Home » Groups » STAR Protected

Table of BHT3 luminosities

Updated on Fri, 2010-01-22 15:07. Originally created by rcorliss on 2010-01-07 14:55.

 I've attached a table of the BHT3 luminosities in the L2W dataset.  This assumes a cross section of 350 nb for BHT3, which is probably not the final value, since it hasn't been corrected for the fraction of the detector that was live during the vernier scans.  When I have an updated table I will post it to this page with a later version number.

The equations used are:  

BHT3=(rawBHT3-SBB)*50

L=BHT3 / (Fdet*xs)

rawBHT3=the total number of L2W random accepts

SBB=120* the sum of the average in each abort gap of the number of L2W random accepts per bunch crossing.  This approximates the assumed constant single beam backgrounds for each beam.

Fdet=the number of towers with good status as extracted from the status table by Jan, meaning the following three calls returned a '1' in the 'stat' argument:

 

    mBarrelTables->getStatus(jBP, softID, statPed,"pedestal");

    mBarrelTables->getStatus(jBP, softID, statOfl);

    mBarrelTables->getStatus(jBP, softID, statGain,"calib");

 

 

xs=the BHT3 cross section as described above

 

Zeroes mean the run did not appear in my output - probably the job failed.

 

Note:  In .v1, I've updated the calculation for the Single Beam Background:  It's now 109*average of the nine-bin gap + 111*average of the eleven-bin gap.  I've also changed the cross section to 369.15nb, which is the average of the result for the two vernier scans.  If a different cross section is wanted, you can re-calculate the last column by BHT3/Fdet/xs.

 

Another Note:  in .v2, I've updated the way L2W random accepts are counted.  Previously, the bit from the L2Result was checked, but due to concerns that the BHT3 thresholds changed during physics running at some point, it now has implemented the L0 condition in software, checking the ADC of each barrel tower to see if it's >30.  This results in an integrated luminosity of 16.302pb^-1

 

Regarding .v3:  Justin pointed out that I was using only the runs with good spin sorting.  I have changed the macro to go over 'setABCD'.  All other notes still apply.

 

 

Regarding 'luminosity.v3' and those that may follow: The cross section has been updated to 501nb and the SBB calculation now takes into account the size and position of each abort gap run-by-run from a static prescription instead of trying to identify each gap dynamically.  Several new columns have been added:

 

RunNumber,      trigID, softL0, L2Wr,   Gap1,   Gap2,   Nblue,  Nyell,  Nsbb,   Nbht3,  Fdet,   L(nb^-1)

L2Wr is the number of L2W random accepts, which should agree with softL0, which is the number of those L2W random accepts that also pass the DSM>=31 condition in software.  For earlier runs, the BHT3 trigger had DSM>=25 (or something close) instead, so applying this condition uniformly is the correct way to recover those runs.
Gap1 and Gap2 are the integrated counts in each abort gap
Nblue and Nyell are the number of bunches in each beam (The yellow beam is the beam that is missing in abort gap 1)
Nsbb is the number of single beam background counts using the updated calculation (see below)
Nbht3 is the number of BHT3 events from subtracting the SBB from sotL0 and correcting for the prescale.  Fdet has not yet been taken into account.
L is the integrated luminosity for the run with Fdet properly taken into account.
 
 Rgarding luminosities.v4:  The calculation for the abort gaps has once again been altered.  it's now a fixed 8 bin window: [32,39] for gap1, [112,119] for gap2.  This avoids concerns (explained here) that the gaps may not be as large as the previous calculation assumed.  This new set also uses a cross section of 481nb, which should be the final value for APS.  The difference was in an approximation used to calculate the number of ions per bunch.  It is now done exactly, using the last cdev values before the start of the vernier scan itself.  details of this will be added to the blog and also to the paper.
There are five runs that seemed to be the furthest outliers in this table:
10076134 is missing, but will be caught on the next iteration
10090004 has a noisy first gap and may need to be removed?
10090046 has a similar feature
10093083 looks reasonable, and is likely just a statistical fluctuation in the gaps.
10093084 has a very noisy second gap and should be removed.
10101083 looks reasonable, and is likely just a statistical fluctuation.
 
The integrated luminosity is now 16.9pb^-1 for this set, which once again seems high.

 

 

In order to better organize what I hope is the final iteration of this, I've made a new blog page which contains the version of the luminosity table that will be used for APS.

Groups: