Date: Fri, 3 Dec 2004 10:54:23 +0100 (MET)
Subject: AI30_1

Dear All,

I have put all the data pertaining to gain history and crab calibrations on
the SDAST Forum:

http://www.dsri.dk/~oxborrow/sdast/sdast_forum.html

Please see the first two entries under `Mission Description Documents'.
The gain history tables contain rev_1 and nrt data from revolution 5, upto
revolution 252. In the case of JEM-X2 the data stop after revolution 239 because
it's been switched off. The tarball of postscript figures contains all the
gain histories and j_cor_gain smoothing fits for the revolutions during which
there were Crab calibrations. These are produced by my ICAL tool, which doesn't
yet include IJD in the output (on to-do list, just before `help negotiate
lasting world peace', or round about OSA5.0, whichever is soonest).

A little note about each of these calibrations:

Colours are: 	calib. source one: red/blue
		calib. source two: green/yellow
		calib. source theee: pink/purple
		calib. source four: orange/turquoise
		
The JEM-X1 figures are split up into Cd (green/yellow, orange/turquoise)  and 
Fe sources  (red/blue, pink/purple)

REVS 44/45: JEM-X1 shows exemplary behaviour, as does the fitting program,
	which handles the unplanned switch off/on very well (even if I say so
	myself), with a handful of small glitches showing up clearly for
	sources 1 and 2. Classic examples of after switch-on decay, but nice
	and flat thereafter. Would that all calibrations could be this
	well-behaved! JEM-X2 is generally the same, but source 4 is
	already very noisy, and later on becomes unfittable in some revolutions
	because program will not allow one source to dictate breaks in the
	fitting curve. Nice medium sized glitch on source 1, rev.45
	
REV 102: Significant decrease in the size of the switch-on decay, but otherwise
	well-fitted and almost unvarying gain for entire revolution. Some nice
	glitches on JEM-X1 source 2. JEM-X2 shows a slight, steady increase in
	gain that is not seen in JEM-X1, presumably due to slightly different
	thermal environments. Which side of the instrument was towards the
	Sun in this revolution? JEM-X source 4 has settled significantly higher
	than all the other sources by this time.
	
REV 170: Well-behaved, well-fitted but a bit of a puzzler. Why do the two
	units have different decay directions, and such strong decays at that?
	All suggestions welcome. Testable ideas will be given preference.
	Otherwise though, another good calibration revolution. JEM-X1 anode
	2 still leading in the glitch stakes: is this anode segment
	particularly sensitive, or just positioned in a noisy place?
	JEM-X1 source 4 continues to be noisier than the other sources in that
	unit. Glitch rate seems to have picked up somewhat in JEM-X2
	
REV 239: A major disappointment! The high voltage drop performed on JEM-X1 is
	simply not picked up by the fitting program, so that one long curve
	(after the linear pre-switch-on) is fitted, where ideally two should be
	used. This means that certain SCWs around the HV switch are unuseable.
	This has happened because the gain drop is simply too small to trigger
	a new curve: j_cor_gain could be tuned up to handle cases like this
	simply based on the small time gap during the step down. This would have
	to be implemented quite carefully to ensure that lots of other
	curve breaks are not introduced in the fittings where they shouldn't be.
	Currently, I don't know of any other revolutions that are poorly or
	improperly fitted like this one (though I'm still looking!). Despite 
	the obvious shortcomings of this fit, the curve provided should be
	okay to use for about 70% of the revolution. I am currently assembling
	a list of SCWs in this revolution and the goodness of the gain fit 
	for each one. It should be noted that the fits provided, do still 
	perform their primary function of smoothing out nasty glitches, i.e.
	towards the end of the revolution for the two Cd sources, and for
	most of the revolution for source 1 (Fe). 
	JEM-X2 is well-behaved and well-fitted, with some nice glitches on
	source 2 again. Again we see the strong up-decay after switch on, which
	can be explained by the fact that the unit is being powered up again
	after a long rest, and here we see the effect of `accelerated aging'
	after the unit had returned (temporarily) to its launch state.
	Several good examples of anti-glitches on source 1 which are artifacts
	of the spectrum fitting program which occassionally fails for this
	source. (SPR-3241)
	
Well that's all the calibration revolution news to date. I'll be sending a list
of unuseable science windows for JEM-X1, rev 239, and hope to complete my search
for other improperly fitted revolutions before Christmas Lunch takes over.

Best wishes,

Carol Anne
	


.DANISH.SPACE.RESEARCH.INSTITUTE.DANISH.SPACE.RESEARCH.INSTITUTE.DANISH.
Dr. Carol Anne Oxborrow
Email: oxborrow@dsri.dk	Homepage: http://www.dsri.dk/~oxborrow
Telephone (direct): +45 35 32 57 33
Telephone (secretary): +45 35 32 57 01
Fax: +45 35 36 24 75