an official list of PHOS bad cells have been updated last time on September, 11th 2010 (see /alice/data/2010/OCDB/PHOS/Calib/EmcBadChannels/Run122372_126437_v5_s0.root). Is it possible that state of PHOS cells has not changed during almost half a year?

I've implemented a simple test. Starting from raw data of quite large p-p run 145355 (recorded 14/03/11, about 3 Mevents) I've created cell multiplicity map using only signals with high gain. Subsequent scan of the map selects two groups of the cells: "empty" - which do not contain any signal, and "noisy" - whose multiplicity more than 15 times exceeds mean multiplicity of that PHOS module where it is minimal. In what follows this approach is designated as "method Mult".

Comparison of the bad cells list from OCDB with Mult list shows that many entries are the same in both lists. But Mult list contains 637 cells (478 "empty" and 159 "noisy") which are absent in OCDB. In particular, OCDB list does not take into account that as far back as 26 Aug 2010 the whole branch A (448 cells) of RCU2 in module 3 (online numbering) has been switched off due to high-voltage breakdown.

Conclusion is: last version of OCDB list only partially reflects current state of PHOS cells. I don't understand how one can reconstruct events using this outdated list of bad cells.

Thanks in advance for comments and criticism.

Best regards Yuri

-- YuriKucheryaev - 03-Apr-2011

• 145355_CellMultMap_HG_m3_online.gif: