Here is the content of the MGS Status Report that described the February 17, 2000, events:
Launch / Days since Launch = Nov 7, 1996 / 1206 days
Start of Mapping / Days since Start of Mapping = April 1, 1999 / 328 days
Total Mapping Orbits = 4305
Total Orbits = 5987
The pace of the new mapping Beta sequence generation problem, a build
every week, and a realization of new operational constraints imposed by
the new mapping scheme, has caused two sequencing errors over the last week.
On 00-048 (2/17/2000) at about 00:00 UTC, the new Beta Supplement sequence, mm014b, was scheduled to kickoff but never went active. Since that new sequence controls the tracking of the HGA to Earth, no downlink was possible. Using a 34-m HEF station, a command was successfully uplinked via LGR-2 to change the uplink rate from 125 bps to 7.8 bps. Then, at 7.8 bps, a mini-sequence was successfully radiated to auto-track the HGA for about 30 minutes per orbit for 4 consecutive orbits, allowing for a 4K real-time downlink. During the first of these orbits, spacecraft telemetry was obtained which showed that MGS was in a healthy condition. During the second of these orbits, the uplink rate was changed back to 125 bps and the new mm014c sequence was uplinked to resume normal operations. The mm014c sequence began execution nominally at 14:20 UTC and has successfully completed execution. The mm015 sequence, was uplinked and began execution on 00-051 (2/20/2000) at 00:00 UTC. The mm015 sequence is scheduled to run through 00-054 (2/23/2000). The net effect of the anomaly was the loss of most of DOY 48 recorded science telemetry.
The cause of the mm014 sequence anomaly was that it was loaded into the incorrect area of the sequence buffer, which was already executing the mm013 sequence. This error was induced by having to split the 7 day beta supplement sequence into two parts due to the size of the sequence and the much reduced uplink time available each orbit due to the complex HGA management. This splitting process introduced an ambiguity in the naming convention: mm013a was used for Part 1 of sequence 013 rather than revision level a, and mm013b was used for Part 2 of sequence 013 rather than revision level b. When a second revision of Part 1 was required, mm013c was used, since mm013b was already taken by Part 2. Due to the ambiguity in the naming convention, mm014 was run through SEQTRAN using the FINCON file from mm013c (Part 1) rather than mm013b (Part 2). The result was that mm014b was loaded into the same buffer area as Part 2 of the previous sequence. Since this previous sequence was still executing when mm014b was uplinked, the new sequence did not go active.
To confound matters, due to the current MGS to Earth range, we are not
able to support two-way data (uplink and downlink data) at the high rate
40 ksps real-time downlink rate. Therefore all uplinks are currently performed
during orbits in which the recorded data from the previous day is being
played back. Therefore confirmation of a successful sequence uplink and
activation requires waiting until the next day's data is played back. For
this case there was insufficient time available to determine that the sequence
load had been rejected and to rebuild a replacement sequence in time. This
was due in part to a half-day loss in the schedule due to a problem with
our spacecraft test lab (STL).
The second anomaly occurred with the development of the mm016a sequence, which was uplinked successfully to the spacecraft on 00-052 (2/21/200) and scheduled to begin execution on 00-055 (2/24/00) at 00:00 UTC. In this event an error in the recording and playback of the science data was discovered after the sequence had been uplinked. The sequence was successfully terminated on 00-054 (2/23/2000) at ~14:00 UTC and a replacement sequence, mm016b, uplinked several orbits afterwards upon confirmation of the termination of mm016a. The root cause for this anomaly was simply a rush to uplink the sequence early enough to verify its activation in order to avoid a repeat of the mm014 sequence anomaly.
Several changes have been made to the sequence development process for the weekly sequence builds. The first change, which was already going to be made for the next sequence anyhow, was to rename the sequences. The split of the weekly sequences into two parts creates two sequences, mm016 and mm017, for example, rather than mm016a and mm016b. The second and probably the most important change made was to increase the sequence development time by three days to provide sufficient time with margin for development, review, STL validation, uplink and verification of the sequences.
One final note, the failure to get a new sequence on-board the spacecraft poses no risk to the health of the spacecraft. The spacecraft will maintain mapping nadir pointing and articulation of the solar array. Communications from the spacecraft and collection of science data will be interrupted until a new sequence can be uplinked.
All subsystems are reporting nominal health. A sixth PDS reset occurred on 00-019. So far no cause has been found for the recurring resets. The resets appear to have minimal if any impacts to the instruments. We continue to forward all reset data to the PDS software experts at JPL.
There have been 14 uplinks to the spacecraft during the last week, including new star catalogs and ephemeris files, instrument command loads, and the mm014, mm015 and mm016 sequences. Total command files radiated to the spacecraft since launch is 4459.
The first fixed-HGA mapping sequence of the beta supplement era is scheduled for March 6 and development of this sequence has begun. One of the limitations of the mapping beta supplement scheme is a loss of South Pole radio science atmospheric data due to HGA gimbal constraints. The project approved a three to four day period each month to perform fixed-HGA operations, in which the HGA is pointed along the spacecraft +X axis and the spacecraft is moved to point the HGA to Earth for an eight-hour period (i.e. four orbits). During the eight-hour Earth pointed period the science data recorded from the previous 24 hours of nadir pointed mapping is played back. This scheme allow radio science observations to be made at Earth occultation ingress and egress, at the expense of collecting mapping data for the eight-hour Earth pointed period. The project also approved monthly MOLA nadir off-pointed polar observations. These observations will be performed during the 24 hour nadir pointing period prior to the first eight-hour fixed-HGA Earth pointed period.