Gain Notes for Revolution 252
	
JEM-X1: An interesting revolution with fine gain smoothing. All data from 
this revolution should be useable, though there may be some loss of energy
resolution during the periods when the gain is changing rapidly.
The first two points of each data set are the partial switch-on with the High 
Voltage 3 steps below its nominal value.
The revolution is broken by a short unplanned switch-off which provides one
of the very best examples of fast relaxation (gain decay due to charge buildup
on the microstrip plate) on all four sources.

The switchoff period was caused by an unexplained DPE crash, as documented
by the NRT report:

 A. General remarks
 ------------------- 
 a) 2004-11-06T04z MOc put JMX1, OMC & IBIS into STD-BY modes due to
    some anomaly detection in IREM. IREM was in standard mode instead of
    INT mode. 
    MOC wrote: "after the GS meeting we have disabled the reaction of JEM-X 
    to the Broadcast Packet as recommended by Soeren. We had this morning
    the 16th crash of IREM. JEM-X did not react to this event as foreseen.
     However, 40 minutes after the IREM crash the JEM-X 1 DPE crashed also
     (JEM-X 2 DPE is fine). It is not clear to us whether these events are  
    correlated. However, I have instructed Salma to reactivate the flag for
    both DPE's until the anomaly is analysed. The recovery of JEM-X 1 will 
    take some time (several hours) because we have to reload all patches."  
    
During the crash the unit temperature dropped drastically as can be seen in
the temperature curve in the `Other data and results' column of the table.
 
Xe Analysis: despite strong changes in the instrument temperature and gain, the
Xe line evolution looks very stable and falls well within +/- 1 keV of the
29.6 keV target value. Consequently we can assume that all the science windows
in this revolution have very good gain correction.


CAO 17/11/2006