How Do You Radiation-Test a Complex IC?

How Do You Radiation-Test a Complex IC?

When a radiation-hardened component exists for your space-based electronics application, the path is relatively straightforward, when budget allows. But when your design calls for a complex mixed-signal integrated circuit (IC), one that combines signal processing, microprocessor, and analogue functions on a single die, radiation-hardened availability and characterisation becomes a different problem entirely. For us, these types of IC devices are a critical part within their small size, weight and power space-based radar modules.

Each functional block on a device like this can fail in different ways under radiation exposure. And those failure modes can interact. A single event effect in the signal processor may behave very differently from one in the analogue front end, and the mitigation strategy for each may be completely different. The question is not just “will this device survive in orbit?” but “how exactly will it misbehave, how often, and in which operating modes?”

In partnership with Radiation Analysis Services Ltd (RASL) the Plextek team completed a radiation characterisation campaign on a complex mixed-signal IC intended for use in Low Earth Orbit (LEO). This article describes the methodology and what made it technically challenging.

Planning the test

RASL led the test planning, producing a formal test plan that defined the test aims, the ions of interest, and the approach to data collection. The choice of ions matters: different ion species deliver different levels of Linear Energy Transfer (LET) to the device, and the test plan needs to cover a range wide enough to characterise the device’s response from low to high LET values.

But the test plan is only half the problem: the other half is making sure the data you collect is actually applicable to the way the device will be used in flight. For a simple memory or a power converter, this is manageable. For a complex mixed-signal device with multiple functional domains, it means exercising the device in each of its operating modes and monitoring the right parameters in each case. Getting that wrong means your test results don’t map to your real application, and your mitigation strategy is built on the wrong data.

Building the test bench

Plextek designed the test bench: a custom setup built around a development board that could exercise the device across its different operating modes while monitoring supply current draw and pulling diagnostic information from the device under test in real time.

Supply current monitoring is particularly important for detecting latch-up, a destructive single event effect where radiation causes a parasitic thyristor structure within the IC to turn on, drawing excessive current. Left undetected, latch-up can permanently damage the device. The test bench needed to catch these events quickly.

The test software was one of the more challenging elements. It had to cycle the device through its various functional modes, capture diagnostic data, and flag anomalies, all while running reliably in a test environment where the device under test is being bombarded with heavy ions and may behave unpredictably at any moment.

The test campaign at RADEF

Testing was carried out at the RADEF (RADiation Effects Facility) in Finland, a facility widely used for this type of characterisation work. RASL arranged access to the facility, ensured testing was conducted on delidded samples, and supported the test campaign.

Delidding removes the device’s package to expose the silicon die directly to the ion beam. Without it, the package material would stop or scatter the ions used during testing before they reach the active circuitry, with the aim of more closely replicating the energy levels which occur in space.

The test used a cocktail of ion species covering a wide range of LET levels. This spans the spectrum from relatively benign particle strikes through to high-energy events that occur less frequently in orbit but can cause serious disruption or damage.

The devices were tested at elevated temperature and with raised supply voltages, both of which increase susceptibility to single event latchup. Testing under these stressed conditions gives a more conservative, and more useful, Single Event Latch-up (SEL) characterisation than testing at room temperature and nominal voltage, which was more appropriate for the Single Event Upset (SEU) & Single Event Functional Interrupt (SEFI) characterisation performed.

Two devices were tested, and the results showed good correlation in behaviour, which increases confidence that the observed effects are characteristic of the device design rather than sample-specific anomalies.

Analysing the results

RASL carried out statistical analysis on the test results, producing predictions of single event effect rates across different orbital environments. The variables here are orbit altitude and inclination, which determine the radiation environment the device will actually experience in operation.

The output is not a simple pass or fail. It is a probabilistic model: for a given orbit, the analysis predicts how frequently different types of single event effect will occur. Some will be non-destructive and recoverable. Others, like latch-up, require active mitigation.

This is what radiation characterisation actually delivers: not a binary answer, but the quantitative data an engineer needs to design the right mitigation for the right mission.

What this enables

Armed with SEE rate predictions for their target orbit, the design team can make informed decisions about mitigation. That might mean current-limiting protection against latch-up, error detection and correction for memory upsets, or watchdog timers to recover from functional interruptions. The specific approach depends on the device, the orbit, and the mission’s tolerance for disruption.

Without this kind of characterisation, engineers are either guessing at mitigation requirements or over-engineering their protection, both of which cost time, mass, power, and money. For missions in low Earth orbit, where the use of complex commercial-grade devices is increasingly common, rigorous radiation characterisation is what makes the difference between a device that operates reliably and one that fails unpredictably.

This project was a collaboration between Plextek, who designed the test bench and test software, and RASL, who led the test planning, facility coordination, and post-test analysis. The combination of device-level engineering expertise and specialist radiation test knowledge is what made it possible to characterise a device of this complexity to the standard required.

About RASL

Radiation Analysis Services Ltd (RASL) is a specialist space radiation effects consultancy with over 20 years of experience in radiation hardness assurance. RASL has supported more than 80 missions for over 60 customers worldwide, from start-ups to large primes, providing radiation testing, space environment modelling, component selection guidance, and design mitigation.