The USA Experiment on the ARGOS satellite was launched in 1999 as part of the Navy’s program in X-ray Astronomy. It is designed not only for basic research on celestial

X-ray sources such as neutron stars and black holes, but it also for study of space technology and applications issues. This talk will cover two aspects of the experiment.

Several features make this experiment well suited for X-ray observation of transient black hole candidates. Its pointing schedule can be revised on a few days notice to permit extensive observations of new transients, extending over many weeks and with coverage up to once or twice per orbit as needed. The sampling frequency is particularly useful when the power spectral density is evolving dynamically on time-scales of hours to days. USA also provides spectral coverage below 3 keV for isolation of disk components of source. To date, USA has been used to observe several sources that are among the prime candidates for black holes in our Galaxy. In particular, XTE J1118+480 has been monitored in this manner during its recent outburst beginning April 2000. Evolution of a low frequency quasi-periodic oscillation (QPO) is observed; it drifts from 0.07Hz to0.15 Hz during the outburst. Possible correlation of the QPO frequency with the flux and spectral characteristics of the source are examined and compared with other low frequency black hole candidates.

The central electronics of the USA Experiment was designed to include a unique experiment on computing in the space environment. Computing on satellites is done for various reasons including navigation and onboard data processing but the space radiation environment is hostile to computing hardware so that special techniques must be used to achieve reliable performances. For many years this has consisted mainly of radiation hardening, which means that the space processors become expensive customized hardware that can be constructed only at special foundries. The test on USA explores whether commercial processors can be made to approach the performance of the hardened processors through the use of fault tolerant software.