For run 16, Jim prepared several bad chip files where the online threshold has been increased :

Threshold = (N*Sigma+4)  , with N= 3 , Constant = 4

OfflineAsicFile _10.txt   contains exactly the same definition of a “bad” chip as was used to collect the online data. 
OfflineAsicFile_10.txt cuts out   8.3% of all channels (not including ladders 2 and 19).
OfflineAsicFile_15.txt cuts out   9.7%
OfflineAsicFile_20.txt cuts out  10.8%
OfflineAsicFile_25.txt cuts out  12.0%
OfflineAsicFile_30.txt cuts out  14.9%
OfflineAsicFile_35.txt cuts out  24.2%
OfflineAsicFile_40.txt cuts out  40.0%

1) From the txt (csv) files



2) tests
below are the number of strips, clusters, hit per side for a) no table used, b) offset=10 c) offset=40
no bad chip table


bad chip table 10


bad chip table 40

 
comments (for myself) :

• there is already a method to mask chips, but it was used in run 14 and run 15 as a second pass of the chipCorrect table
• this table (sstChipCorrect) is no more used in run 16.
• possibilities :
  • use maskChip table as bad chip and do not remask chip
  • make another table (sstBadChip) 

 }
      //chipMask table      
      if( mRunNum >= 14 && mRunNum <=15) {
	if (gStSstDbMaker->maskChip(id_side, ladder, wafer, chip)) continue;
      }

5) pass 2
- Remove the offline threshold (in StSstUtil/StSstWafer.cc) that cut strips having signal>10

no bad chip table


bad chip table 10


bad chip table 15

bla bla