MainPage:Nuclear:NPS:PWO:LightYield

From cua_phy
Jump to navigation Jump to search
⇐ Back to Neutral Particle Spectrometer
⇐ Back to the Main_Page

Instructions to scan LY

(THIS IS A PRELIMINARY VERSION ONLY)

  1. Set the freezer temperature (controller fixed on the wall, on the side of the freezer):
    1. Hold the button SET for 3 seconds
    2. Change the temperature by pressing up/down buttons
    3. Hold the button SET for 3 seconds to save the new temperature
    4. Wait more than 5 hours (?) to stabilize temperature (keep HV on at all time to stabilize PMT also)
    5. Keep one terminal always logging the temperature in the Control computer:
      cd thermometer
      ./MlogT
    6. Check temperature at Control computer:
      cd thermometer
      ./MplotT
      (Ctrl+c to stop the script)
  2. Prepare electronics:
    1. On the NIM module N1470, turn on the trigger (ch0) and the signal (ch2) HV. It is set to keep -760V (trigger PMT) and -2300V (signal PMT), draining approximately 670 and 1035 uA respectively.
    2. If gate width change is required, look in the scope the width of the gate. You change it by turning the hidden screw "width" in the NIM module 794, on the 2nd channel.
    3. Make sure the trigger and the signal cables are connected to the ADC.
  3. Prepare CODA (DAQ computer):
    1. Options -> Limits -> Set Time Limit
    2. Set the time of each run, i.e., the time each step along the crystal where LY will be measured.
  4. Run motor to scan (CONTROL computer) and start CODA when prompted (DAQ):
    1. In a terminal, go to directory motor. Execute the command:
      ./scan_12cm_step1cm_90min (or equivalent for the test you want - be consistent with the time set in CODA!)
    2. If requested, type the password.
    3. Make sure you hear the motor moving
    4. When the terminal prompts that the scan will start in 10 seconds, switch to the DAQ computer and start coda USING:
      Option -> Enable Auto Mode
  5. Converting data
    1. In a terminal at DAQ computer, convert the files created:
      ./convertRange.sh 2000 2010
      (Enter password when prompted to transfer data to Control computer)
    2. In a terminal at Control computer, unzip the transferred file:
      cd data
      unzip 2000.zip
    3. Convert the data to ROOT:
      ./convertRangeHex.sh 2000 2010
  6. Analyze data (important to keep consistency, so we can inter-compare crystals):
    1. Fit the peaks with root (supposing you used the ADC channel 1, otherwise adjust):

root -l 2000.root T->Draw("ch1>>(1201,0,1200)"); [...] (I am going to write a script to do that automatically) - by Marco C.

Light Yield Analysis Instructions (by Salina)

  • Preparing to take light yield data
  1. Wait for freezer to cool at least 4 hours after taking a crystal out/before putting another one in, and BEFORE OPENING THE FREEZER, TURN OFF THE HIGH VOLTAGE OR PMT WILL BE DAMAGED
  2. To turn off the high voltage: turn channel 0 to the middle and wait for it to get to 0, then turn channel 2 to the middle at the same time. When it gets down to 0, you can turn both channel 2 and 0 another notch down.


  • Putting the crystal in the freezer
  1. The crystal will be in a labelled styrofoam holder wrapped with some protection...make sure you keep track of which crystal you are putting in/taking out. Be very careful when handling these crystals, they are very fragile and expensive!
  2. Wrap crystal (white 3 layers teflon and then 2 layers black tape on top) with one end covered with white tape and black tape on top.
  3. Grease the open end of crystal and put crystal inside the holder in the freezer, making sure it is aligned with the source and the PMT. Close the freezer. Generally, for the crystal itself to reach 18°C it should take 4-5 hours. If the freezer was originally at room temperature, then you should wait at least a day for it to cool down.
  4. On the NIM module N1470, turn on the trigger (ch0) and the signal (ch2) HV. It is set to keep -760V (trigger PMT) and -2300V (signal PMT), draining approximately 670 and 1035 uA.
  5. Check the oscilloscope to see if the signal is inside the gate: Plug the gate into channel 1 on oscilloscope, and plug channel 1 into channel 2 or 3 on the oscilloscope.
  6. CHECK THE GATE: The "gate" should look like a square well, and the amplified signal should be inside the gate to accumulate an appropriate amount of incidents read by the PMT. You may have to play with the settings and scaling on the scope to get the desired signal and gate to make sure you have the desired gate width.. (as of 11/29/2016 we are using 100 ns as the gate width).
    • If gate width change is required, look in the scope the width of the gate. Each "square" corresponds to 20 ns (you can see this on the bottom of the screen "...M 20 ns") so for example, if you want a 100 ns gate width, the total gate width should be comprised of 5 of these "squares" .
    • You change the gate width by turning the hidden screw "width" with a mini screwdriver in the NIM module 794, on the 2nd channel (where the gate input is plugged in). To change the position of the signal, you can turn the hidden screw "width" on the 4th channel/where the signal input is plugged in.
  7. Make sure you plug back the signal and gate back into the ADC before you start taking data...once that is done it is time to start data collection.
  8. If measuring as a function of position: the script on computer screen which doesn't have coda running (i.e. step in 1 mm in 20 hours) and switch to the screen with coda and you can click start AS SOON AS THE SCRIPT TELLS YOU YOU HAVE 10 SECONDS TO START COLLECTING.
    • You should make sure if you are measuring as a function of position (with the motor involved), then you should set the time limit on coda
    • For measurement of just light yield (as a function of the peak position) you should set a time or event limit through coda for however many events or time appropriate for the measurement.. for example, to get about 1 million events, you should set a time limit to 440 minutes.
  9. Voila! The data should start collecting, and you can come back to check on it the next day.
  10. ALSO MAKE SURE WHEN YOU ARE TAKING THE CRYSTAL/SWITCHING IT OUT OF THE FREEZER, TO PUT THE WRAPPED CRYSTAL (THAT YOU ARE DONE TAKING DATA WITH) IN THE SAME STYROFOAM HOLDER YOU TOOK IT OUT OF.. you should obviously clean the grease off with ethanol/isopropanol before you put it in there.


  • How to obtain the Light Yield data from our DAQ (Data Acquisition) setup
  1. Save the data by converting the file in the terminal of the DAQ computer: ./convert <run number> (this is saved in a <runnumber>.txt file in the home directory. And then use: emacs <blah> to see the data. Do this to check if there is anything in the file, because if it is empty, something went wrong.
    • This emacs file will open the data in hex, which is why we have to covert it to root to interpret the data.
    • ALSO don't put the .dat file extension for the run number!
  2. Send the .txt file to your computer and make sure you have the HEX2root.C script (ask Marco or Me) - instructions for how to do this (sending files from and to JLAB computer) are below.
  3. On your own computer when you have the data as the .txt file AND you also have the Hex2root.C file you have to convert it to a root file so you can analyze it.
    • < root -l -q "HEX2root.C(3166)" > where the command line is inside the <> 's and the specific run number as an example is 3166
    • Note: the files HEX2root.C and 3166.txt should be in the same directory, and the command should be from that directory too.


  • How to analyze the Light Yield data from our DAQ setup
  1. Open the file in root and find the trees and branches available in it
    • To open a root folder in root, type root infront of file (i.e. if the file name is 3098.root the command is: <root 3098.root>
  2. Print the branches of the tree, so you can have an idea of what you have in it
  3. To find the tree and see branches and leaves: 
    1. < T->Print() >
    • The branches are the channels of the ADC. 
  4. To view and examine a tree
    • < T->StartViewer() >
  5. To see how many entries:
    • < T->GetEntries() >
  6. Draw a histogram of the branch correspondent to channel 1
    • < T->Draw("ch1") >
    • Don't use TTree::Draw() because using root.
  7. Understanding the data analysis and light yield
    • We are only using ch1 for our ADC signal input. There are a total 12 channels.
    • The signal from the PMT is proportional to the amount of light (light intensity), which is proportional to the # of photons. The # of photons is what we are very interested in (refer to 2/18/2016 notes in notebook).
    • Sometimes the PMT signals from the crystal do not always get detected. Or the signal from the source hits the trigger but misses the crystal since there are multiple gamma rays. This, in turn, helps us choose a "zero" for the number of photons or a pedestal. For example, if the signals are 30, 31, 32 and then there's a signal of 254, the 254 is the most important since it indicates that there is communication with the trigger and crystal. You can read more about this here: http://www.vsl.cua.edu/cua_phy/index.php/MainPage:Nuclear:Summer2016:PWO#Light_Yield
The histogram after you "draw" it, of J41 showing the pedestal, SEP, and the signal from the crystal.
  1. Reading the Histogram: SEE FIGURE ABOVE
    • The graph is read as the counts vs. ADC reading or # of signals/bins
    • Next to the pedestal is the single electron peak (SEP) and then to the right of is a hill-type curve, which is the signal from the crystal. If we zoom into these peaks, we can't really analyze the detail since the bin-width is so large (i.e. 10), which is the width of a peak.
    • To calculate the Light Yield = (signal from crystal - pedestal)/(SEP - pedestal), in which the SEP is the single electron peak position
  2. Bin the data with 1200 bins, ranging from 0 to 1199 (this is the resolution of our ADC)
    • <T->Draw("ch1>>(1200,0,1199)") >
      • 1200 - is the number of signals/bins of ADC
      • 0 to 1199 sets the bin width to 1 since 1200-1199 is 1.
      • This improves the resolution significantly.
  3. Put the y-axis in log scale, so you will see the actual signal
    • Right click on 10^3: set logY
  4. Find the position of the peaks you see by fitting each of them with some peak-like curve
    • Do you know what each peak is? Can you extract the light yield from that?
      • Use gaus- A gaussian with 3 parameters: area, width and center of peak
      • Center of peak is given by right clicking on histogram and selecting "FitPanel"
    • Use gaus function implemented within root with all three peaks of the graph.
    • It should be pre-set to gaus. Use the selecting panel located at the bottom of the panel to select the curve you want to fit. Do this for all three peaks: the signal from crystal (the hill-like one), SEP, and the pedestal.
    • The output on Terminal gives you the "Mean" which is the center of each peak:
Screenshot of output on terminal.
  1. Finding the light yield quantitatively
    • We can find the Light Yield by the equation Light Yield = (signal from crystal - pedestal)/(SEP - pedestal)
    • Plugging in values: LY = (4.45690e+02 - 3.33976e+01)/ (8.80272e+01 - 3.33976e+01) = 7.54705141534992 photoelectrons
    • Then to get this result in pe/MeV, just divide by 0.511 MeV which is the energy quanta of the radioactive source in the setup, Na22
    • so --> 7.54705 pe/0.511 MeV = 14.7692 pe/MeV which agrees with the results.
  2. Now we need to some how fit this data as a function of position to see how the light yield is represented along the crystal
    • get graph LY vs. Position (cm)
    • To get error bars:
      • Obtain mean value error values and plug into Light Yield eqn.
      • Then you can plug into the code.
      • Note: The script that you have may already automatically do this for you ... thank you Marco
  3. TO ACTUALLY GET THE LIGHT YIELD GRAPHS AND AS A FUNCTION OF PEAK POSITION (assuming you have the script to work with, which does all the analysis for you)
    • Open and edit the script for the run #. The script is: LYscan_simpleSICCAnofit.C and for a fit- run LYscan_simpleSICCA.C ; you have to edit the code to input the run number
    • < root -l LYscan_simpleSICCAnofit.C > or whatever the name of the script you have with ".C" as the file extension.
  • Copying files to and from JLAB computer
  1. To copy a file from DAQ computer to JLab (for instance, file 3166.txt) - do this in DAQ computer:
    • < scp 3166.txt yourusername@login.jlab.org:/scratch/yourusername/ >
  2. To copy a file from JLab to your computer (for instance, in your Documents/ResearchNP directory) - do this in your laptop:
    • < scp yourusername@login.jlab.org:/scratch/yourusername/3166.txt Documents/ResearchNP >

How to get the Light Yield from the Control Computer Directly

  1. Switch to Control computer
  2. Open a command line in windows:
    • Windows Key + R
  3. Type
    • cmd
    • cd analysis
  4. Copy the data to the Control Computer
    • scp DAQ@136.242.188.135:3166.txt .
  5. To convert the txt file to a root file:
    • root -l -q "HEX2root.C(3166)"
  6. Get the Light Yield:
    • root -l "LY.C(3166)"
Retrieved from "https://www.vsl.cua.edu/cua_phy/index.php?title=MainPage:Nuclear:NPS:PWO:LightYield&oldid=6666"