RU2578053C1 - Method for evaluating resistance of digital electronic equipment to ionising radiation (versions) - Google Patents

Abstract

FIELD: instrument making.

SUBSTANCE: invention relates to investigation of radiation resistance of semiconductor devices (SCD) and integrated circuits, and more of integrated circuits (IC) with series and combinational processing of logic signals.

EFFECT: this invention consists in fact that by comparison and conversion of various data by resistance to dose effects with static or pulse irradiation on simulating units (SU), by resistance to effects of dose rate at pulse irradiation on SU, at resistance to effects of dose rate or during simulation, by resistance to low-intensity space radiation (SR) factors and according to specifications of dynamic parameters of SCD and IC based on general concept of generation of critical charge in volume ν_S, causing SEU and SET effects in digital electronic circuits.

27 cl, 16 dwg, 12 tbl

Description

The invention relates to the field of radiation resistance research of semiconductor devices (IFP) and integrated circuits, and to a greater extent integrated circuits (ICs) with sequential and combinational processing of logical signals. It refers to experimental studies carried out both on simulations of the conditions of exposure to ionizing radiation (II) installations that directly reproduce their types, called simulator installations (MI), and on installations that quantitatively reproduce similar radiation effects, called simulator installations (IE). Among the latter, for example, include pulsed laser sources.

In the modern electronics industry, IC developers have dramatic experience of increasing both the density of transistor structures and the operating frequency. One of the significant reasons for this growth is a decrease in the topological size of the devices included in the IC. Topological size refers to the transistor gate length of the Complementary Metal Oxide Semiconductor (CMOS) technology.

The invention relates, inter alia, to ICs suitable for use in outer space (KP) conditions, such as satellites, interplanetary flights or a working space shuttle. The subject of discussion is the adverse effect of KP factors on IC chips, which is manifested, firstly, in the effects of the total absorbed dose from the action of AI (Total Ionizing Doze = TID) due to the inhibition of electrons and protons, and, secondly, in the effects of single failures, or failures from single particles, (Single Event Upset = SEU) due to interaction with cosmic rays (heavy charged particles (TZZ) high energy), protons and high-energy neutrons. Of all these effects, SEUs present the greatest radiation-induced hazard that must be avoided for microelectronic objects in the control room. The effects of energy loss, charge collection, and failures due to the passage of cosmic rays of semiconductor structures in the IC are considered. If a high-energy ion passes through all materials, it loses energy due to interaction with materials. The ion energy is lost, firstly, due to the interaction of the ion with the orbital electrons of the material, on the ionization of the material and the formation of the track of electron-hole pairs (ehp). The speed at which an ion loses energy is historically called the “braking power” (dE / dx). The differential energy dE is measured in units of [MeV], while the thickness of the material is measured in units of mass thickness [mg / cm ² ]. The term LET (linear energy transfer) is more often used instead of “braking power”. In silicon, an energy of 3.6 eV is required to create one eph. At a silicon density of 2.42 g / cm ^3, one micron of linear thickness is converted to 0.242 mg / cm ² mass. Also, the electron charge is 1.6.10 ^-7 pC. Hence, for the formation of a charge dQ in silicon in a track of length L, when the particle is decelerated, the "deceleration power" LET forms a charge:

Thus, an ion with a LET of the order of 100 MeV-cm ² / mg creates a charge of about 1 pC along its track.

In bulk silicon, eph are not significant because they recombine after all. In the presence of an electric field, however, the ephs are fairly quickly separated by drift in the opposite direction to the field and can be quickly collected by some voltage source that creates the field. In bulk silicon ICs, such an electric field is applied along the pn junction in devices. Each and any of the signals at the output of the IC is usually isolated from the source V _{DD by} one or more such transitions (drain of the p-MOS transistor) and isolated from the common point (“ground”) V _{SS by} one or more such transitions (drain of the transistor n- MOS).

FIG. 1 illustrates the effect of an ion passing through a drain transition in an n-MOS transistor. FIG. 1 shows a cross section of a drain transition (10) having an n ⁺ drain of n-MOS (12) diffusing into a p ^- (epi) film (14) of a p ⁺ substrate ((16) n-MOS). Transition (20) isolates the circuit output for a positive power supply (+ V) from the substrate voltage (V _SS ). An ion flies through transition (10) and creates a track (20) of electron – hole pairs, which correspond to the “-” symbol for electrons and “+” for holes. Electron-hole pairs, in turn, create a plasma whose contour resembles a charge funnel (22).

The indicated current components are observed as signals at the IC terminals in the presence of an electric field at the junction and funnels of separated free charges of electrons and holes. Based on the design of FIG. 1, the electrons are collected on the circuit node, and the holes on the substrate node, as a result, a negative current pulse flows in the diffusion node of the n-MOSFET, which leads to a “discharge” (decrease) in the signal voltage.

This additional current is short-lived, ending in about 100-200 ps.

The delayed component of the current is generated due to the diffusion of electrons and holes from the region where the electric field is equal to zero. These charges could, in the absence of recombination, reach the region where the field exists and where they can recombine. This delayed component could reach a duration of several hundred nanoseconds.

A small charge is collected in the region of the p ⁺ substrate, since the recombination rate is higher here due to the higher concentration of the impurity. If the supply voltage at the drain of the n-MOSFET is zero, the electric field is essentially zero and there is no significant increase in the collected charge.

Similar processes are observed in the vicinity of the p-MOS drain formed by diffusion of the impurity into the n-pocket and shifted by the V _DD source. In this case, an electric field is present, even if the voltage of the bias source is zero. The thickness of the collection area in this case is much smaller (at best half the thickness of the pocket), because the pocket-substrate transition is constantly biased in the opposite direction and can also collect an induced charge.

High-energy neutrons and protons are known to cause similar effects directly in an indirect way through nuclear reactions with silicon. In this case, heavy recoil ions of the accompanying products fly through the transitions and form a current pulse from the collected charge. In the CP, high-energy protons are generated mainly from protons captured by radiation belts and from solar flares. For high-altitude aviation, both high-energy neutrons and protons are associated with the products of reactions formed in space showers, when space TZZ undergo nuclear reactions in the atmosphere.

. Это является верным только в предположении, что схема очень слабо рассеивает заряд в течение последующих сотен пикосекунд). Высокая проводимость транзисторов смягчает этот эффект, так как они рассеивают этот заряд более быстро. Более важная для триггеров и SRAM положительная петля обратной связи по усилению обуславливает «опрокидывание» данных (

) в том случае, когда собранный заряд достигает критического значения Q_CRIT, а напряжение на узле питания возрастает до критического значения. Более детальное обсуждение этих эффектов и, в частности, откликов схем SRAM, можно найти в [1].The components of the additional external current described above, responsible for the effects of SEUs, have been observed in space orbit circuits over the past 10-15 years. More remarkable is that SEUs are found in static triggers (latchs) and SRAM (Static Random Acces Memory). The effect of the appearance of a quasi-stationary current is determined by the dependence of the reaction on the circuit, which, in turn, is determined by the amount of charge collected in the signal node. Essentially, the value of the signal node capacitance determines (in the first case) how large the voltage drop dV is, which is the result of the collected charge dQ in accordance with the equality

. This is true only under the assumption that the circuit dissipates the charge very weakly over the next hundreds of picoseconds). The high conductivity of the transistors mitigates this effect, since they dissipate this charge more quickly. The gain feedback loop, which is more important for triggers and SRAM, causes data to tip over (

) in the case when the collected charge reaches a critical value of Q _CRIT , and the voltage at the power node rises to a critical value. A more detailed discussion of these effects and, in particular, the responses of SRAM schemes can be found in [1].

SEUs become more problematic with a further decrease in topological size. The topological size of the next generations of non-radiation-resistant commercial IMS has decreased in recent years from 1 μm to 0.13 μm (at present) and is continuously decreasing to promising 0.05 μm. To achieve this reduction, several scaling models were used in industry, including lateral (side) scaling, where only the shutter length is scaled, scaling of the DC supply voltage when V _DD remains constant, and scaling of the constant electric field, where V _DD decreases with decreasing thickness of the gate oxide to maintain a constant electric field in the transistor. The electric field scaling model is a proven fact for most practical applications, since it is suitable for describing most of the negative effects of a large electric field (breakdown of a gate insulator and “hot” electrons).

According to [2], the duration of transition pulses SET (Single Event Transient) and the cross sections of errors as a function of technological solutions (0.25, 0.18 and 0.13 μm) and as a function of operating voltage (2.00, 1.50, 1.25 and 1.00 V) at a fixed value of LET = 60 MeV cm ² / mg show a significant increase in the SET pulse duration with a reduction in the topological size and a concomitant decrease in the rated operating voltage and a significant increase in the SET pulse duration, if the devices operate at voltage which is lower than the lower permissible boundary of the workers s specifications for each of the considered technologies. It also contains the results of studies of traditional DICE memory cells (Dual Interlocked Storage Cell = double interlocking memory cell) for checking compliance with the results of measurements according to the PDTL method (Programmable Delay Temporal Latch = PDTL, or programmable delay, or time “latch”) dependence of cross sections SET from voltage functioning. The results of measurements on DICE circuits, in general, correspond to the data according to the PDTL method, i.e. most transients are fixed when the nominal voltage decreases, which is interpreted in such a way that the efficiency of fixing the effects of SETs in DICE cells “improves” due to longer SET pulses. Based on the results of [2], it was determined that the SET pulse durations are a direct proportional function of the rated operating voltage and can potentially continuously increase with a decrease in topological dimensions due to a corresponding decrease in the rated operating voltage. This leads to a potential “double frontal battle” in the respective technologies, so that high operating frequencies are combined with the duration of the extended SET pulses. The progress in future devices towards SET suppression is associated with the use of technologies such as temporary “latch” [2] for bulk silicon (bulk / epi Si), silicon-on-insulator (Silicon-on-insulator = SOI) and / or the need to develop a framework for seriously considering the possibility of manufacturing low-intensity SET instruments suitable for application specific integrated circuits (Application Specific Integrated Circuit = ASIC).

. Самый маленький ток транзистора для постоянного масштабируемого поля требует, чтобы плотность тока металлизации (относительно тока электромигации) уменьшалась с меньшей скоростью, чем масштабирование по постоянному напряжению, при котором ток транзистора остается постоянным. Далее, для систем с малой мощностью, масштабирующих постоянное поле (в которых шкала V_DD пропорциональна), это есть только одна рассматриваемая возможность, так как это является, по существу (квадрат масштабирования), результатом малой мощности рассеивания.To scale the constant electric field, including all the physical dimensions of the device (such as gate length L, gate width W and gate oxide thickness T _OX ), when they are reduced, the applied supply voltage V _DD and the threshold voltage of the transistor V _{TH are} reduced proportionally. This results in a proportional decrease in the drain current (I _D ), load capacitance (C) and a proportional decrease in the capacitance of the gate assembly

. The smallest transistor current for a constant scalable field requires that the metallization current density (relative to the electromigration current) decrease at a lower rate than constant voltage scaling, at which the transistor current remains constant. Further, for systems with low power scaling a constant field (in which the V _DD scale is proportional), this is only one possibility considered, since this is essentially the (squared square) result of low power dissipation.

Orbital microelectronics usually lags behind its ground-based commercial duplicates by one or two generations, since most of the steps of the technological process require that the radiation resistance requirements for TID from KP emissions be met. Most radiation-resistant ICs can be made with a technological size of 0.8 ... 0.7 microns. SEUs in static triggers and SRAMs can occur in devices with topological norms less than 10 μm and for critical charge for failed circuits, below 1 pC (approximately for particles with LET no lower than 1 MeV-cm ² / mg). Since the transverse cross-sectional area of the TCD incident is 20 times smaller, the integrated cosmic ray fluence with LET> 1 MeV-cm ² / mg is 1000 times greater than the particle fluence with LET = 20 MeV-cm ² / mg for the geostationary orbit. This suggests that the error rate from the effects of SEU (by one bit) increases by 50 times. Since the 0.18 micron technology has probably 20 times more triggers than the 0.8 micron technology, the failure rate of an entire IC in this case is 1000 times greater.

To assess the resistance of SPP and IMS to the effects of AI KP, test methods are used on accelerators of charged particles, for example, protons, where, using the selection of targets, the necessary “cocktail” of TZZ is formed [3].

When using simulation methods (IM), picosecond pulsed lasers with a wavelength of 800–1500 nm are usually used [4].

Limitations regarding the use of IM based on pulsed laser irradiation relate to screening the surface of the IC chip by metallization and the need to accurately select the energy of the laser pulse to obtain the equality of the controlled output electrical characteristics of the IFR and IC.

However, pulsed laser exposure makes it possible to detect new effects that are not recorded during radiation experiments at the MU: this is the positioning of malfunctions on the surface of the IC chip and the delay in the development of the malfunction process in time associated with the operation of the clock generator [5, 6]. In this sense, laser MIs are important additional conditions for constructing the cross section dependences of SEU effects as a function of LET.

(средняя энергия спектра эталонного нуклида Co⁶⁰) и величиной порогового значения линейной передачи энергии LET_TH. При этом величина индуцированного радиацией критического заряда Q_C в треке иона и в чувствительном объеме транзисторной гетероструктуры технологии «металл-оксид-полупроводник» (МОП) остается равнойBased on the coulometric method of measuring electrophysical parameters [7] and the device that implements it [8], a method is proposed for determining the coefficient of relative efficiency and the equivalent dose of an x-ray source to obtain the coefficient of relative efficiency (COE), or in the foreign literature RDEF (Relative Dose Enhancement Factor) [9]. RDEF establishes the relationship between the absorbed equivalent dose of x-ray (gamma) radiation with equivalent energy

(average energy of the spectrum of the reference nuclide Co ⁶⁰ ) and the value of the threshold value of the linear energy transfer LET _TH . In this case, the value of the critical charge Q _C induced by radiation in the ion track and in the sensitive volume of the transistor heterostructure of the metal-oxide-semiconductor (MOS) technology remains equal

When simulating SEU effects with pulsed laser radiation, the concept of RDEF can also be introduced.

Then

Previous SEU troubleshooting solutions have focused on the development of SRAM and static triggers. Despite the fact that most of the works reviewed focused on static triggers for use in ASICs, nevertheless, these results can be extended to SRAM developments.

One of these triggers is described in [10]. These trigger designs utilize cross-isolation techniques to ensure that the state of the triggers does not change when an SLC hits any critical site.

Another development is presented in [11]. This development in the form of a DICE trigger also cannot change the logical state when a particle enters a unit node.

Each of these triggers can be switched, apparently, if a single particle of cosmic radiation penetrates the IC in a narrow layer located at an angle parallel to the surface and at the same time crosses at least two p-n junctions. The geometric cross section for this case, although small, can be very significant for applications in near-earth orbits.

Apart from the problem of generating SEUs in the gate and the substrate region, cosmic radiation can induce SETs processes in combinatorial logic, in the global time system and global control lines at the circuit level. SETs generates a minimum of effects in the technologies of 0.8 microns and 0.7 microns, because the propagation speed of the signals in these circuits is insufficient for propagation of SETs of 100 ... 200 ps duration at a noticeable distance within the circuit. However, the less important the technological dimensions of future developments (and their durability) for near-Earth applications, the more insignificant are the effects of SETs for normal signals.

In FIG. Figure 2 shows a graph of the dependence of the critical duration of the transition pulse required for propagation of SET without attenuation through an infinitely long chain of inverters on the technological size. If the pulse length is shorter than the critical pulse duration, the inertial inertial delay of the valve causes the transient single event to decay. The SET effect ends in the crystal after passing through five gates. If the pulse duration is equal to or greater than the critical duration, SET unambiguously propagates in the valves as a normal working signal. The effects of SETs, lasting longer than the critical duration, propagate through all valves without attenuation; SETs, whose duration is equal to half the critical duration, decay in the first valve; SETs with intermediate duration propagate through a variable number of steps.

The graph in FIG. 2 is the result of modeling using SPICE parameters for various topological norms (shown as dots on the graph) in the range from 1.2 μm (1200 nm) and 0.13 μm (130 nm). The generated series of parameters of the SPICE model was obtained based on the known model parameters between 1.2 μm and 0.7 μm, inclusive. The rules of conservation of the electric field were used to generate the model and the dimensions of the transistor to predict the parameters of the model in small topological dimensions. The scaled values of various critical parameters (V _DD , V _TH, and T _OX ) were compared with those published in [12]. The continuous curve contains simulated points, the dashed line extrapolated points to 0.05 μm (50 nm), the design topological norm of commercial technologies in 2012. As discussed above, the duration of SETs processes is 100 ... 200 ps. The graph in FIG. 2 confirms the fact that through one or two generations of orbital systems of microelectronic equipment, which use ICs with topological norms less than 0.35 μm, the effects of SETs will not attenuate within the circuit gates, but will propagate as normal working signals of the circuit. This will have serious, if not catastrophic, consequences for sequential access schemes.

In FIG. Figure 3 shows the layout topology for the sequential access scheme (30). Scheme (30) has a primary trigger (32), a combinational logic block (34), and a secondary trigger (36). In this illustration, triggers (32) and (36) implement the functions of D-triggers. The data of the first trigger (32) is usually fed to the input of the combinational logic along the trailing edge of the clock pulse, which is generated by the time synchronization circuit. From the output of combinatorial logic (34), the signal supports the input of the second trigger (36) for the entire time until the next falling edge of the synchronization pulse arrives. On the trailing edge of this pulse, the second trigger remembers any data received by this time at its input and performs a state change with the transition to standby mode.

If at this time TZC falls into combinatorial logic (34) and the logic is suitable for the propagation of the transient, then eventually SET appears at the input of the second trigger (36), which can be perceived as a useful (true) signal. Which of these two, useful or SET, will be remembered, depends on the time of their arrival and the trailing edge of the clock signal.

FIG. 4 is a timing chart illustrating a time relationship for the case where true data is ignored and positive SETs erroneously arrive at the input of the second trigger. In FIG. 5, the clock signal (50) and four different SET signals (52-58) illustrate four different cases where the SET signal may coincide with the trailing edge of the time signal (50).

The transient can be incorrectly interpreted as a useful signal and subsequently stored in the trigger if it is longer than the time period set for the switching time in front of the falling edge of the clock signal and the holding time after the falling edge of the clock pulse. In FIG. 5, the first of a series of SET signals (52) is observed before the indicated interval, and therefore cannot be switched to the second trigger. The next SET signal (58) is observed after this interval, and therefore again does not give a result in a change in the trigger state. However, the second and third signals SET (54) and (56) come with a certain lead or delay, respectively, with respect to the switching signal, which leads to erroneous storage of the SET signal in the second trigger.

Figure FIG. 6 shows another timing diagram illustrating a timing relationship for other types of SETs that cause incorrect data to be written to a trigger. In this case, SETs are observed within the clock. Figure FIG. 6 shows a data signal (60), a normal clock signal (62), and three different clock signals (64-68), which are distorted by SET (indicated by a dashed line for each case).

True data (60) satisfy the installation and maintenance times near the trailing edge of the clock pulse in case of its correct operation, as corresponds to signal (62). The clock signal (64) contains the negative contribution from the SET along the clock line, which shifts the trailing edge of the clock signal to the lead of the data signal. As a result, the low value “0” of the data signal (60) will be stored. The clock signal (66) shows an intermediate location of the “1” SET signal relative to the “1” data signal. The clock signal does not cause problems, since it arrived and acts after high-level data and can be remembered as long as the data signal remains at level “1”. The clock signal (68) contains a positive SET near the trailing edge of the data signal (60). The clock signal (68) causes the trigger to store a low level error signal as opposed to a previously stored high level signal. Note that combining SET with the trailing edge of the data signal is not a problem. The trigger data can be distorted in any state of the signal on the synchronization line when the trailing edge of the SET pulse follows the trailing edge of the data signal.

Different failure rates in direct access circuits (ie, SEUs in triggers and SETs in combinatorial logic) depend on the clock frequency [3]. A change in state is observed in triggers only when the frequency is low and the trigger is in a hold state. Because Since the clock signal corresponds to less than 50% of the total time, the intensity of the SEU is frequency independent. On the contrary, SETs that are created in combinatorial logic can be accumulated if they coincide with the trailing edge of the clock signal at some input of the trigger, and the number of such coincidences grows linearly with increasing clock frequency. Thus, the intensity of the SEU is independent of the frequency of the clock, while the intensity of the SET is directly proportional to the frequency of the clock. This correlation in error intensity can be illustrated experimentally using pulsed laser irradiation of test structures, when various error intensities are measured depending on the frequency of the clock signals [13].

This correlation of error intensity mixes SET problems with the problem of topological size compression in the IC technology. The result of compression of the topological size and reduction of the delay time in the gate allows the circuits to work with high frequencies of the clock signals. Not only does each combinatorial valve supply the contribution of transient errors to the whole circuit (as long as the transients do not die out), but the likelihood of accumulation also increases (due to the higher clock frequency).

For data processed under the influence of SETs, it is necessary to introduce restrictions on the mechanisms for modeling the processes of electric charge propagation [14] and the development of analytical methods for determining the error intensity in devices with different topological sizes.

Closest to the proposed technical solution (prototype) is a method for predicting the intensity of failures from the effects of SEU / SEE in digital equipment, based on the adoption of the "concept of charge efficiency" as a measure of the effectiveness of radiation-induced charge (RIS) [15]. This method makes it possible to obtain deviations from the approximating step function of the dependence of the fault cross section on the LET value in the region of small values refined by statistical methods of processing experimental data, and to obtain estimates of the threshold value of LET _TH using the digital representation of the characteristics (figure merit = FOM) and the values determined from data on cross sections or from a simulation in a mixed mode: experimentally obtained dependence σ = f (LET) for the TZZ and data on pulsed irradiation azernym radiation. The value of LET _{TH is} used to estimate the saturation cross section σ _SAT , the critical ion fluence F _CRIT, and the failure rate R.

The disadvantage of this method is the fact that the efficiency is measured for sensitivity to SEU effects within one cell of the MOS structure. In addition, the method does not provide for the determination of equivalent AI levels of a pulsed source of gamma-ray radiation, which does not allow it to be used to simulate the effects of SEU / SEE on pulsed sources of gamma-neutron radiation, to determine the level of smooth operation (UBR) and to establish a one-to-one correspondence between UBR and Let _th .

, токи переключения

, напряжения в состоянии логических «0» и «1», V_0L и V_0H, соответственно. Эти данные могут быть использованы для вычисления величины критического заряда Q_C.The proposed technical solution is based on the concept of preserving the critical charge level that causes SEU and SET effects, which allows us to extend this ideology to the data from the specification of semiconductor devices and ICs. So for pulsed diodes and microwave diodes, the specification of the device directly indicates the amount of charge switching from one logical state to another. Semiconductor and IC specifications include switching times

switching currents

, voltages in the state of logical “0” and “1”, V _0L and V _0H , respectively. These data can be used to calculate the critical charge Q _C.

If these processes of critical charge formation are neglected, there is an unconditional need to apply the technology for suppressing errors caused by SEU and SET.

The proposed technical solution relates to ICs, in particular, to both sequential logic (SL) circuits and combinatorial (combinational) logic (CL) circuits.

The technical result of the inventions is to increase the reliability of assessing the resistance of digital electronic equipment to the effects of AI, in particular, assessing the level of trouble-free operation (UBR) and the corresponding TID value, by ensuring, regardless of the irradiation conditions, an unambiguous relationship between the results of tests on simulated pulsed AI installations, the results of simulation modeling, including using pulsed laser radiation, and these specifications on the IFR and IC on dynamic parameters.

The technical result is achieved by comparing and converting various data on the resistance to dose effects during static or pulsed irradiation at the MU, on the resistance to the effects of dose rate during pulsed irradiation at the MU, on the resistance to pulsed laser radiation at the IM effects of dose rate, on the resistance to low-intensity radiation factors KP and according to specifications of the dynamic parameters of the SPT and the IC on the basis of the general concept, based on the concept of equal quantities critical charge Q _C, to tory causes in the sensitive volume ν _S occurrence of external currents causing SEU and SET effects in digital electronic circuits.

This problem is solved by creating variants of a method for assessing the resistance of digital electronic equipment to the effects of ionizing radiation based on the proposed concept.

между эквивалентной дозой D_RS гамма-рентгеновского излучения со спектром RS (Radiation Stress) и

ионов (ТЗЧ),

определяют величину эквивалентной дозы D_RS и по известной длительности импульса ИИ τ_P определяют УБР.The task (option 1) is solved by the fact that to determine the level of trouble-free operation (UBR) of the unit or the entire device of electronic equipment (CEA), in general, containing digital integrated circuits (ICs), according to the results of experimental studies on modeling plants ( MU), in a known method for assessing the resistance of digital electronic equipment to the effects of AI by using experimentally obtained data when modeling RIZ in LSI / VLSI technology CMOS / KND using side generation they are currents from pulsed sources of ionizing radiation (III), which use the accompanying gamma radiation from a pulsed nuclear reactor (INR) or gamma-ray radiation from electrophysical installations (EFs) (linear pulsed accelerators and / or X-ray installations) and from the effects of spaceborne cosmic rays space, then using the corresponding values of the coefficients of relative efficiency

between the equivalent dose D _{RS of} gamma-ray radiation with spectrum RS (Radiation Stress) and

ions (TZZ),

determine the value of the equivalent dose D _RS and the known pulse duration of the AI τ _P determine the UBR.

для каждого цифрового элемента схемы, суммарное поперечное сечение σ_Σ для однотипных элементов последовательной и комбинационной логики, усредненное поперечное сечение σ_D для элементов последовательной и комбинационной логики, полное сечение σ_TOT для всего цифрового устройства.In addition, to determine the cross-section of the failures of the digital device, analyze the structural diagram of the electronic device, determine the values of individual partial cross-sections of failures

for each digital element of the circuit, the total cross section σ _Σ for the same elements of sequential and combinational logic, the average cross section σ _D for elements of sequential and combinational logic, the total cross section σ _TOT for the entire digital device.

In addition, to reduce the number of trials, the total failure cross-sectional value is introduced taking into account the number of digital bistable DFF triggers.

, в

, где c - толщина полупроводниковой пластины, а и b - размеры чипа структуры, в мкм, а Q_C - величина критического радиационно-индуцированного (РИЗ), вызывающего сбои в работе логических устройств.In addition, in order to reduce the number of tests at facilities simulating the impact of the TZZ, to determine the unit failure cross section σ _e for the same elements of the circuit, for most practical cases, the failure rate is determined using the approximating expression for the saturation cross section σ _SAT = a b, ( a , b> c) and the critical threshold value of linear energy transfer (mass loss) $${\text{Let}}_{\text{Th}}^{\text{RS}}={\text{Q}}_{\text{C}}/\text{c}$$

, at

, where c is the thickness of the semiconductor wafer, and a and b are the dimensions of the structure chip, in microns, and Q _C is the critical radiation-induced (RIZ) value that causes malfunctions in the operation of logic devices.

однотипных элементов комбинационной логики вводят с учетом тактовой частоты (частоты синхронизации).In addition, to take into account the frequency dependence of the SEU effects in combinational logic, the value of the partial cross section of failures

the same elements of combinational logic are introduced taking into account the clock frequency (synchronization frequency).

определяют по заданной частоте синхроимпульсов и

, в

, величину сечения сбоев для однотипных элементов последовательной логики определяют с использованием соотношенияIn addition, to take into account the frequency dependence of the SEU effects in the elements of combinational and sequential logic, to take into account the dependence of the fault cross section for the same elements of the combinational logic on the synchronization frequency, on the approximating dependence of the fault cross section for the same elements of the combinational logic

determined by a given frequency of the clock pulses and

, at

, the magnitude of the cross section of failures for the same type of elements of sequential logic is determined using the relation

, где

- парциальное поперечное сечение ошибок i-го элемента, и поперечное сечение сбоев σ_CL для элементов комбинационной логики определяют

, где

- парциальное поперечное сечение ошибок j-го элемента, суммарное сечение ошибок схемы последовательной логики σ_Σ находят с использованием соотношенияIn addition, to unify the results of evaluating the sensitivity to SEU effects of elements of combinational and sequential logic, to determine the total cross-section of faults for elements of serial and combinational logic, the cross-section of faults σ _SL for elements of sequential logic is determined from

where

- the partial cross-section of errors of the ith element, and the cross-section of failures σ _CL for elements of combinational logic determine

where

- the partial cross section of errors of the jth element, the total error cross section of the serial logic circuit σ _{Σ is} found using the relation

where: a _CL is the percentage of combinatorial logic circuits; DFF is the number of fixing elements (triggers) in the circuit, and the total cross section of failures in the circuit is determined using the relation

, чувствительного объема ν_S типовой транзисторной структуры отдельного цифрового элемента и максимального значения величины пространственной диагонали s_MAX (хорды) чувствительного объема, определяют пороговое значение линейной передачи энергии LET_TH и величину интенсивности сбоев

, определяют КОЭ $${\text{RDEF}}_{\text{ION}}^{\text{RS}={\text{Co}}^{\text{60}}}$$

для порогового значения LET_TH и полной поглощенной дозы TID эквивалентного источника гамма-квантов Co⁶⁰ со средней энергией квантов RS=1,25 МэВ и определяют величину TID.In addition, to obtain the result of evaluating the equivalent dose for the reference spectrum of the nuclide source Co ⁶⁰ , based on the spectrum of linear energy transfer

, the sensitive volume ν _{S of the} typical transistor structure of an individual digital element and the maximum value of the spatial diagonal s _MAX (chords) of the sensitive volume, determine the threshold value of the linear energy transfer LET _TH and the magnitude of the failure rate

determine COE $${\text{RDEF}}_{\text{ION}}^{\text{RS}={\text{<figure-callout id="60" label="Co" filenames="00000317.png,00000318.png" state="{{state}}">Co</figure-callout>}}^{\text{60}}}$$

for the threshold value LET _TH and the total absorbed dose TID of the equivalent source of gamma rays Co ⁶⁰ with an average quantum energy RS = 1.25 MeV and determine the value of TID.

, на соответствующей МУ, воспроизводящей пучки ионов (ТЗЧ), выполняют облучение объектов исследования в диапазоне заданных флюенсов ТЗЧ из спектра LET вида

, где dN/dE - распределение числа ТЗЧ по энергиям, определяют чувствительный объем ν_S транзисторной гетероструктуры, в котором формируется критический заряд Q_C, определяют максимальный размер хорды s_MAX (пространственной диагонали) в модели чувствительного объема (ЧО) в виде параллелепипеда (Rectangular Parallel Piped = RPF) из s²=a ²+b²+с², где a=b=L - топологический размер ЧО элемента, с=t - толщина гетероструктуры в ЧО, определяют пороговое значение LET_TH на зависимости

для полного телесного угла, интенсивность сбоев определяют с использованием соотношенияIn addition, to assess the intensity of failures from SEU effects, to obtain the experimental dependence of the cross-section of failures of REA or IC units on LET of the form

, on the corresponding MU reproducing ion beams (TZCH), irradiate the objects under study in the range of specified fluences of TZCh from the LET spectrum of the form

, where dN / dE is the energy distribution of the TCD number, determine the sensitive volume ν _{S of the} transistor heterostructure in which the critical charge Q _{C is formed} , determine the maximum size of the chord s _MAX (spatial diagonal) in the model of sensitive volume (PR) in the form of a parallelepiped (Rectangular Parallel Piped = RPF) from s ² = a ² + b ² + c ² , where a = b = L is the topological size of the HO element, c = t is the thickness of the heterostructure in the HO, determine the threshold value LET _TH on the dependence

for a full solid angle, the failure rate is determined using the ratio

в единицах

, где M - полупроводниковый материал в ЧО транзисторной структуры, а величину эквивалентной дозы гамма-излучения D_RS(γ-кв.RS-экв.-(M)) определяют из соотношенияIn addition, to convert the results of an experimental assessment of the sensitivity to SEU effects of digital equipment when tested at INR and EFU installations simulating the effect of a TCD based on the critical charge identity effect Q _C , the value of the coefficient of relative efficiency is determined

in units

, where M is the semiconductor material in the BH transistor structure, and the equivalent dose of gamma radiation D _RS (γ-square RS-equiv .- (M)) is determined from the relation

ИИ из отношения эквивалентной поглощенной дозы к длительности этого импульсаIn addition, to implement an experiment on a pulsed nuclear reactor (INR) or an electrophysical installation (EFU) generating pulsed gamma-ray radiation, for a known pulse duration of gamma-ray quanta of the reference AI source τ _P , the dose rate is determined for digital equipment

AI from the ratio of the equivalent absorbed dose to the duration of this pulse

which is taken equal to UBR.

сбоев, для определения величины УБР ППП или ИМС при эквивалентном по величине критического заряда Q_C воздействии импульсного ИИ с энергией квантов ~1 МэВ-экв.(М), оценку порогового значения

производят при воздействии импульсного лазерного излучения (ИЛИ), затем с использованием соответствующих значений коэффициентов относительной эффективности

между эффективной дозой D_RS гамма-рентгеновского излучения со спектром RS и

ТЗЧ (ионов),

между энергией сбоев

от лазерного импульса и эффективной дозой D_RS гамма-рентгеновского излучения, определяют величину эквивалентного критического заряда Q_C, величину интенсивности сбоев R, величину эффективной дозы D_RS и по известной длительности импульса ИИ τ_P определяют УБР.This problem is also solved according to option 2, in which the decoupled crystal (chip) of the heteroepitaxial structure of a semiconductor device (IFP) or IC is irradiated with a pulsed laser source, critical threshold energy is determined

failures, to determine the magnitude of the UBR IFR or IC under the equivalent critical charge Q _C exposure to pulsed AI with a quantum energy of ~ 1 MeV-eq. (M), estimate the threshold value

produced when exposed to pulsed laser radiation (OR), then using the corresponding values of the relative efficiency coefficients

between the effective dose D _{RS of} gamma-ray radiation with the RS spectrum and

TZCh (ions),

between energy failures

from the laser pulse and the effective dose D _{RS of} gamma-ray radiation, determine the value of the equivalent critical charge Q _C , the magnitude of the failure rate R, the effective dose D _RS and the known pulse duration of the AI τ _P determine the UBR.

In addition, to obtain adequate conditions with the use of TZZ conditions, when using simulation methods (IM) using OR, the heterostructure of the IFR or IC chip is irradiated with a picosecond laser source pulse with a duration of not more than 350 ps and a wavelength in the range of 800-1500 nm.

, устанавливающего связь между результатом облучения объекта с использованием ТЗЧ и ИЛИ, и

, устанавливающего связь между результатом облучения объекта с использованием импульсного ионизирующего излучения (ИИИ) и ТЗЧ, получают значение

, с использованием которого, в свою очередь, определяют величину эффективной поглощенной дозы гамма-рентгеновского излучения D_RS(γ - кв.RS-экв.-(М)) эталонного источника Co⁶⁰ для материала (М) (например, в качестве M=Si, SiO₂) и при известной длительности импульса гамма-квантов источника эталонного ИИ τ_P УБР цифровой аппаратуры определяют из отношения эквивалентной поглощенной дозы к длительности этого импульсаIn addition, to obtain the equivalent value of the critical charge Q _C , which causes the emergence of external currents in the critical volume ν _S , which cause SEU and SET effects, to realize photoionization of the frequency response of transistor structures, use the value

that establishes a relationship between the result of irradiating an object using a PLC and OR, and

, which establishes the relationship between the result of irradiation of an object using pulsed ionizing radiation (III) and TZCh, get the value

, which, in turn, is used to determine the effective absorbed dose of gamma-ray radiation D _RS (γ - sq. RS-eq. (M)) of the reference source Co ⁶⁰ for material (M) (for example, as M = Si, SiO ₂ ) and for a known pulse duration of gamma rays of the source of the reference AI τ _P UBR digital equipment is determined from the ratio of the equivalent absorbed dose to the duration of this pulse

сбоев в единицах [pJ].In addition, to fix the efficiency of the MI, when using the MI using a pulsed laser source, the critical threshold energy is recorded

failures in units [pJ].

определяют с использованием соотношенияIn addition, for the conversion of MI results, the value of

determined using the ratio

для длины волны лазерного излучения λ_LASER=800-1500 nm принимают равнойIn addition, to fix the effectiveness of the MI, the value

for the wavelength of laser radiation λ _LASER = 800-1500 nm is taken equal

определяют изIn addition, to convert the results of myocardial infarction into the equivalent effect of TZCh, the value of the equivalent value

determined from

In addition, to determine the UBR for MI, in units equivalent to the effects of pulsed AI sources, using

determine the equivalent total absorbed dose of gamma rays D _RS (γ-square RS-equiv. (M)) and for a known pulse duration of gamma rays of the reference ionizing radiation source τ _P UBR digital equipment determine (11) P _γ [rad (Si) · s ^-1 ].

в

.In addition, to use engineering methods for assessing the intensity of failures from the effects of TZZ, transform

at

.

, который вызывает в ЧО ν_S возникновение сторонних токов, генерирующих эффекты SEU и SET в цифровых электронных схемах, в способе оценки стойкости цифровой аппаратуры к воздействию ИИ используют данные о значении динамических электрических параметров спецификаций на элементы ЭКБ.This problem is also solved according to option 3 in that they determine the level of failure-free operation (UBR) based on the use of dynamic electrical parameters of the specification for electronic components to obtain an equivalent critical charge

, which causes the occurrence of external currents in the CH ν _S , generating SEU and SET effects in digital electronic circuits, in the method for assessing the resistance of digital equipment to the effects of AI, data on the value of the dynamic electrical parameters of the specifications for the electronic components are used.

по величине эквивалентного критического заряда

, в качестве динамического параметра используют времена переключения из одного логического состояния в другое

и соответствующее значение напряжения питания ИМС.In addition, to determine the equivalent value

by value of equivalent critical charge

, as a dynamic parameter, the switching times from one logical state to another are used

and the corresponding value of the supply voltage of the IC.

по величине эквивалентного критического заряда

, соответствующего воздействию «ионного коктейля», в качестве динамического параметра используют так же уровни выходных напряжений логических состояний V_0L и V_0H и соответствующее значение емкости на выходе контролируемого сигнала.In addition, to determine the equivalent value

by value of equivalent critical charge

corresponding to the effect of the “ion cocktail”, the levels of the output voltages of logical states V _0L and V _0H and the corresponding value of the capacitance at the output of the controlled signal are also used as a dynamic parameter.

по величине эквивалентного критического заряда

, соответствующего воздействию «ионного коктейля», в качестве динамического параметра используют прямые данные о заряде переключения Q_ПЕР между этими логическими состояниями.In addition, to determine the equivalent value

by value of equivalent critical charge

corresponding to the effect of the “ion cocktail”, direct data on the charge switching Q _PER between these logical states are used as a dynamic parameter.

по величине эквивалентного критического заряда

, соответствующего воздействию «ионного коктейля», используют соотношениеIn addition, to determine the equivalent value

by value of equivalent critical charge

corresponding to the effects of "ion cocktail", use the ratio

, и затем определяют эквивалентное значение

.or

, and then determine the equivalent value

.

между эквивалентной дозой D_RS гамма-рентгеновского излучения со спектром RS и

, соответствующей генерации эквивалентного заряда от параметров динамических переключений логических состояний определяют величину эквивалентной дозы D_RS и по известной длительности τ_P импульса ИИ эквивалентного источника определяют УБР.In addition, for the conversion of the values of the dynamic switching parameters into the effective value of the UBR from the action of AI with the RS spectrum, using the corresponding values of the relative efficiency coefficients

between the equivalent dose D _{RS of} gamma-ray radiation with an RS spectrum and

corresponding to the generation of the equivalent charge from the parameters of dynamic switching of logical states, the equivalent dose value D _RS is determined and the UBR is determined from the known pulse duration AI _P of the equivalent source pulse.

, или уровни выходных напряжений логических состояний V_OL и V_0H, либо из данных о заряде переключения Q_ПЕР. между этими состояниями, затем с использованием соответствующих значений коэффициентов относительной эффективности

между эквивалентной дозой гамма-рентгеновского излучения D_RS и

,

между энергией сбоев от лазерного импульса

и

,

между эффективной дозой гамма-рентгеновского излучения D_RS и

,

между энергией сбоев от лазерного импульса

и эффективной дозой гамма-рентгеновского излучения D_RS, определяют величину эквивалентной дозы D_RS и по известной величине длительности импульса ИИ τ_P определяют УБР.This problem is also solved according to option 4 in that in order to determine the level of trouble-free operation (UBF) of LSI / VLSI under the influence of an equivalent pulsed ionizing radiation source with a quantum energy of 1.25 MeV-eq. (M), the threshold value of linear energy transfer is selectively estimated (LET _TH) either from direct measurements of the cross section depending on the magnitude of failures LET, or by pulsed laser radiation, either from a data value of such parameters on the electrical specifications EKB elements as time switch tions from one logical state to another

, or the output voltage levels of the logical states V _OL and V _0H , or from the data on the switching charge Q _PER. between these states, then using the corresponding values of the coefficients of relative efficiency

between the equivalent dose of gamma-ray radiation D _RS and

,

between the energy of failures from a laser pulse

and

,

between the effective dose of gamma-ray radiation D _RS and

,

between the energy of failures from a laser pulse

and the effective dose of gamma-ray radiation D _RS , determine the value of the equivalent dose D _RS and the known value of the pulse duration II τ _P determine the UBR.

The claimed inventions are connected by a single inventive concept and are a way of assessing the resistance of digital electronic equipment to the effects of AI, in particular, UBR and TID values.

The invention is illustrated by the following figures:

FIG. 1. Cross section of a pn junction formed on a silicon substrate. The effect of the passage of the ion through the transition "drain-body" in the transistor n-MOS

FIG. 2. A graph of the logarithmic dependence of the duration of the transition pulse required for propagation of SET without attenuation through an infinitely long chain of inverters as a function of topological size [17].

FIG. 3. The block diagram of the layout of sequential logic [17].

FIG. 4. Two processes of failure (capsizing). (A) a direct hit in an element of memory or sequential logic; and (B) when it enters a combined logic, a SET is created that gets into the element of memory or sequential logic.

Combinational Logick Network - combinational logic device

PRE - stateful trigger

SET - transient from a single particle

SEU - the effect of switching (changing a logical state) under the action of a single particle

FIG. 5. Timing diagram illustrating the relationship in time for the case when the true data is ignored and the positive SETs erroneously arrive at the input of the second trigger (46).

FIG. 6. Timing diagram illustrating the situation when SET in the clock diagram can lead to a memory error.

FIG. 7. Dependence of the failure rate in serial logic devices (Flip-Flop [17]) and combinational logic (DICE test chips [18]) and their mutual correspondence.

FIG. 8. Comparison of data on the effects of TZZh on two test schemes. The black line corresponds to the saturation cross section for the D-flip chip. Lighter is a linear data approximation for DICE. The point for 1 MHz for D-flip-flops matched exactly, although the chip was in working condition.

FIG. 9. Three options for evaluating UBR digital IC

FIG. 10. The initial structure of the sensitive volume of the MOS transistor for calculating LET.

FIG. 11. Schematic representation of the procedure for assessing the intensity of failures and the parametric probability of maintaining the health of electronic devices KA / 6 /.

FIG. 12. Typical SEE cross-sectional dependence on LET for various PLCs. Y axis: Cross section of SEE (cm ² ); X-axis: LET (MeV ^· cm ² / mg).

.FIG. 13. Typical TZCh range. On the Y axis: the fluence of particles [particles / m ² - sterrad ^· s], on the X axis:

.

FIG. 14. Cross section of the studied CMOS / SSC structures. FIG. 15. The linearized Weibull distribution function as a function of Y = fln (LET-LET _γ )

FIG. 16. Dependence of the saturation cross section of failures on linear energy losses.

In accordance with the claimed, specific to structural features and / or methodological steps (implementation method) should be described, it should be understood that the invention defined in additional links (claims = claims) is not limited to distinctive features or method implementation. On the contrary, the hallmarks and method of implementation are open as the preferred form of use of the claimed links.

The proposed method is implemented according to the general algorithm for determining UBR using three main options and one generalized (Fig. 9) depending on the amount of information available:

The method according to option 1 is implemented as follows.

между эквивалентной дозой D_RS гамма-рентгеновского излучения со спектром RS (Radiation Stress) и

ионов (ТЗЧ),

определяют величину эквивалентной дозы D_RS (9) и по известной длительности импульса ИИ τ_P определяют УБР (10).The task (option 1) is solved by the fact that to determine the level of trouble-free operation (UBR) of the unit or the entire device of electronic equipment (CEA), in general, containing digital integrated circuits (ICs), according to the results of experimental studies on modeling plants ( MU), in the known method for assessing the resistance of digital electronic equipment to the effects of AI by using experimentally obtained data when modeling RIZ in LSI / VLSI CMOS / KND technology for generating external currents from them pulse sources of ionizing radiation (III). As the III use the accompanying gamma radiation of a pulsed nuclear reactor (INR) or gamma-ray radiation of electrophysical installations (EFs) (linear pulse accelerators and / or X-ray installations) and the impact of the TZZ charged particle accelerators. Then, using the corresponding values of the coefficients of relative efficiency

between the equivalent dose D _{RS of} gamma-ray radiation with spectrum RS (Radiation Stress) and

ions (TZZ),

determine the value of the equivalent dose D _RS (9) and from the known pulse duration of the AI τ _P determine the UBR (10).

и

для каждого цифрового элемента схемы последовательной и комбинационной логики, суммарное поперечное сечение (6) σ_Σ для однотипных элементов последовательной и комбинационной логики, усредненное поперечное сечение σ_D для элементов последовательной и комбинационной логики, полное сечение (7) σ_TOT для всего цифрового устройства.To determine the cross-section of failures of a digital device, analyze the structural diagram of the electronic device, determine the values of individual partial cross-sections of failures (5)

and

for each digital element of the serial and combinational logic circuit, the total cross section (6) σ _Σ for the same elements of serial and combinational logic, the averaged cross section σ _D for elements of sequential and combinational logic, the total cross section (7) σ _TOT for the entire digital device.

To reduce the number of tests, the total cross-sectional area of failures is introduced taking into account the number of digital bistable DFF triggers.

, в

, где c - толщина полупроводниковой пластины, а и b - размеры чипа структуры, в мкм, a Q_C - величина критического радиационно-индуцированного (РИЗ), вызывающего сбои в работе логических устройств.In order to reduce the number of tests at facilities simulating the impact of a TZZ, to determine the unit failure cross section σ _e for the same circuit elements, for most practical cases the failure rate is determined using the approximating expression for the saturation cross section σ _SAT = a b, ( a , b> с ) and the critical threshold value of linear energy transfer (mass loss)

, at

, where c is the thickness of the semiconductor wafer, and a and b are the dimensions of the structure chip, in microns, a Q _C is the critical radiation-induced (RI) value that causes malfunctions in the operation of logic devices.

однотипных элементов комбинационной логики вводят с учетом тактовой частоты (частоты синхронизации).To take into account the frequency dependence of SEU effects in combinational logic, the value of the partial cross section of failures

the same elements of combinational logic are introduced taking into account the clock frequency (synchronization frequency).

определяют по заданной частоте синхроимпульсов и

, в

, величину сечения сбоев для однотипных элементов последовательной логики определяют с использованием соотношения (5).To take into account the frequency dependence of SEU effects in the elements of combinational and sequential logic, to take into account the dependence of the fault cross section for the same elements of combinational logic on the synchronization frequency, on the approximating dependence of the fault cross section for the same elements of combinational logic

determined by a given frequency of the clock pulses and

, at

, the magnitude of the cross section of failures for the same type of elements of sequential logic is determined using relation (5).

, где

- парциальное поперечное сечение ошибок i-го элемента, и поперечное сечение сбоев σ_CL для элементов комбинационной логики определяют

, где

- парциальное поперечное сечение ошибок j-го элемента, суммарное сечение ошибок схемы последовательной логики σ_Σ находят с использованием соотношения (6).To unify the results of evaluating sensitivity to SEU effects of elements of combinational and sequential logic, to determine the total cross-section of faults for elements of serial and combinational logic, the cross-section of faults σ _SL for elements of sequential logic is determined from

where

- the partial cross-section of errors of the ith element, and the cross-section of failures σ _CL for elements of combinational logic determine

where

- the partial cross section of errors of the jth element, the total error cross section of the serial logic circuit σ _{Σ is} found using relation (6).

The full cross section of failures in the circuit is determined using relation (7).

, чувствительного объема ν_S типовой транзисторной структуры отдельного цифрового элемента и максимального значения величины пространственной диагонали s_MAX (хорды) чувствительного объема, определяют пороговое значение линейной передачи энергии LET_TH и величину интенсивности сбоев

, определяют КОЭ

для порогового значения LET_TH и полной поглощенной дозы TID эквивалентного источника гамма-квантов Co⁶⁰ со средней энергией квантов RS=1,25 МэВ и определяют величину TID.To obtain the result of evaluating the equivalent dose for the reference spectrum of the nuclide source Co ⁶⁰ , based on the spectrum of linear energy transfer

, the sensitive volume ν _{S of the} typical transistor structure of an individual digital element and the maximum value of the spatial diagonal s _MAX (chords) of the sensitive volume, determine the threshold value of the linear energy transfer LET _TH and the magnitude of the failure rate

determine COE

for the threshold value LET _TH and the total absorbed dose TID of the equivalent source of gamma rays Co ⁶⁰ with an average quantum energy RS = 1.25 MeV and determine the value of TID.

, на соответствующей МУ, воспроизводящей пучки ионов (ТЗЧ), выполняют облучение объектов исследования в диапазоне заданных флюенсов ТЗЧ из спектра LET вида

, где dN/dE - распределение числа ТЗЧ по энергиям. Затем определяют чувствительный объем (ЧО) ν_S транзисторной гетероструктуры, в котором формируется критический заряд Q_C, определяют максимальный размер хорды s_MAX (пространственной диагонали) в модели ЧО в виде параллелепипеда (Rectangular Parallel Piped = RPP) из s²=а ²+b²+с², где a=b=L - топологический размер ЧО элемента, с=t - толщина гетероструктуры в ЧО, определяют пороговое значение LET_TH на зависимости

для полного телесного угла. Интенсивность сбоев определяют с использованием соотношения (8).To assess the intensity of failures from the effects of SEU, to obtain the experimental dependence of the cross-section of failures of REA or IC units on LET of the form

, on the corresponding MU reproducing ion beams (TZCH), irradiate the objects under study in the range of specified fluences of TZCh from the LET spectrum of the form

where dN / dE is the energy distribution of the number of TZZh. Then, the sensitive volume (HO) ν _{S of the} transistor heterostructure in which the critical charge Q _C is formed is determined, the maximum chord size s _MAX (spatial diagonal) is determined in the HO model in the form of a parallelepiped (Rectangular Parallel Piped = RPP) from s ² = а ² + b ² + c ² , where a = b = L is the topological size of the HO element, c = t is the thickness of the heterostructure in the HO, determine the threshold value LET _TH on the dependence

for a full solid angle. The failure rate is determined using relation (8).

в единицах

, где M - полупроводниковый материал в ЧО транзисторной структуры. Величину эквивалентной дозы гамма-излучения D_RS(γ-кв.RS-экв.-(М)) определяют из соотношения (9).To convert the results of an experimental evaluation of the sensitivity to SEU effects of digital equipment when tested at INR and EFU simulating the effects of an SQT by the effect of the identity of the critical charge Q _C , determine the value of the coefficient of relative efficiency

in units

where M is a semiconductor material in a transistor structure. The value of the equivalent dose of gamma radiation D _RS (γ-sq. RS-equiv .- (M)) is determined from relation (9).

ИИ из отношения (10) эквивалентной поглощенной дозы к длительности этого импульса. Эту величину принимают равной УБР.To implement an experiment on a pulsed nuclear reactor (INR) or electrophysical installation (EPF), generating a pulse of gamma-ray radiation at a known pulse duration gamma-ray quanta reference source AI τ _P dose rate determining apparatus for a digital

AI from the ratio (10) of the equivalent absorbed dose to the duration of this pulse. This value is taken equal to UBR.

сбоев. Для определения величины УБР ППП или ИМС при эквивалентном по величине критического заряда Q_C воздействии импульсного ИИ с энергией квантов ~1 МэВ-экв.(М), оценку порогового значения

производят при воздействии импульсного лазерного излучения (ИЛИ), затем с использованием соответствующих значений коэффициентов относительной эффективности

между эффективной дозой D_RS гамма-рентгеновского излучения со спектром RS и

ТЗЧ (ионов),

между энергией сбоев

от лазерного импульса и эффективной дозой D_RS гамма-рентгеновского излучения, определяют величину эквивалентного критического заряда Q_C, величину интенсивности сбоев R, величину эффективной дозы D_RS и по известной длительности импульса ИИ τ_P определяют (10) УБР.This problem is also solved according to option 2, in which the decoupled crystal (chip) of the heteroepitaxial structure of a semiconductor device (IFP) or IC is irradiated with a pulsed laser source, critical threshold energy is determined

failures. To determine the magnitude of the UBR SPP or IC under the equivalent critical charge Q _{C by the} action of a pulsed IR with a quantum energy of ~ 1 MeV-eq. (M), an estimate of the threshold value

produced when exposed to pulsed laser radiation (OR), then using the corresponding values of the relative efficiency coefficients

between the effective dose D _{RS of} gamma-ray radiation with the RS spectrum and

TZCh (ions),

between energy failures

from the laser pulse and the effective dose D _{RS of} gamma-ray radiation, determine the equivalent critical charge Q _C , the magnitude of the failure rate R, the effective dose D _RS, and (10) UBR are determined from the known pulse duration of the AI τ _P.

In order to obtain adequate conditions with the use of TZZ conditions, when using simulation methods (IM) using OR, the heterostructure of the IFR or IC chip is irradiated with a picosecond pulse of a laser source with a duration of not more than 350 ps and a wavelength in the range of 800-1500 nm.

, устанавливающего связь между результатом облучения объекта с использованием ТЗЧ и ИЛИ, и

, устанавливающего связь между результатом облучения объекта с использованием импульсного ионизирующего излучения (ИИИ) и ТЗЧ, получают значение

, с использованием которого, в свою очередь, определяют величину эффективной поглощенной дозы гамма-рентгеновского излучения D_RS(γ-кв.RS-экв.-(M)) эталонного источника Co⁶⁰ для материала (М) (например, в качестве M=Si, SiO₂) и при известной длительности импульса гамма-квантов источника эталонного ИИ τ_P УБР цифровой аппаратуры определяют из отношения эквивалентной поглощенной дозы к длительности этого импульса (11).To obtain the equivalent value of the critical charge Q _C , which causes the appearance of external currents in the PF volume ν _S , causing SEU and SET effects, to realize photoionization of the PF transistor structures, use the value

that establishes a relationship between the result of irradiating an object using a PLC and OR, and

, which establishes the relationship between the result of irradiation of an object using pulsed ionizing radiation (III) and TZCh, get the value

using which, in turn, the effective absorbed dose of gamma-ray radiation D _RS (γ-sq. RS-equiv .- (M)) of the reference source Co ⁶⁰ for the material (M) is determined (for example, as M = Si, SiO ₂ ) and for a known pulse duration of gamma rays of the source of the reference AI τ _P UBR digital equipment is determined from the ratio of the equivalent absorbed dose to the duration of this pulse (11).

сбоев в единицах [pJ].To fix the efficiency of the MI, when using the MI using a pulsed laser source, the critical threshold energy is fixed

failures in units [pJ].

определяют с использованием соотношения (12).For the conversion of MI results, the value of

determined using the relation (12).

для длины волны лазерного излучения λ_LASER=800-1500 nm принимают равной (13).To fix the effectiveness of the MI, the value

for the wavelength of laser radiation λ _LASER = 800-1500 nm is taken equal to (13).

определяют из (14).To convert the results of myocardial infarction into the equivalent effect of the TZC, the equivalent

determined from (14).

To determine the UBR for MI, in units equivalent to the effects of pulsed AI sources, using (15), determine the value of the equivalent total absorbed dose of gamma rays D _RS (γ-square RS-equiv. (M)) and for a known pulse duration gamma rays of the reference ionizing radiation source τ _P UBR digital equipment determine (11) P _γ [rad (Si) · s ^-1 ].

в

.To use engineering methods for assessing the intensity of failures from the effects of TZZ, transform

at

.

, который вызывает в ЧО ν_S возникновение сторонних токов, генерирующих эффекты SEU и SET в цифровых электронных схемах, используют данные о значении динамических электрических параметров спецификаций на элементы ЭКБ.This problem is also solved by option 3 in that they determine the level of failure-free operation (UBR) based on the use of dynamic electrical parameters of the specification for electronic components. To obtain an equivalent critical charge

, which causes the occurrence of external currents in the CH ν _S that generate SEU and SET effects in digital electronic circuits, use data on the value of the dynamic electrical parameters of specifications for electronic components.

по величине эквивалентного критического заряда

, в качестве динамического параметра используют времена переключения из одного логического состояния в другое

и соответствующее значение напряжения питания ИМС.To determine the equivalent value

by value of equivalent critical charge

, as a dynamic parameter, the switching times from one logical state to another are used

and the corresponding value of the supply voltage of the IC.

по величине эквивалентного критического заряда

, соответствующего воздействию «ионного коктейля», в качестве динамического параметра используют так же уровни выходных напряжений логических состояний V_0L и V_0H и соответствующее значение емкости на выходе контролируемого сигнала.To determine the equivalent value

by value of equivalent critical charge

corresponding to the effect of the “ion cocktail”, the levels of the output voltages of logical states V _0L and V _0H and the corresponding value of the capacitance at the output of the controlled signal are also used as a dynamic parameter.

по величине эквивалентного критического заряда

, соответствующего воздействию «ионного коктейля», в качестве динамического параметра используют прямые данные о заряде переключения Q_ПЕР между этими логическими состояниями.To determine the equivalent value

by value of equivalent critical charge

corresponding to the effect of the “ion cocktail”, direct data on the charge switching Q _PER between these logical states are used as a dynamic parameter.

по величине эквивалентного критического заряда

, соответствующего воздействию «ионного коктейля», используют соотношение (16) или

. Затем определяют эквивалентное значение

.To determine the equivalent value

by value of equivalent critical charge

corresponding to the effects of "ion cocktail", use the ratio (16) or

. Then determine the equivalent value

.

между эквивалентной дозой D_RS гамма-рентгеновского излучения со спектром RS и

, соответствующей генерации эквивалентного заряда от параметров динамических переключений логических состояний, определяют величину эквивалентной дозы D_RS и по известной длительности τ_P импульса ИИ эквивалентного источника определяют (10) УБР.To convert the values of the dynamic switching parameters into the effective value of the UBR from the effects of AI with the RS spectrum, using the corresponding values of the relative efficiency coefficients

between the equivalent dose D _{RS of} gamma-ray radiation with an RS spectrum and

corresponding to the generation of the equivalent charge from the parameters of the dynamic switching of logical states, the equivalent dose value D _RS is determined and the UBR is determined (10) by the known pulse duration AI _P of the equivalent source pulse.

между эквивалентной дозой гамма-рентгеновского излучения D_RS и

,

между энергией сбоев от лазерного импульса

и

,

между эффективной дозой гамма-рентгеновского излучения D_RS и

,

между энергией сбоев от лазерного импульса

и эффективной дозой гамма-рентгеновского излучения D_RS, определяют величину эквивалентной дозы D_RS. По известной величине длительности импульса ИИ τ_P определяют (10) УБР.This problem is also solved according to option 4 by the fact that in the known method for assessing the resistance of digital electronic equipment to the effects of AI by using experimentally obtained data when modeling RIZ in the LSI / VLSI CMOS / KND technology from the effects of the high-frequency converter, from the influence of the OR picosecond range, to determine level of uninterrupted operation (UBR) of LSI / VLSI under the influence of an equivalent pulsed source of ionizing radiation with a quantum energy of 1.25 MeV-eq. (M), they selectively evaluate the threshold value of linear energy transfer (LET _TH ) either from direct measurements of the dependence of the failure cross section on the LET value according to option 1, or when exposed to pulsed laser radiation according to option 2, or from the data on the value of the dynamic electrical parameters on the (datasheet) elements of the ECB, or the output voltage levels of the logical states V _OL and V _0H , or from the data on the switching charge Q _PER. between these conditions. Then, using the appropriate COE values:

between the equivalent dose of gamma-ray radiation D _RS and

,

between the energy of failures from a laser pulse

and

,

between the effective dose of gamma-ray radiation D _RS and

,

between the energy of failures from a laser pulse

and the effective dose of gamma-ray radiation D _RS , determine the value of the equivalent dose D _RS . From the known value of the pulse duration of the AI τ _P determine (10) UBR.

An example of a specific implementation.

1. Examples according to option 1. UBR determinations based on the results of experimental studies of the dependence of the failure cross section on the linear energy transfer σ = f (LET) for the TZZ from a given ionic “cocktail”.

A. Critical charge.

Existing trends (i.e., reducing instrument size, power consumption, increasing linear resolution, increasing memory and speed) increase sensitivity to SEU effects. This can be easily understood if we imagine the device as a simple capacitor (C), into which an ionizing particle penetrates, creating a charge Q, as a result of which the voltage (i.e., the logical state) changes. The SEU effect is observed if LET> Q _crit .

With a decrease in the active region of such a device, its capacity also decreases and the same charge contributes to the appearance of SEU effects. The thickness of the device basically remains unchanged, only the length and width of the device undergo changes. If we consider a promising device with a chip of square configuration L × L and thickness of the chip s, then the critical charge sufficient to change the logical state of such a device will be proportional to the square of size L.

Robinson et al. [19] presented a critical charge for the IC of a number of technologies (including NMOS, CMOS / surround, CMOS / SOS, i ² L, GaAs, ECL, CMOS / SOI VHSIC bipolar):

This critical charge leads directly to a switch from a logical “1” state to a logical “O” state or a change in a logical state (counterversion), but it is less than the total radiation-induced charge in the HO. It is significant that Q _crit is the difference between the accumulated charge in the node and the minimum charge necessary for amplification and subsequent correction [20]. In SRAM circuits, Q _crit depends not only on the magnitude of the collected charge, but also on the rate of change of the pulse in time.

An elementary model of SEU can be formulated using the concept of LET using the same thickness of the device in the form of a rectangular parallelepiped (IRPP). It is based on the calculation of the absorbed energy E _dep , representing the trajectory of a particle flying through the volume of the device in the form of a transverse chord.

The accumulated charge depends on the energy of formation of an electron-hole pair w _ehp

where q = 1,60022 × 10 ^-19 Coulombs / e, and the values of w _ehp for a number of elements are given in table. one.

Using this data, it is possible, as a first approximation, to calculate the LET that is minimally necessary to create an SEU (the algorithm is given below).

Below is the procedure for estimating the LET value for a parallelepiped with dimensions a , b, c, where c is the thickness of the device (Fig. 10). The minimum value of LET corresponds to the maximum value of the chord length s _max , which is the diagonal of the parallelepiped

The minimum value of LET will correspond to the case for which the failure can be calculated from the ratio:

It is possible to calculate the minimum length s _min that a particle must pass with given LETs to create the SEU effect

Particles create failures at an angle of θ _with -π / 2: there are two potential cases (for LET _C <LET _TH )

1. If LET> LET _C , then failures occur at all angles of incidence.

2. If LET <LET _C , then there is a critical angle θ _C at which failures are observed. The failure rate was defined as the number of errors per day per chip, [errors / day-chip] or the number of errors per day per bit [errors / bit-day]. The error rate of radiation-resistant circuits is about 10 ^-8 [errors / bit-day], and non-radiation-resistant ones are several orders of magnitude larger.

Several basic steps can be distinguished in calculating the effects of SEU [21], illustrated in FIG. eleven:

(4) - Definition of the HO of the transistor of the MOS structure;

в [отн. ед.] от LET для i-го элемента ЭКБ (

- сечение насыщения);(5) - Determination of the dependence of the relative cross section of the effects of SEE in

in [rel. units] from LET for the i-th element of the ECB (

- saturation cross section);

(6) - Estimation of the mathematical expectation µ and variance D;

(7) - The construction of the physical law of defeat (FZP) for the entire digital circuit.

. Экспериментально зависимость поперечного сечения определяют как функции энергии частиц в виде LET (Фиг. 12).(one). The cross section σ of the SEU / SEE effects is measured as a function of LET for each of the samples at a charged particle accelerator or pulsed IR source [15]. The cross section of device failures (in the sense of IC or transistor structure) is defined as the ratio of the number of failures to particle fluence

. Experimentally, the cross-sectional dependence is determined as a function of particle energy in the form of LET (Fig. 12).

(2) - The determination of the linear energy loss spectrum of LET for TZZh is carried out for a given, based on the actual conditions of existence of digital equipment, spectrum of "ion cocktail" (Fig. 13).

(3) - Estimation of failure rate R _err IC. At this stage, the cross section and the sensitive volume of the device are integrated with the LET spectrum (relation (6)).

(4) - Definition of the HO of the transistor of the MOS structure. CHO is less than the actual physical volume of the device. HO is determined for SEE only for TZh ions and protons, as well as for SEL effects ("thyristor effects"). For gamma and X-ray radiation, these concepts coincide (a PR, such as, for example, an “island” in an n-channel MOS transistor in the p-pocket of a CMOS / KND structure, expands to the geometric dimensions of the entire MOS transistor). The geometry of the sensitive volume and the critical charge are the most difficult parameters to determine.

от LET для i-го элемента ЭКБ (определение

- сечения насыщения и порогового значения LET_TH, определение критического значения флюенса частиц

), аппроксимация зависимости трехпараметрическим распределением Вейбулла [22], оценка параметров распределения (γ-положения, η-масштаба, β-формы) для каждой i-й ИМС;(5) - Representation of the dependence of the relative cross section of SEE in relative units

from LET for the i-th element of the ECB (definition

- cross sections for saturation and threshold value LET _TH , determination of the critical value of particle fluence

), approximation of the dependence by the three-parameter Weibull distribution [22], estimation of distribution parameters (γ-position, η-scale, β-form) for each i-th IC;

(6) - Estimation of the mathematical expectation µ _i and variance D _i for each i-th IC, determination of µ and D for the whole scheme as a whole using [21].

(7) - Construction of the physical law of defeat (FZP) for the entire digital circuit [22].

The following is an example of calculating the comparative LET for CMOS structures of “bulk” silicon technology (v) and CMOS / SPS technology. The LET value for v-technology is taken as 1.

The source data are:

- equality of bias voltage;

- equality of the capacity of the gate unit;

;

;

q is the electron charge;

ρ is the density of the semiconductor material;

V is the bias voltage;

S ² = 2 a ² + c ² is the square of the chord; a = b is the square topology of the MOS transistors;

C is the gate dielectric capacitance.

; σ_SAT - поперечное сечение насыщения SEU в (µm²), a LET_crit - критическое значение LET в единицах [рС/µm].Relations for estimating the failure rate of SEUs were obtained using the distribution function of the cross section of the device failures from the chord length. After integrating the flow and the cross section in the energy range (LET), a formula was obtained for the failure rate at 10% impact (9), where R _err is the SEU intensity in units

; σ _SAT is the saturation cross section of SEU in (µm ² ), and LET _crit is the critical value of LET in units of [pC / µm].

For most practical cases, the failure rate in the above ratio can be determined using the approximating expression for σ _SAT = a b, ( a , b> c) and

where c is the thickness of the silicon wafer, and a and b are its dimensions in microns. The above ratio is applicable for many applications. An example of an evaluation of R _{err is} given in table. 2. This ratio allows a spread in quality assessment (Figure of Merit = FOM) with a value of up to 5 times and a correction factor of 3.5 for GaAs devices.

The main parameters of the MOS structures in the composition of the investigated LSIs that were analyzed are shown in FIG. 14 and tab. PA2 and tab. PA3 from [16].

(20) и (21) можно представить в виде:Relation (18) taking into account

(20) and (21) can be represented as:

or

The results of assessing the sensitivity of MOS structures to the effects of SEU / SEE for CP are summarized in table. 2. From the data table. 2 it follows that low-power transistors are subject to SEZ effects from the current transformer only at a certain angle of incidence of the current transformer. On the other hand, the effect of TZZ can be modeled by the effect of pulsed x-ray or gamma radiation [16]. It is only necessary to understand that in this case, the entire volume of the transistor of the MOS structure is subjected to ionization, while under the influence of an SLC in the HO, a “thread” of charge is generated located locally within the volume of the selected rectangle (Fig. 13).

The index “1)” in indicates the values of LET _TH , which indicate the sensitivity of the TIR structures of this power class to SEE from TZCh. The “2)” index indicates MOS structures that do not require analysis of sensitivity to SEE. This means that low-power transistors are prone to SEZ effects from the current transformer only at a certain angle of incidence of the current transformer. Index “3)” indicates the possibility of total ionization of transistor structures, regardless of the value of LET _TH . Criteria for assessing the sensitivity of devices to the action of KP factors are taken from table. 3 [16].

In the table. 2 in the last column shows the possibility of total ionization of transistor structures regardless of the value of LET _TH . Approximating the dependence in FIG. 12 of the cross-section of the sensitivity of the instruments to the effects of ions from LET by a step function, let us determine for the case LET <LET _{TH the} fluence of particles sufficient for malfunctioning to within 6 × 10 ^-4 ppm / cm ² , or from 1 to 6 × 10 ⁴ p. / μm ² .

B. Results of assessing the stability of spacecraft equipment to the effects of SEU / SEE and SEL according to the results of experimental studies of IC components at the source of TZCh.

A hypothetical spacecraft was considered, the electronic equipment of which included all types of ICs that were tested on a charged particle accelerator. Data processing of the results of measuring the dependence σ = f (LET) using approximation by the Weibull law is given in Table. 3 and 4, taking into account the “critical charge” model discussed above.

For all types of ICs shown in the table, the saturation cross section for malfunctions was calculated using formulas (5) - (7), the remaining parameters were obtained on the basis of qualification tests.

An algorithm was used to obtain parameter values for an IC of type “D”, which is based on the linearization of the Weibull distribution law. Using the experimental data of the dependence σ = f (LET), a linearized Weibull distribution function was constructed in the form Y = f (ln (LET-LET _γ )) (Fig. 15) and the dependence of the fault section on linear energy losses σ = f (LET) ( Fig. 16).

From the slope of the straight line Y (Z) = AZ + B in FIG. 15 determine the shape parameter β = A = arctgφ, where the angle φ determines the degree of inclination of the line to the Z axis. From FIG. 15 shape parameter η = 0.33. If Y (Z) = B at Z = 0, then the value of the constant B from the graph of the linearized Weibull function is -13.81. The scale parameter is η = 9.9 · 10 ⁵ .

The threshold value of linear energy losses at the level σ = 0.25σ _{sat is} determined from the graph of the dependence of the failure cross section on LET. At σ = 4.75 · 10 ^-8 cm ^{2, the} value of LET _TH = 54 MeV · cm ² / mg. Thus, all the parameters of the IC type "D" are defined. Now, according to (8), it is possible to calculate the failure rate R = 3.25 · 10 ^-20 failure / bit-day.

, для всего устройства можно записать:After determining all the parameters of the IC constituting the electronic equipment KA, and taking into account

, for the whole device you can write:

. Величина

принята максимальной из всех возможных значений для полученных в ходе экспериментов с 6-ю типами ИМС. Максимальные значения параметров масштаба 7, получены для ИМС «D», что свидетельствует о сдвиге максимума плотности распределения σ(LET) в область более высоких значений аргумента (более тяжелых ионов), и об уменьшении значения σ_sati для конкретной ИМС. Значения LET_TH>120 МэВ/см²-мг свидетельствует об отсутствии необходимости проведения испытаний на устойчивость к воздействию эффектов SEE. Отсюда следует:As a result of calculations according to (25), taking into account the data in Table 3 and tab. four

. Value

accepted the maximum of all possible values for those obtained during experiments with 6 types of IC. The maximum values of the parameters of scale 7 were obtained for the “D” IC, which indicates a shift in the maximum distribution density σ (LET) to the region of higher argument values (heavier ions), and a decrease in σ _sati for a specific IC. Values of LET _TH > 120 MeV / cm ² mg indicate that there is no need to test for resistance to the effects of SEE. This implies:

1) The value of the critical particle fluence for the hypothetical apparatus of the spacecraft in this case will be

2) Calculations of the mathematical expectation for the entire system from

in view of (27)

where β _Z is the shape parameter for the whole system as a whole, γ _Z is the position parameter for the whole system as a whole, z is a random variable (including the radiation load level), they allow us to represent the relationship between the value lnlnΣ,

, его математическим ожиданием

, и дисперсией

, в видеWhere

his mathematical expectation

, and dispersion

, as

by analogy with (29)

obtained by expanding ln (Σ) in a Taylor series near the average value of ln (x _i ), namely, µ _i . From the data table. 4, it follows that the upper bound for the value of the function Y _V = 118, and the lower bound - Y _U = -30. As follows from the data table. 5 in this case, for a value of Y _{V, the} value of the argument X = ln (LET-LET _γ ) → ∞, and for Y _{U, the} quantity X → 0.

This means that for Y _{V the} quantity LET → ∞, and for Y _{U the} quantity LET → 0, which means that LET = LET _γ , i.e. position parameter. It follows that the maximum particle fluence leading to SEE effects from the action of the TLC for the value of Y _U is equal to F _UV = 1 / exp (3.5) = 3.6 × 10 ^-2 cm ^-2 , and the same value for F _UU = 1 / exp (16) = 5.3 × 10 ^-4 (see table 3 and table 4).

Then the lower acceptable limit of particle fluence is in the range

The ratio µ _Y / σ = 0.5774. In accordance with

we obtain the probability of maintaining operability P for each value of the damage mode using the relation (28), where Ф (х) is the generalized standard normal distribution function (quantile)

which is tabulated, for example, in [23]. Thus, P is assigned to the normal distribution, which follows from the fact that ln (x _i ) is a normal distribution, and the representation of lnln (Σ) as a sum of normally distributed random variables makes lnln (Σ) also normally distributed. The probability of maintaining the operability of a spacecraft with such ERI from (29) is equal to the quantile of the normal distribution

Thus, the entire procedure for assessing the intensity of failures and the parametric probability of maintaining the health of electronic devices of the spacecraft described above can be represented in the form of the following flowchart (Fig. 11).

B. Assessment of the equivalent absorbed dose D _{R of the} source of III and UBR to simulate the effects of SEU / SEE.

Below, we analyze the sensitivity of the CMOS / KND LIS technology to the possibility of SEE using the procedure proposed above.

Two methods are used for this. The first of them (B. 1) provides for the use of the cross-section of the faults for the entire apparatus in FIG. 16. The second method (B. 2) involves the use of design ratios to determine the threshold value LET _TH (Fig. 9)

When calculating by the method (B. 1), the algorithm shown in FIG. 9, implemented in Appendix “A”. The evaluation results are contained in table. PA.3 Annex "A".

When calculating by the method (B. 2) using the values of LET _TH from table. 2 and substituting their values in (5) for s _max = 28.4 μm and the corresponding RDEF values from table. PA.2 of Appendix “A” is obtained taking into account the criteria of the table. 3 [16] the values of the equivalent doses of D _R (γ-sq. 1 MeV-eq .- (Si)), which are summarized in table. PB.1 of Appendix "B".

, где

- длительность импульса МУ, P_R - предельная мощность дозы для квантов с энергией ~1 МэВ, обеспечивающая эквивалентную ионизацию ЧО для плотности потока F_crit=1,9·10⁶ част. см^-2 (или УБР).Value

where

- the duration of the MI pulse, P _R is the limiting dose rate for quanta with an energy of ~ 1 MeV, providing equivalent ionization of the HO for the flux density F _crit = 1.9 · 10 ⁶ part. cm ^-2 (or UBR).

[16] ее значение рассчитывают из соотношения (ПА. 10), которое в дальнейшем используют для определения величин

и эквивалентной дозы D_R(γ-кв.RS-экв.-(М)) моделирующего SEE в БИС технологии КМОП/КНД аналогичной технологии и конструктива источника гамма-рентгеновского излучения. Так при подстановке в (ПА.10) данных, соответствующих структурам («L»↓↑) и («М»↓↑): K_Σ и

из табл. ПА. 11 Приложения «A»; a=L, b=B, s_max из табл. 2, c=t_ПС=3,5·10^-5 см; w=3,6·10^-6 МэВ;

; LET_TH из (22), получают расчетные значения D_R или P_R для аналогичных структур МДП по конструкции и электрическим режимам эксплуатации. При s_max=28,8 мкм, L=17,5 мкм, B=3 мкм величина

, а величина

для структур («L»↓↑) и

для структур («M»↓↑).In order to increase the reliability of determining the integral value of constants

[16] its value is calculated from the ratio (PA. 10), which is then used to determine the values

and the equivalent dose D _R (γ-sq. RS-eq .- (M)) of the SEE simulating in LIS CMOS / LPC technology of a similar technology and the design of a gamma-ray source. So when substituting in (PA.10) the data corresponding to the structures (“L” ↓ ↑) and (“M” ↓ ↑): K _Σ and

from table PA 11 Annex "A"; a = L, b = B, s _max from the table. 2, c = t _PS = 3.5 · 10 ^-5 cm; w = 3.6 · 10 ^-6 MeV;

; LET _TH from (22), the calculated values of D _R or P _R for similar TIR structures according to their design and electrical operating conditions are obtained. With s _max = 28.8 μm, L = 17.5 μm, B = 3 μm, the value

, and the value

for structures (“L” ↓ ↑) and

for structures (“M” ↓ ↑).

An example of implementation according to Option 2 of determining the value of UBR when using MI using pulsed laser radiation

сбоев при воздействии импульсного лазерного излучения, определяют значение величины

с использованием из соотношения (3)

, величину

определяют из соотношения

[14]. В табл. 6 приведены значения переводных

для трех значении длин волн импульсного пикосекундного лазера: λ=590 nm, 600 nm, 800 nm для ИМС IDT 71256 (R-MOS) SRAM [14].To determine the UBR value when using MI using pulsed laser radiation, the heterostructure of the IFR or IC chip is irradiated with a picosecond laser source pulse with a duration of not more than 350 ps and a wavelength in the range of 800-1500 nm, the critical threshold energy is fixed

failures when exposed to pulsed laser radiation, determine the value of

using from relation (3)

, value

determined from the ratio

[fourteen]. In the table. 6 shows the values of the translated

for three wavelengths of a pulsed picosecond laser: λ = 590 nm, 600 nm, 800 nm for the IDT 71256 (R-MOS) SRAM ICs [14].

(ИМС IDT 71256 (R-MOS) SRAM), величина

, а величину эквивалентного значения

определяют из (14)

, и с учетом данных о σ_TOT[cm²]=1,26·10^-5 cm² (5V HM-6504 4K CMOS RAM, среднее значение сбоев

) и

[14], и s_max[µm] определяют величину интенсивности сбоев

из (9)For λ = 800 nm, the value

(IC IDT 71256 (R-MOS) SRAM), value

, and the equivalent value

determined from (14)

, and taking into account the data on _TOT [cm ² ] = 1.26 · 10 ^-5 cm ² (5V HM-6504 4K CMOS RAM, the average value of failures

) and

[14], and s _max [µm] determine the magnitude of the failure rate

from (9)

что соответствует lerr-chip/3,45 year. С использованиемand for the information capacity of IC 256 K, the intensity will be

which corresponds to lerr-chip / 3.45 year. Using

Для транзисторов малой мощности «L» из табл. ПА.2 значение

для источника X-Ray(10keV) и

, и для источника Co⁶⁰(1,25MeV) получают

, соответственно, при ориентации структуры "↑↕" по отношению к падающему излучению. Тогда D_R(γ-кв.RS-экв.-(М))=41,4…5,35rad(Si), и при известной длительности импульса гамма-квантов источника эталонного ионизирующего излучения τ_P УБР цифровой аппаратуры определяютdetermine the value of the equivalent total absorbed dose of gamma rays

For low power transistors "L" from the table. PA.2 value

for X-Ray source (10keV) and

, and for the source Co ⁶⁰ (1.25 MeV) receive

, respectively, with the orientation of the structure "↑ ↕" with respect to the incident radiation. Then D _R (γ-sq. RS-equiv .- (M)) = 41.4 ... 5.35rad (Si), and for a known pulse duration of gamma rays of the reference ionizing radiation source τ _P UBR digital equipment determine

") в приборном слое транзисторных структур ИМС; б) поперек ("↑↓")приборного слоя. Время пролета равно

, где E_ION[J] - энергия иона, m_ION[g] - масса иона.The UBR value can be calculated in two ways: 1) using the time of flight of ions through the HO; 2) according to the time of setting the equivalent dose on MU. When using the ion flight time, two cases are also considered: a) longitudinal flight ("

") in the instrument layer of the IC transistor structures; b) across (" ↑ ↓ ") the instrument layer. The flight time is

where E _ION [J] is the ion energy, m _ION [g] is the mass of the ion.

In the table. 10 shows the data used to determine the speed of a typical “cocktail” of ions. To obtain the smallest value of t _TRANS , a maximum energy E _{ION is} required with a minimum mass m _ION . Therefore, in the table. 7 shows the data on the estimates of the rate of He and Au ions.

In the table. 8 summarizes the results of UBR calculation for estimating the flight time of two TZZ for modern MOS technologies.

An example of implementation according to option 3 of determining the value of UBR by using official data of specifications for semiconductor devices and ICs.

. Критический заряд переключения равен

. Для определения величины хорды использовали эмпирическую формулу Q_C=0,35[рС]·s²[µm²], откуда

. Для определения величины порогового значения LET_TH использовали (36), при подстановке численных значенийAs an object of analysis, we chose the IC HX6656 Honeywels′s SOI RAM technology RICMOS ™ IV (Radiation Insensive CMOS) with the organization 32K × 8 bit. From the specifications on the IC, the values of the input capacitors of the IC are C _IN = 7pF, output C _OUT = 9pF. Level of switching of logical states by voltage

. The critical switching charge is

. To determine the value of the chord, the empirical formula Q _C = 0.35 [pC] · s ² [µm ² ] was used, whence

. To determine the threshold value of LET _{TH, we} used (36) when substituting numerical values

, величина

из (8). Тогда эквивалентная доза квантов эталонного источника Co⁶⁰ составит

.Value

, value

from (8). Then the equivalent dose of quanta of the reference source Co ⁶⁰ will be

.

.When the pulse duration of the test source X-Ray τ _P ≈ 1µS, the UBR value will be

.

Thus, the results of studies of radiation resistance to the effects of gamma-ray radiation from INR and EFU can be used to determine the resistance to SEE from TZZ using the appropriate RDEF coefficients. Therefore, the proposed method, which involves only irradiation with gamma-ray radiation of INR or EFU, can significantly reduce the cost of the tests, reduce their volume and increase the reliability of the test results.

, что позволяет выбирать режимы испытаний БИС технологии КМОП/КНД с учетом конструктивных особенностей и электрического режима эксплуатации на основании как расчетных, так и экспериментальных значений

.The dependence of the equivalent dose of gamma-ray radiation of INR or EFU was established according to the criterion of equivalence of critical RHE Q _C with the value

that allows you to select the test modes LSI technology CMOS / KND taking into account the design features and electrical operation based on both calculated and experimental values

.

The value of the equivalent dose of gamma-ray radiation of INR or EFU is reduced to the energy of quanta E _{γ, X-Ray} = 1 MeV-eq., Which makes it possible to use MU with an arbitrary spectrum of FS II.

APPENDIX A: Estimation of the equivalent absorbed dose D _{R of the} source of III and UBR to simulate the effects of SEU / SEE using the dependence of the fault cross-section for the entire equipment.

, суммарное сечение сбоев для элементов комбинаторной логики составит с учетом зависимости сечения от частоты

. Полное сечение для всей аппаратуры KA равно σ_TOTσ_SL+σ_Σ=(0,76+1,14)·10^-7=1,9·10^-7 cm². Из зависимости σ=f(LET) на Фиг. 16 следует, что ее аппроксимация ступенчатой функцией дает для σ_SAT=σ_TOT=1,9·10^-7 cm² значение LET_TH=53 MeV·cm²/mg.Suppose that the spacecraft apparatus contains a _SL = 40% of the same type of sequential logic ICs and a _SL = 60% of combinatorial logic circuits performed according to the standard CMOS technology. The number of sequential logic circuits N _i = 1000 units, the number of combinatorial logic circuits N _J = DFF = 114 units Clock frequency F = 10 MHz. The total fault section for elements of sequential logic will be

, the total cross section of failures for elements of combinatorial logic will be taking into account the dependence of the cross section on frequency

. The total cross section for the entire KA apparatus is σ _TOT σ _SL + σ _Σ = (0.76 + 1.14) · 10 ^-7 = 1.9 · 10 ^-7 cm ² . From the dependence σ = f (LET) in FIG. 16 implies that it gives an approximation of a step function σ _SAT = σ _TOT = 1,9 · 10 ^-7 cm ² value LET _TH = 53 MeV · cm ² / mg.

Instead of irradiating the CMOS / KND semiconductor LSI technology with electronic, proton radiation, or an SLC stream for SEE modeling, a limited sample of LSIs is irradiated with pulsed gamma-neutron ionizing radiation of the INR or pulsed X-ray radiation of an EFU with an equivalent dose that causes the generation of RIZ Q _t equal to that of the SLC in the black and white lens. The average energy of gamma-neutron radiation from the INR or X-ray radiation of an EFI is chosen to be 1.0-3.0 MeV [16].

To determine the equivalent dose D _R (γ-sq. RS-equiv .- (M)) of radiations of INR or EFU, causing in the material (M) equal to linear energy loss LET _TH for TZH value RIZ Q _t in the range LET _TH from units up to hundreds of MeV · cm ² / mg use the value of the RDEF coefficient of the dose of x-ray or gamma radiation in relation to the value of LET _TH from (2).

With equivalent ionization of the total volume of the MOS structure from gamma-X-ray radiation of AI and the local volume from KP ions, one can calculate the equivalent dose of gamma-ray radiation D _R providing the same ionization as the critical ion flux density (dose)

Here, the “RS” spectrum of the AI (Radiation Spectrum) is MU, and “M” is the material in which the energy of the AI (M∝Si or M∝SiO ₂ ) is absorbed. Equality (32) is also represented as

- длительность импульса ИИ МУ; t_ПР. - время пролета ТЗЧ через чувствительный объем, которое определяют из соотношенияwhere P _{R, I} is the absorbed dose rate of the corresponding sources of AI;

- pulse duration of AI MU; t _PR - time flight TZCh through a sensitive volume, which is determined from the ratio

where E is the energy of the TZZh; m _H - mass TZCh; s _max - the maximum length of the chord in the sensitive volume along the track of the primary particle. With an isotropic distribution of particles in space, the flight time will vary depending on the length of the chord, which varies from the thickness of the instrument layer t _PS at normal particle incidence to s _max with practically longitudinal particle propagation in a thin layer. The dose D _I from (PA.1) is determined by the ratio

The equivalent dose rate is the ratio of (PA.2) to (PA.1)

Coordination of the total absorbed dose of pulsed AI D _R (γ-square RS- (M)) in the MIS structure from MU with the RS spectrum with the equivalent dose from the effect of TZZh D _I (γ-square 1 MeV-eq. (M)) from (PA.4) is produced using the ratio (PA.6)

можно получить, отнеся величину критического заряда Q_crit=Q_C, создаваемую ТЗЧ в ЧО структуры МДП, к величине поглощенной дозы D_I от воздействующей ТЗЧ:Value of the constant

can be obtained by relating the value of the critical charge Q _crit = Q _C created by the TZC in the HO of the MDP structure to the absorbed dose D _I from the acting TZC:

where: ρ _c is the critical charge density; ν = a × b × c = A × t _PS - the volume of HO; A = a × b is the surface area of the HO; t _PS = s = s _min is the thickness of the instrument layer (Fig. 14). The value of Q _{C is} taken into account (2) from (PA.7)

where s - corresponds to the length of the spatial chord when a single TZCh falls at arbitrary angles to the front (inverse) surface of the chip structure (Fig. 10). The value of the absorbed dose from one TZCh take reduced to the entire volume of HO with an arbitrary value of the length of the chord 5:

Substituting (PA.8) and (PA.9) in (PA.4), we obtain

из (36) при подстановке (37) и (39) в видеand the value of the constant

from (36) when substituting (37) and (39) in the form

A similar equality can be written for dose rates P _R (γ-eq. RS- (M)) and P _I (γ-1MeV-eq .- (M)).

для источника рентгеновского излучения 10 кэВ и

для нуклидного источника Co⁶⁰ [16].In (PA.6) it is possible to accept

for an X-ray source of 10 keV and

for the nuclide source Co ⁶⁰ [16].

Given the known values of the parameters of the MOS structures and geometric parameters, the results of irradiation at one source can be corrected for the conditions of irradiation at another and based on (2) and (PA.1) the concept of relative efficiency coefficient (RDEF) [9]

in units

,

,

, а также преобразовать их в соответствующие значения для нуклидного источника Co⁶⁰:

,

,

(табл. ПА.1).Based on these data, you can determine the values

,

,

, and also convert them to the corresponding values for the nuclide source Co ⁶⁰ :

,

,

(tab. PA.1).

,

,

принимают следующие значения констант: ρ=2,33 [g/cm³] для Si;

- максимальное значение хорды в чипе структуры МДП; а - ширина чипа структуры МДП; b - длина чипа; с - высота чипа, [µm];

- постоянную генерации радиационно-индуцированных носителей заряда для МУ и одиночной ТЗЧ, предельное значение доли нерекомбинированного радиационно-индуцированного заряда в структуре МДП при воздействии излучения МУ и одиночной ТЗЧ в присутствии приложенного электрического поля напряженностью Е, [MV·cm^-1], из табл. ПА.1 Приложения «А» [16];

- фактор дозового накопления в структуре КНД гамма-рентгеновского излучения МУ со спектром квантов RS и одиночной ТЗЧ из табл. ПА.1 Приложения «А» [16].When calculating constants

,

,

take the following constant values: ρ = 2.33 [g / cm ³ ] for Si;

- the maximum value of the chord in the chip of the TIR structure; a - chip width of the TIR structure; b is the length of the chip; c is the height of the chip, [µm];

- the constant generation of radiation-induced charge carriers for MU and single TZC, the limit value of the fraction of unreinforced radiation-induced charge in the MIS structure when exposed to MU radiation and a single TZCh in the presence of an applied electric field of intensity E, [MV · cm ^-1 ], from table . PA.1 of Appendix “A” [16];

- the factor of dose accumulation in the structure of the directivity gain of gamma-x-ray radiation MU with a spectrum of quanta RS and a single TZZ from table. PA.1 Annex "A" [16].

определяют для энергии рентгеновских квантов E_X-Ray=10-300 кэВ и для энергии гамма-рентгеновских квантов ИЯР или рентгеновских квантов ЭФУ с энергией Е_ЯР,Х-Ray=1…6 МэВ [16]. Для учета конструктивных особенностей транзисторной структуры КНД, для оценки величины фактора дозового накопления ВЦ используют зависимость величины

от толщины подзатворного оксида t_ox для энергии квантов 10 кэВ и напряженности электрического поля E=1…6 МВ^·см^-1 [16]. Для ИИ ИЯР, ЭФУ, одиночной ТЗЧ с энергией квантов Е_{ИЯР,ЭФУ,ТЗЧ}≥1 МэВ принимают величину

.To take into account the kinetics of accumulation and relaxation of the charge of radiation-induced carriers in an electric field, the coefficient

determine for the energy of X-ray quanta E _X-Ray = 10-300 keV and for the energy of gamma-ray quanta of INR or X-ray quanta of EFU with energy E _{NR, X-Ray} = 1 ... 6 MeV [16]. To take into account the design features of the transistor structure of the KND, to assess the magnitude of the factor of dose accumulation of the CC use the dependence of the value

from the thickness of the gate oxide t _ox for the quantum energy of 10 keV and the electric field strength E = 1 ... 6 MV ^· cm ^-1 [16]. For AI INR, EFU, single TZCh with energy of quanta E _{INR, EFU, TZH} ≥1 MeV take the value

.

для источника рентгеновского излучения с энергией 10 кэВ и

для ИЯР или ЭФУ. Величину

для ТЗЧ вычисляют с использованием соотношения (ПА.23) Приложения «А» [16], а величины

и

для соответствующих значений энергий из соотношений, приведенных в табл. ПА.1 Приложения «А» [16].In the manufacture of a dielectric of a gate oxide of a MIS structure based on silicon dioxide (M = SiO ₂ ), the radiation generation constants of electron-hole pairs in it are taken equal to

for an X-ray source with an energy of 10 keV and

for INR or EFU. Value

for TZCh calculated using the ratio (PA.23) Appendix "A" [16], and the values

and

for the corresponding energies from the ratios given in table. PA.1 Annex "A" [16].

, для источников гамма-рентгеновского излучения ИЯР, гамма-квантов нуклидного источника Co⁶⁰ приведены в табл. 4 [16] для транзисторных структур МДП различного класса мощности: низкой (low - «L»); средней (middle - «M»); большой (big - «B»).Data on the calculation of the RDEF value for X-ray radiation of an EFU with a quantum energy of 10 KeV keV, TZh s

, for gamma-ray sources of INR, gamma-quanta of the nuclide source Co ⁶⁰ are given in table. 4 [16] for MOS transistor structures of various power classes: low (low - “L”); middle (middle - “M”); large (big - "B").

и

удобнее использовать тестовые транзисторы МДП при аналогичных параметрах А и t_ox, что и у транзисторов МДП основной схемы.For determining

and

it is more convenient to use MIS test transistors with the same parameters A and t _ox , as for MIS transistors of the main circuit.

(ПА.12) для двух источников ИИ: Х-Ray и ИЯР (Co⁶⁰) с энергией гамма-квантов E_γ-кв.≥1 MeV для транзисторных структур МДП различного класса мощности: низкой (low - «L»); средней (middle - «М»); большой (big - «В»).In the table. PA.2 shows RDEF and a translation factor

(PA.12) for two sources of AI: X-Ray and INR (Co ⁶⁰ ) with gamma-ray energy E _γ-sq. ≥1 MeV for MOS transistor structures of various power classes: low (low - “L”); middle (middle - "M"); large (big - "B").

The results of the sensitivity assessment of the spacecraft equipment from the experimental dependence in FIG. 16 to the effects of SEU / SEE are summarized in table. PA. 3.

ступенчатой функцией, определяют поток частиц для случая LET<LET_TH, достаточный для образования SEE.Approximating the dependence in FIG. 16 cross-section σ [cm ² ] the sensitivity of the devices to SEE from TZCh from

step function, determine the particle flux for the case LET <LET _TH , sufficient for the formation of SEE.

To do this, use the fact that the cross section A _t HO of the TIR structures of power "Z" and "M" according to the table. 3 is equal to the cross section σ _SEE (LET _TH ) for SEE in FIG. 12, i.e. A _t = L × B = σ _SEE (LET _TH ). The reciprocal cross section σ _SEE (LET _TH ) allows us to estimate the critical particle flux, for which only one particle with probability 1 falls into the BH of the MIS structure. Using the data table. 3 E = 17.5 μm = 1.75 · 10 ^-3 cm and B = 3.0 μm = 3.0 · 10 ^-4 cm, an estimate of the cross section gives the value of σ _SEE (LET _TH ) = 5.25 · 10 ^{- 7} cm ² , and the critical fluence of the particles is F _crit = 1.9 · 10 ⁶ part ^· cm ^-2 .

APPENDIX B: Estimation of the equivalent absorbed dose D _{R of the} source of III and UBR to simulate the effects of SEU / SEE using calculated ratios to determine the threshold value LET _TH

Using this data, we can make a first approximation of the calculation of the threshold value LET _TH , the minimum necessary to create SEU / SEE ..

The minimum value of LET _TH will correspond to the case for which the failure can be calculated from relation (2) in the form:

Then equality (PA.4) can be written in the form:

or

or

In FIG. Figure 9 shows the most common forms of relationships for determining the threshold value of LET _TH for a wide class of topological dimensions of modern MOS structures

In this case, the equality of the radiation-induced charge Q _t in the sensitive LSI volume is maintained under the influence of both the radiation of the ME and linear energy losses LET.

, где

- длительность импульса МУ, P_R - предельная мощность дозы для квантов с энергией ~1 МэВ, обеспечивающая эквивалентную ионизацию ЧО для плотности потока F_crit=1,9·10⁶ част.см^-2 (или УБР).When calculating according to option (B.2), the LET _TH values from Table 2 and, substituting their values in (5) for s _max = 28.4 μm and the corresponding RDEF values from table. PA.2 of Appendix “A” is obtained taking into account the criteria of the table. 3 [16] the values of the equivalent doses of D _R (γ-square MeV-eq .- (Si)), which are summarized in table. PB.1. Value

where

- the duration of the MI pulse, P _R is the limiting dose rate for quanta with an energy of ~ 1 MeV, which provides equivalent _HO ionization for the flux density F _crit = 1.9 · 10 ⁶ part.cm ^-2 (or UBR).

ее значение рассчитывают из соотношения (ПА.11), которое в дальнейшем используют для определения величин RDEF и эквивалентной дозы D_R(γ-кв.RS-экв.-(М)) моделирующего SEE источника гамма-рентгеновского излучения БИС технологии КМОП/КНД аналогичной технологии и конструктива. Так при подстановке в (ПА.11) данных, соответствующих структурам («L»↓↑) и («M»↓↑): K_Σ и RDEF из табл. ПА.2 Приложения «A»; a=L, b=B, s_max из табл. 5, c=t_ПС=3,5·10^-5 см; w=3,6·10^-6 МэВ;

; LET_TH из (ПА.29) Приложения «А» [16], получают расчетные значения D_R или P_R для аналогичных структур МДП по конструкции и электрическим режимам эксплуатации. При s_max=28,8 мкм, L=17,5 мкм, B=3 мкм величина

, а величина

для структур («L»↓↑) и

для структур («M4»↓↑).In order to increase the reliability of determining the integral value of constants

its value is calculated from the relation (PA.11), which is then used to determine the RDEF values and the equivalent dose D _R (γ-sq. RS-equiv .- (M)) of the SEE simulating source of gamma-ray radiation of the BIS CMOS / KND technology similar technology and design. So when substituting in (PA.11) the data corresponding to the structures (“L” ↓ ↑) and (“M” ↓ ↑): K _Σ and RDEF from Table PA.2 Annex "A"; a = L, b = B, s _max from the table. 5, c = t _PS = 3.5 · 10 ^-5 cm; w = 3.6 · 10 ^-6 MeV;

; LET _TH from (PA.29) of Appendix A [16], the calculated values of D _R or P _R for similar TIR structures in terms of design and electrical operating conditions are obtained. With s _max = 28.8 μm, L = 17.5 μm, B = 3 μm, the value

, and the value

for structures (“L” ↓ ↑) and

for structures (“M4” ↓ ↑).

List of cited sources

1. Dodd R.E., Sexton F.W. "Critical Charge Concepts for CMOS SRAMs" // IEEE Transaction on Nuclear Science, Vol. 42.-No. 6. - December 1995. - pp. 1764-1771.

2. Benedetto J.M., Eaton P.H., Mavis D.G., Gadlage M., Turfinger T. Digital Single Evet Trasient Trends With Technology Node Scaling // IEEE Transaction on Nuclear Science, Dec. 2006. - vol. 53. - No. 6. - pp. 3462-3465.

3. Berg M., Wang J.-J., Ladbury R., Buchner S., Kim H., LaBel K., Phan A. and Friendlich M. An Analysis of Single Event Upset Dependencies on High Frequency and Architectural Implementations within Actel RTAX-S Family Field Programmable Gate Arrays / IEEE Transactions on Nuclear Science, Dec. 2005. - Vol. NS-52. - No. 6. - pp. 3569-3574.

4. Chugg AM, Jones R., Moutre MJ, Duncan PH, Sorensen R. Harboe, Mattisson S., Larsson S., Fitzgerald R. and O'Shea T. Laser Simulation of Single Event Effects in Pulse Width Modulators / IEEE Transactions on Nuclear Science, Dec. 2005. - Vol. NS-52. - No. 6. - pp. 2487-2494.

5. Ferlet-Cavoris V., Paillet P., McMorrow D., Torres A., Gaillardin M., Melinger JS, Knudson AR, Cambell AB, Schwank JR, Vizkelethy G., Shaneyfelt MR, Hirose K., Faynot O. , Jahan C., Tosti L. Direct Measurement of Transient Pulses Induxed by Laser and Heavy Ion Irradiation in Deca-Nanometr Devices // IEEE Transactions on Nuclear Science, Dec. 2005. - Vol. 52. - No. 6. - pp. 2104-2113.

6. McMorrow D., Melinger J.S., Bucher S., Scott T., Brown R. D., Haddad N.F. Application of pulsed laser for evacuation and optimization of SEU-Hard designs // IEEE Transactions on Nuclear Science, Jun. 2000 .-- vol. 47. - No. 3. - p. 559.

7. Kabalnov Yu.A., Kachemtsev A.H., Kiselev V.K. Method for coulometric measurement of electrical parameters of nanostructures of an n-MOS transistor in CMOS / KNS and CMOS / KNI / Patent invention RU 2439745 C1 with priority. from 10/18/2010.

8. Kachemtsev A.N., Kiselev V.K., Palitsyna T.A. A device for coulometric measurement of the electrophysical parameters of the nanostructures of the n-MOS transistor in the CMOS / KNS and CMOS / KNI / Patent invention RU 2456627 C1 with priority. from 11/31/2011.

9. Kachemtsev A.N., Kiselev V.K., Skupov V.D., Torokhov S.L. A method for determining the coefficient of relative efficiency and equivalent dose of an x-ray source / Patent for invention RU 2480861 C1 with priority. from 08/31/2011.

10. Dooley J.G. SEU Immune Latch for Gate Arrey, Standarte Cell and other ASIC Application / US. Pat. No. 531070.

11. Galin T., Nicolaidis M. and Velazco R. Upset Hardened Memory Design for Submicron CMOS Technology / IEEE on Transactions, Vol. 43. - No. 6. - December 1996. - pp. 2874-2878.

12. U.S.A. National Technology Roadmap for Semiconductors.

13. Busher S.M., Daze D., Brown D., McMorrow D. and Melinger J. Comparision of Error Rates in Combinatorial and Sequential Logic // IEEE on Nuclear Science, Vol. 44. - No. 6. - December 1997. - pp. 2209-2216.

14. Gadlage M.J., Eaton P.H., Benedetto J.M. and Turfinger T.L. Comarison of Heavy Ion and Proton Inducced Combinatorial and Sequental Logic Error Rates in a Deep Submicron Process / IEEE Transactions on Nuclear Science, vol. NS-52. - No. 6. - Dec. 2005 .-- pp. 2120-2124.

15. Petrsen E.L., Pouget V., Messengill L.W., Buchner S.P. and McMorrow D. Rate Predictions for Single-Event Effecs - Critique II / IEEE Nransactions on Nuclear Science, Dec. 2005. - Vol. 52. - No. 6. - pp. 2158-2167.

16. Kachemtsev A.H., Kiselev V.K., Torokhov S.L. Test method for semiconductor LSI CMOS / KND technology for resistance to the effects of single failures from exposure to heavy charged particles (TZZ) // Patent for invention RU 2495446 C2 with priority. from 10.17.2011.

17. Buchner S., Baze M., Brown D. McMorrow D. and Melinger J. Comparision of Error Rates in Combinatorial and Sequental Logic / IEEE Transactions on Nuclear Science, Dec. 2004. - Vol. NS-51. - No. 6.- pp / 3480-3485.

18. Benedetto J., Eaton P., Avery K., Mavis K., Gadlage M., Turfinger T., Dodd P. and Vizkelethyd G. Heavy Ion Induced Digital Single-Event Transients in Deep Submicron Processes / IEEE Transactions on Nuclear Science, Dec. 2004. - Vol. NS-51. - No. 6. pp. 3480-3485.

19. Robinson P., Lee W., Aguero R., Gabriel S. Anomalies Due to Single Event Upsets / Journal of Spacekraft and Rockets, Mar-Apr., Mar.-Apr. 1994. - Vol. 31. - No. 2 - pp. 166-171.

20. Pickel J.C., Blanford J.T. Cosmic Ray Induced Errors in MOS Memory Cells / IEEE Transactions on Nuclear Science, Dec. 1978. - Vol. NS-25. - No. 6. - pp. 1166-1171.

21. LaBel K. Single Event Effects Specification ″, radhome.gsfc.nasa.gov/radhome/papers/seespec.htm, 1993.

22. Kachemtsev A.H., Kiselev B.K., Korsakova N.G. A method for determining the resistance of electronic components and blocks of electronic equipment to the effects of ionizing radiation / Patent for invention RU 2504862 C1 with priority. from 07/17/2012.

23. Gnedenko B.B., Belyaev Yu.K., Soloviev A.D. Mathematical methods in the theory of reliability. - M .: Science. Home Ed. FML. 1965 .-- 524 s.

24. Dozier S.M., Brown D.B. The Use of Low Energy X-Ray for Device Testing - a Comparision Whith Co-60 Radiation // IEEE Transactions on Nuclear Science, 1975. - Vol. NS-30. - No. 6. pp. 4382-4387.

Claims (27)

1. A method for assessing the resistance of digital electronic equipment to the effects of ionizing radiation (option 1), in which experimental studies are performed of the dependence of the failure cross section on the linear energy transfer σ = f (LET) for each heavy charged particle (TZ) from a given ionic “cocktail” ", is determined from this dependence of the cross-sectional area σ _SAT saturation fault, determine the minimum flow HCP causing malfunction effects determined threshold value linear energy losses LET _TH, corresponding to guise σ _SAT, determine the magnitude of the critical charge Q _C generation failure effect, are used to estimate the value R _SEU faults intensity, characterized in that to determine the level bessboynoy work (UBR) unit or the whole device electronics (CEA), as a whole, comprising digital integrated circuits (IMS) in its composition, according to the results of experimental studies on modeling plants (MU), determine the coefficient of relative efficiency (COE) $$RDE{F}_{ION}^{RS=C{o}^{60}}$$

for the threshold value LET _ТН and the total absorbed dose D _{RS of the} equivalent source of gamma quanta Co ⁶⁰ with an average quantum energy E _γ = 1.25 MeV, and for a given pulse duration τ _{P of} ionizing radiation, the dose rate of ionizing radiation (AI) is determined

for MU, which corresponds to the level of trouble-free operation (UBR). 2. The method according to p. 1, characterized in that to determine the cross-section of the failures of the digital device analyze the structural diagram of the electronic device, determine the values of individual partial cross-sections of failures

for each digital element of the circuit, the total cross section σ _Σ for the same elements of sequential and combinational logic, the average cross section σ _D for elements of sequential and combinational logic, the total cross section σ _TOT for the entire digital device. 3. The method according to p. 2, characterized in that to reduce the number of tests, the value of the total cross-section of failures is introduced taking into account the number of digital bistable DFF triggers. 4. The method according to p. 2, characterized in that in order to reduce the number of tests at the plants simulating the impact of the current transformer, to determine the unit failure cross section σ _e for the same elements of the circuit, for most practical cases the failure rate is determined using an approximating expression for the saturation cross section σ _SAT = ab, (a, b> c) and the critical threshold value of linear energy transfer (mass loss)

where c is the thickness of the semiconductor wafer, a and b are the dimensions of the chip structure, in microns, a Q _C is the value of the critical radiation-induced (RIZ), which causes malfunctions in the operation of logic devices. 5. The method according to p. 2, characterized in that for taking into account the frequency dependence of the effects of SEU in the combinational logic, the value of the partial cross section of failures $${\sigma }_{CL}^i$$

the same elements of combinational logic are introduced taking into account the clock frequency (synchronization frequency). 6. The method according to p. 5, characterized in that to take into account the frequency dependence of the SEU effects in the elements of combinational and sequential logic, to take into account the dependence of the fault cross-section for the same elements of combinational logic on the synchronization frequency on the approximating dependence of the fault cross-section for the same elements of combinational logic

value $${\sigma }_{SL}^i$$

determined by a given frequency of the clock pulses and

the magnitude of the fault section for the same type of elements of sequential logic is determined using the relation

7. The method according to any one of paragraphs. 3 or 6, characterized in that to unify the results of evaluating the sensitivity to SEU effects of elements of combinational and sequential logic, to determine the total cross-section of faults for elements of serial and combinational logic, the cross-section of faults σ _SL for elements of sequential logic is determined from $${\sigma }_{SL}={\displaystyle \sum _i{\sigma }_{SL}^i}$$

Where $${\sigma }_{SL}^i$$

- the partial cross section of errors of the ith element, and the cross section of failures σ _SL for elements of combinational logic determine $${\sigma }_{SL}={\displaystyle \sum _j{\sigma }_{SL}^j}$$

where $${\sigma }_{SL}^j$$

- the partial cross section of errors of the jth element, the total error cross section of the serial logic circuit σ _{Σ is} found using the relation

where a _CL is the percentage of combinatorial logic circuits; DFF is the number of fixing elements (triggers) in the circuit, and the total cross section of failures in the circuit is determined using the relation
$${\sigma }_{TOT}=a_{SL}[\frac{\%}{\mathrm{one\; hundred}}]{\sigma }_{SL}[\mathrm{from}{m}^2]+{\sigma }_{\sum },$$

8. The method according to p. 1, characterized in that to obtain the result of the assessment of the equivalent dose for the reference spectrum of the nuclide source Co ⁶⁰ , based on the spectrum of linear energy transfer

the sensitive volume ν _{S of the} typical transistor structure of a single digital element and the maximum value of the spatial diagonal s _MAX (chords) of the sensitive volume, determine the threshold value of the linear energy transfer LET _TH and the magnitude of the failure rate

determine COE $$RDE{F}_{ION}^{RS=C{o}^{60}}$$

for the threshold value LET _TH and the total absorbed dose TID of the equivalent source of gamma rays Co ⁶⁰ with an average quantum energy RS = 1.25 MeV and determine the value of TID. 9. The method according to p. 8, characterized in that for assessing the intensity of the failures from the effects of SEU, to obtain an experimental dependence of the cross-section of the failures of the REA or IC units on the LET of the form

on the corresponding MU reproducing ion beams (TZCH), irradiate the objects under study in the range of specified fluences of TZCh from the LET spectrum of the form

where dN / dE is the energy distribution of the TCD number, the sensitive volume ν _{S of the} transistor heterostructure, in which the critical charge Q _{c is formed} , is determined, the maximum size of the chord s _MAX (spatial diagonal) is determined in the sensitive volume (PR) model in the form of a {Rectangular Parallel Piped = RPP) from s ² = a ² + b ² + c ² , where a = b = L is the topological size of the HO element, c = t is the thickness of the heterostructure in the HO, determine the threshold value LET _TH on the dependence

for a full solid angle, the failure rate is determined using the ratio

10. The method according to p. 1, characterized in that for the conversion of the results of an experimental evaluation of the sensitivity to SEU effects of digital equipment when tested in INR and EFU simulating the impact of a current-frequency response on the basis of the identity of the critical charge Q _C , the coefficient of relative efficiency is determined $$RDE{F}_{ION}^{RS}$$

in units

where M is the semiconductor material in the black hole of the transistor structure, and the equivalent dose of gamma radiation D _RS (γ - sq. RS-equiv .- (M)) is determined from the relation

11. The method according to p. 10, characterized in that for the implementation of the experiment on a pulsed nuclear reactor (INR) or electrophysical installation (EFU) generating pulsed gamma-ray radiation, with a known pulse duration of gamma-ray quanta of the source of the reference AI τ _P for digital instruments determine dose rate

AI from the ratio of the equivalent absorbed dose to the duration of this pulse

which is taken equal to UBR. 12. A method for assessing the resistance of digital electronic equipment to the effects of ionizing radiation (option 2), in which a decoupled crystal (chip) of a heteroepitaxial structure of a semiconductor device (IFP) or an IC is irradiated with a pulsed laser source, the critical threshold energy is determined $$E_{TH}^{LASER}$$

failures, characterized in that to determine the magnitude of the UBR of the SPP or IC with the equivalent critical charge Q _{with the} action of a pulsed AI with a quantum energy of ~ 1 MeV-eq. (M), an estimate of the threshold value $$E_{TH}^{LASER}$$

produced when exposed to pulsed laser radiation (OR), then using the corresponding values of the relative efficiency coefficients $$RDE{F}_{ION}^{RS}$$

between the effective dose D _{RS of} gamma-ray radiation with the RS spectrum and $$LE{T}_{TH}^{ION}$$

TZCh (ions), $$RDE{F}_{LASER}^{RS}$$

between energy failures $$E_{TH}^{LASER}$$

from the laser pulse and the effective dose D _{RS of} gamma-ray radiation, determine the value of the equivalent critical charge Q _c , the magnitude of the failure rate R, the effective dose D _RS and the known pulse duration of the AI τ _P determine the UBR. 13. The method according to p. 12, characterized in that for assessing the intensity of failures during MI, taking into account data on σ _TOT [cm ² ], s _MAX [μm] on the dependence σ [cm ² ] = f (LET) determine the intensity failures using ratio

14. The method according to p. 12, characterized in that in order to obtain adequate conditions with the effect of TCD, when using simulation methods (IM) using OR, the heterostructure of the IFR or IC chip is irradiated with a picosecond laser source pulse with a duration of not more than 350 ps and a wavelength of the range of 800-1500 nm. 15. The method according to p. 12., characterized in that to obtain the equivalent value of the critical charge Q _c , which causes in the critical volume ν _{S the} occurrence of external currents that cause the effects of SEU and SET, to implement photoionization of the transistor structures, use the value $$RDE{F}_{LASER}^{ION}=\frac{LE{T}_{TH}^{ION}[\frac{MeV\cdot c{m}^2}{mg}]}{E_{LASER}[pJ]}{|}_{Q_C=const}$$

that establishes a relationship between the result of irradiating an object using a PLC and OR, and $$RDE{F}_{ION}^{RS}=\frac{D_{RS}(\gamma -\mathrm{to}\mathrm{at}.RS-\mathrm{uh}\mathrm{to}\mathrm{at}.-(M))}{LE{T}_{TH}^{ION}}{|}_{Q_C=const}$$

, which establishes the relationship between the result of irradiation of an object using pulsed ionizing radiation (III) and TZCh, get the value $$RDE{F}_{LASER}^{RS}=\frac{D_{RS}(\gamma -\mathrm{to}\mathrm{at}.RS-\mathrm{uh}\mathrm{to}\mathrm{at}.-(M))}{E_{LASER}(pJ)}{|}_{Q_C=const}=RDE{F}_{ION}^{RS}\times RDE{F}_{LASER}^{ION}$$

using which, in turn, the effective absorbed dose of gamma-ray radiation D _RS (γ-sq. RS-equiv .- (M)) of the reference source Co ⁶⁰ for the material (M) is determined (for example, as M = Si, SiO ₂ ) and for a known pulse duration of gamma rays of the source of the reference AI τ _P UBR digital equipment is determined from the ratio of the equivalent absorbed dose to the duration of this pulse

16. The method according to p. 15, characterized in that for fixing the effectiveness of the MI, when using the MI using a pulsed laser source, a critical threshold energy is fixed $$E_{TH}^{LASER}$$

failures in units [pJ]. 17. The method according to p. 15, characterized in that for the conversion of the results of the MI, the value of $$E_{TH}^{LASER}$$

determined using the ratio

18. The method according to p. 17, characterized in that in order to fix the effectiveness of the MI, the value $$RDE{F}_{LASER}^{ION}$$

for the wavelength of laser radiation λ _LASER = 800-1500 nm is taken equal

19. The method according to p. 18, characterized in that for converting the results of the MI into the equivalent effect of the TZCh, the value of the equivalent value $$LE{T}_{ION}^{E\mathrm{TO}\mathrm{AT}.}$$

determined from
$$LE{T}_{ION}^{E\mathrm{TO}\mathrm{AT}.}[\frac{MeV\cdot c{m}^2}{mg}]=0.67\times LE{T}_{TH}^{LASDER}[\frac{MeV\cdot c{m}^2}{mg}]$$

. 20. The method according to p. 19, characterized in that for determining UBR with MI, in units equivalent to the effects of sources of pulsed AI, using
$$D_{RS}(\gamma -\mathrm{to}\mathrm{at}.RS-\mathrm{uh}\mathrm{to}\mathrm{at}.-(M))=RDE{F}_{ION}^{RS}\times RDE{F}_{LASER}^{ION}\times LE{T}_{TH}[\frac{M\mathrm{uh}B\cdot \mathrm{from}{m}^2}{mg}]$$

determine the value of the equivalent total absorbed dose of gamma rays D _RS (γ-squared RS-eq .- (M)) and for a known pulse duration of gamma rays of the reference ionizing radiation source τ _P UBR digital equipment determine

21. The method according to any one of paragraphs. 12 or 19, characterized in that for the use of engineering methods for assessing the intensity of failures from the effects of TZCh, transform
$$LE{T}_{ION}^{E\mathrm{TO}\mathrm{AT}}[\frac{MeV\cdot c{m}^2}{mg}]$$

at $$LE{T}_{ION}^{E\mathrm{TO}\mathrm{AT}}[\frac{pF}{\mu m}]$$

. 22. A method for assessing the resistance of digital electronic equipment to the effects of ionizing radiation (option 3), characterized in that they determine the level of smooth operation (UBR) based on the use of dynamic electrical parameters of the specification for electronic components to obtain an equivalent critical charge $$Q_{\mathrm{FROM}}^{E\mathrm{TO}\mathrm{AT}}$$

, which causes the occurrence of external currents in the CH ν _S , generating SEU and SET effects in digital electronic circuits, in the method for assessing the resistance of digital equipment to the effects of AI, data on the value of the dynamic electrical parameters of the specifications for the electronic components are used. 23. The method according to p. 22, characterized in that for determining the equivalent value $$LE{T}_{TN}^{E\mathrm{TO}\mathrm{AT}}$$

by value of equivalent critical charge $$Q_{\mathrm{FROM}}^{E\mathrm{TO}\mathrm{AT}}$$

, as a dynamic parameter, the switching times from one logical state to another are used

and the corresponding value of the supply voltage of the IC. 24. The method according to p. 22, characterized in that for determining the equivalent value $$LE{T}_{TN}^{E\mathrm{TO}\mathrm{AT}}$$

by value of equivalent critical charge $$Q_{\mathrm{FROM}}^{E\mathrm{TO}\mathrm{AT}}$$

corresponding to the effect of the “ion cocktail”, the levels of the output voltages of the logical states V _0L and V _0H and the corresponding value of the capacitance at the output of the monitored signal are also used as a dynamic parameter. 25. The method according to p. 22, characterized in that for determining the equivalent value $$LE{T}_{TN}^{E\mathrm{TO}\mathrm{AT}}$$

by value of equivalent critical charge $$Q_{\mathrm{FROM}}^{E\mathrm{TO}\mathrm{AT}}$$

corresponding to the effect of the “ion cocktail”, direct data on the switching charge Q _PER between these logical states are used as a dynamic parameter. 26. The method according to any one of paragraphs. 23, 24, 25, characterized in that for determining the equivalent value $$LE{T}_{TN}^{E\mathrm{TO}\mathrm{AT}}$$

by value of equivalent critical charge $$Q_{\mathrm{FROM}}^{E\mathrm{TO}\mathrm{AT}}$$

corresponding to the effects of "ion cocktail", use the ratio

or $$LE{T}_{TN}^{E\mathrm{TO}\mathrm{AT}}=Q_{\mathrm{FROM}}^{E\mathrm{TO}\mathrm{AT}}/\mathrm{from}$$

, and then determine the equivalent value $$LE{T}_{TN}^{E\mathrm{TO}\mathrm{AT}}$$

, and the conversion of the values of the dynamic switching parameters into the effective value of the UBR from the action of AI with the RS spectrum is performed using the corresponding values of the relative efficiency coefficients $$RDE{F}_{Q_{\mathrm{FROM}}^{E\mathrm{TO}\mathrm{AT}}}^{RS}$$

between the equivalent dose D _{RS of} gamma-ray radiation with an RS spectrum and $$LE{T}_{TN}^{E\mathrm{TO}\mathrm{AT}}$$

Corresponding to generate an equivalent dynamic parameters of the charge switching logic states, then determine the value of the equivalent dose D _RS and the known pulse duration τ _P AI equivalent source UBR determined. 27. A method for assessing the resistance of digital electronic equipment to the effects of ionizing radiation (option 4) by using experimentally obtained data when simulating REE in LSI / VLSI CMOS / LPC technology from the effects of TZZh KP, from the effects of the OR picosecond range, characterized in that to determine the level failure-free operation (UBR) of LSI / VLSI under the influence of an equivalent pulsed ionizing radiation source with a quantum energy of 1.25 MeV-eq. (M), the threshold value of linear transmission is selectively estimated energy (LET _VT ), is produced either from direct measurements of the dependence of the fault cross section on the LET value, or when exposed to pulsed laser radiation, or from data on the value of such electrical parameters of the specification on ECB elements as the times of switching from one logical state to another

, or the output voltage levels of the logical states V _OL and V _0H , or from the data on the switching charge Q _PER. between these states, then using the corresponding values of the coefficients of relative efficiency $$RDE{F}_{ION}^{RS}$$

between the equivalent dose of gamma-ray radiation D _RS and $$LE{T}_{TN}^{ION}$$

, $$RDE{F}_{LASER}^{ION}$$

between the energy of failures from a laser pulse $$E_{LASER}^{SEU}$$

and $$LE{T}_{TN}^{ION}$$

, $$RDE{F}_{ION}^{RS}$$

between the effective dose of gamma-ray radiation D _RS and $$LE{T}_{TN}^{ION}$$

, $$RDE{F}_{LASER}^{RS}$$

between the energy of failures from a laser pulse $$E_{LASER}^{SEU}$$

and the effective dose of gamma-ray radiation D _RS , determine the value of the equivalent dose D _RS and the known value of the pulse duration II τ _P determine the UBR.

Images