About ROSAT	ROSAT Home Page	ROSAT Images

D.3 Performed Investigations

 

Initially, the investigations were based on the assumption that the cause of the observed attitude inaccuracies must be sought in the performance of the star trackers. Later on also the accuracy of the X-ray system was studied. The following tasks were carried out:  

1. A theoretical study was made of the effects that errors in the star tracker correction lens have on the spot size in the image, i.e., , on the scale factor in the star tracker. Possible errors may be a result of neglecting the diffraction index of air in ground calibrations or manufacturing and positioning inaccuracies (DASA).
2. The effect of a modified spot size on the position determination was estimated analytically (DASA).
3. The scale factor was re-determined (GSOC and DASA).
4. The new scale factor was incorporated into the attitude solution (GSOC). 
5. New boresight studies based were performed on observations for which the attitude solution was produced with the new scale factor. No reduction of the attitude inaccuracies was found (MPE).   
6. Several studies of the detailed movement of guide stars in the star tracker were carried out. The variation of the separation of pairs of stars was studied as well as the movement of single stars with respect to the mean position of the stars. In particular, correlations with the spacecraft wobble were found and also with the CCD pixel structure (see next item). However, the discovered effects turned out to be much smaller than the attitude inaccuracies (DASA, GSOC, and MPE).
7. The shifts which affect the measured star positions in both star trackers were inferred. For this purpose the distribution of a large number of guide star measurements in the sub-pixel range of the star tracker CCD was determined (so called ``modulo flatfield''). A characteristic pattern appeared in the distribution which can be explained by a modified scale factor (MPE).
8. A correction procedure for removing the inhomogeneities in the ``modulo flatfield'' was developed. However, it again turned out that the discovered systematic shifts of single star positions (maximum value 2.5'') are to small to make a significant contribution to the observed position errors (MPE).
9. The off-axis dependence of the ``modulo flatfield'' was studied. The shift effects were more clearly apparent, but they are still small. In an extreme case 4'' were found, but on the average the shifts are much smaller (MPE).
10. Test slews were carried out with ROSAT which were specially devised to investigate the star tracker measurement effects and to make a comparison with the gyro measurements. However, these tests turned out to be very difficult for operational reasons so that no significant results could be obtained. Also the accuracy of the gyros measurements was not good enough (DASA).
11. A study of the (otherwise neglected) influence of guide star colour on the derived position of the star in the star tracker. The variation of the spot size with colour results in shifts of the stars by . So it is by far too small to be a major contributor to the observed error (DASA).
12. Ground measurements with the ASTRO SPASS star tracker.  They showed a performance comparable to the ROSAT star trackers (DASA).
13. Development of a correction algorithm for the improvement of the attitude solution. This algorithm uses the a priori known separations of the guide stars. No significant improvement was obtained (DASA). 
14. An intercomparison of the guide stars observed in both star trackers. The scatter in the measured separations and the scatter in the relative movements of the stars were determined. Both are much smaller than the attitude inaccuracies (DASA).
15. Several studies consisting of a comparison of the positions of the star tracker guide stars with the positions of selected bright X-ray sources in the X-ray system.  This comparison was made for the PSPC as well as for the HRI; for the HRI it was made both with and without wobble. These studies showed the attitude inaccuracies to their full extent (DASA and MPE). 
16. Search for a possible time shift between star tracker data and data from the X-ray system which could explain effects correlated with the wobble. No time shift was found (GSOC, DASA, and MPE).
17. Extensive studies of the accuracy of the system of X-ray telescope and detectors.  Small image rotations were found for PSPC and HRI. However, it turned out that the total error budget of the X-ray system is too small to be considered as the major contributor -- systematic errors greater than 3'' can be excluded (MPE and DASA). 
18. An analysis of the precision of the entries in the used star catalogue. The position accuracy turned out to be sufficient and do not lead to significant errors in the attitude solution (GSOC). 
19. First evaluation of in-orbit measurements performed with the ASTRO SPASS star tracker indicate comparably small star tracker inaccuracies (DASA).