TA07-191: Radiation Hardened LCL for Space Applications (PSWCITAR)

https://radnext.web.cern.ch/ https://www.linkedin.com/company/radnext EDMS NO. 3337058 VALIDITY Released REV. 1.0 DOI: 10.5281/zenodo.17287822 RADNEXT Transnational Access Summary Report Project title Radiation Hardened LCL for Space Applications (PSWCITAR) Project TA identifier TA07-191 General application Space Type of test TID, SEE Group leader, Institute Saulo Finco, Center for Information Technology Renato Archer – CTI. Co-authors, Institutes Agnaldo Vieira Dias and Sílvio Manea, Brazilian National Institute for Space Research - INPE. Ronald Hassib Galvis Chacón, Center for Information Technology Renato Archer – CTI. Date(s) of the experiment 29/08/2024 – 02/09/2024 Facility TRIUMF – UBC - Canada Amount of access granted 24h Objectives of the experiments A Latching Current Limiter (LCL) is a crucial component in satellite power distribution systems, acting as an "intelligent switch". Its primary function is to protect the distribution systems and electronic loads from current anomalies, such as those caused by Single Event Latch-up (SEL). When an over-current is detected, the LCL first limits the current to a predefined safe value (ILimit) for a specific "trip-off" time (ToT). If the anomaly persists beyond this time, the LCL isolates the faulty load. LCLs are vital in space applications because the radiation environment can cause SELs, where semiconductor devices draw abnormally high currents, potentially leading to malfunction or permanent damage. Unlike traditional fuses which are irreversible, LCLs can automatically reset or power cycle the affected system, enhancing reliability and extending the operational lifetime of critical components. An LCL was designed with both the control circuitry and the power transistor integrated using 0.6 µm SOI (Silicon-on-Insulator) technology. Its primary characteristic lies in the implementation of a robust set of Radiation Hardened by Design (RHBD) techniques, aiming to significantly enhance its tolerance to radiation environments. Specifically, the LCL employs basic cells constructed with Enclosed Layout Transistors (ELT). This is crucial, as the ELT geometry eliminates drain-to-source leakage currents that would be caused by charge accumulation in the "bird's beak" region in standard transistors, thereby conferring greater robustness to the circuits against cumulative radiation effects. Complementarily, redundancy was applied to each device (standard PMOS and NMOS transistors from the Process Design Kit - PDK), consisting of a series/parallel configuration of four ELT transistors for each transistor. This technique was adopted with the central objective of reducing sensitivity to Single Event Effects (SEE). Furthermore, the SOI technology was selected to eliminate Single Event Latchup (SEL) effects in the control circuits, due to substrate isolation. EDMS 3337058 v.1 status In Work access Restricted PDF from TA07_191-Zenodo.docx modified 2025-10-07 15:41 https://radnext.web.cern.ch/ https://www.linkedin.com/company/radnext EDMS NO. 3337058 VALIDITY Released REV. 1.0 In the past, the LCL switch was tested for total ionizing dose using a Cobalt-60 source, where it demonstrated resilience up to several hundred krad. Tests were also conducted using heavy ions (oxygen, chlorine, and copper) to assess the LCL's susceptibility to Single-Event Effects (SEE). In the test performed, the switch showed robustness; however, the energy levels of the test facility were limited (oxygen – 43.5 MeV; chlorine – 63 MeV; copper – 84 MeV), which prevented full exploration of the LCL’s limits. The main goal was to verify the functionality of LCL under irradiation conditions, ensuring proper operation as specified. The objective of the experiment was to test the behavior of the LCL using highenergy protons (480 MeV) at different flux levels to evaluate circuit performance and robustness, aiming to observe the occurrence of SEE. The experiment was also used to verify the behavior of the LCL when subjected to high dose rates (100 krad/h). These accelerated total ionizing dose effects were particularly relevant in the internal references of the LCL switch controller. Different references are essential in the power switch system primarily because the various internal subcircuits require different voltage biasing levels for their proper operation. Specifically, the system incorporates three distinct references: Reg_5V, Reg_5VUP, and current limiting reference. The understanding of TID influence on these references are crucial to improve the performance of the LCL switch. Experiment test report The LCL is designed to operate with an input voltage (VIN) ranging from 22 V to 36 V and a maximum limiting current (ILimit) of 6 A. The need for voltage regulators within the structure is critical because, although the system input voltage varies across a high-voltage range (22 V to 36 V), the LCL's internal control circuits are predominantly implemented with low-voltage (LV) devices that support only up to 5V. Consequently, the system generates two regulated 5 V voltages: one referenced to GND (V5), essential for biasing the control logic and telemetry circuits, and another tied to VIN (V5UP), utilized for biasing the feedback loop that controls the pass transistor. Figure 1 shows the test board and Figure 2 highlights the controller and power transistor of the LCL, both implemented using RHBD techniques. Figure 1 – Test board of LCL switch https://radnext.web.cern.ch/ https://www.linkedin.com/company/radnext EDMS NO. 3337058 VALIDITY Released REV. 1.0 Initially, 33 boards were assembled and identified (Figure 3). The LCLs were calibrated to ensure consistent values for ILimit (6 A) and ToT (2 ms) across all units. Additionally, both regulators were precisely adjusted to maintain a nominal output voltage of 5 V on every board. Two boards were chosen as “control group” and were not intended to be exposure to radiation. Figure 2 – Controller and power transistor of LCL switch Figure 3 – Assembled and identified boards for testing • Test bench setup To properly check the behavior of the LCL circuit and its susceptibility to heavy ions, a test setup is necessary. To perform this task, an automated test bench has been developed. The need for automation arises from the various test conditions, which can be a time-consuming process and the impossibility to conduct manual tests during radiation exposure. The automated test setup was designed for conducting tests on LCLs. The core of this test environment is a computer running NI Labview, which controls and monitors the devices through scripts developed specifically for this purpose. The instrumentation includes an oscilloscope with a current probe for current https://radnext.web.cern.ch/ https://www.linkedin.com/company/radnext EDMS NO. 3337058 VALIDITY Released REV. 1.0 measurements, a DC electronic load utilizing, a DC power supply, and a data acquisition board to control the LCL and read analog and digital signals. Another computer was used to monitor the experiment from outside the room. Figure 4 shows the test bench setup. The equipment on the blue trolley was placed inside the beam room and computer (Outside computer) was placed in the control room. All that equipment is interconnected, forming an integrated test bench. During tests, the system continuously monitors the LCL's output current, while the electronic load periodically applies short circuits. Short circuits are generated every 50 ms with a duration of 400 us, with the current set to 1 A for the remainder of the time. A test mask was created to verify if the switch maintains current control during the short circuits, adhering to this mask. This test mask can be seen in Figure 5. The test bench also checks if, during the intervals where current limitation is not in operation, the switch's conduction loss stays within defined limits. If any limits are violated, the failure waveform is saved, and a mask failure counter is incremented. One possible radiation effect on the LCL switch is its shutdown due to particle incidence. If the switch turns off during the test, the setup attempts to restart it. If the switch resumes operation, the test continues, and a switch shutdown counter is incremented. If the switch turns off, and the test setup fails to restart the LCL after three attempts, the test is terminated. The test setup can monitor and verify masks for 90% of the switch's radiation exposure time, meaning that in 10% of the time, events of this nature may be missed. Shutdown events are continuously monitored. Furthermore, some signals from the controller, such as voltage references and current consumption, are measured and stored to check if the switch is undergoing any changes in its operating point. For radiation tests, concrete blocks and other materials were employed to protect the measurement instruments from the radiation source (Figure 6) and only the LCL board was exposed to the beam. • The radiation test issues The tests were scheduled to August 29th with a start time of 14:00 (12 hours of duration) and September 2nd with a start time of 8:30 (12 hours of duration). The test setup was placed in the beam room, the LCL board was properly positioned and the test was started. However, after a few hours, the automated setup stopped functioning due to a crash of the computer located inside the beam room, which was controlling the experiment. A new computer was configured for use in the test setup, and the test was restarted. In less than an hour, the setup stopped working again. It was concluded that the radiation level inside the beam room was too high and was interfering with the test setup. The tests were halted, and a new configuration was assembled. In this new setup, all equipment was placed outside the beam room, and only the board under test remained inside. After this modification, the test setup operated normally until the end of the test period. Due to the need to modify the test setup, we lost beam time during the initially allocated period. However, the lab staff kindly allowed us to use the facility during time slots when no other users were scheduled. In total, more than 33 hours of testing were performed on the LCL switches. https://radnext.web.cern.ch/ https://www.linkedin.com/company/radnext EDMS NO. 3337058 VALIDITY Released REV. 1.0 Figure 4 – Automated test bench setup Figure 5 – Test mask showing a failure Figure 6 – Test setup protected using concrete blocks and other materials https://radnext.web.cern.ch/ https://www.linkedin.com/company/radnext EDMS NO. 3337058 VALIDITY Released REV. 1.0 • Test results and discussion In the tests, 16 switches were exposed to the beam, while two others were used as control samples. The main objective of the test was to verify the immunity of the LCL switch to Single-Event Effects (SEE). Throughout all these tests, where the flux varied between 10⁷ particles/cm2/s and 10⁹ particles/cm2/s and fluences exceeded 10¹¹ particles/cm2, no SEE was observed, confirming the robustness of the switch under proton bombardment. The accelerated total ionizing dose effects were tested in six boards (B1, B2, B3, B4, B5 e B6) and the applied TID was 100 krad(Si). Effects were observed in the internal references of the LCL switch controller. The test results for the Reg_5V reference are presented in Figure 7, with the values normalized to facilitate comparison. During irradiation, the reference voltages across the different boards exhibited no uniform behavior. Some boards showed an increase in voltage (e.g., board B6), while others experienced a decrease (e.g., board B5), and the remaining boards (B1, B2, B3, and B4) showed fluctuations around the initial voltage level. The maximum observed variation was approximately 5%, and on average (black curve in Figure 7), the values remained relatively close to their initial levels. No significant recovery was observed in the reference voltages after the annealing period. Figure 7 - Test results for references of Reg_5V Figure 8 shows the normalized test results for the Reg_5VUP references. During irradiation, the reference voltages across the boards exhibited a noticeable trend, even on average (black curve in Figure 8). Initially, a decrease in voltage was observed, continuing up to around 35 krad. After this point, the values began to increase, surpassing the initial voltage near 60 krad. Toward the end of the irradiation, a saturation effect was observed. Following the annealing period, a strong recovery in the reference voltages occurred, with the final values stabilizing at approximately 90% of the original level. Figure 9 presents the normalized test results for the current limiting references. During irradiation, the reference voltages across all boards exhibited a clear and consistent downward trend. Voltage reductions were observed from the beginning of the exposure, with no periods of recovery throughout the test. Toward the end of the irradiation (after 90 krad), a saturation effect appears to occur. The maximum voltage reduction reached approximately 10%, while the average decrease (represented by the black curve in Figure 9) was around 7%. No significant recovery was observed in the reference voltages after the annealing period. https://radnext.web.cern.ch/ https://www.linkedin.com/company/radnext EDMS NO. 3337058 VALIDITY Released REV. 1.0 It was observed that the voltage reference used in the Reg_5V was the most stable during the tests, with a maximum variation close to 5% and an average variation that was virtually zero. The other references showed greater variations during the tests and, after the annealing period, did not return to their initial values, exhibiting a permanent change in their behavior. Despite the variations in electrical parameters, all irradiated boards remained functional after the radiation tests. A future study should be conducted to investigate the reason for the different levels of degradation observed among the various reference circuits when subjected to irradiation. Understanding the radiation interaction mechanisms in different topologies will enable the design of more radiation-resistant references, resulting in more reliable circuits for use in space missions. Figure 8 - Test results for references of Reg_5VUP Figure 9 - Test results for references of current limiting https://radnext.web.cern.ch/ https://www.linkedin.com/company/radnext EDMS NO. 3337058 VALIDITY Released REV. 1.0 • Conclusion It was observed that the LCL design was resilient and had no issues concerning SEE. These results also provide a preliminary validation of the proposed RHBD technique, which consists of a series/parallel configuration of four ELT transistors for each transistor. Also, it was observed the voltage reference used in the Reg_5V was the most stable during the tests, with a maximum variation close to 5% and an average variation that was virtually zero. The other references showed greater variations during the tests and, after the annealing period, did not return to their initial values, exhibiting a permanent change in their behavior. Despite the variations in electrical parameters, all irradiated boards remained functional after the radiation tests. A future study should be conducted to investigate the reason for the different levels of degradation observed among the various reference circuits when subjected to irradiation. Understanding the radiation interaction mechanisms in different topologies will enable the design of more radiation-resistant references, resulting in more reliable circuits for use in space missions. Outcome of the experiments Please indicate what the experiment is likely to lead to by putting an ‘X’ next to one or more of the possible outcomes below. Journal publication X Data for Thesis X Follow-up experiment at same facility Follow-up experiment at another facility X Other As a RADNEXT user, we encourage you to submit the scientific results of your experiments to journals as well as to the NSREC and RADECS data workshops. Please remember to include the RADNEXT acknowledgment into your publications! RADNEXT acknowledgment: This project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 101008126.