Status bit summary for BSMD status table for pp500, 2009 data (EDIT: final as of 06/08/09). 

Note that the cuts are applied sequentially (bit 1 first, then bit 2, etc.) and if a strip fails a fatal cut (i.e., it is marked as bad), none of the other cuts are tested.  Also, bits 1-3 are calculated using non-zero-suppressed data from fills 10399, 10402, 10403, 10404, 10407, 10412, 10415, while bits 5-7 are calculated using L2W data from those fills.

Bit 0: Set to 1 if the strip has failed a fatal cut or if all of bits 5-7 are 1.

Bit 1: Set to 1 if abs(MPV)>50.  This cut is fatal.

Bit 2: Set to 1 if RMS<1 or RMS>15.  RMS is calculated over the range MPV-23 to MPV+23.  This cut is fatal.

Bit 3: Set to 1 if the ratio of the integral over a window to the integral over all ADCs < .95.  The definition of the window is a little complicated due to the presence of some strips that have extremely wide pedestal peaks, in some cases running over 150 or more channels, starting at -99 (or possibly lower, but in my monitoring code I don't save anything below that limit).  In this case the fact that the RMS is calculated over a window of width 46 rather than the whole ADC distribution causes the RMS to be significantly underestimated.  Therefore, if RMS*3>23, I use the window MPV-8.5*RMS to MPV+8.5*RMS to ensure that such channels pass this cut.  Otherwise the window used is the same as that used to calculate the RMS.  This cut is fatal.

Bit 4: Set to 1 if RMS>11.5.  RMS is calculated as in bit 2.  This cut is not fatal.

Bit 5: Set to 1 if Par300<.00003 or Par300>.02.  Par300 is the ratio of the integral from 300-500 to the total integral.  This cut is not fatal.

Bit 6: Set to 1 if Par500<.00003 or Par500>.02.  Par 500 is the ratio of the integral from 500-800 to the total integral.  This cut is not fatal.

Bit 7: Set to 1 if Par800<.00003 or Par800>.02.  Par 800 is the ratio of the integral over all ADCs>800 to the total integral.  This cut is not fatal.

Note, however, that if a strip fails all three cuts then it is marked as bad.