NGC 6940 was observed with the Position Sensitive Proportional Counter (PSPC) on board ROSAT between 1993 Nov 30th 00:44 UT and 1993 Dec 4th 02:22 UT, for a net observation time of 7365 s. A description of the satellite and the instrument can be found in Trümper (1983) and Pfeffermann et al. (1986). The data were analyzed using the EXSAS package (Zimmermann et al. 1994). Due to the degrading of the Point Spread Function at large off-axis angles, we limited our analysis to the inner 20 of the field of view. We followed the standard procedure within EXSAS for the detection of sources. First we produced a background map by removing all possible sources and smoothing with a spline-filter technique. Then we ran a Maximum Likelihood (ML) technique to test for deviations from a pure background distribution (Cruddace et al. 1988). The ML threshold for detection was set at 12, corresponding to a single-trial probability of a chance detection of . A second run of the ML algorith with a lower threshold (ML=10, P=4.5 ) was also run in order to increase the sensitivity. We applied the procedure described above for three PSPC channel bands: 11-240 (total band T, corresponding roughly to 0.1-2.4 keV), 11-40 (soft band S, 0.1-0.4 keV) and 41-240 (hard band H, 0.4-2.4 keV).
In the first ML run, we detected 9 sources in the T band and 15 in the H band. No sources were detected in the soft band. The selected threshold is so high that none of these sources is considered spurious. In the second run at a lower threshold, three additional sources were detected in the hard band. These sources have carefully been checked visually: one of them resulted an obvious spurious detection, while the other two were included in our list of detections (sources 16 and 17 in Table 1). After cross-correlating the hard and total lists, we produced a final list of 17 sources detected in the inner 20 of the PSPC detector. A summary of the sources with their positions, positional errors and detection band is given in Table 1. The reported positions have been corrected for the offset ( ) between the X-ray detector and the optical star sensor by means of a cross-correlation with stars in the Space Telescope Guide Star Catalog (GSC: Lasker et al. 1990). Since six sources could be identified with GSC entries, the boresight correction obtained with this procedure is robust. The 90% error radius includes a systematic error of 3 to keep into account residual uncertainties.
Table 1. Sources detected in the NGC 6940 field. The columns give source ID number, position of the source, 90% confidence radius, countrate, and channel band to which the count rate corresponds (T=11-240, H=41-240).
Although we restricted our analysis to the central region of the PSPC, we also considered a strong source (0.59 0.01 cts/s) detected at a large off-axis angle (source # 18 in Table 1). Given the uncertainties on the point spread function in this region of the detector, we estimate the positional error to be of the order of 1-2.
© European Southern Observatory (ESO) 1997
Online publication: October 15, 1997