Known Issues and Limitations

1. The JEM-X lightcurves are deadtime corrected. DEADC in the lightcurve files are set to 1.0 (for XRONOS compatibility).

2. Due to changes of the on-board configuration, the detection efficiency has changed significantly several times during the mission history. In particular for pointings between revolutions 26 to 45, this means that the measured fluxes of stable sources - in particular at low energy - will strongly depend on the time when the data was taken. These changes are not corrected for in flux units (counts/cm2/s in the given energy interval) but taken into account in spectral responses.

3. The JEM-X detector gain varies significantly for a few hours after the instrument has been switched on. This mostly affects the beginning of each revolution but can also happen if the instrument was shut down, e.g., for solar flares. The pattern is very similar each time and modeled in the gain correction step even in complicated cases. Nevertheless, it could in principle fail, in which case linear-interpolation gain correction values would be used, which could lead to distorted spectra. Users are advised to check this possibility in case of highly unusual source spectra e.g. by consulting

   http://outer.space.dtu.dk/users/oxborrow/sdast/GAINresults.html

4. If the gain correction step fails then take a look at the gain history table. Gain correction failure is often signaled by all corrected events having a non-zero STATUS value due to bad gain determination (64). If the gain history for your revolution shows multiple switch on/offs, this may be confusing j_cor_gain. Then remove all gain history values up to the switch on/off just before your SCW being analyzed. For help fitting data in these complicated revolutions contact Dr. Carol Anne Oxborrow at oxborrow@space.dtu.dk .

5. The source coordinates found by j_ima_iros may deviate a little from the true positions and this can occasionally cause inaccurate flux reconstructions from j_src_spectra or j_src_lc. If a good source position is available, it is better to force these coordinates by use of a user catalogue. An example is given in the cookbook (but see also point 8 below).

6. Lightcurves from weak sources may be contaminated by counts from stronger sources in the FOV. This happens because the source extraction does not take into account the presence of the other sources.

7. If you mix FULL and REST data then be sure to give chanMin/Max that match REST channel limits, for example:

           chanMin:   64  128  160 192
           chanMax:  127  159  191 223
   

8. In OSA v.7.0 and later, the source position reported in columns RA_OBJ and DEC_OBJ of JMXi-SRCL-RES will always be the one found by j_ima_iros. Columns RA_CAT and DEC_CAT reflect the catalog position if a user catalog has been defined. The SPE and LCR levels will read the RA_OBJ and DEC_OBJ columns and do the extraction using those. In order to force the use of the catalog positions - which is recommended - the JMXi-SRCL-RES table must be manipulated e.g. by an ftool, to update columns RA_OBJ and DEC_OBJ.

9. Light curve extraction is unchanged in OSA 10 compared to previous versions in order to allow the easy generation of short-bin light curves. However, long-term stability is not assured in this case; the user interested in long-term light curves or who doesn't need time bins shorter than the length of a science window is advised to generate light curves from the imaging step, as explained in the cook book.

10. It has been noticed that in mosaics of JEM-X images a plus-like depression in the background around certain sources can occur. This can happen for sources that are too weak to be noticed in the search for sources in the individual science windows. The cleaning process excludes (known) source areas. It operates horizontally and vertically since the systematics are strongest in these directions. However, adding many images can amplify the effect of an unnoticed source since the distribution of position angles is quite narrow, in particular for the sources near the galactic center, which is also where the probability to find a source in the depression caused by a neighboring source is highest. If a source is situated in such an indentation its peaksize is reduced accordingly, whereas there is no change for the source causing the feature. This is solely an image feature so j_ima_iros flux determinations are unaltered.

11. A count-limiting mechanism, the grey filter, is actived, when sources corresponding to more than 0.75 Crab on-axis are in the field of view. The grey filter is adjusting itself automatically, according to the rate of events accepted as X-rays and the filling level of the onboard telemetry buffer. Ideally, a grey filter should randomly reject events. However, the mechanism implemented is only pseudo-random. Therefore some care should be taken in interpreting power spectra of arrival times of events from very bright sources with a very significant grey filter, as QPO artifacts may show up. Normally, the automatic grey filter is varying over a science window. This fortunately has the effect of "averaging" out power spectra artifacts, as they are specific to a particular grey filter setting. Therefore, if noticing transient features in the power spectra of very strong sources it should be checked if this is limited to a period of a specific grey filter setting. Please check the User Manual for further explanations.

12. Since 30 March 2012, there is a new instance of the Instrument Model Group (IMOD files version 22) produced by the JEM-X Team. The usage of these new IMOD files is highly recommended, and will be automatic upon update of your copy of the Instrument Characteristics files. Note however that the data are fully reliable only above 5 keV.

13. For the time being it is not trustable to extract spectra of strong sources with "mosaic_spec" from images obtained with the PIF option (6.6.3).

14. The flux of a given source can be obtained either with the ``standard'' extraction or with mosaic_spec. In case the fluxes obtained with the two methods would differ, it is suggested to consider the one obtained from the standard extraction (SPE level).

15. When analysing very-Near Real Time data (i.e. within a few hours from the data transmission to ground, and well before the entire revolution is completed) JEM-X gain calibration might fail. This happens when there are not enough data collected as yet to describe the gain evolution for the current revolution. In this case we suggest to run the analysis with parameter "COR_gainModel=2" to force the use of a simplified fitting model. The default model (COR_gainModel=-1) can be used again at revolution completion.