Due to a problem with an ICS in the AOT logic, SWS Fabry-Perot observations were temporarily suspended in early April. The problem affected the SW grating used by simultaneous grating observations, causing the mission planning software to crash. The corrected AOT logic was delivered on April 10, tested by April 23rd with AOT 7 observations released to the planning system soon after. No skybins were lost.
In Rev 528, SWS worked nominally until the last observation. During it all detector signals suddenly jumped in value. During the next three formats the analogue housekeeping values went to raw values out of conversion table (therefore shown on the Instrument Controllers screen as "?"'s). Further anomalous behaviour was seen for the next few formats and in view of this behaviour SWS was switched off. When it was restarted (in a safe mode) a few minutes later all values appeared to be nominal and since this event operations have continued nominally. More information on this anomaly can be seen on the Real-Time Report page.
After investigations by the SWS Support team (SIST) the most probable cause of this anomaly has been identified as a corruption of one or more on-board RAM locations by a particle hit, although since SWS was switched off a RAM dump could not occur. Whatever changes in RAM may have occurred, they were not of a permanent nature: the results of the on-board power-on memory tests at the start of revolution 528 and after the anomaly both indicate no memory errors.
In the unlikely event of this occurring again (it has only happened once in 550 revolutions), the instrument switch off procedure has been modified to include a RAM dump for later analysis.
Possible Detector Memory Effects in SWS Grating range spectra (AOT S06) due to SWS Reference Scans