[About EXOSAT] [Archive] [Software] [Gallery] [Publications]


A number of separate correction factors which must be applied to the LE data in order to determine flux intensities are summar ised below:

1) The 'Flap'

2) Sum Signal Distribution

  • Required to correct for counts below the effective lower threshold of the sum signal distribution.
  • determined from the observed sum signal distribution by fitting a Pearson type I function (prog. EFCOR at ESOC).
  • typical correction: divide by 0.98.
  • A fixed correction of this value is made in the Observatory software as for (1).
  • refer to FOTH Sect. 8.1.3. for notes on the sum signal calibration.

3) Sample Rate Dead Time Correction

  • Loss of events caused by the restriction of ~< 1 event per sample interval (1.92 ms at the nominal rate of 512 Hz) determined by normal Poisson statistics.
  • multiply count rates by a factor C defined as C = 512loge(1-T/512)/T where T is total no. of counts sampled per second. T is the observed count rate corrected for point (5) below.
  • correction is typically 4% for the nominal background count rates.
  • This factor has recently (3.7.85) been included in the image accumulation software (ADIMA) and count rates derived from these images are corrected. The LCURV program has always included this effect.

    Note that with the use of a diamond filter to reduce telemetry loading, T, the total no. of counts, must be inferred from the average local background.

4) 'Vignetting'

  • Variation of mirror efficiency across the telescope FOV.
  • Typically +5-10% for pointed sources - Note that the spacecraft attitude is normally such that the target of observation is expected to appear at 8-10 arcmin fron the optical axes of the LE1 telescope. Refer to CCF for other off-axis values.
  • Included in observatory analysis software (SOSTA) since Sept. 85.
  • See FOTH Sect. 8.1, fig. A.

(5) Telemetry Dead Time (Exposure fraction)

  • Because of telemetry limitiations imposed by the requirements of the ME/GS data and the background variability in the CMA it is generally not possible to transmit all events collected by the telescope and the experiment therefore has an effective 'ON' time somewhat less than elapsed time.
  • Calculated fron the buffer collection time in the packets and the difference between packet reference times of adjacent packets.
  • very variable.
  • this has always been included in the observatory telemetry reading software (ADIMA, LCURV).

(6) Internal Electronic Dead Times

  • Examination of the time tag distribution of LE events shows that a small fraction of events with specific values of the time tag related to the electronic reset are 'lost'. This effect of the order of 0.75% is negligible for all count rates and is therefore not included in the Observatory analysis software.

J. Osborne

[About EXOSAT] [Archive] [Software] [Gallery] [Publications]