
Date: Wed, 28 Mar 2001 10:06:23 +0200 (MET DST)
From: Carol Anne Oxborrow <oxborrow@dsri.dk>
Subject: j_performance
To: smaisala@astro.helsinki.fi
Cc: peter.kretschmar@obs.unige.ch, njw@dsri.dk, stefan@astro.su.se, juhani.huovelin@astro.helsinki.fi
MIME-Version: 1.0
Content-MD5: KwHb9vII0zPHc9OY6hxPGA==

Dear Sami,

I don't think I ever gave you a formal list of things that should be monitored
by j_performance that come from in-flight calibration. Since V3 in-flight
calibration is slowly growing into a 4-executable monster, I don't think this
list can be complete until j_ical is fully designed. Anyway here's as good a
list as I can think of:

Output from j_calib_gain_fitting:

- the offset and linear gain parameters for the four anodes of the detector
	as listed in JMXi-GAIN-CAL-IDX. There will be 8 new values (2 parameters
	for each of 4 anodes) every 4.3 minutes. These should be monitored:
	- short term: say for the one whole revolution, to detect any
	sudden malfunctions in the detector. If just 3 days  of data are to be
	viewed, we can probably plot all 8 values on just two graphs - the
	large distance in sizes of the offset and slope of the gain parameters
	probably requires two graphs, as do the fact that we have to accomodate
	4 anodes on each graph. One revolution will give just over 1000 values
	for each quantity to be plotted, so you may want to do a running
	average over 3-4 values at a time.
	- over many months: to monitor for slow degradation of the detector
	performance (changes in quantum efficiency, gas contamination etc.).
	In this case you'll probably want to average over all the anodes since
	the single revolution plots can show us whether one particular anode
	is performing (or not performing!) worse than all the others. An average
	of each parameter also over each revolution will also probably be 
	needed to give a clear picture of the general behaviour of the
	instrument. This will give us a single plot with two lines, one for
	gain offset and one for gain slope, averaged over the four anodes and
	averaged over each revolution: about 120 points for each plot in a
	year.
	-for both instruments: While we are concerned with monitoring the 
	details of each instrument, there should be a way of seeing say the
	longterm, averaged data, for both instruments together: they may well
	both behave a little differently but in a consistent way and when their
	relative behaviours differ it could be a sign of something BAD happening
	with one of them.
- Source intensities for each of the four FRSS sources as listed in
	JMXi-GAIN-CAL-IDX. This will be needed to monitor the detector
	 efficiency both in the long term and the short term: the same remarks
	 apply as above for short term and long term monitoring, with the
	 difference that there is only one parameter to monitor for each source
	 (instead of two for the gain) - which should make it easy to create
	 an instructive plot of the source intensities for both instruments
	 together.
- Energy resolution at 8Kev and 22Kev for each instrument, stored in
	JMXi-GAIN-CAL-IDX. Same remarks apply for short-term, long-term and
	the two instruments, as for the gain parameters.

Output from j_***_electronic:

- Not currently sure where this component will be, but the outcome will be the
	same: the results of a linear fit to the ADC channel outputs and
	residuals of the fit, which will allow us to see if one or more of
	the channels is playing up: this will only be significant if the channel
	falls on one of the telemetry energy binning boundaries.
- Offset and slope parameters for the linear fit: longterm display only needed
	here since there's only two values produced every 3 days: about 100
	points per value every year.
- Residuals to the fits: A user should be able to call up a plot of  the
	residuals to the fits wherever the offset and slope parameters look a
	little unusual.


Output from j_ical:

- At the moment it is assumed that only the instrument team will use j_ical
and this will be done whenever the scientific data indicates that interactive
calibration is required (i.e. not regularly). The one exception is the
electronic calibration that takes place once per revolution, which may or may
not be in j_ical. All the output from j_ical is expected to be monitored by
the instrument team at DSRI and is never fed directly back into the ISSW, hence
I do not think this output needs to be part of j_performance.

I hope this makes output from my executables a little clearer.

Best wishes,

Carol Anne


.DANISH.SPACE.RESEARCH.INSTITUTE.DANISH.SPACE.RESEARCH.INSTITUTE.DANISH.
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

