HV change vi_A --> vi_B towers only
HV used : set vi_A
250K minB events, details of slopes extraction.
RELATIVE GAIN CHANGE
Procedure of changing HV is caried separately for every eta bin:
* rejet 3 largest and 3 smalles slopes
* project slopes in to histo using weights=1/sigSlope^2
* fit gaussian to find mean of the histo
* correct gains/HV for all slopes exceededing +/- 5%
* NOT correct gain for other towers
* Formula for HV:
float gainCorr= tw->sl/avrSl;
float hvCorr=pow(gainCorr,1./8.3);
Example of analysis for eta bin 6 is below.
PDF for all 12 eta bins.
Left: slopes vs. phi. Black towers will be corrected in this iteration.
Right: weighted, truncated distribution of slops & fit.
Table with gain & HV correction.
Intended usage newHV = oldHV *hvCorr
Distribution of gain corrections vs. phi-index ~ TPC sectors, all towers.
Distribution of HV corrections vs. phi-index ~ TPC sectors, all towers.
ABSOLUTE GAIN CHANGE
was applied on the top of the relative gain correction described above.
Based on MIP response from pp data compared to the expected MIP d\response from M-C (the blue to green ratio on the plot below), DETAILS
The HV were multiplied by the eta dependent factor as follows:
set absHV(01) 1.00
set absHV(02) 1.00
set absHV(03) 1.00
set absHV(04) 1.00
set absHV(05) 1.009
set absHV(06) 1.011
set absHV(07) 1.012
set absHV(08) 1.006
set absHV(09) 1.007
set absHV(10) 1.009
set absHV(11) 1.005
set absHV(12) 1.015
RESULTS
Input HV HVset_vi_A/
Output HV HVset_vi_B/
list of bases with changed HV A2Bdiff/
Used script slope2gain.C ,
mergeHVcorr.tcl