* National DefenseDefence nationale

PULSE-WIDTH DEPENDENT RADIATION EFFECTSCON ELECTRONIC COMPONENTS

by

T. CousinsNuclear Effects Section

Electronics Division

DTIC

S ELECTEMR2Z19;

1 1

DEFENCE RESEARCH ESTABLISHMENT OTTAWATECHNICAL NOTE 89-33

PCN NowaW M041 LSOt"

Applovod for pubit rer~xtyibudon UnUSIId~ Z -!_

ABSTRACT

The simulation of the prompt gamma-ray pulse effects onelectronics with an electron linear accelerator (LINAC) has beenperformed by DREO and other groups. However, the use of a LINACnormally entails wider pulses than those expected on thebattlefield. This rcport examines the effects of the variationof pulse width on electronic response using both theoretical andexperimental examples. The conclusions are that pulse-widthfidelity is important for a number of possible scenarios, andthat for complete understanding of electronic performance, avariable pulse-width simulator is essential.

RESUME

La simulation de l'action des rayons gamma rapides sur lescomposantes electroniques & l'aide d'un accelerateur lineaire(LINAC) a ete realisee par le CRDO et d'autres groupes. Parcontre, l'emploi du LINAC donne normalement des impulsions pluslarges que celles normalement eprouvees sur un champ de bataille.Ce rapport examine les effets de la variation de la largeur desimpulsions sur les composantes electroniques en se servantd'exemples theoriques et experimentaux. Les conclusions sont quela precision de la largeur des impulsions est tres importantepour un nombre possible de scenarios et que pour unecomprehension complete du rendement des composantes electroniquesun simulateur a impulsions largeur variable est indispensable.

Aooession For

NTIS GRA&IDTIC TAB 0Unannounced 0Justifioatio-

ByDistributi

AvalablitY Codes

I . Avail and/oriii. Det Special

- I

EXECUTIVE SUMMARY

In order to realistically simulate the effects of the promptgamma-ray pulse associated with a nuclear weapon on electronics,an electron linear accelerator (LINAC) is often used. The pulsewidths available from most LINACs are longer than the typicalbattlefield pulse. This report examines the effects of varyingpulse widths on selected electronic devices summarizing somerecent DREO work at a variable pulse-width facility located atChalk River Nuclear Laboratories.

V

TABLE OF CONTENTS

PAGE

ABSTRACT/RESUME .......... ................... ii

EXECUTIVE SUMMARY ......... .................. iii

TABLE OF CONTENTS ........ .................. iv

1.0 INTRODUCTION .................... 1

2.0 EXPERIMENTAL ....... ................. 1

2.1 PROPAGATION DELAY TIME IN BIPOLARTRANSITIONS .......... ................. 1

2.2 SATURATION TIMES IN OP AMPS ..... ......... 4

2.3 CHARGE TRAPPING EFFECTS IN GaAs .... ....... 5

2.3.1 PULSE WIDTH DEPENDENCE OF CHARGE TRAPPING 5

2.3.2 TEMPERATURE DEPENDENCE OF CHARGE TRAPPING 7

3.0 CONCLUSIONS .......... .................. 9

4.0 BIBLIOGRAPHY ........ .................. 19

TABLES

TABLE 1 Bipolar Transistor Response toPulse Width Variation ...... ........... 3

TABLE 2 OPAMP Parameters as Function ofPulse Width ......... ............... 4

vii

FIGURE CAPTIONS

Figure 1 Propagation delay time characteristicsof a bipolar transistor showing thecomponents described in the text .. ..... 10

Figure 2 Measured response of 2N4401 transistorto 80 ns wide pulsc. The radiationstorage time and fall time parametersare evident ..... ............... 11

Figure 3 Measured response of TIP122 to 350 nswide pulse. Note that tsr and tf aregreater than for the 2N4401. Theincrease in tsr was expected due tolarger electrical storage time ...... 11

Figure 4 Measured response of 741C operationalamplifiers to 84 ns and 2.5 gs widepulses, showing wide voltage swings,varying with pulse width . ........ 12

Figure 5 Pulse width dependence of positive andnegative saturation times for 741C op-ampNote that the saturation times decreaseas the pulse width decreases toward the morerealistic value of two shakes ...... 13

Figure 6 Idealized GaAs MESFET response tothree pulse widths for traps havingtime constants of 10 ns, 0.5 gs and10 As. Note that these effects aretruly pulse width - as opposed to doserate - dependent ... ............ 14

Figure 7 Realistic predicted GaAs MESFET responseto incident square wave irradiations ofvarious pulse widths, incorporating devicesaturation. Note that in order to observea particular trap, the pulse width shouldbe similar to the trap time constant 15

ix

Figure 8 Response (lower curve) of NEC270 GaAsMESFET to 3 gs wide pulse (upper curve).Note the similarity of the response to thatpredicted - i.e. a saturation which mimicsthe pulse, followed by an exponential deLiydue to charge trapping .. ......... 16

Figure 9 Response of TOS710 MESFET to 2 gs widepulse clearly showing superposition ofthe two effect discussed in the text . . . 17

Figure 10 Response of NEC270 MESFET at twodifferent temperatures, 300 K and 333 K.The enhanced de-trapping at highertemperatures is readily apparent ..... 17

Figure 11 T2T vs 1/T plots for both types of MESFETS.From the slopes and intercepts of theseplots the trap depths and cross sectionscan be ascertained as described in thetext ....... ................... 18

xi

1.0 INTRODUCTION

The concept of using an electron linear accelerator (LINAC)to simulate the radiation damage in electronics arising from theprompt gamma-ray component of a nuclear burst (causing subsequentphotocurrent generation) has been demonstrated valid by manyauthors (eg 1) and has also been successfully demonstrated byDREO at the Chalk River Nuclear Laboratories (CRNL) 10 MeV LINAC(PHELA) facility (2). The LINAC, as one advantage over a flashX-ray facility, generally offers a variability in pulse widths.Some arguments have been made (3) that dose-rate (photocurrent)effects should be to a large extent pulse-width independent.PHELA was originally capable of 1 As to 6 As pulse widths, andhas recently been modified, under DND contract (4), to change itspulse width range to span 150 ns to 3 As. Thus the effects ofpulse width, for the same dose rate, could be directly examined.

In view of the fact that recent DREO work has suggested apulse width of 20 ns (2 shakes) (5) for an actual weapon burst isreasonable - knowledge of this pulse width dependence isessential to the design and subsequent possible purchase of aDREO-based LINAC dedicated to Transient Radiation Effects onElectronics (TREE) work.

The work presented here represents a brief examination of thepulse-width dependence for some of the devices previously testedby DREO. These results represent only one day's work at PHELAdue to severe time constraints, however some interesting trendsare observed.

2.0 EXPERIMENTAL

2.1 Propagation Delay Time in Bipolar Transitions

One area in which a definite pulse width effect has beennoted is in the propagation delay time in bipolar transistorscaused by the radiation pulse. Following along the lines of thediscussion in Messenger and Ash (6), this propagation delay timetp is given (see Fig. 1) as

tp = td+ tr+ tsr + tf (1)

where td = delay time to bring device from its on state to 10%of saturation current

tr = rise time to 90% of saturation

-2-

tsr = time device spends in saturation following pulsecessation (radiation storage time)

and tf = fall time back to 10% of saturated current.

Of these, tsr has shown some pulse width dependence as (1)

t t in ( K, (l-e - tp/o.3ts) - K2 ) (2)

where t, = electrical storage time

tp pulse width

and K, and K2 are device dependent constants.

Messenger and Ash imply no direct pulse width dependence inthe remaining three terms of equation 1.

In order to examine these trends, two bipolar transistortypes were exposed to varying pulse widths, for the same doserate (4 z 3 X 109 Rads (Si)/s). They were 2N4401 (t, = 225 ns)and TIP122 (t, = 8.5 gs) (where the electrical storage times arefrom the manufacturers). Examples of their measured responses toLINAC pulses appear in Figs. 2 and 3 respectively. To expeditedata analysis, a Tektronics 230 oscilloscope output of deviceresponse was fed to a PC (in LOTUS format) enabling accurateanalysis simply by numerical interpolation of the LOTUS files.The results of these analyses appear in Table 1.

Immediately obvious is the previously observed (2)proportionality between radiation storage time and electricalstorage time. Explicitly the ratios are in somewhat reasonableagreement at:

It srI2N4 01 1.04 X 10-1 (3)(tsr) TlP122

and {tsL2N401 2.65 X 10-2 (4)(ts) TIP122

As shown in table 1, for both transistors studied here, allparameters proved to be pulse-width independent - within

-3-

experimental error. (Note that td and tr are not evaluated for theTIP122 transistor due to the fact that this device provedsensitive to the (6 gs) RF pulse associated with the electronpulse. However, in view of the very fast rise time at the actualpulse arrival, it is felt that td and tr are both insignificant incomparison to tr). For these two cases then, LINAC testing atvirtually any pulse width should yield reliable results whenextrapolated to the battlefield timeframe.

Table 1

Bipolar Transistor Response To Pulse Width

Variation (Dose Rate - 3 X 109 Rads(Si)/s)

(a) 2N4401 (t5 = 225 ns)

LINAC Pulse Width(ns) td tr tsr tf tp

(FWHM) (ns) (ns) (ns) (ns) (ns)1600±100 <10 50±5 950±100 390±50 1400±110540±50 <10 50±5 1054±100 300±50 1414±110283±20 <10 50±5 1132±100 360±50 1552±110137±10 <10 50±5 1083±100 470±50 1553±11078±5 <10 50±5 1064±100 380±50 1504±110

(b) TIP122 (t5 = 8.5 ns)

LINAC Pulse Width(ns) td tr tsr tf tp

(FWHM) (ns) (ns) (us) (us) (as)1600±100 - - 9.6±U.1 4.3±0.1 13.9±0.1350±30 - - 9.5±0.1 4.4±0.1 13.9±0.1

-4-

2.2 Saturation Times in OP AMPS

A 741C operational amplifier was powered and biased, withgain of 100, and exposed to various pulse widths (dose rate- 7 X 108 Rads (Si)/s). The measurement circuit used here was thesame as reported in previous DREO work (2). Typical responsesare shown in Fig. 4. Note the alternate negative and positivesaturations typical of op-amp dose-rate performance. The timethe device spends at the negative and positive rails as afunction of pulse width appears in Table 2.

Table 2741C OP-AMP Parameters as Function of Pulse Width

LINAC Pulse Time At Time AtWidth Negative Rail Positive Rail(ns) (As) (as)

(±10%) (±. lus) (±.l"s)

2500 5.1 10.51500 5.0 10.0650 4.3 10.0500 4.3 9.4430 4.4 9.3200 4.1 9.0112 3.7 5.084 3.4 4.5

There is a large pulse width-dependent effect in thisparticular op-amp. The response of the op-amp may be consideredto be dominated by (bipolar) transistors being driven intosaturation and subsequent on-board amplification. Thus therewould seem to be definite evidence for a radiation storage timedependence on pulse width for the transistors used in this op-amp. Extrapolation down to the two shake pulse width would givepositive and negative saturation times of less than 1 gs. (seeFig. (5)). So when designing/building circuitry capable ofwithstanding the op-amp radiation-induced voltage swings producedby gs wide pulses, protection against the actual (20 ns) inducedswings will be achieved but will result in much unnecessary over-design (and accompanying cost). The op-amp is an example of adevice which must be tested in the expected battlefieldtimeframe.

-5-

2.3 Charge Trapping Effects in GaAs

Charge trapping is the withholding of normal charge motion ina material for any number of physical reasons. In the case ofGaAs MESFETS, electron charge trapping near the FET channelregion has been observed by numerous authors (eg 2,7,8).

For the case of a LiNAC pulse irradiating a GaAs device,electron hole pairs are created by the pulse, however theelectrons may be trapped at an energy level ET below theconduction band. These electrons may then decay from this levelwith a characteristic time constant depending upon, among otherfactors, trap depth and temperature as d scussed later in thissection.

2.3.1 Pulse Width Dependence of Charge Trapping

This immediate discussion centres on the pulse widthdependence of charge trapping. Although the total number ofelectrons which are trapped at any level for a given irradiationis directly proportional to incident radiation fluence, thetimeframe over which the radiation is given is also a key factorin the determination of this parameter. The radiation pulsewidth will determine which traps (corresponding to a particulartime constant) will he preferentially filled - and subsequentlyemptied. Thus the radiation pulse width will determine whetheror not a specific charge trapping reaction may be experimentallyobserved.

The mathematical and pictorial analysis below should clarifythis, however as a rough rule of thumb in order to observe a trapwith a characteristic tire constant r, the radiation pulse widthmust be at maximum of the order of T. This explains why chargetrapping effects are impossible to observe with steady stateradiation as no trap depths coriesponding to very long timeconstants are possible (due to very large ET exceeding theconduction band gap).

MacLean, in his excellent review article on MOS devices (9),has referred to short-pulse charge trapping as an apparent doserate effect. We consider this terminology to be somewhatmisleading as dose rate effects usually refer to photocurrentgeneration, which is not a factor in the charge trappingmechanisms. Rather the terminology "pulse-width dependentionization effects" is proposed as a more accurate description.

- 6 -

In order to mathematically simulate the charge trapping anddecay mechanisms we use a simple mathematical approach. Sincethe traps are observed to have a characteristic time constant,then the trapping and de-trapping may be expressed by simpleexponential functions - analogous to neutron capture andradiative decay - for a constant steady-state fluence.

N(t) a0 (E) a (Z) (1 - e tI ); t < tw (5)

N (t) a N (tw) e-t/'; t > tw (6)

where N(t) = number of electrons in trap at energy ETp(E) = energy fluence of incident radiationa(E) = energy-dependent trapping cross sectionT = characteristic time constant of the traptw. = radiation pulse width

N (t) = number of traps filled at t=t.

Of course photocurrents will also be generated due to thehigh dose rates. For high electron mobility devices such asGaAs, one need not consider the diffusion component ofphotocurrent, but only the drift component (i.e. neglectradiation storage time). Then the photocurrent component willmirror the radiation pulse P(t) giving the total transistorresponse R(t) as,

R(t) = N(t) + C P(t) t < t, (7)

R(t) N(t) t > tw (8)

where C = proportionality constant

Now, consider the case where more than one trap is filled,each having its own characteristic time constant Tr then

R(t) a C P(t) + p (E) a (E) ZE (l-exp (-t/r1 ); t < t, (9)

R(t) cx N (tw) Ej exp (-t/r1 ); t > t, (10)

To simplify the above expressions, consider the cases ofthree traps with time constants of 10 ns, 1 ps and 100 gs (oneshould note that the cited literature has observed trapping timeconstants from a few ns to 70 s). As an "idealized" response wefirst consider C vanishingly small, and thus the first term in

-7-

eq'n 9 to be insignificant when compared to the second. Forsimplicity we also set 0 = a = 1. The total idealized responsefor pulse widths of 10 ns, 0.5 As and 10 As is given in Fig. 6.Note the vast difference in recovery time for the three cases,dictating which trap levels may be observed.

Now, for a more realistic case, we allow the transistor to bedriven into saturation very quickly by the pulse and held thereuntil pulse cessation. Stating this another way, since there aremany more states in the conduction band than traps, thenphotocurrent generation will always dominate, obscuring trappingand allowing only (post pulse) de-trapping to be observed. Thismodification, for square wave pulses, results in the possibleresponses in Fig. 7 (Here the saturation levels have beenshifted to allow easier observation of the effects). Note thesharp drop-off immediately after the pulse followed byexponential decay. Fig. 8 (from (2)) shows this behaviourexperimentally, i.e. the device mimics the pulse until the valueof Nctw, is reached, when charge de-trapping occurs.

Fig. 9 show the response of the TOS710 MESFET to a 2 As wideLINAC pulse, as discussed later. Clearly contributions fromcharge de-trapping and the very fast GaAs photocurrent generationare observed.

Thus to predict the behaviour of GaAs devices on thebattlefield, it is imperative that an accurate pulse width beemployed in any simulator. However, to completely understand thenature and origin of trapping in GaAs, a variety of pulse widthsis necessary. A variable pulse width LINAC is essential then, tounderstand all possible reactions.

2.3.2 Temperature Dependence of Charge Trapping

The PHELA experiments conducted for this work also involvedan analysis of the temperature dependence of trapped charge decayin two GaAs MESFETS - the NEC 270 and TOS710. The temperaturewas controlled both during and after the radiation pulse by theuse of a TATS 400 (10) control system. A thermostat was placednear the device to achieve accurate readout. In order tominimize the number of interfering traps and stay within time

-8-

constraints, only one pulse width was used. This was the longestpossible (-2gs) at the closest possible distance (20cm) to thebeam exit window in order to maximize total dose and guaranteemaximum possible trapping.

Consider, for simplicity, only one single trap (or onedominant trap) corresponding to one single energy level at depthET below the conduction band. Then one may write the post-irradiation emission probability, p, from the ET trap (at anytemperature T) as

p = a Vth N, exp [ - ET/kT] (11)

where a = capture cross section

k = Boltzman's constant

Vth = thermal velocity of electrons

and N, = effective density of electron states in theconduction band

The associated emission lifetime from ET is simply

T = /p = [a vth N] -1 exp (ET/kT) (12)

If the constants in square brackets were completelytemperature independent then a semi-log plot of i vs I/T wouldyield ET. However, only the trapping cross-section satisfies thiscriterion. The others may be evaluated from (6)

Vth = [3kT/m]1 /2 and N, = 2/h3 (27rm kT) 3/2 (13)

where m = electron mass

h = Planck's constant

Substituting into eq'n 12 then the result is

T2 r = (1/aA) exp [ET/kT] (14)

-9-

where A = temperature independent constant - 54.6 k'm/h 3

Thus plotting (semi-log) T2 T vs l/T will give the trappingdepth ET from the slope, and a from the intercept.

This analysis was used on the data from the PHELA work. Thetemperature dependence of the emission lifetime is clearlyevident in Fig. 10 for the NEC 270 device. Fig. (11) shows theT 2T plots for the two devices. From the slopes, the evaluatedtrap depths are ET = (0.3 ± .02) eV for the NEC 270 device and ET= (0.09 ± .02) eV for the TOS710 device. Bellem (8) has doneextensive investigations on charge trapping in GaAs CCD devices,and has reported trap levels at 0.36 eV (labelled E3 or EL5) and0.1 eV (labelled El), in good agreement with the values reportedhere.

The associated cross section analysis would yield extremelylow values of - 10-19 cm2 for both devices.

The room temperature de-trapping time constants for the two

devices were 330 As for the NEC 270 and 105 As for the TOS710.

3.0 CONCLUSIONS

The use of accurate pulse widths to simulate the promptcomponent of a nuclear weapon burst has proven to be of greatimportance for a number of devices. It is suggested that anyDREO - based machine have the capabilities to span a wide rangeof pulse widths, with a 20 ns wide pulse being of primeimportance. A span of 10 ns to 10 As would be extremely useful,provided the same dose rate is available for all widths.

- 10 -

t

RADIATIONPULSE

SATURATED

CURRENT t p d +tr +tsr+tf

TRANSISTORRESPONSE

t trtd tr

Figure 1 Propagation delay time characteristicsof a bipolar transistor showing thecomponents described in the text

- 11 -

170

zDi 160 -

* 1 i0-

140

ft130

120

110 I

(T,-m5. oE-, )11ML,

Figure 2 Measured response of 2N4401 transistorto 80 ns wide pulse. The radiationstorage time and fall time parametersare evident

210

200-

170

Iea

0

T50

140 -

130 '

0 0.2 0.4 0.A 0.8 1 1.2 1.4 1.8 1.8

Figure 3 Measured response of TIP122 to 350 nswide pulse. Note that tar and tf aregreater than for the 2N4401. Theincrease in tar was expected due tolarge electrical storage time

- 1.2 -

43t)

20

210 -i

200

190L

+U U

4-

14 4-40 +

LJ7 +

120 4-La +

+

92 4

70-

(Tim I OE-5)TIME (sec)

84 ns pulse + 2Buo M, pulse

Figure 4 Measured response of 741C operationalamplifiers to 84 ns and 2.5 gs widepulses, showing wide voltage swings,varying with pulse width

-13

ClOZ co 'O

4.)r- 4-)C/ *.4 Wn

%emoo 4-) t 0a 4) 0..I0L + 0 * w W41

4- + WS-iWcn0 .)W 4~)

0 0 0 * 0 ) 4a) -4 'd 0)U

-)4.) ((S

+ -r4 4i (d

0 -)

a4r. z t

,. H

I--sw~ 0l cc~J1 E.-OOG (A

-14 -

0

0

0A

a))(p4) 0)

0 90(f > U))too wU

- V 0 4-4O4~ ~ ~ ~ -- -c --- M v 4

oE ) fk4-4r-4J

0 '10 ti 1 -

r2'4 wI 41) .-4f* ( r 4-) WG: .0 (0 4-) Q)

n 410 In 0

/- r-40 (f 4

000

U)~~ to) 4) C4) -t-) 0CO~ ~ ~ > 9) 9 14

qmr 0-1 4e:w0L~w~0 w.6 O~ HWL4i) 4J.Ur-4:c4-iwe

- 15 -

01

0a')

C!Cu>'0 LU o-oS(a 4-4> :

Co o. 0 0) n 0,, 0 1o I

co a L 0 Lo 0.c)

4-0. 4) 0r.0 r. (a

oo7I .4 4J J4

-1 41 G 0

I .4 -H J I J-,I.-.,0.4 .W 0 . -

Vw 00

n It0 w

' l l,-4 -r- - , ---

--------- 4* W4J

co> >I100Mimi~ M o4t

o 0 -4 4)0r

1H &I -S 0' I4 4

Q4w~ a"00

~WCOLL4-) -HOOC6 M 4J -H -A

- 16 -

r.Hi izomv S A Zus -46.SmV EXTiCHZ SOrnV50OHM

Cfitqndf- Z. -i 1 H

T

HI

0

k - cI T _ ---- CHZ

CIZ gn&{- .

Figure 8 Response (lower curve) of NEC270 GaAsMESFET 3 pss wide pulse (upper curve). Notethe similarity of the response to thatpredicted - i.e. a saturation which mimics

the pulse, followed by an exponential decaydue to charge trapping

-17 -

190

170-

160-

LIMz 10

0 150 -

140

so1,0

120

S 110

100

90

80-

70 ,0 5 l- I

(Tnmes IOE-6)TIME(sec)

Figure 9 Response of TOS71O MESFET to 2 gs widepulse clearly showing superposition ofthe two effect discussed in the text

180 - -

170 -

iso

140

13o

120 -

1 10| I I I I I I I I I I

0000 19 0.00021 000023 000026 000u27 0 00029 0.00031 0.00033 U U0036 U.00037 0.00039

0 T-300K + T-333K

Figure 10 Response of NEC270 MESFET at twodifferent temperatures, 300 K and 333K.The enhanced de-trapping at highertemperatures is readily apparent

-18-

0 0 (

co M MM

0D 04a ,>,flQ)

o rotir. Q.

o coc-0IZW

Z 04 ( 4-

0) . (n 4-U)W

wCL

w '

I-- 2N -le oa-

- 19 -

4.0 REFERENCES

1) Harrity J.W. and P.E. Gammill "Upset and Latchup Thresholdsin CD 4000 Series of CMOS devices", IEEE Annual Conference onNuclear and Space Radiation Effects, Cornell University, Ithaca,N.Y., July 1980.

2) Cousins T. and B.E. Hoffarth, "The Measured Response ofSelected Electronic Components to Fast Pulse Electrons (U)", DREOTechnical Note 88-13, May 1988 (confidential).

3) IRT Corporation Report #4251-009, "Guidelines for RadiationEffects Testing", IRT Corporation, 1978.

4) Cousins T. and R.A. Gravelle, DREO Trip Report "Visit toChalk River Nuclear Laboratories PHELA Facility, 02-03 February1989", 14 February 1989.

5) Cousins T. "The Use of the Computer Code ATR to Relate DREOExperimental Results to Nuclear Battlefield Threats", DREOTechnical Note (in press).

6) Messenger G.C. and M.S. Ash, "The Effects of Radiation onElectronic Systems", Von Nostrand Reinhold Company, N.Y., 1986.

7) Simons M. and E.E. King, "Long Term Radiation Transients inGaAs FETS", IEEE Trans. Nuc. Sci., Vol NS 26, No. 6, 1979, pp.5080--5086.

8) Bellem R.D. and W.C. Jenkins, "Radiation Effects on GaAscharge coupled devices with high resistivity gate structures",IEEE Trans. Nuc. Sci., Vol NS-33, No. 4, 1986, pp. 1084-1089.

9) MacLean F.B. and T.R. Oldham "Total Dose Ionization Effects"IEEE Trans. Nuc. Sci. Short Course Notes, 1987.

10) Tausch H.J., R. Wemhoner, R.L. Pease, J.R. Schworl: and R.J.Maier, "A Programmable Test System for Transient AnnealingCharacterization of Irradiated Mosfets" IEEE Trans. Nuc. Sci.,Vol NS-34, No. 6, 1987 pp. 1763-1768.

UNCLASSIFIED -21-

SECURITY CLASSIFICATION OF FORM(highest classification of Title. Abstract, Keywords)

DOCUMENT CONTROL DATAiSecurity classification of title, body of abstract and indexing annotation must be entered when the overall document is classified)

1. ORIGINATOR (the name and address of the organization preparing the document. 2. SECURITY CLASSIFICATIONOrganizations for whom the document was prepared, e.g. Establishment sponsoring (overall security classification of the document.a contractor's report, or tasking agency, are entered in section 8.) including special warning terms if applicable)

Defence Research Establishment OttawaOttawa, Ontario UNCLASSIFIED

KIA OZ4

3. TITLE (the complete document title as indicated on the title page. Its classification should be indicated by the appropriateabhreviation (S.C.R or U) in parentheses after the title.)

Pulse-Width Dependent Radiation Effects on Electronic Components (U)

4. AUTHORS (Last name, first name, middle initial)

Cousins, Thomas

5. DATE OF PUBLICATION (month and year of publication of 6a. NO OF PAGES (total 6b NO o REFS (total cited indocument) containing information. Include document)

October 1989 Annexes, Appendices, etc.)

19 107. DESCRIPTIVE NOTES (the category of the document, e.g. technical report, technical note or memorandum. If appropriate, enter the type of

report, e.g interim, progress, summary, annual or final. Give the inclusive dates when a specific reporting period is covered.)

Technical Note

8 SPONSORING ACTIVITY (the name of the department project office or laboratory sponsoring the research and development. Include theaddress.)

Defence Research Establishment Ottawa

Ottawa, Ontario

KlA OZ49a. PROJECT OR GRANT NO. (if appropriate, the applicable research 9b. CONTRACT NO. (if appropriate, the applicable number under

and development project or grant number under which the document which the document was written)was written. Please specify whether project or grant)

041LS

10a. ORIGINATOR'S DOCUMENT NUMBER (the official document 10b. OTHER DOCUMENT NOS. (Any other numbers which maynumber by which the document is identified by the originating be assigned this document either by the originator or by theactivity. This number must be unique to this document.) sponsor)

DREO TECHNICAL NOTE 89-33

1 1. DOCUMENT AVAILABILITY (any limitations on further dissemination of the document, other than those imposed by security classification)

IX) Unlimited distribution

( ) Distribution limited to defence departments and defence contractors; further distribution only as approvedI ) Distribution limited to defence departments and Canadian defence contractors; further distribution only as approved

I Distribution limited to government departments and agencies; further distribution only as approved) Distribution limited to defence departments; further distribution only as approved

I Other (please specify):

12. DOCUMENT ANNOUNCEMENT (any limitation to the bibliographic announcement of this document. This will normally correspond tothe Document Availabilty (11). However, where further distribution (beyond the audience specified in 11) is possible, a widerannouncement audience may be selected.)

UNCLASSIFIED

SECURITY CLASSIFICATION OF FORM

-22- UNCLASSIFIED

SECURITY CLASSIFICATION OF FORM

13. ABSTRACT ( a brief and factual summary of the document. It may also appear elsewhere in the body of the document itself. It is highlydesirable that the abstract of classified documents be unclassified. Each paragraph of the abstract shall begin with an indication of thesecurity classification of the information in the paragraph (unless the document itself is unclassified) represented as (S), (C), (R). or (U).It is not necessary to include here abstracts in both offical languages unless the text is bilingual).

-The simulation of the prompt gamma-ray pulse effects on electroaicb with au

electron linear accelerator (LINAC) has been performed by DREO and other groups.

However, the use of a LINAC normally entails wider pulses than those expected on

the battlefield. This report examines the effects of the variation of pulse width

on electronic response using both theoretical and experimental examples. The con-

clusions are that pulse-width fidelity is important for a number of possible

scenarios, and that for complete understanding of electronic performance, a vari-

able pulse-width simulator is essential.,

1 4. KEYWORDS. DESCRIPTORS or IDENTIFIERS (technically meaningful terms or short phrases that characterize a document and could behelpful in cataloguing the document. They should be selected so that no security classification is required. Identifiers, such as equipmentmodel designation, trade name, military prolect code name, geographic location may also be included. If possible keywords should be selectedfrom a published thesaurus. e.g. Thesaurus of Engineering and Scientific Terms (TEST) and that thesaurus-identified. If it is not possible toselect indexing terms which are Unclassified, the classification of each should be indicated as with the title.)

Alransient Radiation Effects on Electronics-

Electron Linear AcceleratorRadiation Dose Rate*

Electron Pulse Width-Biopolar Transistors

Op-Amp-

Gallium Arsenide

Charge Trapping

UNCLASSIFIED

SECURITY CLASSIFICATION OF FORM