FGT Electronic QA
Updated on Tue, 2016-05-24 09:02. Originally created by amani on 2016-05-23 08:16.
Using FGTpedestal runs to find the good and bad FGT channels, and for the good channels find the Mean ADC and the RMS vlaues, and save in the FGT pedestal status table.
Start with the pedestal runs, you can find them by using STAR RunLog Browser, and select the TRG SETUP ->fgtPedAsPhys
We have 281 runs, removed the runs that marked as bad runs, the final list have 178 runs. ( you can find a list of the runs in the attachment pedestal_run_index_run_number Sheet1.pdf).
Now for each run make an ADC distribution plot for each channels, in RUN 13 FGT instaled completely so we have 6 disks each disk have 4 quadrants (A,B,C,D), each quadrant have 10 APV chips, and each chip have 128 channels, so the total number of channels we have is (6*4*10*128= 30720 channels),
attached a copy of run #14161025 plots from June, you can find the plots in the attachment 14161025.pdf.
From those plots the information I need is the mean and the RMS vlaues for each channel for each run. The time period I cover is from 3/18/2013-6/10/2013.
Information I need from histogram I :
1.The mean pedestal RMS for each channel(Mean y from histogram II).
Start with the pedestal runs, you can find them by using STAR RunLog Browser, and select the TRG SETUP ->fgtPedAsPhys
We have 281 runs, removed the runs that marked as bad runs, the final list have 178 runs. ( you can find a list of the runs in the attachment pedestal_run_index_run_number Sheet1.pdf).
Now for each run make an ADC distribution plot for each channels, in RUN 13 FGT instaled completely so we have 6 disks each disk have 4 quadrants (A,B,C,D), each quadrant have 10 APV chips, and each chip have 128 channels, so the total number of channels we have is (6*4*10*128= 30720 channels),
attached a copy of run #14161025 plots from June, you can find the plots in the attachment 14161025.pdf.
From those plots the information I need is the mean and the RMS vlaues for each channel for each run. The time period I cover is from 3/18/2013-6/10/2013.
I made a text file for each run and inside it I save the channel ID, Mean, and RMS (see ped.14161025.txt)
Now make a two histograms for each channel and fill the first one with the Mean of the plot above for all the runs, and fill the second histogram with the RMS of the plot above for all the runs.
so for each channel we will have :
Now make a two histograms for each channel and fill the first one with the Mean of the plot above for all the runs, and fill the second histogram with the RMS of the plot above for all the runs.
so for each channel we will have :
1.Pedestal mean for each channel(Mean y from histogram I).
2.The RMS of the pedestal distribution for each channel(RMS y from histogram I).
a text file where I saved the channel Id, the mean ADC, and the RMS from the histogram (allchanneladc.txt)
Information I need from histogram II :a text file where I saved the channel Id, the mean ADC, and the RMS from the histogram (allchanneladc.txt)
1.The mean pedestal RMS for each channel(Mean y from histogram II).
2.The RMS of the pedestal RMS distribution for each channel(RMS y from histogram II)
a text file where I saved the channel Id, the mean RMS, and the RMS of thre RMS from histogram II (allchannelrms.txt)
Save these values in a text file, and apply some conditions on them to find the bad channels.
What do we expect for a good channels???????
•The mean ADC for each channel to be stable and within a certain range (300-800).(mean of histogram I) .
•The RMS of the RMS distribution of each channel to be stable.(RMS of histogram II).
•The RMS of the pedestal distribution(RMS of histogram I) to be comparable to the mean RMS of the pedestal distribution(the mean of histogram II). If the RMS of the pedestal distribution for each channel is larger than the mean pedestal RMS it indicates that the pedestal move more in average during the runs than is statistically meaningful as determined by the mean pedestal rms.
so we can consider a channel to be bad if :
•Pedestal Mean > 800 --> this cut finds 1574 bad channels you can find them in the newhighadc.txt(channel id, Mean ADC).
•RMS of Pedestal RMS distribution > 12 --> this cut finds 6521 bad channels you can find them in newhighrmsrms.txt (channel id, rmsoftherms)
•Pedestal RMS/Mean RMS > 2 -->this cut finds 2316 bad channels you can find them in newhighratio.txt
The plots below show the cuts I used :
Apply all the cuts I find :
• 23824 good channels out of 30720 channels, you can find a list of all the bad channels bad_channels_list.xlsx.
• 23824 good channels out of 30720 channels, you can find a list of all the bad channels bad_channels_list.xlsx.
•Means we have 77.55 % good channels and 22.44 % bad channels.
The plot below shows the 6 disks and show the good and bad quadrents in each disk:
»
- amani's blog
- Login or register to post comments