[Date Prev][Date Next][Thread Prev][Thread Next][Search] [Main Index] [Thread Index] [HEASARC Archives]

ASCA updates

ASCA Users,

This message contains 6 items:

	1. ASCA AO-6 proposal deadline reminder
	2. 4-CCD and 2-CCD mode observations during AO-6
	3. Rev 2 (re)processing news
	4. 'Why can't ascascreen find my files'?
	5. Bug fixes in addascaspec and ascascreen
	6. An article on GIS gain

1. Important reminder

ASCA AO-6 proposals are due by COB on Thursday, August 13th.

2. 4-CCD and 2-CCD mode observations during the AO-6 period

The SIS team has recently produced an updated prediction of the
hot/flickering pixel rate in different chips in different modes,
through 1998 December (the end of the AO-6 period).  Together with
considerations of other secular changes (in detection efficiency
and spectral resolution), here are some predicted performances
and some recommendations.

(1) Although the spectral resolution continues to slowly degrade,
1-CCD mode remains otherwise unaffected.  This is the mode we
recommend for point source observations, particularly 1-CCD Faint
mode if that does not lead to telemetry saturation.

(2) The combination of 2-CCD Faint mode at High bit rate and
2-CCD Bright mode at Medium bit rate remains viable in terms
of hot/flickering pixels.  However, if the CCDs are warm, data
below ~0.48 keV may have to be thrown away using the lower level
discriminator in the on-board processing.  The RDD effect
will have degraded the detection efficiency to the 80-90% level
the end of the AO-6 period; although the spectral resolution will
also have degraded, the effect of non-uniform CTI effect (which
is independent of CCD clocking mode) is believed to be dominant
for 2-CCD mode.  The use of 2-CCD mode for point sources is
not highly recommended, unless an accurate determination of the
local background is of overriding scientific importance.

The tool 'correctrdd' allows the users to recover 1-CCD mode
like resolution and efficiency for Faint mode data.  However,
2-CCD Faint mode at Medium bit rate requires data below 0.7 keV
(nominal temperature) or below 1.1 keV (warm) be thrown out.

(3) Without an RDD correction, the quality of 4-CCD mode data
has declined to such an extent that we no longer recommend its use.
While RDD corrected 4-CCD mode data are believed to be of comparable
quality to 1-CCD mode data, its use during Medium bit rate requires
a level discrimination set at ~1.5 keV (0.55-0.7 keV at High bit rate).

* * *

Mixing of different clocking modes or of different level discriminations
within a single observation is not recommended because both affect the
response of the SIS.

The nominal CCD temperature is -61.7C; however, the on-board temperature
regulation is not always able to achieve this, depending on the orbit and
the attitude of the satellite.  For this study, the SIS team designated
the temperature range of -62 - -61C to be 'nominal', which applies to ~half
of all observations so far; and the range of -61 - -60C 'warm', even though
excursions up to ~-59C are also known.  The hot/flickering pixel rate when
the CCDs are warmer then -60C will be studied in the future.

3. Rev2 (re)processing news

In late June, we have started Revision 2 (rev2) processing of ASCA data.

If you receive your observation processed this way, there is a pink sheet
of paper informing you of this.  For archival data, the latest processing
version of a given sequence, and other info, can be found in the
Processing Status web pages at:


The highlights of Rev2 are given in the new getting started guide:


The rev1 getting started guide is also available at 


for comparison, as well as for the users of rev1 data.

We are reprocessing in reverse chronological order (most recently processed
sequences first), and we expect to complete rev2 reprocessing in about a year
(July 1998)

4. 'Why can't ascascreen find my files'?

If ascascreen fails to recognize your ad*.unf files in the unscreened
directory, you are probably using FTOOLS v3.6.1 or older.  Try
'ascascreen -e unf' instead.

If ascascreen fails to recognize your ft*.fits files in the raw
directory, you are probably using FTOOLS v4.0.  Try
'ascascreen -e fits' instead.

The reason we switched the default is that the Rev2 reprocessed sequences
will not have the raw/ft*.fits files.  They are too numerous to be convenient,
and unscreened/ad*.unf files exactly duplicate all the information in a small
number of files.  The minimum set of files necessary to carry out your own
ASCA data screening is these unscreened/ad*.unf files and the aux/ft*.mkf
files; ascascreen will assume that the mkf file(s) can be located as:

5. Bug fixes for addascaspec and ascascreen

* addascaspec (v1.21): The name used for the first command-line parameter
     for 'addarf' is not compatible with the latest 'addarf' in ftools
     release v4.0. The parameter name should be 'list', instead of
     'in_ARFs'. Replace $FTOOLS/scripts/addascaspec with the fixed version
     (v1.23). A few more improvements have also been made (type 'addascaspec
     -h' to see details).
* ascascreen: If only one file is read into ascascreen and SIS or GIS
     cleaning is requested then the resulting events list is not mkf
     filtered. This problem is most likely to show up when analyzing SIS
     BRIGHT2 data. Ascascreen runs on raw (ft*.fits) files are unlikely to
     have been affected, because there are generally dozens of input files;
     even with unscreened (ad*.unf) files, if the datamode selected mixes
     together High and Medium bit rate then the bug will not be triggered.
     The fix is to replace $FTOOLS/scripts/ascascreen with the new version
     (Thanks to Vinay Kashyap for helping us spot this one).

Note that you may well have to edit the first lines of these scripts to point
to your local version of perl.

6. An article on GIS Gain

An article by Idesawa et al, intended for ASCA Newsletter vol 5.,
"Calibration of Temporal and Spatial Variations of the GIS Gain"
was not included in the printed version due to a technical problem.
The article is now available as a gziped Postscript file at


It will be incorporated into the WWW version of ASCANews no5 in the near future.