PSPC Calibration Tasks
(Note: This section is in progress)
Point Spread FunctionThe primary in-flight calibration datasets related to the psf are the Ar Lac pointings. Ar Lac was chosen for this as it has no discernable extended emission, yet it is a bright X-ray source. The early AR Lac pointings were taken at various positions across the detector face, these observations are contained in sequences 110586-110602. Details of these sequences are listed in Appendix B. There was no successful early calibration observation with this source centered in the PSPC, thus in the Hasinger et al psf memo (CAL/ROS 92-001) many proprietary sequences belonging to the authors were used as a test of the psf model.
Other bright Galactic sources were similarly observed across the face of the PSPC, but most of these had some extended component to their X-ray emission. For example Cyg X-2 was observed over sequences 110001-110012, but is thought to have an extended X-ray emission component due to a dust scattering halo.
Ghost ImagingAs this effect is related to the softest X-ray photons detectable by the PSPC, very soft X-ray sources were used to quantify the effect. HZ 43 was observed at various x,y with no wobble, over sequences 110014- 110066, in the same way as the psf observations were carried out (above). The primary use of these sequences is to illustrate the severe ghost imaging effect.
The electronics processing the signals from the PSPC cathode wires have a lower level discriminator setting which allows distinction between electronic noise and soft X-ray events. For soft X-rays the number of photoelectrons and ions created is small. This means that there is a substantial probability that the sum of the cathode induced signal plus electronic noise may never produce a trigger large enough to pass the lower level discriminator. If one of the cathode signals is lost the processing unit assumes that the event occurred directly over tha opposite cathode wire. This results in the formation of point images not at the true X-ray location, but at four places corresponding to the projection of the source position on the nearest cathode wires. In addition there is a chance that if one signal is missed that the signal in the perpendicular coordinate will be similarly missed, resulting in a location inferred on the intersection of the nearest cathode wires.
(Source: Article by Nousek and Lesser, appearing in full in the ROSAT Newsletter --issue 8, page 13.)
Spectral ResponseThe earliest Crab spectral data are PI data (Trumper) these now reside in the ROSAT archive (sequence 500065F is a boron filter observation of the Crab). Boron filter observations of HZ43 (100113F); N132D (100130F); Ar Lac ( 100588F) and other sources are important to spectral studies. Observations of constant X-ray sources such as white dwarves (HZ43 and meaty) and SNR (N132D) allow the project to monitor variations in the response of the detector. calibration observations such as these are combined with calibration line data from the on-board Al K_alpha calibration source are used in the spectral calibration.
Temporal ResponseTiming studies have proven particularly challenging on timescales of a few tens of seconds to minutes, due to the modulations produced in the light curves of a constant source, as the detector wobble periodically occults the source. It has not been possible to date to effectively remove the effect of the wobble from the source light curve. In part, the problem is that the wobble is not strictly along a single vector, and it varies slightly in amplitude and timescale. The accuracy to which the PSPC grid is mapped is also a limiting factor in removing the effect.
Please use the Feedback link if you have questions on ROSAT.