Abstract
The single event effect caused by space heavy ion radiation is one of the important factors affecting the safety and operation of spacecraft on orbit. In the research and evaluation of the frequency, spatial distribution and time characteristics of single event effects, linear energy transfer (LET) spectra of space radiation play an important role. On the Beidou navigation M15 and M16 satellites, a single event upset (SEU) and LET monitor was developed to obtain the upsets of the memory device and the LET spectra of space radiation which passes through the device. Through the measurement results from this monitor, the correlation between the device’s SEUs and the LET spectra could be studied.
1 Introduction
In space, the major radiation includes heavy ions of galactic cosmic rays, high energy protons in the radiation belt and high energy particles from the sun. The space radiation and their secondary particles produced can cause single event upset (SEU), single event latch-up (SEL), single event gate rupture, single event burnout and other single event effects of the spacecraft’s electronic components, and these effects could lead to the abnormal operation or even failure of the spacecraft. In the operation of spacecraft, handling and maintenance of various fault abnormalities caused by SEU and SEL are the most common (Li et al. 2019). Statistics on the failures of the satellites found that SEUs are the main cause of on-orbit failures, and the failures caused by SEUs accounted for 85.7% of the total number of failures (Zhou et al. 2012). The SEL is considered to be accountable for the failure of the Russian Forbes-Soil Mars rover (Han et al. 2015). In the design and operation of satellites, there is an urgent need to understand the diagnosis, frequency and distribution of the occurrence of the single event effects.
Measurements and studies on the single event effects have been conducted previously. Analysis of the SEL of a low-earth orbit satellite found that SELs mainly occurred in the South Atlantic Anomaly (SAA) region, and the SELs accounted for 50% of the total failures, while in polar regions, the SELs accounted for 30% failures. The single event effects in the SAA are mainly caused by the high energy protons in the earth’s radiation belt, while in the polar regions they are mainly caused by various high-energy heavy ions. The occurrence of single event effects has short-term and long-term characteristics (Li et al. 2019). The SEUs of a 4M-gates field programmable gate array (FPGA) and a 0.6M-gates FPGA were monitored on a remote sensing satellite (Hou et al. 2014). The upset probability of the device and the spatial distribution of SEUs were obtained and mainly occurred in the SAA (Noeldeke et al. 2021, Li et al. 2018).
The rate of the single event effect of a component could also be calculated from the component’s cross section and the linear energy transfer (LET) (Koontz et al. 2011, He et al. 2016, Sajid et al. 2017). LET is the average energy imparted to the medium by the charged particles per unit length of track. The LET spectra of charged particles in space have been calculated with theories (Timoshenko and Gordeev 2021) and measured with the active and passive detectors (Akopova et al. 1990, Dudkin et al. 1992, Doke et al. 1995, Badhwar et al. 1995, 1996, Badhwar and O’Neill 2001, Doke et al. 1996, 1999, Pázmándi et al. 2006, Dachev et al. 2015, Looper et al. 2020). The particle radiation effect monitor on an SSO satellite measured the radiation LET spectrum, and can also measure the SEU rate of the memory on the monitor. Comparisons of the SEU rates from measurement and from the estimation with the LET spectrum could be made. Results showed that the estimation is nearly four times higher than the measurement (Chen et al. 2016, 2019). LET spectrum of Heavy ion radiation were also measured in the low earth orbit and the medium earth orbit, and the measurement result is 50% higher than that calculation with the CREME96 model (Yuan et al. 2018). A silicon detector telescope was designed for the measurement of LET spectra in deep space, and the measurement range of LET spectra is 0.01–100 MeV/(mg/cm2) (An et al. 2020).
Though monitoring of the SEUs, detections of the LET spectrum of the heavy ions and energy spectrum of the high energy protons have been carried out on orbit, these monitoring and detections are carried out separately and it is difficult to accurately determine the cause and probability of SEU of the device. In addition, in failure diagnosis of the spacecraft, due to the lack of on-orbit data support, when the spacecraft containing components with a low threshold of SEU can resume normal operation after being switched off and on, this abnormality is generally considered to be caused by the SEU, which makes SEU account for a high proportion of failure analysis (Zhou et al. 2012). There is an urgent need to improve the accuracy and timeliness of the failure diagnosis.
Research indicates that the prediction accuracy of SEU rates can be improved by providing the real-time measurement of LET as an input parameter for the prediction model (Chen et al. 2022). SEU and LET were developed for the M15 and M16 Beidou navigation satellites. In addition to measuring the energy spectrum of high-energy protons, the monitor can also measure LET spectrum of the heavy ions and SEUs of the device to determine the correlation between the SEU and the radiation LET data.
2 SEU and LET monitor
2.1 Principle of the SEU and LET monitor
The basic principle of SEU and the radiation LET spectra monitor is shown in Figure 1, using a “sandwich” like structure a memory chip is placed between the two silicon detectors. The memory chip is used to monitor the SEUs. The silicon detectors above and below the memory chip are used to measure the LET spectra of heavy ions. When the heavy ion passes through the detector and the memory chip, the LET of this ion could be measured and whether it induces any SEU in the memory, and the SEU information such as upset bits and cells are recorded.
Basic scheme for monitor of SEU and radiation LET spectra.
2.2 SEU monitor
The chip that monitors the SEUs is a 16Mbit SRAM memory produced by ATMEL, and its model is AT68166HT. The memory chip consists of four identical 4Mbit AT60142F chips, and its layout is shown in Figure 2, which was taken by X-ray imaging device. The size of each AT60142F chip is 12 mm × 6 mm.
AT60142F chip layout in AT68166HT chip taken by X-ray imaging.
Previous research shows that AT60142F chip has excellent SEU performance, the SEU probability of each address is consistent, and it is not affected by temperature. The total radiation dose hardness is up to 120 krad (Si), and the SEL threshold is larger than 80 MeV/(mg/cm2); the chip’s SEU cross section is known for comparison and analysis. When the LET value of the heavy ion is greater than 3.5 MeV/(mg/cm2), the SEU cross section of the AT60142F chip is greater than 1 × 10−9 cm2/bit. The SEU cross section is about 1 × 10−7 cm2/bit. For LET value of the heavy ion less than 3.0 MeV/(mg/cm2), the SEU cross section decreases rapidly to less than 1 × 10−10 cm2/bit. The test on ground found that few multi-cell upsets occurred, and only two-cell upsets occurred when the LET of irradiation heavy ion is 106 MeV/(mg/cm2). The chip has undergone a large number of ground tests and in-orbit flight tests. The European Space Agency took the AT60142F chip as the reference chip for SEU monitor (Harboe-Sorensen et al. 2005, 2008, D’Alessio et al. 2013). The chip is also used for the verification of single event effect evaluation model (Galimov et al. 2019).
The AT68166HT chip is connected to the FPGA, and the FPGA reads it to determine whether the chip has a SEU.
2.3 LET monitor
The detection scheme of LET spectra of heavy ions adopts the universal silicon detector telescope method. LET is defined as the energy loss per unit path length of charged particles, namely,
where ΔE is the energy lost by the charged particles (MeV), and Δx is the mass length through which the charged particles pass (mg/cm2). From the definition, the principle of LET detection is to measure the energy loss ΔE (MeV) of particles in the detector. The length of particles passing through the detector varies with the angle of incidence, and the average length d × sec(θ/4) is usually used, where d is the thickness of the detector, and θ is the detection opening angle of the telescope, as shown in Figure 1. The silicon detector used for the LET measurement is from the Micron Semiconductor Ltd. The type of the detector is MSD026-300, which has a thickness of 300 µm and with a diameter of 26 mm.
In this design, the electronic circuit analyzes the signal amplitude of the two detectors. Signals from the two detectors within 5 µs interval are considered to be produced from the same particle.
The LET spectra of MEO orbit is calculated by CREME96 model as shown in Figure 3. The number of particles with LET greater than 1 MeV/(mg/cm2) is 2 orders of magnitude lower than the number of particles greater than 0.1 MeV/(mg/cm2). The fluence with LET greater than 4 MeV/(mg/cm2) is about 6 × 10−6 cm−2 sr−1 s−1. The range of LET spectra is designed to be measured within 0.1–100 MeV/(mg/cm2), and LET values of heavy ions with atomic number Z = 2–92 could be measured.
Simulated LET spectra with CREME96 of the M15/M16 satellites’ orbits.
2.4 Circuit and software design
The electronics principle of the instrument is shown in Figure 4. The signal generated by the silicon detectors that measure the LET spectra is processed by pre-amplification, main amplifier amplification and shaping, and then converted to digital signal and sent to the FPGA. The FPGA reads the data of the memory chip AT68166HT, determines if there is any upset of the data, and then resets the data of the memory. The FPGA packs the collected LET spectra information and the SEU information of the memory chip and transmits the information to the satellite platform.
Circuit scheme of SEU and LET monitor.
After the monitor is powered on, it measures LET spectra of heavy ion radiation continuously. The signal from detectors is processed to obtain the A/D value of its amplitude. For each sensor, FPGA records the information of heavy ions incident on the detector, including the incident time (accurate to 10 ms) and the signal amplitude it generates. FPGA can record information of up to six heavy ions incident on each detector per second. Due to the low flux of heavy ions in space, the average number of particles incident on each detector is less than one per second, and usually the particle incident on the detector per second is less than six.
In addition to the control analysis and processing of LET spectra detection, FPGA also performs SEU monitor of the memory chip at the same time. After the instrument is powered on, FPGA initializes the value of each address of the AT68166HT chip to 0X55, and then reads the value of each address of the memory chip every second, and judges each value. If the value is not 0X55, it is considered that a SEU has occurred in the address data. The time, address and its upset value, etc., are recorded and then the value of the address is set to 0X55 again. According to the MEO orbital heavy ion radiation LET spectra and the SEU cross section of the memory chip, it is estimated that the frequency of SEU events caused by heavy ions is about 1 time per day when the memory chip is working in orbit. In the design, FPGA can record the information up to 7 SEUs per second, which will not cause loss of SEU events.
The FPGA transmits the information of the LET spectra and SEUs in the memory to the satellite platform every second, which is packaged and downloaded to the ground receiving station.
3 Conclusion
This research carried out a comprehensive design of a monitor for the detection of LET spectra and measurement of the SEU with a “sandwich” like structure. With this monitor, the LET spectra of heavy ions and SEUs in the memory could be measured at the same time. For each particle passing through the monitor, one can get the LET value of this particle, determine if there is any SEU induced in the memory by this particle and the upset bits. Moreover, each SEU induced in the memory could be associated with the LET of the heavy ion. The SEU cross sections of the memory could also be obtained in space. This would improve the assessment of the SEU risk.
Acknowledgements
We thank the following facilities and personnel for supporting in the calibration of the monitor: Irradiation facility dedicated for heavy-ion-induced single event effect, China Institute of Atomic Energy (CIAE), China and Heavy Ion Research Facility (HIRFL) in Lanzhou, China. We thank Guo Gang (CIAE) and Liu Jie (HIRFL) for providing service of calibration.
-
Funding information: This work is supported by the Innovation Academy for Microsatellites of Chinese Academy of Sciences via the assignment contract No. QMT08-03.
-
Author contributions: All authors have accepted responsibility for the entire content of this manuscript and approved its submission.
-
Conflict of interest: The authors state no conflict of interest.
References
Akopova AB, Magradze NV, Dudkin VE, Kovalev EE, Potapov YV, Benton EV, et al. 1990. Linear energy transfer (LET) spectra of cosmic radiation in low Earth orbit. Int J Radiat Appl Instrum D Nucl Tracks Radiat Meas. 17(2):93–97.10.1016/1359-0189(90)90189-5Search in Google Scholar PubMed
An H, Wen X, Li DT, Wang Y, Li CH, Yang SS, et al. 2020. Detection technology and design analysis of LET spectrum of space radiation particles. Nucl Tech. 43(10):100501.Search in Google Scholar
Badhwar GD, Atwell W, Benton EV, Rank AL, Keegan RP, Dudkin VE, et al. 1995. A study of the radiation environment on board the space shuttle flight STS-57. Radiat Meas. 24(3):283–289.10.1016/1350-4487(95)00007-2Search in Google Scholar PubMed
Badhwar GD, Konradi A, Atwell W, Golightly MJ, Cucinotta FA, Wilson JW, et al. 1996. Measurements of the linear energy transfer spectra on the MIR orbital station and comparison with radiation transport models. Radiat Meas. 26(2):147–158.10.1016/1350-4487(95)00290-1Search in Google Scholar PubMed
Badhwar GD, O’Neill PM. 2001. Response of silicon-based linear energy transfer spectrometers: implication for radiation risk assessment in space flights. Nucl Instrum Methods Phys Res A. 466:464–474.10.1016/S0168-9002(01)00285-6Search in Google Scholar PubMed
Chen A, Chen HF, Xiang HW, Yu XQ, Shi WH, Shao SP, et al. 2019. Predictions of the single event upset in space application. Chin J Space Sci. 39(4):460–468.10.11728/cjss2019.04.460Search in Google Scholar
Chen HF, Yu XQ, Shao SP, Shi WH, Cui ZG, Xiang HW, et al. 2016. Direct measurement of the linear energy transfer of ions in silicon for space application. Sci China (Technol Sci). 59:128–134.10.1007/s11431-015-5773-8Search in Google Scholar
Chen J, Lange T, Andjelkovic M, Simevski A, Lu L and Krstic M. 2022. Solar particle event and single event upset prediction from SRAM-based monitor and supervised machine learning. IEEE Trans Emerg Top Comput. 10(2):564–580.10.1109/TETC.2022.3147376Search in Google Scholar
Dachev TP, Semkova JV, Tomov BT, Matviichuk YN, Dimitrov PG, Koleva RT, et al. 2015. Overview of the Liulin type instruments for space radiation measurement and their scientific results. Life Sci Space Res. 4:92–114.10.1016/j.lssr.2015.01.005Search in Google Scholar PubMed
D’Alessio M, Poivey C, Ferlet-Cavrois V, Evans H, Harboe-Sorensen R, Keating A, et al. 2013. SRAMs SEL and SEU in-flight data from PROBA-II spacecraft. 2013 14th European Conference on Radiation and Its Effects on Components and Systems. 2013 Sep 23–27; Oxford, UK. IEEE, 2014. p. 1–8. 10.1109/RADECS.2013.6937398.Search in Google Scholar
Doke T, Hayashi T, Kikuchi J, Hasebe N, Nagaoka S, Kato M, et al. 1995. Real time measurement of LET distribution in the IML-2 Space-Lab (STS-65). Nucl Instrum Methods Phys Res A. 365:524–532.10.1016/0168-9002(95)00523-4Search in Google Scholar
Doke T, Hayashi T, Hasebe N, Kikuchi J, Kono S, Murakami T, et al. 1996. A new silicon detector telescope for measureing the linear energy transfer distribution over the range from 0.2 to 400 keV/μm in space. Jpn J Appl Phys. 35:6241–6247.10.1143/JJAP.35.6241Search in Google Scholar PubMed
Doke T, Hayashi T, Kikuchi J, Nagaoka S, Nakano T, Sakaguchi T, et al. 1999. Application of real-time radiation dosimetry using a new silicon LET sensor. Mutat Res. 430:191–202.10.1016/S0027-5107(99)00130-XSearch in Google Scholar
Dudkin VE, Karpov ON, Potapov YV, Akopova AB, Magradze NV, Moiseenko AA, et al. 1992. Experimental and calculated LET distributions in the Cosmos-2044 biosatellite orbit. Int J Radiat Appl Instrum D Nucl Tracks Radiat Meas. 20(1):143–147.10.1016/1359-0189(92)90092-ASearch in Google Scholar
Galimov AM, Galimova RM, Zebrev GI. 2019. GEANT4 simulation of nuclear interaction induced soft errors in digital nanoscale electronics: Interrelation between proton and heavy ion impacts. Nucl Instrum Methods Phys Res A. 913:65–71.10.1016/j.nima.2018.10.039Search in Google Scholar
Han JW, Feng GQ, Yu YT, Ma YQ, Shangguan SP, Chen R, Zhu X. 2015. Analysis of single-event effects rate of K6R4016V1D chips applied in low Earth orbit. Chin J Space Sci. 35(1):64–68.10.11728/cjss2015.01.064Search in Google Scholar
Harboe-Sorensen R, Guerre TX, Roseng A. 2005. Design, Testing and Calibration of a “Reference SEU Moniter” System. 2005 8th European Conference on Radiation and Its Effects on Components and Systems. 2005 Sep 19–23; Cap d’Agde, France. IEEE, 2005. p. 1–7. 10.1109/RADECS.2005.4365561.Search in Google Scholar
Harboe-Sorensen R, Poivey C, Guerre FX, Roseng A, Lochon F, Berger G, et al. 2008. From the reference SEU monitor to the technology demonstration module on-board PROBA-II. IEEE Trans Nucl Sci. 55(6):3082–3087.10.1109/TNS.2008.2006896Search in Google Scholar
Hou JW, Zhang AB, Zheng XZ, Yu QL. 2014. Research on in-orbit detection of SEU of FPGA. J Astronaut. 35(4):454–458.Search in Google Scholar
He W, Wang YK, Xing KF, Deng W. 2016. Single event effect vulnerability analysis and on-orbit error rate prediction. 2016 IEEE International Conference on Signal and Image Processing. 2016 Aug 13–15; Beijing, China. IEEE, 2017. p 471–7. 10.1109/SIPROCESS.2016.7888307.Search in Google Scholar
Koontz S, Reddell B, Boeder P. 2011. Calculating spacecraft single event environments with FLUKA: investigating the effects of spacecraft material atomic number on secondary particle showers, nuclear reactions, and linear energy transfer (LET) spectra, internal to spacecraft avionics materials, at high shielding mass. 2011 IEEE Radiation Effects Data Workshop. 2011 Jul 25–29; Las Vegas (NV), USA. IEEE, 2011. p. 1–8. 10.1109/REDW.2010.6062528.Search in Google Scholar
Li J, Choutko V, Xiao L. 2018. Single event upset analysis: On-orbit performance of the alpha magnetic spectrometer digital signal processor memory aboard the international space station. Nucl Instrum Methods Phys Res Sect A: Accel Spectrometers Detect Associated Equip. 885:98–104.10.1016/j.nima.2017.12.035Search in Google Scholar
Li Q, Xu BL, Wang DP, Li HF, Zhu J, Zhang BH, et al. 2019. SEL analysis and related operations in the TT&C for near Earth satellite. Spacecr Environ Eng. 36(2):112–119.Search in Google Scholar
Looper MD, Mazur JE, Blake JB, Spence HE, Schwadron NA, Wilson JK, et al. 2020. Long-term observations of galactic cosmic ray LET spectra in lunar orbit by LRO/CRaTER. Space Weather. 18(12):e2020SW002543. 10.1029/2020SW002543.Search in Google Scholar
Noeldeke C, Boettcher M, Mohr U, Gaisser S, Rua MA, Leslie JEM, et al. 2021. Single event upset investigations on the “Flying Laptop” satellite mission. Adv Space Res. 67(6):2000–2009.10.1016/j.asr.2020.12.032Search in Google Scholar
Pázmándi T, Deme S, Láng E. 2006. Space dosimetry with the application of a 3D silicon detector telescope: response function and inverse algorithm. Radiat Prot Dos. 120(1–4):401–404.10.1093/rpd/nci539Search in Google Scholar PubMed
Sajid M, Chechenin NG, Torres FS, Gulzari UA, Butt MU, Zhu M, Khan EU. 2017. Single Event Upset rate determination for 65nm SRAM bit-cell in LEO radiation environments. Microelectron Reliab. 78:11–16.10.1016/j.microrel.2017.07.084Search in Google Scholar
Timoshenko GN, Gordeev IS. 2021. Computation of linear energy transfer of space radiation in biological tissue analog. Planet Space Sci. 10.1016/j.pss.2021.105190.Search in Google Scholar
Yuan B, Zhang SY, Zhang BQ, Chang Z, Shen GH, Sun Y, et al. 2018. Analysis of space particle radiation LET spectra detecting result in MEO/LEO and its error. At Energy Sci Technol. 52(2):334–339.Search in Google Scholar
Zhou F, Li Q, Xin TL, Wei XF, Zhang H. 2012. Analyses and countermeasures of in-orbit satellite failures caused by space radiation environment. Spacecr Environ Eng. 29(4):392–396.Search in Google Scholar
© 2023 the author(s), published by De Gruyter
This work is licensed under the Creative Commons Attribution 4.0 International License.
