CalDB Public Release Notes
Version: 3.0.3 (patch)
Effective Date: 2005-05-08T16:00:00

CalDB 3.0.3 (patch) contains only one additional new file beyond
the most recent version 3.0.2. However the last public release of
the CalDB was version 3.0.1, in Feb. 2005. Hence, the public release
of CalDB 3.0.3 includes the 3.0.2 patch as well. See the release notes
of 3.0.2 for details on that upgrade.

I. Summary of Changes:

A. ACIS phase 2 response file repaired (p2_resp)
	Filename: acisD2000-01-29p2_respN0003.fits
	Location: $CALDB/data/chandra/acis/cpf/p2_resp/

We have discovered an incompatibility between the mkacisrmf software and
the currently released p2_resp file (CalDB 3.0.0, December 2004), namely
acisD2000-01-29p2_respN0002.fits. The problem exists in Block 4 of the file, 
the CTI-induced scattering matrices for the FI chips.

Specifically there are 21 regions (tiles)
of the ACIS front illuminated chips that have data that has not been
sorted in increasing PHACHAN order. The result of this is that on these
21 tiles, an RMF with zero values over most of the bandpass will be generated. 
For individual tiles, the zero RMF will yield obviously bad fitting results; 
however for weighted RMFs where a WMAP is employed, the result is less 
obvious, because the amplitude of the resulting RMF will be to some degree 
reduced. In forward fitting of the response to observation spectra, the result 
is an overestimate of the source flux for any given model.

The new file, version N0003, contains the same data points as the old one, but 
for the 21 tiles, the PHACHAN sorting has been corrected. For most of these 
tiles, the RMF will now be acceptable below 10 keV, but will be truncated 
above that value. In a future release of the CalDB, a new p2_resp file will be 
included that contains a broader grid, and will use the same grid for all 
tiles of ACIS.

Threads affected:
Creating ACIS RMFs with mkacisrmf
http://cxc.harvard.edu/ciao/threads/mkacisrmf/

II. Technical Details:

A. ACIS p2_resp version N0003

The 21 tiles with the mis-sorted scattering matrix data are as follows:

REGNUM	CCD_ID	chipx_lo	chipx_hi	chipy_lo	chipy_hi
7	0	1		256		193		224
18	0	1		256		545		576
30	0	1		256		929		960
39	0	257		512		193		224
93	0	513		768		897		928
144	1	1		256		481		512
146	1	1		256		545		576
221	1	513		768		897		928
272	2	1		256		481		512
274	2	1		256		545		576
295	2	257             512             193             224
349	2	513		768		897		928
400	3	1		256		481		512
402	3	1		256		545		576
423	3	257             512             193             224
477	3	513		768		897		928
528	6	1		256		481		512
530	6	1		256		545		576
551	6	257             512             193             224
573	6	257             512             897             928
605	6	513		768		897		928

The only way a user can identify if any of these regions is included in
their RMF calculation is by watching the standard output of mkacisrmf, or
by using the logfile option, and checking this for any of the above region 
numbers. The CAVEAT for this has been posted on "mkacisrmf bugs" at
http://cxc.harvard.edu/ciao3.2/bugs/mkacisrmf.html#sorterr

The CXC has elected to do an emergency patch of the CalDB to alleviate
this problem as soon as possible. Note that this repair does NOT fix
the SEGV problem that occurs on Chip I1, Node 1 for the entire
node. That problem requires a patch to the mkacisrmf software,
currently anticipated for release in June 2005.