1、Designation: F 744M 97 (Reapproved 2003)METRICStandard Test Method forMeasuring Dose Rate Threshold for Upset of DigitalIntegrated Circuits Metric1This standard is issued under the fixed designation F 744M; the number immediately following the designation indicates the year oforiginal adoption or, i
2、n the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. Asuperscript epsilon (e) indicates an editorial change since the last revision or reapproval.1. Scope1.1 This test method covers the measurement of the thresh-old level of radiation dose
3、 rate that causes upset in digitalintegrated circuits under static operating conditions. The radia-tion source is either a flash X-ray machine (FXR) or an electronlinear accelerator (LINAC).1.2 The precision of the measurement depends on thehomogeneity of the radiation field and on the precision of
4、theradiation dosimetry and the recording instrumentation.1.3 The test may be destructive either for further tests or forpurposes other than this test if the integrated circuit beingtested absorbs a total radiation dose exceeding some predeter-mined level. Because this level depends both on the kind
5、ofintegrated circuit and on the application, a specific value mustbe agreed upon by the parties to the test (6.8).1.4 Setup, calibration, and test circuit evaluation proceduresare included in this test method.1.5 Procedures for lot qualification and sampling are notincluded in this test method.1.6 B
6、ecause of the variability of the response of differentdevice types, the initial dose rate for any specific test is notgiven in this test method but must be agreed upon by the partiesto the test.1.7 This standard does not purport to address all of thesafety concerns, if any, associated with its use.
7、It is theresponsibility of the user of this standard to establish appro-priate safety and health practices and determine the applica-bility of regulatory limitations prior to use.2. Referenced Documents2.1 ASTM Standards:E 665 Practice for Determining Absorbed Dose VersusDepth in Materials Exposed t
8、o the X-Ray Output of FlashX-Ray Machines2E 666 Practice for Calculating Absorbed Dose from Gammaor X Radiation2E 668 Practice for Application of Thermoluminescence-Dosimetry (TLD) Systems for Determining Absorbed Dosein Radiation-Hardness Testing of Electronic Devices2F 526 Test Method for Measurin
9、g Dose for Use in LinearAccelerator Pulsed Radiation Effects Tests33. Terminology3.1 Definitions of Terms Specific to This Standard:3.1.1 determined integrated circuitintegrated circuitwhose output is a unique function of the inputs; the outputchanges if and only if the input changes (for example, A
10、ND-and OR-gates).3.1.2 dose rateenergy absorbed per unit time and per unitmass by a given material from the radiation to which it isexposed.3.1.3 dose rate threshold for upsetminimum dose rate thatcauses either: (1) the instantaneous output voltage of anoperating digital integrated circuit to be gre
11、ater than thespecified maximum LOW value (for a LOW output level) orless than the specified minimum HIGH value (for a HIGHoutput level), or (2) a change of state of any stored data.3.1.4 nondetermined integrated circuitintegrated circuitwhose output or internal operating conditions are not uniquefun
12、ctions of the inputs (for example, flip-flops, shift registers,and RAMs).4. Summary of Test Method4.1 The test device and suitable dosimeters are irradiated byeither an FXR or a LINAC. The test device is operating butunder static conditions. The output(s) of the test device and ofthe dosimeters are
13、recorded.4.2 The dose rate is varied to determine the rate whichresults in upset of the test device.4.3 For the purposes of this test method, upset is consideredto be either of the following:4.3.1 An output voltage transient exceeding a predeter-mined value, or1This test method is under the jurisdic
14、tion of ASTM Committee F01 onElectronics and is the direct responsibility of Subcommittee F01.11 on Quality andHardness Assurance.Current edition approved Feb. 10, 1997. Published April 1997. Originallypublished as F 744 81. Last previous edition F 744 92.2Annual Book of ASTM Standards, Vol 12.02.3A
15、nnual Book of ASTM Standards, Vol 10.04.1Copyright ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.4.3.2 For devices having output logic levels which are notunique functions of the input logic levels, such as flip-flops, achange in the logic st
16、ate of an output.4.3.3 For nondetermined integrated circuits, a change ofstate of an internal storage element or node.4.4 A number of factors are not defined in this test method,and must be agreed upon beforehand by the parties to the test:4.4.1 Total dose limit (see 1.3),4.4.2 Transient values defi
17、ning an upset (see 4.3.1),4.4.3 Temperature at which the test is to be performed (see6.7),4.4.4 Details of the test circuit, including output loading,power supply levels, and other operating conditions (see 7.4,10.3, and 10.4),4.4.5 Choice of radiation pulse source (see 7.7),4.4.6 Radiation pulse wi
18、dth (see 7.7.2),4.4.7 Sampling (see 8.1),4.4.8 Need for total dose measurement (see 6.8, 7.6, and10.1),4.4.9 Desired precision of the upset threshold (see 10.8),and4.4.10 Initial dose rate (see 1.6 and 10.5).5. Significance and Use5.1 Digital integrated circuits are specified to operate withtheir in
19、puts and outputs in either a logical 1 or a logical 0 state.The occurrence of signals having voltage levels not meetingthe specifications of either of these levels (an upset condition)may cause the generation and propagation of erroneous data ina digital system.5.2 Knowledge of the radiation dose ra
20、te that causes upsetin digital integrated circuits is essential for the design, produc-tion, and maintenance of electronic systems that are required tooperate in the presence of pulsed radiation environments.6. Interferences6.1 Air IonizationA spurious component of the signalmeasured during a test c
21、an result from conduction through airionized by the radiation pulse. The source of such spuriouscontributions can be checked by measuring the signal whileirradiating the test fixture in the absence of a test device. Airionization contributions to the observed signal are generallyproportional to the
22、applied field, while those due to secondaryemission effects (6.2) are not. The effects of air ionizationexternal to the device may be minimized by coating exposedleads with a thick layer of paraffin, silicone rubber, or noncon-ductive enamel or by making the measurement in a vacuum.6.2 Secondary Emi
23、ssion4Another spurious component ofthe measured signal can result from charge emission from, orcharge injection into, the test device and test circuit. This maybe minimized by shielding the surrounding circuitry andirradiating only the minimum area necessary to ensure irradia-tion of the test device
24、. Reasonable estimates of the magnitudeto be expected of current resulting from secondary-emissioneffects can be made based on the area of metallic targetmaterials irradiated (see Note 1). Values generally rangebetween 1011and 1010As/cm2Gy, but the use of a scatterplate with an intense beam may incr
25、ease this current (7.7.2).NOTE 1For dose rates in excess of 108Gy(Si)/s, the photocurrentsdeveloped by the package may dominate the device photocurrent. Careshould be taken in the interpretation of the measured photoresponse forthese high dose rates.6.3 OrientationThe effective dose to a semiconduct
26、orjunction can be altered by changing the orientation of the testdevice with respect to the irradiating beam. Most integratedcircuits may be considered “thin samples” (in terms of therange of the radiation). However, some devices may havecooling studs or thick-walled cases that can act to scatter th
27、eincident beam, thereby modifying the dose received by thesemiconductor chip. Care must be taken in the positioning ofsuch devices.6.4 Dose EnhancementHigh atomic number materialsnear the active regions of the integrated circuit (package,metallization, die attach materials, etc.) can cause an enhanc
28、eddose to be delivered to the sensitive regions of the device whenit is irradiated with bremsstrahlung. Therefore, when an FXR isused as the radiation source, calculations should be performedto determine the possibility and extent of this effect.6.5 Electrical NoiseSince radiation test facilities ar
29、e in-herent sources of r-f electrical noise, good noise-minimizingtechniques such as single-point ground, filtered d-c supplylines, etc., must be used in these measurements.6.6 TemperatureDevice characteristics are dependent onjunction temperature; hence, the temperature of the test shouldbe control
30、led. Unless the parties to the test agree otherwise,measurements shall be made at room temperature (23 6 5C).6.7 Beam Homogeneity and Pulse-to-Pulse RepeatabilityThe intensity of a beam from an FXR or a LINAC is likely tovary across its cross section. Since the pulse-shape monitor isplaced at a diff
31、erent location than the device under test, themeasured dose rate may be different from the dose rate towhich the device was exposed. The spatial distribution andintensity of the beam may also vary from pulse to pulse. Thebeam homogeneity and pulse-to-pulse repeatability associatedwith a particular r
32、adiation source should be established by athorough characterization of its beam prior to performing ameasurement.6.8 Total DoseEach pulse of the radiation source impartsa dose to both the device under test and the device used fordosimetry. The total dose deposited in a semiconductor devicecan change
33、 its operating characteristics. As a result, theresponse that is measured after several pulses may be differentfrom that characteristic of an unirradiated device. Care shouldbe exercised to ensure that the total dose delivered to the testdevice is less than the agreed-upon maximum value. Care mustal
34、so be taken to ensure that the characteristics of the dosimeterhave not changed due to the accumulated dose.7. Apparatus7.1 Regulated d-c Power Supplies, with floating outputs toproduce the voltages required to bias the integrated circuitunder test.4Sawyer, J. A., and van Lint, V. A. J., “Calculatio
35、ns of High-Energy SecondaryElectron Emission,” Journal of Applied Physics, Vol 35, No. 6, June 1964, pp.17061711.F 744M 97 (2003)27.2 Recording DevicesA single dual-beam, or two single-beam oscilloscopes, equipped with cameras; or transient digi-tizers with appropriate displays. The bandwidth capabi
36、lities ofthe recording devices shall be such that the radiation responsesof the integrated circuit and the pulse-shape monitor (7.6) areaccurately displayed and recorded.7.3 Cabling, to complete adequately the connection of thetest circuit in the exposure area with the power supply andoscilloscopes
37、in the data area. Shielded twisted pair or coaxialcables may be used to connect the power supplies to the biaspoints of the test circuit; however, coaxial cables properlyterminated at the oscilloscope input are required for the signalleads.7.4 Test Circuit (see Fig. 1)Although the details of testcir
38、cuits for this test must vary depending on the kind ofintegrated circuit to be tested and on the specific parameters ofthe circuit which are to be measured, Fig. 1 provides theinformation necessary for the design of a test circuit for mostpurposes. The capacitor, C, provides an instantaneous source
39、ofcurrent as may be required by the integrated circuit during theradiation pulse. Its value must be large enough that thedecrease in the supply voltage during a pulse is less than 10 %.The capacitor, C, should be paralleled by a small (approxi-mately 0.01 F) low-inductance capacitor to ensure thatpo
40、ssible inductive effects of the large capacitor are offset. Bothcapacitors must be located as close to the integrated circuitsocket as possible, consistent with the space needed forconnection of the current transformer and for any shielding thatmay be necessary. The switch, S, provides means to plac
41、e theoutput of the integrated circuit (here a NAND gate) in either alogic LOW or a logic HIGH state. The arrangement of thegrounding connections provides that only one ground exists, atthe point of measurement. This eliminates the possibility ofground loops and reduces the common-mode signals presen
42、t atthe terminals of the measurement instruments. The resistor, R0,is the termination for the coaxial cable and has a value within2 % of the characteristic cable impedance. All unused inputs tothe test device are connected as agreed upon between theparties to the test. The output of the test device
43、may be loaded,as agreed upon between the parties to the test. To preventloading of the output of the test device by the coaxial cable,one may use a line driver that has a high input impedance andadequate bandwidth and voltage swing to reproduce accuratelyat the output end of the coaxial cable, the w
44、aveforms appearingat the line-driver input.7.5 Radiation Pulse-Shape MonitorUse one of the fol-lowing to develop a signal proportional to the dose ratedelivered to the test device:7.5.1 Fast Signal Diode, in the circuit configuration of Fig.2. The resistors, R1, serve as high-frequency isolation and
45、 mustbe at least 20 V. The capacitor, C, supplies the charge duringthe current transient; its value must be large enough that thedecrease in voltage during a current pulse is less than 10 %.The capacitor, C, should be paralleled by a small (approxi-mately 0.01 F) low-inductance capacitor to ensure t
46、hatpossible inductive effects of the large capacitor are offset. Theresistor, R0, is to provide the proper termination (within 62%)for the coaxial cable used for the signal lead. This is thepreferred apparatus for this purpose.7.5.2 P-I-N Diode, in the circuit configuration of Fig. 2 asdescribed in
47、7.5.1. Care should be taken to avoid saturationeffects.7.5.3 Current Transformer, mounted on a collimator at theoutput window of the linear accelerator so that the primaryelectron beam passes through the opening of the transformerafter passing through the collimator. The current transformermust have
48、 a bandwidth sufficient to ensure that the currentsignal is accurately displayed. Rise time must be less than 10 %of the pulse width of the radiation pulse being used. The lowfrequency cutoff of some commercial current transformers isFIG. 1 Test Circuit Example for a NAND GateF 744M 97 (2003)3such t
49、hat significant droop may occur for pulse widths greaterthan 1 s. Do not use a transformer for which this droop isgreater than 5 % for the radiation pulse width used. Whenmonitoring large currents, care must be taken that the current-time saturation rating of the current transformer is not ex-ceeded. It may be required that the signal cable monitoring thecurrent transformer be matched to the characteristic impedanceof the transformer, in which case R0would have this imped-ance (within 62 %), as specified by the manufacturer of thecurrent transformer.NOTE
