The PCS included the hardware and flight software necessary for precision telescope pointing, stabilization, slewing, tracking, and safe mode functions. The PCS performed the initial attitude acquisition of the spacecraft following launch vehicle separation. It provided periodic boresight calibration for the telescope relative to the star trackers on the spacecraft. The PCS provided the capability for both rapid large angle slews and small maneuvers to place and reposition science targets within the science instrument apertures; it maintained the solar array orientation toward the Sun; and, it pointed the high gain antenna toward Earth for downlink. The PCS also contained Wide Angle Sun Sensors (WASSs) that act as a second check on the spacecraft orientation to ensure that the hard pointing constraints (see section 3.5) are not violated. High-level fault protection would have placed the telescope in a safe mode if a violation was detected. The performance numbers presented in this and the following sections are based on our present understanding and measurements of the on-orbit PCS performance.
The PCS was a celestial-inertial, three-axis stabilized control system. A high performance star tracker/inertial reference unit (ST/IRU) package provided attitude determination and reconstruction capabilities. On-board pointing commands and variables used the J2000 coordinate system. Reference to the J2000 celestial sphere was implemented within the ST through autonomous identification of stars carried in an on-board catalog of 87,000 Tycho stars down to 9th visual magnitude. [The on-board catalog actually used the ICRS coordinate system, so, in fact, the on-board pointing system is really ICRS. However, the differences between ICRS and J2000 are so small (≤120 mas) that no conversions are made, and Spitzer is effectively considered to use the J2000 coordinate system.]
The ST was used to point an instrument boresight to a desired location on the sky with an initial accuracy of at least 0.5 arcsecond (1σ radial). The ST field of view is 5°×5°, which ensured that Spitzer can point to any part of the sky and have the ST meet its pointing requirements. Typically 40 stars were used simultaneously. The gyros provided pointing stability when not using the ST as a pointing reference; the pointing drift derived from the gyros was <3 mas/sec over 8 hours. The drift rate when using the IRU-only mode should generally be better than 1 mas/sec over 200 seconds.
All telescope pointing was defined and calibrated relative to redundant PCRSs located in the focal plane. During the course of the mission, the PCRS was periodically (about every 12 hours) used to calibrate the telescope-to-star tracker boresight alignment that may drift due to thermo-mechanical effects. Each PCRS detector is a Si PIN photodiode array divided into two 4×4 subarrays for redundancy. Each pixel is 250 µm square, with a plate scale of 10 arcseconds per pixel. The PCRS calibration measured the star position with an accuracy of 0.1 arcseconds (1σ per axis), and was sensitive down to 10th visual magnitude at a wavelength of 550 nm.
Spitzer also has WASSs (Wide Angle Sun Sensors), which measure the Suns position with respect to the spacecraft. These sensors were used during initial attitude acquisition after launch, as well as for Sun avoidance, fault protection, and safe mode during the mission. Each wide-angle Sun sensor provides a field of view of 2π sr with an accuracy of ±0.1° at null. They were placed at the top and the bottom of the solar panels to maximize the coverage, with their boresights aligned to the spacecraft Z-axis. After the pitch angle of downlinks exceeded operational angle of the WASSs in September of 2016, they were disabled before each downlink to Earth. After the downlink completed, the spacecraft was pointed to a spot in the continuous viewing zone (CVZ) near the north ecliptic pole before fault protection were re-enabled, and then science observations could recommence.
Four reaction wheels provided the primary control actuation for all modes of operation. They were mounted in a pyramid orientation about the X-axis; each canted at 30° towards the X-axis. Over time, angular momentum accumulates in the reaction wheels, due primarily to the small offset between the center of mass and the center of (radiation) pressure. Unlike an observatory in low Earth orbit, which can dump this momentum magnetically, Spitzer has a Reaction Control System (RCS), which used cold nitrogen gas thrusters to provide the reaction wheel momentum unloading capability. Opportunities to dump momentum were scheduled approximately every 12 hours, but used only if sufficient momentum had accumulated in the reaction wheels. The nitrogen supply was sufficient to accommodate the entire mission lifetime from launch to decommissioning with room to spare.
On-orbit measurements show that the PCS is capable of slewing the telescope 180° in 900 seconds, 1° in 60 seconds, and 1 arcminute in 6 seconds, while maintaining its inertial pointing knowledge. These times include the acceleration and deceleration of the telescope, but do not include the time it takes for the PCS to stabilize after the slew has completed. The pointing system had several operating modes, and the AOTs are designed to use the pointing mode most appropriate for each observing mode. Settling time varies with operating mode and slew magnitude. For IRU-only slews, slews less than 30 arcseconds settle to within 0.2 arcseconds rms within 10 seconds. (Settling may have taken longer in some cases.) The AOTs made use of on-board slew completion and stabilization indicators to proceed with the observation as soon after a slew as is possible. Note that the time required for small slews, dithers, offsets, settling, etc., within an AOR is considered part of the observation.
3.7.2 Pointing Accuracy and Stability
The blind pointing accuracy was the same as the on-board attitude knowledge, <0.5 arcseconds (1σ radial rms) with a stability of <0.1 arcseconds (1σ radial rms over 200 sec). In the incremental pointing mode, the PCS performed controlled repositioning of the boresight with an offset accuracy usually no worse than 0.2 arcseconds and less than 0.6 arcseconds for across angular distances of up to 30 arcminutes. This accuracy was sufficient to move a source from the IRS Peak-Up array to one of the spectrometer slits. MIPS observers should note that the initial blind pointing accuracy for the 160 micron array was expected to be significantly worse than for the shorter wavelengths (~3.9 arcseconds). See the MIPS Instrument Handbook for more information.
3.7.3 Scanning Stability and Performance
Spitzer could execute linear scans at selected rates from 0.01 arcseconds/second to 20 arcseconds/second. The MIPS Scan Map mode used rates from 2 arcseconds/second to 20 arcseconds/second.
3.7.4 Tracking Capabilities
Spitzer does not have a true tracking capability for Solar System Objects (SSOs). However, Spitzer simulated tracking by scanning in linear track segments at rates up to 1 arcsecond/second. The linear track segments are linear in equatorial coordinate space; they were commanded as a vector rate in J2000 coordinates, passing through a specified RA and Dec at a specified time.
The SSO ephemerides were maintained on the ground, rather than on-board Spitzer, but an observer may specify flexible scheduling constraints, and the linear pseudo-track specification (start point, rate and direction, in equatorial J2000 coordinates) was calculated at the time of scheduling. The observation was executed at the scheduled time (within a window of +3, -0 seconds), and the PCS follows the track as specified, assuming the given start point corresponds to the time given in the track command. All other PCS movements could be superposed on a specified track, including dithers and scans.
PCS measurements indicate that the blind track acquisition accuracy was ≤ 0.5 arcseconds, independent of the track rate, which is consistent with expected performance on fixed targets. The track stability was better than scan mode requirements: ~0.5 arcseconds in 1000 seconds. The offset accuracy during tracking was better than ~0.55 arcseconds, also consistent with the required performance on fixed targets.
3.7.5 Pointing Reconstruction
Pointing reconstruction refers to the post-facto determination of where the telescope was pointed with a greater accuracy than was known by the flight system. Spitzer had a requirement to perform pointing reconstruction to 1.4 arcseconds. Based on in-orbit results, Spitzer met this requirement, except for observations scheduled immediately after an instrument changeover (in which case the inaccuracy may be as much as ~2 arcseconds). The nominal pointing is reported in the data products. Based on post-BCD processing and comparison of extracted sources with 2MASS sources, users will receive data with IRAC data pointing reconstruction generally <1 arcseconds, and MIPS scan pointing reconstruction generally <1.5 arcseconds, with respect to the 2MASS coordinate system. Relative pointing (relative separation of two objects within an AOR) is good to <0.5 arcseconds.