Optimization of Al_poly FLD Filter

Optimization of the Open/Al_poly FLD (flare detection) Filter.

(10-Jul-2022, initial version by Aki Takeda)

Summary

We started to use a new version of FLD table (#001A: Open/Al_poly at 4 msec) since 23-June-2022, and on 8-July-2022, XRT observed the first major flare ( M2.5 event from AR13053) with the flare mode triggered by the new FLD table:

Flare mode detected:  8-Jul-2022 20:21:48.008 -  8-Jul-2022 20:45:47.061
[Bad_Photo entry/exit: 20:46:00 - 21:22:00]
Flare mode detected:  8-Jul-2022 21:23:09.394 -  8-Jul-2022 21:30:28.364

#Event  Begin   Max    End  Obs  Q  Type Loc/Frq  Particulars       Reg#
6050 +  2007   2049   2128  G16  5   XRA  1-8A    M2.5    7.9E-02   3053

At least as long as this event is concerned, the new FLD table is working properly : the timing of flare trigger on/off is reasonable, causing no excess triggering after the main phase, as shown below (click image to see movie).

Review of Chronological Record

2017/05/27 : In response to the 3rd pre-filter failure event, the FLD filter was changed from Open/Ti_poly (8 msec) to Al_poly/Open (8 msec).
2017/06 -- 2017/10 : The performance of the new FLD filter had been monitored and concluded that there was no problem in the current settings of the FLD parameters (see this report ).
2018 -- 2021 : Years of low solar activity (no problem reported).
2021/12 -- 2022/05 : Started to see false (excess) flare triggers: Flare mode was triggered after every Bad_Photo events or start of a new XOB. Typically occured after the major flare events and lasted for a few days.
2022/05/18 : Double "FLD reset" command was tried in the XOB (but not Bad_Photo event) in an attempt to ensure the reset of previous FLD image.
2022/06/02 : It turned out that the root cause of the false flare triggering was NOT the imcomplete FLD reset, BUT the saturation of the FLD images on the bright ARs. The shorter exposure (4msec) version of FLD table (#001A) was created, but not used because of low solar activity. The attempt of Double FLD reset was called off.
2022/06/23 : FLD table was switched to #001A from #0019(Al_poly 8msec).
2022/07/08 : The new FLD table turned out working as it should be for the observation of the M2.5 flare on 7/8.