Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/calibration/documentation/epic_cal_meetings/200502/mk_frametimes.pdf Дата изменения: Fri May 16 13:29:04 2014 Дата индексирования: Wed Apr 13 13:27:13 2016 Кодировка: Поисковые слова: storm

EPIC

ESAC

A time jump analysis on EPIC-pn with some side results

Marcus G. F. Kirsch M.J. Freyberg, E. Kendziorra
Marcus Kirsch, XMM-Newton European Space Astronomy Center Page 1


menu

EPIC

ESAC

· Do we still have time jumps? · How can we measure those? · Can a time jump detection be implemented in a sort of routine check (Qcheck)? · Are there time issues that we still do not understand ?

Marcus Kirsch, XMM-Newton European Space Astronomy Center Page 2


What is a time jump
Photon a Photon b

EPIC

ESAC

CCD frames

n*frame time

If the time difference between the two arrival times of photon a and photon b is not a multiple of the frame time TIME JUMP
Marcus Kirsch, XMM-Newton European Space Astronomy Center Page 3


How to detect the jump
· · · · ·

EPIC

ESAC

Take all frame times in array: T =(T(i+1)-T(i))/frametime =difference of to next full frametime Plot () If >tolerance Юtime jump
0

Time jumps
Counting mode


Marcus Kirsch, XMM-Newton European Space Astronomy Center Page 4


Mega processing

EPIC

ESAC

· Past: only known problem cases have been checked at MPE · Now: processing of all public and non public available archive data at ESAC · Some characteristic results:
FF LW SW

Marcus Kirsch, XMM-Newton European Space Astronomy Center Page 5


findings
· Slope:

EPIC

ESAC

­ Is the clock drifting? ­ Are the frame times correctly used?

· Diagonal patterns: numerical effects

Marcus Kirsch, XMM-Newton European Space Astronomy Center Page 6


slopes for all ODF ODF
Relative difference of measured frame time
·

EPIC

ESAC

Different slopes for different modes · LW similar to Burst ­1E-5 · Rest similar around 1E-5 · Errors:
­ FF group: 1E-5 frametimes per frame

·

Slight drift in frame times

Oscillator stable, otherwise relative differences should show same shift
Marcus Kirsch, XMM-Newton European Space Astronomy Center Page 7


resulting absolute error
·

EPIC

ESAC

Measured error of frame time [sec]

·

Burst, Timing and Small Window mode do show negligible absolute errors FF, eFF and LW show room for improvement Ю frame times are not fully correct Ю can be calibrated using the result of that analysis

Not critical for exposure time and timing Analysis can be used to refine frame times with very high accuracy
Marcus Kirsch, XMM-Newton European Space Astronomy Center Page 8


Numeric effects
· mjf

EPIC

ESAC

Marcus Kirsch, XMM-Newton European Space Astronomy Center Page 9


How to proceed now

EPIC

ESAC

· Determine frametimes · Repeat analysis with correct frametimes and search for time jumps

Marcus Kirsch, XMM-Newton European Space Astronomy Center Page 10