The features of the spacecraft of most relevance to the acquisition of the astronomical data were the control of the satellite, the method of executing the observational program, and the storage and transmission of data. To understand these, it is necessary to discuss the onboard computers and their associated software, the attitude control system, data recording and the communications links.
The random access memory contained the routines for executing a complete Satellite Operations Plan (hereafter denoted an SOP) for the ten to fourteen hour period between passes over the ground station and for generating the scientific data stream. Although 64,000 words of memory were available, this capacity was sometimes insufficient to store as large a program of observations as could be carried out during the observation period; this was a result of the high efficiency of the ground system at filling the observation time and applied particularly to the third coverage of the sky during the last four months of the mission (Section III.C.2).
The satellite attitude was controlled by three orthogonal reaction wheels; excess momentum was dumped via magnetic coils to the Earth's magnetic field as necessary. The attitude, and changes in attitude, were sensed by a combination of an horizon sensor, a sun-sensor and three orthogonal gyros. The z-axis gyro was used in all modes of control and was duplicated to provide a redundant backup.
Figure II.B.1 Spacecraft control axes labeled x, y and z. The axes
x(s),y(s) and z(s) are fixed with respect to the Sun and the
north ecliptic pole.
The spacecraft control axes are shown in Fig. II.B.1. In observational modes, the y-axis was always kept perpendicular to the satellite-Sun vector while the x-axis corresponded closely to the telescope boresight. Only two of the many possible control modes are described here. During normal operations, the signals to control rotations about the x- and y-axis were obtained from a two-axis Sun-sensor with 3.5" x 7" resolution. Signals from the z-gyro were used to control the rotation about the z-axis at the rate necessary to achieve the desired rate of scan, d/dt, across the sky (see Section III.C.3). Towards the end of the mission, the Sun was eclipsed by the Earth for a time during each orbit and the Sun-sensor could not be used. At such times, all three gyros were used to control the satellite, although with a marked loss of control accuracy. No scientific data were taken during eclipses (see Section III.B.9).
Onboard attitude updates and ground attitude reconstruction were made using a two-axis star-sensor of the V-slit type in the focal plane of the telescope. Section V.B describes in detail the attitude reconstruction process. The absolute pointing accuracy for control purposes of the system was approximately 30". The accuracy of reconstructed positions is discussed in Sections V.B and VII.C.
The data for successive SOPs were recorded alternately on two tape recorders. In record mode, earlier data were erased. During a ground-station pass, the data recorded during the previous SOP were transmitted to the ground from one recorder while the other was commanded into its record mode ready for the data from the next SOP. This procedure protected data from being immediately over-written on the occasions when it proved impossible to transmit all the data to the ground during the prime station pass.