Single Event Upset Cross Section Research Articles

SEU cross sections derived from a diffusion analysis

A simple theoretical prediction of single-event upset (SEU) cross section versus linear energy transfer (LET) is derived from a diffusion analysis, and the result is compared to some real device curves. It was found that at least some real device curves show two regimes. One regime (high-LET) is characterized by a very good fit to the theoretical prediction, and the other (low-LET) is characterized by a very bad fit. The existence of a high-LET regime provides additional credibility for the increasingly popular postulate that diffusion has an important effect on the shape of the cross-sectional curve.

Single event effects and performance predictions for space applications of RISC processors

Proton and ion single event phenomena (SEP) tests were performed on 32-b processors including R3000A's from all commercial manufacturers along with the Performance PR3400 family, Integrated Device Technology Inc. 79R3081, LSI Logic Corporation LR33000HC, and Intel i80960MX parts. The microprocessors had acceptable upset rates for operation in a low earth orbit or a lunar mission such as CLEMENTINE with a wide range in proton total dose failure. Even though R3000A devices are 60% smaller in physical area than R3000 devices, there was a 340% increase in device single event upset (SEU) cross section. Software tests of varying complexity demonstrate that registers and other functional blocks using register architecture dominate the cross section. The current approach of giving a single upset cross section can lead to erroneous upset rates depending on the application software.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A comparison of two different types of geosynchronous satellite measurements during the 1989 solar proton events

The proton telescope aboard the GOES-7 satellite continuously records the proton flux at geosynchronous orbit, the therefore provides a direct measurement of the energetic protons arriving during solar energetic particle (SEP) events. Microelectronic devices are susceptible to single event upset (SEU) caused by both energetic protons and galactic cosmic ray (GCR) ions. Some devices are so sensitive that their upsets can be used as a dosimetric indicator of a high fluence of particles. The 93L422 1K SRAM is one such device. Eight of them are on the TDRS-1 satellite in geosynchronous orbit, and collectively they had been experiencing 1–2 upset/day due to the GCR background. During the large SEP events of 1989 the upset rate increased dramatically, up to about 250 for the week of 19 Oct, due to the arrival of the SEP protons. Using the GOES proton spectra, the proton-induced SEU cross section curve for the 93L422 and the shielding distribution around the 93L422, the calculated upsets based on the GOES satellite data compared well against the log of measured upsets on TDRS-1.

Heavy ion microscopy of single event upsets in CMOS SRAMs

The single event upset (SEU) imaging has been applied at the GSI heavy ion microprobe to determine the sensitivity of integrated circuits (IC) to heavy ion irradiation. This method offers the possibility to directly image those parts of an IC which are sensitive to ion-induced malfunctions. By a 3-dimensional simulation of charge collection across p-n-micro-junctions, we can predict SEU cross-sections. For a MHS65162 2k x 8bit CMOS SRAM we found two regions per bit with different sensitivity and measured a total cross-section of (71/spl plusmn/18) /spl mu/m/sup 2/ for a bitflip per cell and simulated 60 /spl mu/m/sup 2/ with an argon beam of 1.4 MeV/nucl. (LET of 19.7 MeV/(mg/cm/sup 2/)).< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

The shape of heavy ion upset cross section curves (SRAMs)

An expression for SEU (single event upset) cross section for a unidirectional, monoenergetic ion beam is developed. The expression incorporates both the stochastics of the energy deposition process and the variability of the response of an array of memory cells. The energy deposition stochastics result from the distribution of path lengths the ion may take through a memory cell and the statistics of the charge production process. The variability in the memory response results from the dependence of the sensitive volume size on deposited charge and cell-to-cell variations due to processing. Thus the shape of a measured upset cross section curve is the result of the complex interplay of a number of physical processes whose relative contributions vary with experimental conditions, memory technology, and processing. A procedure for the separation of the basic memory response from the stochastics of the energy deposition process using heavy ion SEU measurements is given. This can be used to predict the SEU vulnerability in any radiation environment. Results of the model applied to data for bipolar and CMOS/SOS SRAMs (static random-access memories) are presented. >

Effects of process parameter distributions and ion strike locations on SEU cross-section data (CMOS SRAMs)

The effect of statistical parameter distributions and stochastic ion strike locations in hardened memory arrays on observed SEU (single event upset) cross-section data is discussed. Application of numerical analysis to the parasitic bipolar gain distributions in SIMOX (separation by implantation of oxygen) SRAMs (static random-access memories) and the introduction of an effective critical charge based on sensitivities to ion strike locations explain the nonsaturating behavior of measured cross-section curves in SOI (silicon-on-insulator) and other hardened memories. A direct, analytical relationship between the measured cross-section curves seen in contemporary SOI CMOS SRAMs and the distribution of the parasitic bipolar beta values controlling each cell critical charge is presented. Results of this technique clearly indicate that distributed processing parameters and random ion strike locations lead to nonideal cross-section curves. The results also indicate that if the distribution for a critical parameter is known, the LET (linear energy transfer) threshold and shape of the upset cross-section curves can be anticipated.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Simple model for proton-induced latch-up

Computer simulations show that the strong angular dependence exhibited by proton-induced single-event latch-up can be explained by a simple mechanism. Latch-up occurs if, and only if, more than some threshold amount of energy is deposited within the sensitive volume. A procedure for determining the SEU (single event upset) parameters by comparing SEU cross sections and CUPID simulations at different incident energies and angles of incidence is described. The thickness of the sensitive volume and the value of the critical charge determined for the NEC 4464, a 64-kb CMOS SRAM (static random-access memory) agrees with the measured thickness of the p-well and the value of the threshold LET (linear energy transfer) determined using heavy ions. >

Rate prediction for single event effects-a critique

The authors review various single event effects (SEE) testing and rate prediction methodologies and recommend standard approaches. This discussion is limited to single event upset (SEU) rate prediction for direct-ionization-induced effects. The standard approach being recommended is based partially on a different way of viewing the results of SEU cross-section measurements. The measurements are not measuring a distribution of cross-sections. They are measuring a distribution of device sensitivities, due to differences of sensitive region critical charges and to differences of charge collection. The linear energy transfer (LET), at which 50% of the cell population upsets, corresponds to the charge deposition necessary to upset the median cell in the circuit array. The threshold LET corresponds to the most sensitive region being hit in its most sensitive location, and does not represent the entire array. The shape of the cross-section curve is described by an integral Weibull distribution. The upset rate for a device should then be calculated using the differential rate of each sensitive region, combined with an integral weighting given by the Weibull distribution that describes the measured cross-section curve. >

High energy proton SEU test results for the commercially available MIPS R3000 microprocessor and R3010 floating point unit

Proton single event upset (SEU) cross sections and proton total dose hardness of commercially available MIPS R3000 microprocessors (CPU) and R3010 floating point units (FPU) were obtained by exposing these parts to 256 MeV protons from the linear accelerator (LINAC) at the Los Alamos Meson Physics Facility (LAMPF). Parts from several manufacturers were tested. The CPUs and FPUs were tested dynamically during radiation exposure with specially designed assembly language codes which exercised a subset of the available instructions. Cross sections derived from the SEU data were used to calculate expected upset rates for a 500-km, 60-degree inclination orbit during quiet solar conditions and during the August 4. 1972, King solar fire event modeled by J.H. Adams et al. (1981).< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Variation in proton-induced upsets rates from large solar flares using an improved SEU model

Proton integral flux measurements during solar energetic particle (SEP) events, from the CPME instrument onboard the IMP-8 satellite, and the proton detector on the GOES-7 satellite, are utilized to calculate proton-induced single-event upset (SEU). An improved two-parameter SEU cross section model is used for three RAM devices. The log of the 239 actual RAM upsets on the TDRS-1 satellite during the October 19, 1989 SEP event allowed a comparison between the recorded number of upsets/chip per day and those predicted based on the GOES-7 data, the two-parameter model, and the shielding distribution curve. >

Depletion region geometry analysis applied to single event sensitivity

The effects of device geometry on single-event-upset (SEU) rate in space, beginning with the ground test measurements, are explored. Analysis of sensitive volume geometry is used to establish a relationship between it and SEU cross section. This relationship is used as a basis for geometric models wherein measured cross sections determine model shape parameters. Fits of one of these models to a range of experimental cross sections are exhibited. These fits and physical argument suggest that the actual sensitive volume shape differs from that assumed in most current calculations, i.e. a rounded shape with no lower corners is more appropriate than the box shape currently used. A correction that justifies a current heuristic practice of relating critical charge to an LET (linear energy transfer) on the shoulder of the cross section rather than to onset LET is devised. A further correction is obtained by a calculation using the omnidirectional chord distribution for a hemisphere. Together the two corrections show that nearly an order of magnitude error results from assuming that the sensitive volume is box shaped when calculating upset rate due to isotropic fluences.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A model for proton-induced SEU

The authors present a method for predicting proton-induced single-event upset (SEU) rates in devices exposed to given proton fluxes, within a particular spacecraft shielding. The approach uses experimental heavy-ion cross-section data, combined with nuclear reaction calculations, in order to determine the proton-induced SEU cross section versus proton energy relationship. Calculations for two devices, the Fairchild 93L422 RAM and the Intel 2164A dynamic RAM, for which heavy-ion test data were available, are presented and compared with other theoretical results and with ground-based experimental data. Available on-orbit SEU data are then compared with the present predictions. Predicted SEU rates for both protons and cosmic rays at various 60 degrees circular orbits are also compared. Good agreement is found in all cases.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

An exact, closed-form expression of the integral chord-length distribution for the calculation of single-event upsets induced by cosmic rays

In a previous paper [J. Appl. Phys. 50, 3799 (1979)] Bradford derived a distribution function for the chord lengths through a rectangular parallelepiped (a×b×c, a≤b≤c). This chord-length distribution describes the path-length distribution of cosmic-ray particles in an object exposed to the isotropic cosmic-ray flux outside the earth’s magnetosphere. It is an essential quantity needed for computing the rate of single-event upsets (SEUs) induced in space-borne random-access memory cells by cosmic rays. While Bradford’s distribution is exact for chord length x within the range 0&amp;lt;x≤b, it is only approximate for b&amp;lt;x≤w, where w=(a2+b2+c2)1/2. In this paper we present an exact distribution which is valid for all values of chord length (0&amp;lt;x≤w). The extent to which the two distributions differ is explored. It is found that the two distributions agree exactly for 0&amp;lt;x≤b, but are very different for b&amp;lt;x≤w, and that this difference can lead to serious errors when Bradford’s distribution is used for computing the SEU rate. We also present an integral for computing the SEU rate in which the actual shape of the SEU cross-section curve is explicitly taken into consideration. Our formulation is much easier to evaluate numerically than the one proposed recently by Shoga et al. [IEEE Trans. Nucl. Sci. NS-34, 1256 (1987)].

A method for characterizing a microprocessor's vulnerability to SEU

A system has been developed that tests microprocessors for single-event upset (SEU) at the specified clock speed and without adding wait or hold states. This system compiles a detailed record of SEU-induced errors and has been used to test the Sandia SA3000 microprocessor and prototypes of its commercial equivalent, the Harris H80C85 at the Lawrence Berkeley Laboratory 88-inch cyclotron facility. Using appropriate test programs and analyzing the resulting upset data, the authors have established the SEU cross section of the major functional elements of the hardened processors. With these cross sections and the estimated duty factors of a 'typical' program, they computed the expected upset rate in a parallel, normally incident, beam as a function of linear energy transfer and measured the rate in several cyclotron beams. Good agreement between the measured and calculated rates was obtained.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Single Event Upset Measurements of Gaas E-JFET RAMS

This paper reports the first single event upset measurements on a GaAs RAM. These measurements on fully operational 256-bit GaAs static RAMs were performed with 40-MeV protons at the NRL Cyclotron. The average upset cross section was 8 × 10-12 cm2/proton-bit. The single event upset cross section for these GaAs memories is comparable to the single event upset cross section measured in 40-MeV proton beams for representative Si memories. An estimate of the expected single event upset rate due to protons in 600 orbits for these GaAs memories is compared to representative Si memories.