Pair Spectrometer Low Granularity Calibration

From GlueXWiki
Revision as of 16:52, 29 July 2015 by Dolan (Talk | contribs) (PSC Counter Occupancy)

Jump to: navigation, search

PSC Calibration

Figure 1.Top Down View of the Hall D Pair Spec. from original schematics

For the better part of the last five years (since appx. 2010) Hall D in conjunction with UNCW has been constructing a Pair Spectrometer (pair-spec). The purpose of which is to convert coherent Bremsstrahlung photons into electron/positron pairs. These charged pairs can be separated and extracted for data about the beam. This is particularly useful data because it is being collected after it has traveled from the tagger magnet, and through the collimator.

In 2013 the counters for the low-granularity section of the pair spectrometer were constructed. In 2014 they were tested, and then installed in Hall D. In fall of 2014, and spring of 2015 test beam was run through the hall as part of a 12 GeV commissioning run. We can take data from the commissioning runs to calibrate the pair-spectrometer to give us data within acceptable accuracy. That is the purpose of this page, to explain, and walk through the calibration process.

PSC Data Recognition and Analysis

Fully calibrating the Pair Spectrometer involves a multitude of data from different sources. The first step is simply accessing the data. Jefferson lab uses a data storage and scientific computing array in CEBAF. It is collectively known as ifarm. Each hall and respective departments store the data from each run on the ifarm server back. When accessing ifarm via secure shell (ssh) we gain access to the repository of data collected.

PSC Raw Data vs Skims

When calibrating the PSCs it is important to distinguish between raw data and data skims. Each have their uses, and are essentially the same thing. However, skims are a selection of raw data that we know have hits events.

PSC Raw Data

PSC raw data is data straight from the flash ADC and TDC. There is an associated time signature in the case of TDC data, and 100 small samples of signal characteristics in the case of ADC data that is broken into 4ns segments. You can think of ADC data as the raw digital version of the peak on an oscilloscope (an analog instrument). This is the case for both raw and skimmed data. What distinguishes them is raw data is events that have not been selected for. It is a measurement of every event through the detector, regardless of its usefulness.

PSC Skims

Figure 2.A page of runs with events selected for (PSC Skim)

PSC skims simply mean that events have been selected for. The skims we used are simply the work of someone who searched in the database for runs with known PSC hit event data. The convenience of a data set that can be easily accessed, and has the data we require is enormously helpful.

Plugins

In order to do data analysis we first need to be able to build our data into histograms within the ROOT analysis package. ROOT is a powerful object-oriented analysis package developed by Rene Brun et al. at CERN, and incorporated into the ifarm server system. In order to obtain data from runs and turn that data into graphs we need to run the ROOT package and source the environmental variables.

Source

In ifarm65 located at /w/halld-scifs1a/home/user/work/HDBuildManager/builds/dev/scripts/env/env_halld_Sp15.csh

  • Has been moved to '/w/halld-scifs1a/home/user/work/hdpm/pkgs/dev/scripts/env/env_halld_Sp15.csh'

First we have to source the environment so that we can run ROOT and our plugins with the data we desire. Sourcing provides all the environmental variables needed to fulfill the actions of the analysis package.

PSC Multiplicity Count

Figure3. Example of Flash ADC multiplicity for PSCs

In this case multiplicity count refers to how many hits were recorded at the same time. We except that when we pair produce we should get one electron and one positron. However, multiplicity counts tell us if the those particles hit one counter, or two. Simple conclusions we can draw from the multiplicity numbers.

  • Two Hits
    • Electron and Positron hit the center of the PSC scintillator
  • Three Hits
    • One electron or positron hit the center of a scintillator and the other clipped the edge of two overlapping scintillators.
  • Four Hits
    • Both electron and position clipped the edge of overlapping scintillators
Due to the models not being complete, some events were listed as 'zero' hits. However, the general trend should be correct.

PSC Counter Occupancy

ADCoccupancy000.png

During the analysis of the PSC events and data from each run it is beneficial to known the number of events that occurred to each detector. Not only does this information allow us to check to see if every detector is functioning properly, but it also gives us an idea of the approximate hit distribution.

Given the energy of the particles being

PSC ADC Average Amplitude

PSC TDC Average Time