Preemptive Testing Can Mitigate Cosmic Radiation Effects

Space travel and exploration are popular topics for all ages — and continue to be subjects of international debate, pride, strength, and domestic security. It has been 40 years since the U.S. put men on the moon. However, space is not kind to electronic systems, so testing must be done on the ground to ensure appropriate performance and protection against the effects of radiation. Those of us who work in this industry know that space is also a deadly frontier for both humans and electronics. Without a lot of work, neither will survive very long.

As you may surmise, our sun is a continuous, self-sustaining nuclear reaction that discharges a vast amount of energy and deadly radiation toward the Earth. Fortunately, the Earth's magnetic field protects us from most, but not all, of this radiation. It causes the radiation to flow around us, much like water around a rock in a stream of water. You can get some sense of this radiation by viewing the aurora borealis, or northern lights, where incoming radiation due to a weaker magnetic field at the Earth's poles ionizes the atmosphere (Fig. 1).

While our atmosphere provides additional shielding from incoming radiation, it produces both beneficial and harmful effects. One benefit is that incoming radiation constantly rebuilds our ozone layer to shield us from the sun's ultraviolet (UV) radiation. But incoming radiation also produces destructive terrestrial neutrons that are part of the human aging process, causing cellular damage over time.

Exposure to terrestrial neutrons occurs at a rate of about 10 neutrons/cm² at ground or sea level. These neutrons can cause circuit or logic upsets in electronic designs. When these neutrons pass into semiconductor chips, they collide with other atoms, releasing small amounts of energy that causes atomic-size nuclear fissions.

COSMIC COMPUTING PROBLEMS

These fissions produce enough energy to cause glitches in ICs, and users of memory systems or highly complex ICs are concerned about these upsets. For example, the real culprit of many computer crashes is not necessarily Microsoft. Instead, it could be the reaction of a neutron strike at a critical circuit node inside one of the ICs. In his “Sun Screen” article in Forbes Global from 2000, Daniel Lyons notes that the reaction of a neutron striking at a critical circuit node inside an integrated circuits can cause computer problems [1].

Commercial aircraft flying at 40,000 feet are subject to 300 times more radiation from terrestrial neutrons than on the ground. Therefore, their avionic systems must be tolerant to these effects. Controversy surrounds how much of this radiation can be tolerated by people who fly frequently.

As semiconductor feature sizes shrink, making circuits immune to these effects is increasingly difficult. Complex, modern-day digital systems are more sensitive to neutron strikes, so they typically employ some form of mitigation techniques, such as error correction or redundancy.

Unlike the shielded environment in which we live, space is a very dangerous environment. Of particular concern are solar storms, or flares, (see a video at www.space.com/common/media/video/player.php?videoRef=sun_storm) that can discharge deadly particle showers of protons and electrons that can kill you quickly or slowly, depending on the strength of the storm. Apollo astronauts were fortunate they did not encounter a solar flare during their typical 12-day missions.

These protons and heavy ions from the space-radiation environment also expose low Earth orbiting (LEO) spacecraft and satellites. Electrons, as well as protons and heavy ions, play a large role in geostationary orbiting (GEO) satellites. In the course of a 10- to 15-year mission, satellites encounter large doses of radiation. Even with shielding, they receive enough radiation to kill humans in a few months.

Human tolerance to radiation is approximately 450 rads or less. A rad is a unit of measurement that determines the amount of energy absorbed from radiation by the material, being defined as 100 ergs (g·cm²/s²) per gram of material. Typically, ICs used in space must survive 10,000 to 100,000 rads or greater over their mission lifetime. Shielding helps to a limited degree but there is a point of diminished return, where additional shielding is no longer effective.

Due to secondary reactions, adding additional shielding material after a thickness of 300 mils of aluminum does little to help. The effect is similar to Newton's cradle toy (Fig. 2), or collision balls, where one ball strikes a series of balls on strings causing a ball at the opposite end to shoot off.

Incoming radiation can react in a similar way. High-energy electrons in cosmic rays collide with the shielding material to produce a secondary radiation effect. This is just as damaging as it exits the shielding material on the opposite side. Electronics manufacturers must design systems to operate in this deadly and harsh environment. Several design techniques and special semiconductor processes can be used to ensure that circuits operate reliably over long periods in space.

In our daily lives, whether for international communication, music, DTV, GPS, weather gathering, surveillance, national security, etc., we are more dependent on satellite systems than ever before. Electronic content in space is growing rapidly. There are thousands of satellites out there, i.e., with most of them being space junk. Satellite systems are becoming more sophisticated; they're not the simple bent-pipe systems used in the past. Most people do not realize the number of satellites in space. But, a quick look at Google Earth with an overlay of a satellite database shows their pervasiveness: http://science.slashdot.org/article.pl?sid=08/09/05/1231228.

PROTECTING MEDICAL ELECTRONICS

Designing circuits that can tolerate the many radiation effects is necessary — not only for space and aircraft, but also for today's medical applications. Diagnostic and security equipment rely increasingly upon radiation-tolerant chips. Today's dental, medical X ray, and CT scanner diagnostic equipment typically use electronic-imaging and data-conversion chips instead of film to capture images.

The equipment produces images in near real time, as opposed to waiting for technicians to develop film then determine if they captured an accurate image. The effects of radiation accumulate over time which, for example, can occur in X-ray screenings at airports, where this equipment is used continuously.

POWER ELECTRONICS AND HEAVY IONS

Power electronics, especially higher-voltage electronics, are very susceptible to heavy ions in space applications. Heavy ions pass through semiconductor material, deposit a charge, and cause miniature temporary conduction paths. This is similar to a miniature lightning bolt.

Once a small conduction path is established, a high-current discharge can occur from the positive high-voltage supply-to-ground. In many cases, this can occur through the dielectric, which causes gate punch-through and burnout in many power semiconductor devices. Power component manufacturers typically derate their devices by 20% to 50% of their normal safe operating supply voltage to circumvent this issue.

PREEMPTIVE AND PREDICTIVE TESTING

TI's High-Reliability Group develops and tests circuits against the various radiation effects that electronic systems can encounter in medical, avionics, and space applications. Repairing a satellite is not an option if a problem occurs after launch and deployment. Even the backup systems often deployed in satellites have limitations.

Continue to next page.

Testing parts for space applications can be difficult. First, test machines on Earth can't duplicate the exact effects encountered in space due to limited machine energy. Also, parts can't be tested over years to determine if they are usable in space. So, accelerated testing techniques are often used.

Accelerated testing techniques attempt to duplicate the effects of low-dose radiation rates where a part may be subject to over a 10- to 15-year mission in space. Techniques similar to those used in space can be misleading. In some cases, the same part subjected to low-dose rates can produce unexpected results compared with those tested at high-dose rates — a problem referred to as enhanced low-dose-rate sensitivity (ELDRS). This issue remains a discussion topic in many of today's radiation conferences and standards meetings.

Circuit technology also plays a major role in the effects of circuit performance in a high-radiation environment. Process technologies such as CMOS, Bipolar, GaAs, SiGe, and others all have pros and cons in the way they behave under various radiation effects. Understanding these behaviors is critical to the design of a high-reliability product with high tolerance to the various effects. When real-life measurements of certain conditions are not possible, measurements can be predicted based on only statistics and limited data.

Since test machines are limited in the type of particle energy they can produce, it is important to understand the environment in which your products will operate, as well as to ensure the circuits are reliable and tolerant to specific threats. Using HiRel products in systems that must be reliable helps to reduce the risk of system failure. Cutting corners to reduce costs potentially can jeopardize millions of dollars of space or avionic hardware. In the end, it really boils down to more than rocket science.