The detector plates in the JEM-X units carry microstrips that run up and down along the plate. Occassionally, excess charge in the plate can damage these conducting strips so that they no longer transmit the electron cloud charge into the digital front end electronics as well as they once did.
This damage leads to a reduction in the number of events coming from a particular strip, either partially (unstable anodes), completely (dead anodes) or transiently (undead anodes). Since the analogue data from the detector plate is binned into 256 x-position values before telemetry transmission, the death of a single physical anode strip may appear as the loss of more than one x-position value.
The events from dead anodes are flagged as coming from an area of unknown activity, so that these can be removed from the processing pipeline after calibration and correction. This is necessary, for flux determination since areas of reduced sensitivity cannot be counted as active detector areas.
The list of dead, unstable and undead anodes is maintained by Jerôme Chenevez, based on his offline analysis of the recent event counts on each anode. It can be seen here: List of dead anodes. The bad anodes in the list are labelled in 3 categories:
1) dead: very low event activity. These can arise in-flight or already be present prior to launch. Dates are given in the list for the death of each anode. 2) unstable: have been seen to have low activity 3) neighbours: either side of dead anodes, these may have lower than usual activity if the `anode' x-position is actually partially binned with a dead microstrip, or enhanced event activity if electrons that should be discharged through a dead strip are deflected to a neighbouring live strip.
Earlier in the mission, a spatial gain variation table was used to correct for gain differences caused by the presence of dead anodes, but these effects were found to be very small in comparison with differential gain aging effects, and this table of corrections was dropped from the instrument model tables.
Dead anodes can easily be seen in science data shadowgrams as vertical strips with few or no events. Both units have their `valley of death', x-positions 73-78 for JEM-X1 and 140-150 for JEM-X2. These show up best in empty fields with lots of background or very full fields where the entire plate is illuminated quite evenly.
CAO 26/4/2006