Guy Berger

From the Reference SEU Monitor to the Technology Demonstration Module On-Board PROBA-II

The reference SEU Monitor system designed and presented in 2005 (R. H. SOslashrensen, F.-X. Guerre, and A. Roseng ldquoDesign, testing and calibration of a reference SEU monitor system,rdquo in Proc. RADECS, 2005, pp. B3-1-B3-7) has now been used by many researchers at many radiation test sites and has provided valuable calibration data in support of numerous projects. As some of these findings and results give new insight into improved inter-facility calibrations and provide additional inputs into ongoing SEE research, a few of the more interesting cases are presented. Furthermore the dasiadetector elementpsila, the Atmel AT60142F SRAM, now in a hybrid configuration, will form the key detector element in the Technology Demonstration Module (TDM) to be flown on-board the PROBA-II satellite, to be launched at the beginning of 2009. This flight opportunity extends the Reference SEU Database with both ground and space data, taken on the same device under identical operating conditions. Additionally, the Reference SEU Monitor concept is employed as the basis for the new Reference SEL Monitor system, currently under characterization and preparation for integration on the TDM. Ground SEU/SEL characterization of this latch-up experiment is also presented as well as the basic concept of the TDM, the PROBA-II Radiation Monitor module.

Linear Energy Transfer of Heavy Ions in Silicon

Researchers performing radiation testing on electronic components often rely on semi-empirical prediction codes for determining the linear energy transfer (LET) (or electronic stopping force) of ions, without paying much attention to their reliability. However, it is seen that estimations calculated with different codes can have over 10% discrepancies, especially in the case of heavy ions with higher LET (e.g., xenon). As a consequence of the modern component fabrication techniques this has become an important issue when studying the radiation durability of electronics. In order to clarify this inconsistency, LET measurements for 131Xe and 82Kr in silicon have been undertaken and obtained results are presented, discussed and compared with earlier predicted data.

The heavy ion irradiation facility at CYCLONE - a dedicated SEE beam line

After a two years evaluation and assessment period of the CYClotron of LOuvain la NEuve (CYCLONE) usability for heavy ion SEE testing, the European Space Agency has now initiated the set-up of a permanent beam line dedicated for SEE testing. This paper describes the CYCLONE accelerator, the beam line used so far, presents some experimental results which are compared with data obtained at the Tandem accelerator at Brookhaven National Laboratory (BNL) and details present and future developments.

Experimental Linear Energy Transfer of Heavy Ions in Silicon for RADEF Cocktail Species

Experimental linear energy transfer values of heavy ions in silicon are presented with comparison to estimations from different semi empirical codes widely used among the community. This paper completes the experimental LET data for the RADEF cocktail ions in silicon.

Simultaneous Evaluation of TID and Displacement Damage Dose Using a Single OSL Sensor

The possibility to evaluate simultaneously the total ionizing dose and the displacement damage dose using a single OSL integrated sensor is demonstrated. Results obtained during the real time monitoring of both proton and electron beams are presented

Acceleration and mass-separation of radioactive ion beams in an isochronous cyclotron

The authors report on first operational results obtained with the two-cyclotron concept for the production and acceleration of a /sup 13/N radioactive beam (T/sub 1/2/=10 min.). A pure beam of 70 particle pA of /sup 13/N at 0.63 meV/amu has been accelerated and used for cross-section measurements of the /sup 1/H(/sup 13/N, gamma )/sup 14/O reaction which is involved in astrophysical processes. Although the /sup 13/N beam is accelerated with a /sup 13/C beam at least 10/sup 3/ times more intense, the specific characteristics of a cyclotron make it possible to separate clearly these two species whose relative mass difference is only 1.8*10/sup -4/, without loss of intensity for the /sup 13/N beam. The /sup 13/C contamination on the final target is less than 1 particle pA. The advantages of a cyclotron over a linac as a post-accelerator for low-energy radioactive beams are discussed.<<ETX>>

Efficient single event upset simulations of a tolerant PD SOI CMOS D Flip-Flop

A simple way for modeling the single event upset (SEU) in Partially Depleted (PD) SOI CMOS circuits by Spice simulations is presented. A Verilog-A module connected to the body contact of PD SOI MOSFET is implemented to describe the transient current generated by an ion-track crossing the transistor. Verilog-A module is given by a physical based compact model and hence accounts for all variations in MOSFETs physical parameters (i.e. mobility, lifetime, etc.) caused by irradiation, temperature, bias conditions, etc. Good agreement with mixed-mode numerical simulations is observed at different conditions. Both kinds of simulations show that PD SOI CMOS D Flip-Flop is tolerant to high LET energies, whereas bias supply reduction as well as an increase in the clock frequency of the flip-flop can degrade this tolerance.

High-energy neutrons effect on strained and non-strained SOI MuGFETs and planar MOSFETs

A comparative investigation of high-energy neutrons effect on strained and non-strained devices with different geometries is presented. Both single-gate planar and multiple-gate (MuG) silicon-on-insulator (SOI) devices are considered. Device response to the neutron irradiation is assessed through the variations of threshold voltage and transconductance maximum. The difference between strained and non-strained device response to the high-energy neutrons exposure is clearly evidenced. The reasons for such a difference are discussed. Analysis of the experimental results allows for suggesting that strain relaxation is one of the probable causes.

SEE evaluation of a low-power 1μm-SOI 80C51 for extremely harsh environments

In this paper, we present the SEE characterization of an 80C51 microcontroller optimized for high temperature and low-power applications. Its microarchitecture has been completely redesigned with deep low-power optimizations. It has been manufactured in a 1μm SOI process with tungsten metallization layers ensuring a high sustained operating temperature of 225 °C. The SEE characterization presented here have been carried out in the Cyclotron Research Centre of Louvain-la-Neuve. It shows a very high threshold LET of 90 MeV/mg/cm2 at a nominal Vdd of 5 V, while it is latchup immune due to SOI. Thanks to the power-optimized microarchitecture, its power consumption is only 8.5 mW/MIPS, which is close to a LEON2 FT manufactured in a far more sensitive 0.18μm process, making it a very low-power 80C51 for extremely harsh environments. Hardening by using a harsh process allows to use standard design tools and advanced aggressive low-power techniques. Even if the SOI 1μm technology used here is very big, results obtained in this work show that it can compete for low-power microcontrollers with a 0.18μm sub-micron technology hardened by using power hungry tripple modular redundancy. Whereas this harsh process approach offers quite better SEE tolerance.