Subject: JEM-X Monday Meeting 2006-09-18

Participants: JC, CBJ, AH, SB, CAO, NL, NJW. ILR parttime.

JEM-X status (SB, JC): Works OK, but some problems in TM. Probably fixed by
delivery of consolidated data.
Dead anode 234 has now shown a real decrease (already noticed
in August 2006). The loss rate is a little worrying. Perhaps we should lower the
voltage. SB will check gain situation.

Spare model exercise (SB): 80 hours of activity with test of switch-on procedure and
view of gain increase after HV switch-on. The conclusion is that keeping the lowest
voltage (180 V) during 'off' reduces the relaxation time considerably.
This observation confirms that the test with continued HV during perigee passage
in the coming Crab calibration is worthwhile.

OSA 6 status (NL, CAO, NJW, JC):
j_ima_mosaic (JC): Delivered an adapted version that matches j_ima_iros-1.6.3
                   Can we design some tests that show how well this functions?
                   Both on the input data (from j_ima_iros) and mosaicking.
j_deadtime_calc (CAO): Made and delivered last week. CSSW-HRW ORHK-HRW kinds
                   of housekeeping are mixed. Many rows with only zeros
                   has started to appear and the previous version had
                   difficulties handling this.
                   Perhaps this should be discussed at the coordination meeting.
j_cor_gain (CAO):  Redelivered and running in NRT pipeline.
j_cor_position (CAO) : Redelivered with randomized DETX and DETY
                   The situation with excluded anode segment(s) should be handled.
j_ima_iros (NJW):  Version 1.6.3 is about to be delivered with 'neighbor' correction
                   as default. New IMATYPE names.
                   (A ghost source has been in the j_ima_iros images, which we
                   haven't noticed before. Not all strong sources in the edge
                   of the FOV will produce ghosts. Energy bins where detector
                   resolution is low are more likely to allow ghosts).

Crab calibration (ILR): Be careful with DFEE memory update. (separate meeting).

Shadowgram studies (CBJ): Showed a Sco X-1 observation shadowgram overlayed with
the mask pattern. Some mask holes that are open have no corresponding counts.
Closed by support structure?
(NL): Studies of what pixels are included as background and as source.

/NJW