Subject: JEM-X calibration wishes


Hello Peter:

Your response to our discussion on the Crab calibration has prompted 
us to look again at what we want and how to achieve it.

We have identified the need for two different calibration observation
sequences. Both aims at improving our understanding of the working
of the JEM-X discriminators for small signals.

1) Crab observations. We would like to have (for both JEM-X units)
two reference Crab spectra, one taken at nominal high voltage, and
another taken at nominal high voltage minus 40 V (4 steps on the raw
HV-command values). We need to have reasonable good statistics so we
ask for two on-axis observations of 10 ksec duration each, with high 
telemetry allocation (15+15 packets). We want to use these two spectra
as test cases for the ARF's we will derive from the second observation
sequence described below. Both units should be activated at belt exit 
with their nominal high voltage and the voltage should be lowered by 
4 steps for the second 10 ksec JEM-X observation period. After the
second JEM-X obsevation the voltage of JEM-X1 should be set back to
normal and the JEM-X2 should be switched off. The specific JEM-X 
observations should not start earlier than 24 hours after radiation 
belt exit. We suggest to give JEM-X1 7 packets and JEM-X2 1 packet 
for those parts of the orbit which are outside the specific JEM-X 
observations. To assure the best thermal stability it is preferable to
switch on the JEM-X2 DFEE (without high voltage activation) also in 
the orbit preceeding the Crab calibration orbit.

2) Engineering session without strong sources in the field - may very
well be combined with OMC flat field session. There are no pointing
requirements from our side - not even a requirement for a stable 
pointing. Here we will concentrate the analysis on the counts from our 
calibration sources. We want to operate both JEM-X units at seven 
different voltage levels: for JEM-X1: nominal and nominal minus 1, 2,
3, 4, 5 and 6 voltage steps. This will sweep the gain over a factor 
2.2 such that the 6 keV Fe-55 photons will cover a signal range 
corresponding to 2.8 to 6 keV at nominal gain. 
For JEM-X2, where we only have the Cd-109 sources at 22 keV, we need 
to cover the voltage range between nominal minus 9 to nominal minus 15 
steps, sweeping from 7 to 3.2 keV equivalent signal. The duration of 
each exposure should be at least 1000 s and we need 9+9 packets of 
telemetry. 
We will for this engineering session move the calibration signals into
the normal event data stream, this gives us much more detailed 
information about each event - and eliminate concerns about the 
start and stop times for the integration of the calibration spectra. 
The procedure for moving the calibration events into the nominal data 
stream has been tested previously and worked without problems. 
(See Operations Change Request #159, date 23-11-2003).
There is no requirement to have the JEM-X2 DFEE switched on during 
the preceeding orbit, but the engineering session should not start 
earlier than  hours after exit from the belt and switch-on of the 
high voltage for both JEM-X units (nominal minus 6 steps for JEM-X1, 
nominal minus 15 steps for JEM-X2).

We can start the development of our new spectrum extraction software
on the basis of the data Soren already have extracted for JEM-X1 from 
the gain variations seen during normal operations, but we need the
more complete data set to verify our interpretation and to put JEM-X2 
safely on the map as well.

with best regards

niels