DPA-B Anomalous Shutdown (BEP-A Active)¶
What is it?¶
The DPA-B shuts down anomalously, presumably due to a spurious command.
Note
IMPORTANT: The diagnosis and response to this anomaly presented in this document assumes that BEP A, powered by DPA side A, is the active BEP.
This anomaly requires human intervention to recover to science operations.
When did it happen before?¶
The DPA-B has shut down once:
- December 13, 2007: 2007:347:17:50, obsid 58072
Will it happen again?¶
It appears likely that the anomaly will occur again if the mission continues.
How is this anomaly diagnosed?¶
Within a major frame (32.2 seconds), one should see:
- 1DPPSB (DPA-B Power Supply On/Off) change from 1 to 0 (On to Off)
- 1DPP0BVO (DPA-B +5V Analog Voltage) drop to 0.0 +/- 0.3 V
- 1DPICBCU (DPA-B Input Current) drop to < 0.2 A (this value is noisy, so take an average)
- DPA-B POWER should go to zero
- 1DP28BVO (DPA-B +28V Input Voltage) is expected to have a small uptick, ~0.5 V, consistent with the load suddenly dropping to zero. May require more than 1 major frame.
- If 1STAT2ST = 0, the DPA-B shutdown has caused a watchdog reboot of the BEP in use. This will happen if the DPA-B shuts down during an observation which is using B-side FEPs. If a watchdog reboot occurs, the software will revert to Version 11.
All other hardware telemetry should be nominal. The current values for these can be found on our Real-Time Telemetry pages. Older data can be examined from the dump files or the engineering archive.
To extract information from the dump data, run ACORN on it as per the instructions in “Running ACORN on data dumps in the case of an anomaly (04/06/16)”. Information from the tracelog files written by the ACORN tools can be plotted using the Python or command-line interfaces to ACISpy, see below for details.
What is the response?¶
Our real-time web pages will alert us and the Lead System Engineer will call us. We need to:
Send an email to the ACIS team at the official anomaly email address,
acis-anomaly -at- googlegroups -dot- com. If it is off-hours, another ACIS Ops team member should call Peter, Bob, and Jim Francis.Send an email to
sot_red_alert@cfaannouncing that the ACIS team is aware of the DPA-B shutdown and is investigating, and that a telecon will be called when more information is available.Contact the GOT Duty Officer to inform that we need the dump data as soon as possible and to email or call us when the dump file is available.
Process the dump data and make sure that there is nothing anomalous in the data BEFORE the shutdown. We want to know if a new occurrence looks just like the single previous occurrence. If yes, it should appear as if in one frame the DPA-B turned off.
Once analysis of the dump data is complete, convene a telecon at the next reasonable moment with the ACIS team and review the diagnosis. The MIT ACIS team (Peter Ford, Bob Goeke, Mark Bautz, and Bev LaMarr) should also be included in the discussion, either in the telecon or via email. If the diagnosis is consistent with previous DPA-B anomalies, proceed with the recovery. If the diagnosis is not consistent with previous DPA-B anomalies, stop and start a more involved analysis with the ACIS team.
As soon as the time of the DPA-B shutdown is known, inform
sot_yellow_alert@cfa.Identify whether or not additional comm time is needed and if so ask the OC/LSE to request it.
Send an email to
sot_red_alert@cfaand call a telecon with the FOT, SOT, and FDs to brief them on the diagnosis and the plan to develop CAPs to recover.Prepare the CAPs and submit them for review to capreview AT ipa DOT harvard DOT edu, and cc: acisdude. Call the OC/CC to determine which number should be used for the CAPs.
The steps in the main CAP will depend on whether or not the active BEP has executed a watchdog reboot. This may happen if the shutdown occurs during an observation that utilitizes the side B FEPs (DPA side B powers FEPs 3-5), or if a subsequent observation requests them. Note that this implies that a watchdog reboot of the BEP will be avoided only if it occurs during an observation using only 1 or 2 CCDs, and until an observation occurs using 3 or more CCDs.
If BEP A - the active BEP - has not executed a watchdog reboot, the steps should be:
- Turn on DPA side B (
SOP_ACIS_DPAB_ON, this can be executed even if a science run is currently in progress, see note below).
Otherwise if, in addition to this anomaly, the BEP A has executed a watchdog reboot, the steps should be:
- Stop the science run and power down the FEPs and video boards (
SOP_61021_STANDBY) - Turn on DPA side B (
SOP_ACIS_DPAB_ON) - Warm-boot the BEP and start a DEA housekeeping run (
SOP_ACIS_WARMBOOT_DEAHOUSEKEEPING).
Unique circumstances occur during perigee passages. As of now, an ECS measurement will use no less than 4 chips. This means that a watchdog boot will occur should the anomaly occur during an ECS measurment. Presently, after the inbound ECS measurement is complete, a
WSPOW00000is issued followed by aWSPOW0002Aan hour later. Timing of the shutdown could mean that either or both of the WSPOW commands were not executed. In this case the WSPOW0002A command should be added to the CAP, if time permits.A CAP to update txings values from their defaults to most-recent settings should follow the main recovery CAP if possible. A template for this is in
acis_docs/CAPs:CAP1622_TXINGB_SETPARAMS.- Turn on DPA side B (
Execute the recovery CAPs at the next available comm.
Write a shift report and distribute to
sot_shiftto inform the project that ACIS is restored to its default configuration.
Note
At this point in the mission (Jan 2017), it is standard practice to power off unused FEPs to reduce power consumption and keep the electronics temperatures lower. For this reason, it is believed that it is safe to power on the DPA side B during a science run that does not use these FEPs. If this practice is changed later in the mission, this procedure may have to be revisited.
Impacts¶
- Until the DPA-B is powered back on, science operations which require the use of the side B FEPs will be affected.
- If it is necessary to warm boot the BEP, this will reset the parameters of the TXINGS patch to their defaults. If not updated during initial recovery as above, txings settings should be updated as soon as possible via CAP (see CAP 1622) or SAR to prevent undesired radiation shutdown.