Pedestal

• pedestal = 0 adc
• pedestal = 255 adc
• rms = 0 adc
• rms = 255 adc

Comments

• 255 adc represents the maximum of our adc range
• most of the entries are multiple of 128 : it is because for each wafer the FEE is constituted by 6 ALICE128 chips, so that each reads the 128 channels/strips
• the maximum of entries for each (wafer,ladder) is the 768.
• the rms is also coded to 255 but further in the soft we divided it by 16 to get a real rms.

Pedestal

Zero value

Figure 1 : wafers on p-side	Figure 2 : wafers on n-side

Max value

Figure 3 : wafers on p-side	Figure 4 : wafers on n-side

rms

Zero value

Figure 5 : wafers on p-side	Figure 6 : wafers on n-side

Max value

Figure 7 : wafers on p-side	Figure 8 : wafers on n-side

How to retrieve these histograms

• open root4star -b
• .x bfcread_hist_list_all.C("my_file.hist.root") : you see that the branch for the ssd pedestal histograms is labelled "SpaStrip"
• close root4star
• re-open root4star and .x bfcread_hist_extract.C("my_file.hist.root","SpaStrip")
• it has created "my_file_SpaStrip.hist.root"

Basic bfc chain to produce ssdStripCalib.root file

Option : fill the Pedestal tuple

• cvs co StRoot/StSsdDaqMaker
• (this way will give you the recent version of StSsdPointMaker, to get a previous, according to librairies, type star***old,pro,new***)
• put mPedOut=1 in the Init()
• the ouptut is : PedestalNTuple.root (size ~2.5MB)
• cons +StSsdDaqMaker

• side : 0 for p-side, 1 for n-side
• ladder : from 1 to 20
• wafer : from 1 to 16
• strip : from 1 to 768 : all strips in a given wafer
• pedestal, noise : pedestal and noise for a given strip
• chip : from 1 to 6 : each wafer has 6 FEE chips
• id_wafer : from 7100 to 8620
• tot_strip : from 1 to 12288 : all strips in a given ladder

• root.exe [5] PedestalNTuple.Draw("pedestal:tot_strip","side==0 && ladder==1 ") : will plot a 2D histogram that represent the pedestal for all the p-side strips of ladder 1

  Option : fill the tuple in from PointMaker

  Actually there is a flag in StSsdPointMaker to store tuple (because it takes space) and cannot be done in cvs version.
  So there are many steps to get these tuples :
  • get the code from cvs : create a 'StRoot' folder in your work dir and type cvs co StRoot/StSsdPointMaker.
  • edit the PointMaker and remove in the init the line related to Debug for DeclareNTuple :
    instead of : 'if (Debug() > 1) DeclareNtuple(); ' just do 'DeclareNTuple()'.
  • Remove all the lines related to StBFChain in DeclareNTuple() and put an explicit output : myfile = newTFile("myresults.root","RECREATE")
    the tuples below will then be stored in this file.
  • Don't forget to declare TFile *myfile in the header.
  • in the Make() of the PointMaker, comment the Debug for filling the tuple (~ lign 507).
  • in the finish, put myfile->Write(); and myfile->Close();

  How to run with your own ssdStripCalib

  first step : produce the ssdStripCalib

  • run the bfc chain with a pedestal file , it wil produce in your work dir a file name : ssdStripCalib.TimeStamp.root
  • the timeStamp is : YYYYMMDD.HHMMSS

  second step : set your environment

  • create a StarDb-->svt-->ssd arborescence in your directory.
  • put the ssdStripCalib in there.
  • in the PointMaker, replace (in the init) :
    m_noise2 = (St_ssdStripCalib*) GetDataBase("Calibrations/ssd/ssdStripCalib"); with
    
    TDataSet *svtparams = GetInputDB("svt");
    TDataSetIter local(svtparams);
    m_noise2 = (St_ssdStripCalib*)local("ssd/ssdStripCalib");
    
    
    You see that instead of looking in the directory StarDb of the database, it will look before in your own StarDb.

Strips

Clusters

• side : 0==p-side ladders ; 1==n-side ladders
• ladder : from 1 to 20
• wafer : from 1 to 16
• nstrip : size of the clusters(number of strips)
• snratio : signal to noise ratio of this cluster : sum(signal_strip)*cluster_size/sum(noise_strip)
  The sums are over the strips in this cluster.
• noise : sum(noise_strip)/nstrip
• first_strip : id of the first strip of this cluster
• TotAdc : sum(signal_strip)
• FirstAdc : signal of the first strip of this cluster
• LastAdc : signal of the last strip of this cluster
• TotNoise : sum(noise_strip)

• one for all the clusters reconstructed (result of the cluster finder)
• a second which stored only the clusters that are use din hits, eg those which are matched (result of the cluster matcher)
• note for me : with these 2 tuples, one can define the cluster efficiency.

  map

  These 2 plots show the distribution of cluster per wafer, ladder for the p and n side.
  It is independantly of the matching between the 2 sides and it represent what we get at the end of the cluster finder.
  We saw for these data (CuCu62) the 2 ladders that were not working : in reality, there were taking data but you can see from the previous plot, that the noise of their strips are zero or at the maximum of the adc range.
  

  Figure 9 : wafers on p-side	Figure 10 : wafers on n-side

  snratio

  Space points

  Jonathan Bouchet
  Last modified: Mon Mar 26 22:11:32 EDT 2007