PDS_VERSION          = PDS3
RECORD_TYPE          = STREAM
OBJECT               = TEXT
PUBLICATION_DATE       = 1996-11-23
NOTE                   = "MGS RST Instrument Health Report #2"
END_OBJECT           = TEXT
END



Engineering Data
================

     Following statistics were derived from 1996/327 spacecraft engineering
data using query2plot:

CHANNL    TITLE      EU      EU      EU      EU    DN  DN    DN     DN RCRDS A
                    LOW    MEDIAN AVERAGE   HIGH  LOW MDN  AVERAGE  HI
------ ---------- ------- ------- ------- ------- --- --- -------- --- ----- -
L-0200 USO_REG_V   4.06    4.08    4.0791  4.1    203 204 203.9544 205 21297 0
L-0201 USO_OVEN_V  2.58    2.62    2.6122  2.64   129 131 130.6099 132 21297 0
T-0315 USO_T      22.3056 23.4724 23.3876 24.3575 138 134 134.2925 131  5327 0


Open Loop Data
==============

  Data Acquisition
  ----------------

     No new open loop data have been delivered to Stanford since the last
report.  Analysis of 1996/326 data (covering TCM-1) are in progress.

  Errata
  ------

     Last entry in the table of the previous report should have shown
1996/326 data coming from DSS 15 rather than DSS 45.

  Retransmission Test
  -------------------

     Trish Priest requested retransmission of open loop data acquired on
1996/323 at SPC 40.  The original transmission showed 73 missing records
during about 1.5 hours of 5000 sps operation and 4 missing records during
about 5 hours at 200 sps.  We hypothesized that retransmission would be more
systematic, and that we might see fewer errors.

     Trish reported that the 200 sps retransmission appeared to have been
successful, but that (from volume/time estimates) she was not sure any of
the 5000 sps data had been resent.

     We issued new queries to the NERT Cache for both acquisition periods.
The total number of records delivered was slightly less (by as many as 3
records), but the number of missing records was exactly the same; the
the records which were missing were identical to the records which were
lost in the original transmission(s).

     Explanations include:

       (1) The retransmissions never took place.  We have added additional
           code to our software to list GIF Record Creation Time.  This MAY
           allow us to distinguish between original and second transmissions.
           At present, we cannot make that distinction.  We should run
           additional tests in the future when we are certain that
           retransmissions have taken place.

       (2) Records were lost during storage to disk at the SPC.  In the
           past, data stored on tape have tended to show fewer errors.
           We should schedule tests during which both disk and tape
           storage is enabled, then compare the missing record results.

       (3) Records were lost during data acquisition.  If tape and disk
           records are the same, then the only explanation for missing
           records is that the RSSD/DSP system is losing them.  This is
           the most serious scenario from our point of view as there is
           no in-place work-around.


Closed Loop Data
================

     At Stanford we have not looked at any closed loop data yet.


Figure Captions
===============