Subject: JEM-X Monday Meeting 2006-08-28
Date: Mon, 28 Aug 2006 14:23:19 +0200 (MEST)

Present: CAO, AH, JC, SB, NL, CBJ, NJW.

JEM-X status (SB, CBJ): Very weak hotspot activity. Else normal.

Coming Crab calibration rev 484 (NL): Remember c.o. Eric Kuulkers and others at ISOC.
No direct response to our request. There could be an issue with respect to telemetry
packets (background is still high).

- check JMX2 performance with old rejection criteria
- check if data can be analyzed when an anode segment is switched off

It would be a good idea to have a plan on what to do in the case a giant solar flare
or a strong long duration flare (to several tens of Crabs) from a source in the FOV
emerges. In that case the TM rate will be insufficient. The JEM-X approach could be
to switch off one or two anode segments. How would IBIS and SPI get around that
situation?

The IMOD detector map can be used to flag the missing anode segment(s).
A simulation of the missing anode segment by removing counts artificially
from a sample data set could be attempted, but there are other issues such as
how the gain calibration will react to missing calibration counts.

Comparison of PIF and shadowgrams (NJW): The Crab spectrum from j_src_properties
deviates from the expected shape. Activity is ongoing to find the root for this
problem and the first guess is that the detector resolution has not been correctly
determined.
Can we at all change to the spectrum determination by images in narrow energy
bins and get the intensity from gauss fit to the source peak? This was discussed.
Perhaps the use of j_ima_fine_resol is a possibility.

(JC): The on-axis Crab flux (3-35 keV) derived from j_ima_iros images gives a number
twice the expected value. Are the counts counted more than once?

j_cor_gain delivery status (CAO): j_cor_gain has been delivered for OSA6. New IC
table (JMXi-GAIN-SCP) with smoothed gain can be produced. Note that the PI
assignment is unchanged with the respect to the previous one on an average level.

j_ima_iros (NL): Use of malloc to assign very large arrays is better than declaring
large arrays but a strange error has crept in.

Burst (SB): We have a case with a burst from on of two sources separated by
3 arcminutes. JEM-X can actually tell from which sources it came.

/NJW