- wleight's home page
- Posts
- 2012
- 2011
- 2010
- December (1)
- November (3)
- October (4)
- September (5)
- August (4)
- June (1)
- May (1)
- April (2)
- March (2)
- February (2)
- January (3)
- 2009
- December (1)
- November (1)
- September (2)
- August (2)
- July (8)
- June (4)
- May (3)
- March (1)
- February (2)
- January (1)
- 2008
- 2007
- 2006
- My blog
- Post new blog entry
- All blogs
BPRS/BSMD Online QA
I have expanded my BPRS monitoring code to include the BSMD. For each of the BPRS, BSMDE, and BSMDP, the code takes as inputs 2d histograms plotting pedestal-subtracted ADC vs. softId and pedestal-subtracted ADC vs. capdId and a 1d histogram plotting status vs. softId. The production of these histograms is entirely decoupled from the analysis portion of the code, so that I can generate these histograms from muDst's in order to test the analysis portion of the code now and then switch to receiving files from Tonko later.
Plot information:
- Run number: 9052001_1010001
- Number of events: 5000
- Pedestals: taken from the database for all detectors (note that the previous BPRS plots were produced with Jan's pedestals), so no capacitor dependence
Figure 1: For each BPRS crate, the code first produces 3 sets of plots that cover 400 BPRS channels each:
On the right is a plot of pedestal-subtracted ADC vs. channel id. The red lines are channels with bad status. On the left is the projection of the right-hand plot onto the y-axis. Note that the projection includes the channels with bad status. The blue line marks 0, where we expect to see the peak.
Figure 2: Then each of these 2d plots of 400 channels is turned into a plot of most probable ADC value vs. channel:
The green lines show the (largely arbitrary, at this point) cutoff value of abs(MPV)=5. Any channel with abs(MPV)>5 is regarded as being a bad channel.
Figure 3: The number of channels with good status but bad MPV is counted for each crate and plotted:
Figures 4-6: The same thing is then done for each BSMD Phi and Eta crate, except that each crate is split up into 5 blocks of 450 channels (=3 modules):
Figure 7: Finally, the plots which are considered important are aggregated into a summary page which the shift crew will be able to look at and hopefully to use to identify problems without digging into the details. Currently the summary page has one plot, which combines the plots counting the number of bad channels (the BSMDE and BSMDP plots are summed):
Finally, the whole thing is combined into a large pdf, file attached. The file is formatted as follows:
Page 1 | Summary Page | See Figure 7 |
Page 2 | Pedestal-subtracted ADC vs. Cap ID by BPRS Crate | See Figure 8 |
Page 3 | BPRS MPV Summary Page (includes plot of # of channels with good status vs. crate number) | See Figure 3 |
Page 4-6 | BPRS PSD19W, MPV+RMS of Pedestal-Subtracted ADC | See Figures 1, 2, 9 |
Page 7-9 | BPRS PSD1E, MPV+RMS of Pedestal-Subtracted ADC | See Figures 1, 2, 9 |
Page 10-12 | BPRS PSD1W, MPV+RMS of Pedestal-Subtracted ADC (the green line in the top plot of p. 7 indicates a jump in softId) | See Figures 1, 2, 9 |
Page 13-15 | BPRS 20E, MPV+RMS of Pedestal-Subtracted ADC | See Figures 1, 2, 9 |
Page 16-17 | Pedestal-Subtracted ADC vs. Cap ID by BSMD Crate | See Figure 8 |
Page 18 | BSMD MPV Summary Page (includes plots of # of eta and phi channels with good status vs. crate number) | See Figure 6 |
Page 19-21 | BSMD-Eta-RDO-0, northwest top, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 22-24 | BSMD-Eta-RDO-1, southwest top, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 25-27 | BSMD-Eta-RDO-2, northwest bottom, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 28-30 | BSMD-Eta-RDO-3, southwest bottom, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 31-33 | BSMD-Eta-RDO-4, northeast top, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 34-36 | BSMD-Eta-RDO-5, northeast bottom, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 37-39 | BSMD-Eta-RDO-6, southeast bottom, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 40-42 | BSMD-Eta-RDO-7, southeast top, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 43-45 | BSMD-Phi-RDO-0, northwest top, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 46-48 | BSMD-Phi-RDO-1, southwest top, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 49-51 | BSMD-Phi-RDO-2, northwest bottom, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 52-54 | BSMD-Phi-RDO-3, southwest bottom, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 55-58 | BSMD-Phi-RDO-4, northeast top, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 58-60 | BSMD-Phi-RDO-5, northeast bottom, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 61-63 | BSMD-Phi-RDO-6, southeast bottom, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Page 64-66 | BSMD-Phi-RDO-7, southeast top, MPV+RMS of Pedestal-Subtracted ADC | See Figures 4, 5, 9 |
Update:
Figure 8: I have added plots of pedestal-subtracted ADC vs. capId for each crate. The above table of contents has been edited to reflect this.
Looking at the pdf file you see that for some crates these plots have horizontal bands: the current hypothesis is that these are due to bad channels. At the moment, this plot includes all channels with good status.
Update 2
Figure 9: For each block of channels (400 for BPRS, 450 for BSMD) the pdf now includes, in addition to the plots of ADC vs. channel and ADC MPV vs. channel, a plot of ADC RMS vs. channel. The table of contents above has been modified to reflect this change. Here is an example of a plot for one group of 400 BPRS channels (but a different run than the plots above).
- wleight's blog
- Login or register to post comments