Having created BSMD status tables based on fills 10399, 10402, 10403, 10404, 10407, 10412, 10415 (see this blog page for a description of how the status was calculated), the next step is to look at BSMD strips that passed the status cuts vs. time.

In order to have enough statistics, this is done by fill, and only for the pedestal, not the tail.

First, a pedestal correction is obtained by calculating the MPV and RMS: if MPV>RMS, then the MPV is saved as a correction factor.  For this reason no cut is imposed on the MPV.

Next the usual cuts for bits 2 and 3 are imposed (see the blog page).  Then for each module the results for the pedestal correction, RMS, and status are plotted vs. softId and fill.  The list of fills used is as follows, with notes that explain some features on these plots:

10383
10398
10399
10402
10403
10404
10407
10412
10415
10426
10434: For this fill the monitoring was broken such that no pedestals were subtracted.
10439: Same as fill 10434
10448: No monitoring data exists for this fill (probably due to my error)
10449: Same as fill 10448
10450: Crate 2 (first 15 modules) was off for this fill.
10454: Ditto
10455: Ditto
10463: Ditto
10464: Ditto
10465: Ditto
10471
10476
10478
10482
10486
10490
10494
10505
10507
10508
10517: This fill had no non-zero-suppressed BSMD events, hence no monitoring data.
10525
10526
10527
10528
10531
10532
10535
10536

Plots for all modules are available here.  Attached are two examples, modules 3 (for which a lot of correction will be required) and 6 (which is fairly clean).  Below is a plot of the number of strips that need to be corrected per module and fill.