Brown, lead-underpant moments
The space industry wants to exploit the benefits offered by the latest, leading-edge, commercial-grade components. Many technologies are initially not designed with space applications in mind, and in most cases, radiation testing is necessary to de-risk parts before or during device qualification. Some of these technologies require the most recent test equipment to characterise their perfomance.
Six years ago on a cold winter night in Paris, some colleagues and I were testing the proton sensitivity of a new component. The device-under-test was strategically positioned in front of the beam and our expensive test equipment was placed 5 metres away behind some paraffin-impregnated, shielding blocks as shown below.
Just before midnight and two hours into the test campaign, seated in the adjoining control room, I realised that one of the instruments monitoring SEEs (single-event effects) had stopped communicating. We entered the radiation chamber, with our dose meters attached, to discover that a very expensive item of test equipment had been affected by secondary radiation and was completely unresponsive and 'dead.' This was a 'Brown, Lead-Underpant Moment!'
Using the other monitoring instruments, we were only just able to continue and collected a sufficient amount of data by working throughout the night. The following month, when the item of test equipment had 'cooled' and returned to us, it had annealed and worked first time!
Last February on a very cold winter night in Switzerland, we were asked to test the proton sensitivity of a new, high-frequency component. Based on my previous Parisian 'Brown, Lead-Underpant Moment,' I was understandably reluctant to place an expensive $80k signal generator anywhere near the beam line. After much discussion, I was begrudgingly persuaded to place it behind some paraffin-impregnated blocks as shown below.
Testing started at 200 and then 100 MeV/mg/cm2, but ninety minutes into the campaign, at a LET (linear energy transfer) of 60 MeV/mg/cm2, our beloved signal generator died. Being engineers, we rolled up our sleeves, retrieved the instrument from the radiation chamber, ignored the 'Calibration Void if Removed' seal and unscrewed its lid. It didn't take long to diagnose that its impressive power supply had been affected by secondary radiation!
We searched the facility high and low for a similar signal generator but the best we could find was 300 times slower and had a different type of electrical output. Testing had to be abandoned, most of the expensive beam time was wasted and my $80k baby was dead! This really was a 'Brown, Lead-Underpant Moment!'
Three months later (and still grieving), with a shiny new power supply in the signal generator, we were asked to repeat low-energy proton testing: I was determined never to experience another 'Brown, Lead-Underpant Moment.'
Before traveling to Belgium, I ensured there were cabling ducts within the 1-metre shielding wall between the radiation chamber and the adjoining control room as shown below.
A large physical separation between the device-under-test and the test equipment results in long cable runs, which can be challenging for high-frequency signals. However, our expensive instruments were protected from scattering effects and secondary radiation. Testing was completed successfully and we delivered a comprehensive report much to the satisfaction of our understanding and patient customer.
The space industry is baselining the latest, highly-integrated, advanced components which are being used in everyday, consumer-grade electronics. In the current economic climate, some suppliers are eager to diversify their product offering and support the space industry. Astrium partners with some of the leading electronics companies helping them to de-risk their parts. For some vendors, this is a new experience and their evaluation boards are simply not suitable for irradiation or the underlying technology is inherently soft.
The following short video shows the output of a space-grade device-under-test during heavy-ion irradiation. On the same evaluation board, there is a neighbouring commercial part supplying the part-under-investigation. Within the vacuum chamber, the commercial-grade device becomes too hot after ten seconds, loses lock, corrupts the interface to the device-under-test adversely affecting the output and hence the ability to detect SEEs. Some SEUs (single-event upsets) and SETs (single-event transients) can be seen at the start of the clip.
These examples demonstrate that only by working together - i.e., the agencies, radiation-test facilities, component suppliers, radiation-test houses and the standards committees - can we assess the latest enabling technologies in a timely and cost-efficient manner to meet strict product schedules.
I'd love to hear if you have had any 'Brown, Lead-Underpant Moments' at the cyclotrons or during your career in space electronics. I'm travelling to northern Finland this weekend to do some heavy-ion testing and as you can see from the crate below, I've packed the necessary kit!