2. Observations and data reduction
The BeppoSAX narrow field instruments consist of one low-energy concentration spectrometer (LECS, Parmar et al. 1997), three medium energy concentrator spectrometers (MECS, Boella et al. 1997), a high pressure gas scintillation proportional counter (HPGSPC, Manzo et al. 1997), and a phoswich detector system (PDS, Frontera et al. 1997) covering the 0.1-10 keV, 1.3-10 keV, 4-120 keV and 13-300 keV energy ranges, respectively. HPGSPC data will not be considered in the present paper since the source is too faint for a correct background subtraction and we restricted the spectral analysis to the 0.1-4.5 keV and 1.5-10 keV energy bands for the LECS and MECS, respectively, where the latest released (September 1997) response matrices are best calibrated. The LECS data were ignored below 0.6 keV, i.e. below the lowest energy at which Mkn 3 is detected at the 3 level.
Mkn 3 was observed for about two days using all the above instruments during the period April 16-18, 1997. The total effective exposure was 39433 s in the LECS, 113690 s in the MECS and 51386 s in the PDS. Light-curves and spectra were extracted (using the ftools v4.0 software package) from within a region of 4´ radius centered on the source for both LECS and MECS instruments. Standard blank-sky files provided by the BeppoSAX Science Data center (SDC) were used for the background subtraction. Similar results were obtained using the background spectra extracted directly from the instruments fields of view (FOVs). The 3 MECS instruments gave consistent timing and spectral results if taken individually and were, thus, added together. In the LECS and MECS FOVs, we also detected emission in the direction of the BL Lac object MS 0607.9+7108 about 6-7´ NW of Mkn 3, that certainly lies also in the PDS FOV (which has a triangular response with FWHM of 1.3o, Frontera et al. 1997). Although the statistics are too poor for a detailed spectral analysis of this source, we can, however, exclude the possibility of contamination for the present analysis because a) it is weak (about 1/100 the 2-10 keV flux of Mkn 3) and b) its spectrum is steep as indicated by the fact that its detection is clear in the LECS FOV but marginal in the MECS FOV. Similar considerations were also reported by I94 and G98. Searches in high-energy catalogues reveal no other likely contaminating source within the 1.3o FOV of the PDS.
In total, the source count-rates were (6.80.5) 10-3 cts s-1 in the LECS and (6.950.08) 10-2 cts s-1 in the MECS, the background contributing 20% at 6 keV and 7% at 8 keV, respectively. No significant variability was found within the statistical fluctuations of 15% (MECS data), thus all data were accumulated over the whole observation. The PDS data reduction was performed using both XAS (v.2.0, Chiappetti & Dal Fiume 1997) and SAXDAS (v.1.3.0.) software packages and yielded consistent results. In the following, we'll refer to the results obtained with the XAS package. The source count-rate in the PDS was 1.16 0.02 cts s-1 which corresponds to a total of 50 detection in the 13-150 keV energy range and 7 between 100-150 keV. No significant variability was found but we cannot exclude variations up to a factor 2 (corresponding to the statistical fluctuations) within the entire observation.
For the spectral analysis, the LECS and MECS data were rebinned such as to sample the instrument resolution with a number of 1 and 5 channels per energy bin, respectively (about 20 and 50 cts bin-1). The PDS data were grouped logarithmically between 13-200 keV, with a S/N ratio 3 per bin. The spectral analysis was performed using version 10.00 of the XSPEC program (Arnaud 1996).
© European Southern Observatory (ESO) 1999
Online publication: March 29, 1999