AMANDA-DAQ - Reader2001 Page

---

The current reader release is reader2001 from DAQ98-2.1.x/a.
This release is to handle 1998-2001 data.
Test-Phase.
Please, send input from 1998/1999 analysis test results with this test release.

Comments:

• This is the first release that unifies data handling for the 2000/2001 and 1998/99 data sets.
  Note, that in January 2000 a major daq98-upgrade took place, integrating the DMADD-2000 trigger system. Correspondingly, the detector.cnf files structure was severely upgraded, and changes in the trigger-readout being done.

  Therefore, when running the reader with f2k-output option -f, the user must specify the flag --detector-setup. It can be set to 1998,1999,2000,2001, e.g. --detector-setup=1998,
  where 1998 counts from station close 98 until jan/feb 99, when the new daq-setup was comissioned, etc.

• Trigger information reported to F2K are the seven trigger bits as used by February 2001, and the Main trigger TDC time.
  See here for a description of the bits for the different years.

• Some OMs are double-readout, these readout-channels are mapped to a different OM-number. See here for details for the different years.

• Tests on backward compatibility for 1998 and 1999 have been done on a few thousand events for the following files
  • 1998: ab_196_109_31.data.gz.put_to_tape_2.analed.gz
  • 1999: ab_1999_102_0213_018.data.min_bias.gz

  No unexpected effects were found:
  1. The VME-TDC Overflow bit is now correctly reported to F2K, i.e. as the ">" preceeding the number_of_edges item in the HT-line.
  2. GPS times deviated by some fractions of usecs for a few events.
  3. TDC hits with "disordered primary buffer" are now reported with negative TOTs (had been mapped to very large TOT before; e.g. >32usec, see also discussion at Bartol-Meeting).
  This needs tests from larger statistics analysis to decide on whether to reject the hits or the re-order them.

• One backward incompatibility for 1999 data was found (however irrelevant for F2k-Muon analysis):
  The 0.5sec/1sec time stamp information for the Counter-events in 1999 (SIS3800 Counter: Trigger rates and SN for str11-13) are currently not available. Should be corrected soon.
  

• Remapping of the OM-allocation to ADC/TDC-Readout channels with the reader can be done using the flag --use-config=detector.cnf , where detector.cnf is an detector configuration file that
  • reflects precisely all true hardware settings for that particular run/file,
  • and additionally does the ADC and/or TDC mapping changes included.
  The reader writes an F2K HI-line with the filename of the detector.cnf file being used.
  Remapping with the reader is routinely used for 2000 data analysis already.
  A word of caution: using the --use-config option with a wrong detector.cnf file can cause, beyond the obvious effect of "wrong re-mapping" more serious effects.

  To say it differently:
  Using this option one has to be sure for every (!) datafile it is applied on, that the true hardware configuration and the corresponding labels are correctly reflected in the supplied detector.cnf file (for all boards: ADC/TDC/GPS/Register/Counter/...).

  For 1998 and 1999, the the detector.cnf files have a quite different syntax, and the ad-hoc request (as of March 2001) of the --use-config option being used with old-style raw-data files means additional complications, and in any case you have to test this option carefully. In particular, using the --use-config - option for 1998/1999 needs a detector.cnf file that is not only simply derived from the raw data as for 2000/20001 (by --extract-config), but also to be edited (this restriction holds with the current release). A perl script can be obtained on request.
  Here, we give 2 example detector.cnf files for
  1998 (file ab_196_109_31)
  and for
  1999 (file ab_1999_102_0213_018) .

  The main drawback of this way to re-map raw data is the need to have a detailed control about which runs/files were running with which original detector.cnf file (e.g. a change in a GPS module or a SN-Counter will force you to use a different detector.cnf file to correct.)
  

  The bottom line:

  • The reader allows to use the --use-config - option for 1998-2001.
    
  • The correction detector.cnf file has to exactly reflect all details of the orignal detector.cnf file used at data taken. I.e. the user has to generate an appropriate specific detector.cnf file at any moment when the orignal configuration has even slightly changed.

Distribution of the code for non-developers is done via tarball and via executables.

• Distribution tarball .
  To be added.

• Executables:
  
  • Hbook-Version, statically linked
    (histogram option available, but filling done only on request).

  • NoHbook-Version, statically linked (no histogramming possible).

Return to previous page .

Author of this page:
Ralf Wischnewski
15-September-1999