Last revision 4 April 2000

 TJD 11634-11637 = 31 March - 3 April 2000

The MQT for LA detector 3 stopped again.  It was discovered several days
later when it was noticed that the SPHIST-FIT program for the LADs 
had been failing and that the count rates for det. 3 in LOCBURST-CONT
were zero.  The occurrence during a weekend plus inexperienced
undergraduate student operations personnel led to the delay in 
discovering the problem.

The MQT for LA detector 3 stopped during a data gap; the time was after
   TJD/SoD = 11634/10977.0174 (time of last good packet before the gap)
and before
   TJD/SoD = 11634/13459 (time of start of HER_DATA.3 interval read out
after the gap; the interval had no counts).  (The CONT data resumed, i.e.,
the data gap ended, at TJD/SoD = 11634/13799.1614, showing no counts.)

The problem was fixed by turning off detector module 3 by command BAUB3OFF after
   TJD/SoD = 11637/66857.7214 (time of last good packet based on DISCLA data) 
and turning it on by command BAUB3ON (new version with voltages adjusted) before
   TJD/SoD = 11637/68270.8414 (time of first good packet based on CONT data).


Therefore, the MQT for LA detector 3 (affects CONT and HER data) was out
   between  11634/11000 (arbitrarily set to just after beginning of gap)
   and  11637/68268.

The entire module 3 (also affects DISCLA, DISCSP, and SHER data) was out
   between  11637/66858  and  11637/68268.
Flags 73 and 56 (also temporarily flag 50 for CDFITS) were set.


The gain for LA detector 3 was out of balance (no flag 64 was set)
   between  11634/11000 (arbitrarily set)  and  11637/70000.


The affected triggers were numbers 8058-8064 which were
 Bursts: 8058-8059, 8061-8064
 Flare: 8060