In May a satellite hovering high above Kansas failed, causing the first-ever national paging outage, affecting up to 45 million pager users in the country. The PanAmSat Corp.-owned Galaxy IV satellite, which was insured for more than $160 million, may never be returned to service, leaving a spacecraft weighing almost three tons and with a wingspan of 81 feet spinning above the central US.
Dan Fleetwood, a physicist in Radiation Technology and Assurance Dept. 1727, heard of the satellite failure with interest, wondering if it was caused by an electronics component gone haywire due to long-term radiation exposure. Whatever the cause, reliability tests that he and his colleagues have developed might prevent future such problems.
For the past seven years a Sandia team, headed by Dan and Peter Winokur, Manager of Dept. 1727, has been doing studies and developing tests to determine if bipolar electronics, commonly used for analog signal processing in satellites, can survive long-term radiation exposure in space. The tests were adopted earlier this year as the official test standards for satellite electronics parts by the American Society for Testing of Materials (ASTM), an agency that sets test standards for the electronics industry.
Sandia�s work on test development is funded primarily by the Defense Special Weapons Agency (DSWA) and is part of a consortium made up of the DSWA, Mission Research Corp., Naval Surface Warfare Center-Crane, Jet Propulsion Laboratory, The Aerospace Corporation, and Vanderbilt University.
Bipolar electronics found in many satellites are commercially manufactured without special consideration for their possible application in space. Until 1991, it was generally thought that bipolar electronics could withstand radiation exposure on the basis of testing performed by satellite manufacturers.
�The electronics were tested using high doses of radiation delivered in short times at room temperature, and they functioned just fine under those circumstances,� Dan says. �Many parts easily survived testing to levels of radiation well past the point where people would be concerned, and it was thought that the longer times over which the doses would be delivered in space would only improve their response due to self-healing effects.�
But confidence in bipolar electronics� reaction to long-term radiation began to erode seven years ago following a startling discovery by researcher Ed Enlow of Mission Research Corp. in Albuquerque. Using a standard radiation test developed at Sandia for CMOS (complementary metal oxide semiconductor) electronics � which are different from bipolar electronics � he found that bipolar electronics failed prematurely at considerably lower radiation levels than laboratory testing would suggest.
The problem was that the active �base� region of the bipolar transistor typically is covered by a thick layer of insulation that is adequate for performance in applications like consumer or automotive electronics where there is no radiation exposure, or even in some environments where radiation exposure occurs so rapidly that the amount of damage to the device is limited by radiation-induced space-charge screening within the insulator. At the lower dose rates encountered in space this protective charging effect does not occur, and the active base of the transistor is damaged. These results have been described in a series of papers at the annual Institute of Electrical and Electronic Engineers (IEEE) Nuclear and Space Radiation Effects conference by Dan, Peter, and colleagues from Vanderbilt, The Aerospace Corporation, and other members of the DSWA consortium.
�Because of Sandia�s 40-year history in radiation effects testing, the DSWA tapped the Labs to explore tests to determine if bipolar electronics could withstand long-term radiation exposure,� Peter says.
Challenge to predict response
The challenge was to come up with tests that would predict a 10-year response in a few weeks.
�We do this by irradiating components in three ways,� Dan says. �First, we zap them quickly with high doses of radiation at room temperature, taking only a few hours. We compare their response to a second group that receives the same dose of radiation delivered over a few weeks. And we also compare them to a group that receives a high dose of radiation at higher temperatures. By combining the information, we were able to come up with a simple screening procedure that satellite manufacturers can do themselves to determine the reliability of the parts they are putting into their systems.�
Peter says that seven years from discovery of the problem to the development and acceptance of test standards is a �pretty fast cycle,� which was primarily due to the fact that the �satellite industry really needed this kind of a test.�
The Sandia efforts have already paid off for some space systems. For example, one manufacturer has reconfigured its system design to avoid potential failures in a critical bipolar voltage regulator that were identified with the ASTM screening procedure. In fact, up to about 30 percent of commercial bipolar electronics may be vulnerable to these kinds of effects.
�Satellite system designers are paying a lot of attention to these problems now in the design phase of their systems,� Dan says. �When a satellite fails in space, it�s hard to send a repair crew to see what broke. You need to put in as reliable parts as possible from the beginning to prevent future problems.�