

35

ASET Testing (Heavy Ions)

Pulsed Laser ASET Testing Technique• Use light instead of particles to generate free

carriers (electrons and holes).

Coulomb interactionbetween nucleus ofincident particle andbound electrons of Si

AbsorptionParticles

Absorption of photons bybound electrons of Si

-4 -2 0 2 410

8

6

4

2

0

w(z)

1/e Contour

Distance, µm

Dep

th in

Mat

eria

l, µm

ASET Testing (Pulsed Laser)

-4 -2 0 2 410

8

6

4

2

0

w(z)

1/e Contour

Distance, µm

µ

λλλλ=590 nm λλλλ=800 nm

30 MeV Ar

Comparison of Ion and Laser-light Induced Charge Tracks

36

ASET Testing (Heavy Ions)

Pulsed Laser ASET Testing Technique• Can focus light to a diameter of ~ 1 µµµµm 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.


One Photon:600 nm, 800nm, 1.06 µmAbove band gapSingle-photon absorption

400 600 800 1000 1200 140010-1

100

101

102

103

104

105

1.06 µm

1.26 µm

800 nm

600 nm

Abs

orpt

ion

Coe

ffici

ent,

cm-1

Wavelength, nm

Two Approaches:

Two Photon:λ > 1.15 µmSub-bandgapTwo-photon absorption

37

ASET Testing (Pulsed Laser)Two Approaches:

600 nm

Ec

Ev

Ec

Ev

1260 nm

+- -

-

--- -

1-Photon 2-Photon

Eg=1.1eV

+

-

• 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:


ωα

),( ),( zrIdt

zrdN =

Carrier generation equation:

zoeIzrI α−=),(

- 4 - 2 0 2 41 0

8

6

4

2

0

w ( z )

1 /e C o n to u r

P o s it io n , µ m

One-Photon Absorption SEE Experiment

38


• Carriers are generated by nonlinear absorption at high pulse irradiances by the simultaneous absorption of two photonsω

βω

α 2

),(),( ),( 22 zrIzrI

dtzrdN +=

Carrier generation equation:

-4 -2 0 2 4

030

1020

N α I2

w (z )

1 /e C o n tou r

P os itio n , µm

Dep

th in

Mat

eria

l, µm

Two-Photon Absorption SEE Experiment

• 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 -2 0 2 4

030

1020

N α I2

w(z)

1/e Contour

Position, µm

Dep

th in

Mat

eria

l, µm µ

Carrier Density Distribution:1-Photon vs. 2-Photon Absorption


-4 -2 0 2 410

8

6

4

2

0

w(z)

1/e Contour

Distance, µm

Dep

th in

Mat

eria

l, µm 800 nm

39


ps or fs laser source

Polarizer

λ/2

DUT

xyz

laser pulse

ccdcamera

PD2

PD1

Pulsed Laser SEE Experimental Apparatus

Validate Technique


40

Q2Q3 Q4

Q5

Q19Q15Q16

Q14

Q13

Q11

Q10Q8

Q7Q12

Q9

Q6

Q21Q17

Q18 Q20 R1

R2

C1

Vin(-)

Vin(+)

V+

V-

Vout

Q2Q3 Q4

Q5

Q19Q15Q16

Q14

Q13

Q11

Q10Q8

Q7Q12

Q9

Q6

Q21Q17

Q18 Q20 R1

R2

C1

Vin(-)

Vin(+)

V+

V-

Vout


Devices Sensitive to SETs (LM124)


0 5 10 15

1.0

1.1

1.2

1.0

1.1

1.2

40 MeV Cl 590 nm LaserQ6

Time (µs)

Vou

t (V)

Q2

0 5 10 15 20

0.9

1.0

1.1

1.2

1.2

1.3

1.4

1.5

Q20

Time (µs)

40 MeV Cl 590 nm Laser

Q18

LM124Inverting Configuration:

Vdd = +/-6 VVin = 60 mV

Buchner et al. IEEE TNS 2003

SETs: Comparison of Pulsed Laser Light and Low LET Heavy Ions

41

0 5 10 1556789

10

5.0

5.5

6.0

6.5

7.0

Q20

Time, ms

Q18

0 5 10 150.0

2.0

4.0

6.0

5.0

5.5

6.0

6.5

Vou

t (V)

Time, µs

Q9Q19

Q2Q4Q5


LM124 Voltage Follower:Vdd = +/-15 VVin = 5 V

Pease et al. IEEE TNS 2002

SETs: Comparison of Pulsed Laser Light and High LET Heavy Ions

How do we handle these transients?


42


Generate Plots of Amplitude vs Width

Q17

Q21

Q19

C

Q16 Q15

R1

Q9

Q2

Q3

Q4

Q5

Q20, C1

0 5 10 15 20-20

-15

-10

-5

0

5

10

Q20 V∆∆∆∆t Plot

SET

Am

plitu

de, V

Pulse Width, µs

0 10 20 30

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs

ASET Testing (Pulsed Laser)590 nm Pulsed Laser ASET Data: Q20 (C1)

43

0 5 10 15 20-20

-15

-10

-5

0

5

10


SET

Am

plitu

de, V

Pulse Width, µs

0 10 20 30

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs

590 nm Pulsed Laser ASET Data: Q20 (C1)


0 5 10 15 20-20

-15

-10

-5

0

5

10


SET

Am

plitu

de, V

Pulse Width, µs

0 10 20 30

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs

590 nm Pulsed Laser ASET Data: Q20 (C1)


44

0 5 10 15 20-20

-15

-10

-5

0

5

10


SET

Am

plitu

de, V

Pulse Width, µs

0 10 20 30

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs


0 5 10 15 20-20

-15

-10

-5

0

5

10


SET

Am

plitu

de, V

Pulse Width, µs

0 10 20 30

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs


45



Q17

Q21

Q19

C

Q16 Q15

R1

Q9

Q2

Q3

Q4

Q5

Q20, C2

0 5 10 15 20-20

-15

-10

-5

0

5

10


SET

Ampl

itude

, V

Pulse Width, µs


46

0 5 10 15 20-20

-15

-10

-5

0

5

10


SET

Ampl

itude

, V

Pulse Width, µs

0 10 20 30

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs


0 5 10 15 20-20

-15

-10

-5

0

5

10


SET

Ampl

itude

, V

Pulse Width, µs

0 10 20 30

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs

ASET Testing (Pulsed Laser)590 nm Pulsed Laser ASET Data: Q20

47

0 5 10 15 20-20

-15

-10

-5

0

5

10


SET

Ampl

itude

, V

Pulse Width, µs

0 10 20 30

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs


0 5 10 15 20-20

-15

-10

-5

0

5

10


SET

Ampl

itude

, V

Pulse Width, µs

0 10 20 30

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs


48

0 5 10 15 20-20

-15

-10

-5

0

5

10


SET

Ampl

itude

, V

Pulse Width, µs

0 10 20 30

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs




Q17

Q21

Q19

C

Q16 Q15

Q9

Q2

Q3

Q4

Q5

R1

49

0.0 0.5 1.0 1.5 2.0 2.5 3.0-10

-5

0

5

10Q20 V∆∆∆∆t Plot

SET

Am

plitu

de, V

Pulse Width, µs0 5 10 15

0

5

10

Out

put S

igna

l, V

Time, µs

590 nm Pulsed Laser ASET Data: R1


0.0 0.5 1.0 1.5 2.0 2.5 3.0-10

-5

0

5


SET

Am

plitu

de, V


0

5

10

Out

put S

igna

l, V

Time, µs



50

0.0 0.5 1.0 1.5 2.0 2.5 3.0-10

-5

0

5


SET

Am

plitu

de, V

Pulse Width, µs

0 5 10 15

0

5

10

Out

put S

igna

l, V

Time, µs





Q17

Q21

C

Q15

Q9

Q2

Q3

Q4

Q5

Q19

Q16

51

0 5 10 15 20 25 30

-20

-15

-10

-5

0


SET

Am

plitu

de, V


-20

-15

-10

-5

0

Out

put S

igna

l, V

Time, µs


590 nm Pulsed Laser ASET Data: Q19

-5 0 5 10 15 20 25 30 35

-20

-15

-10

-5

0

5

10

Out

put S

igna

l, V

Time, µs

0.0 5.0 10.0 15.0 20.0 25.0

-20

-15

-10

-5

0

5

10


SET

Ampl

itude

, V

Pulse Width, µs


590 nm Pulsed Laser ASET Data: Q16

52

0 5 10 15 20 25 30

-20

-15

-10

-5

0

5

10

Q2

Q18Q5

Q4

Q9,Q16,Q19Q20

Q20

R1,Q6,Q16SE

T Pu

lse

Ampl

itude

, V



590 nm Pulsed Laser ASET Data: All Nodes

0 5 10 15 20 25 30

-20

-15

-10

-5

0

5

10

Q2

Q18Q5

Q4

Q9,Q16,Q19Q20

Q20

R1,Q6,Q16

SET

Puls

e Am

plitu

de, V



Comparison of Heavy Ion and Laser Data

0 5 10 15-15

-10

-5

0

5

10

SET

Am

plitu

de, V


53


Sub-bandgap 2-photon absorption induced SEE:

• Deposit charge at different depths• Backside irradiation


-0.1 0.0 0.1 0.2 0.3 0.4 0.5

0

1

2

3

4

5

Out

put S

igna

l, V

Time, µs

LM119 Q6 1260 nm 590 nm

Comparison of 1-Photon and 2-Photon SET

54

C1C2

E

B

OverlayersP (C1)P (E) N+ (B)

P+ (iso)P+ (iso)

N (base)

P (substrate)

12 µm

P (C2)

N+ (buried layer)

P


LM124 Q20: General Characteristics

McMorrow et al. IEEE TNS 2003

OverlayersP (C1)

P+ (iso)P+ (iso)

N (base)

P (substrate)

12 µm

P (C2)

N+ (buried layer)

P

0 10 20 30

-2

0

2

4

6

Z = -15 µm

Out

put S

igna

l, V

Time, µs

Z” Dependence: LM124 Q20 C1-epi Junction(Inverting Configuration; gain of 20)



55

OverlayersP (C1)

P+ (iso)P+ (iso)

N (base)

P (substrate)

12 µm

P (C2)

N+ (buried layer)

P

0 10 20 30

-2

0

2

4

6

Z = -6 µm

Out

put S

igna

l, V

Time, µs




OverlayersP (C1)

P+ (iso)P+ (iso)

N (base)

P (substrate)

12 µm

P (C2)

N+ (buried layer)

P

0 10 20 30

-2

0

2

4

6

Z = -2 µm

Out

put S

igna

l, V

Time, µs




56

OverlayersP (C1)

P+ (iso)P+ (iso)

N (base)

P (substrate)

12 µm

P (C2)

N+ (buried layer)

P

0 10 20 30

-2

0

2

4

6

Z = 0 µm

Out

put S

igna

l, V

Time, µs




OverlayersP (C1)

P+ (iso)P+ (iso)

N (base)

P (substrate)

12 µm

P (C2)

N+ (buried layer)

P

0 10 20 30

-2

0

2

4

6

Z = 4 µm

Out

put S

igna

l, V

Time, µs




57

OverlayersP (C1)

P+ (iso)P+ (iso)

N (base)

P (substrate)

12 µm

P (C2)

N+ (buried layer)

P

0 10 20 30

-2

0

2

4

6

Z = 7 µm

Out

put S

igna

l, V

Time, µs




OverlayersP (C1)

P+ (iso)P+ (iso)

N (base)

P (substrate)

12 µm

P (C2)

N+ (buried layer)

P

0 10 20 30

-2

0

2

4

6

Z = 13 µm

Out

put S

igna

l, V

Time, µs




58

OverlayersP (C1)

P+ (iso)P+ (iso)

N (base)

P (substrate)

12 µm

P (C2)

N+ (buried layer)

P

0 10 20 30

-2

0

2

4

6

Z = 20 µm

Out

put S

igna

l, V

Time, µs




OverlayersP (C1)

P+ (iso)P+ (iso)

N (base)

P (substrate)

12 µm

P (C2)

N+ (buried layer)

P

0 10 20 30

-2

0

2

4

6

Z = 25 µm

Out

put S

igna

l, V

Time, µs




59

0 10 20 30

-2

0

2

4

6

Z = 38 µm

Out

put S

igna

l, V

Time, µs

OverlayersP (C1)

P+ (iso)P+ (iso)

N (base) 12 µm

P (C2)

N+ (buried layer)

P

P (substrate)




OverlayersP (C1)

P+ (iso)P+ (iso)

N (base)

P (substrate)

12 µm

P (C2)

N+ (buried layer)

P

0 10 20 30

-2

0

2

4

6

Z = 44 µm

Out

put S

igna

l, V

Time, µs




60

0 10 20 30

-2

0

2

4

6

Z = 47 µm

Out

put S

igna

l, V

Time, µs

OverlayersP (C1)

P+ (iso)P+ (iso)

N (base) 12 µm

P (C2)

N+ (buried layer)

P

P (substrate)




0 10 20 30

-2

0

2

4

6

Z = 53 µm

Out

put S

igna

l, V

Time, µs

OverlayersP (C1)

P+ (iso)P+ (iso)

N (base) 12 µm

P (C2)

N+ (buried layer)

P

P (substrate)




61

0 10 20 30-1

0

1

2

3

4

5

6

OverlayersP (C1)

P+ (iso)P+ (iso)

N (base)

P (substrate)

P (C2)

N+ (buried layer)

P

0 10 20 30

-3

-2

-1

0

0 10 20 30-1.5

-1.0

-0.5

0.0

“C1” Transient

“C2” Transient

C1-Sub“Shunt”



Sub-bandgap two-photon absorption - backside irradiation

62


Surface elements opaque to optical excitation

Tightly focused two-photon excitation

source

Substrate transparent to single photon

sub-bandgap excitationCircuit Layer(s)

Region of 2 PhotonCarrier Generation

Backside “Through-Wafer” TPA Illumination



Back Side Front Side

Photomicrograph of Q20 in LM124 Captured with NIR InGaAs Camera

63


0 5 10 15 20

0.0

1.0

2.0Q18

Out

put S

igna

l, V

Time, µs

0.0

0.5

1.0

1.5 Q18

Front Side Back Side

Backside “Through-Wafer” TPA IlluminationLM124 Operational Amplifier

McMorrow IEEE TNS 2004


X-Y Scanning……..

64


2D Scan of LM6144 Op-Amp

Area 1

V. Pouget, IXL, France




0

1

2

3

4

5

6

7

8

9

10

0.00 10.00 20.00 30.00 40.00 50.00

Time (µµµµs)

Am

plitu

de (V

)

65




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.

66

5. Case Studies

ASETs Originating in Resistors in LM119 Voltage Comparator

67

Case Studies – LM119

Sternberg, IEEE TNS 2002

LM119 Circuit



Resistor R11 shows ASET sensitivity

Laser

Laser

Simulation

68



Comparison of experimental and simulated results

Spreading Resistance for Input Transistors in LM111 Voltage

Comparator

69



Validating the SPICE Model for the LM111

I(t)

I(t)

I(t)

C/B

E/B

E/C

CASE STUDIES LM111


Location of ASETs in LM111Induced by ion microprobePhotomicrograph

Of LM111 input


70




Base

Collector

Emitter


Pease, IEEE TNS 2002


71

ASETs originating in Resistor in LM124

Case Studies – LM124Validating the SPICE Model for the LM124

Heavy Ion

72


“R1”


-1

0

1

2

3

4

-10 -8 -6 -4 -2 0 2 4 6 8 10

Time (µµµµs)

Am

plitu

de (V

)

Pulsed LaserHeavy Ion

73


Q2Q3 Q4

Q5

Q19Q15Q16

Q14

Q13

Q11

Q10Q8

Q7Q12

Q9

Q6

Q21Q17

Q18 Q20 R1

R2

C1

Vin(-)

Vin(+)

V+

V-

Vout

Q2Q3 Q4

Q5

Q19Q15Q16

Q14

Q13

Q11

Q10Q8

Q7Q12

Q9

Q6

Q21Q17

Q18 Q20 R1

R2

C1

Vin(-)

Vin(+)

V+

V-

Vout


Pease, SEE Symposium 2004


R1 has a transistorstructure but thebase is left floatingand so it acts as aresistor.

SPICE

74

Long-Duration Pulses(LDPs)

ASET width determined by device bandwidth.

Case Studies

Unity gain bandwidth = 35 kHz Unity gain bandwidth = 300 MHz

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

0 50 100 150 200 250 300 350 400

Time (ns)

Am

pltiu

de (V

)

LMH6628OP293

300 us

30 ns

75

Case Studies – LM6144Long-Duration Pulses observed in Heavy-Ion Testing of LM6144 Op-Amp.

Boughassoul et al. IEEE TNS 2004

Unity-gain bandwidth - 17 MHz

((((∆∆∆∆t < 1µµµµs)

Only observed for LETs > 50 MeV.cm2/mg


Laser Scan identified the origins of the LDPs

-9 V to –9.25 V 9.2 V to 9.8 VBoughassoul et al. IEEE TNS 2004

76


Laser Scan identified the origins of the LDPsEffect 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 msVdd = ± 7.5V, ∆t=45 msVdd = ± 5.0V, ∆t=100 ms



Used SPICE to model effects

• Able to simulate all effects• Pulse Length vs supply voltage


77


Used SPICE to model effects

• Suggested hardeningapproach – adding capacitors across C/Bof two transistors in the bias/startup circuit.


6. ASET Mitigation

78

ASET MitigationApproaches

• 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

Van Vonno, IEEE TNS 2001

Triple Modular Redundancy

Cin

Rin

CinRin

79

ASET Mitigation

Van Vonno, IEEE TNS 2001

Minimum Input Overdrive for ASET-free Operation

- Depends on Input Resistance and Capacitance

0

20

40

60

80

100

120

140

160

0 20 40 60 80 100 120

Input Resistance (Kohms)

Inpu

t Ove

rdriv

e (m

V)Cin = 0 pF

Cin = 100 pF

9 V15 V30 V

9 V15 V30 V

7. Summary & Conclusions

80

Summary & Conclusions

R=3.6k

LM139

LM139LM139

R=5.1k

R=10kR=1.5k

R=1k

R=100k

C=10 µµµµFVref=2.5V Vref=2.5V

Vref=2.5V

+5V

+5V

EEP_RESET (∆∆∆∆=2.5V)

HW_RESET (∆∆∆∆=2.5V)

+5V

Poivey RADECS 2002

2.9V

2.9V 5V

Which LM139 Caused Reset on MAP?

LM139

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 focused-beam 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.

Single Event Transients in Linear Integrated Circuits€¢ Each ASET type also has a variety of amplitudes. • Not all ASETs pose a threat. • Transient widths often related to circuit

Documents