Physics 116C Lab on Radioactive Decay, Geiger Counter, Counting Statistics

V. 1.0, 4/20/2010

Introduction
The goals of this experiment are to investigate: 1. radioactive decay measurements with a Geiger counter; 2. basic data acquisition techniques with LabVIEW; 3. counting statistics (Poisson and Gaussian distributions as limits of the binomial distribution). The experiment exemplies a nuclear or particle physics experiment at the simplest level. A subatomic particle enters a detector and, through some interaction with the material of the detector, produces an electrical pulse. The pulse is amplied and recorded as an event in a counter read out by a computer. The events are entered into histograms for immediate monitoring and stored for further analysis. The eventual result may be a publication in a scientic journal or, as in this case, a lab report. In the experiment, you will use a LabVIEW program to record histograms of the number of detected radioactive decay events from a 137 Cs source in a specied time interval when the measurement is repeated many times. When the source is far from the detector and the count interval is short, the average number of decays per interval is small. This should yield a Poisson distribution. Moving the source closer and using a longer interval results in a larger mean number of counts per interval and a distribution which should approach a Gaussian. Statistical tests are performed on the resulting distributions to test these hypotheses. Material necessary for understanding the experiment (particles, interactions, detectors, radioactive decay) is summarized in Chapter 8 of the text by Melissinos and Napolitano.[1] Statistical analysis is covered in Bevington [2] plus Chapter 10 of Melissinos and Napolitano. Also refer to the specications of the Geiger counter tube.[3] Some possible questions to be addressed are as follows. What are the energies of the particles resulting from 137 Cs decay? What physical effects allow these particles to be detected? How is the Geiger counter constructed and how is the electrical pulse produced? What are the advantages and disadvantages of a Geiger counter as a radiation detector? Here, X-rays and gamma rays may considered to be particles (photons) as they are emitted, scattered and absorbed in discrete events. 1

Experiment
The block diagram of the circuit is shown in Fig. 1. The idea is to use a continuously running counter which can be read at regular intervals by the computer to take a set of data samples {ni } (number of counts in ith interval). The computer program (LabVIEW VI) then makes a histogram of {ni } which can be read out manually or saved to a le for comparison with statistical theory. The timing accuracy is at least of the order of a few ms, which is sufcient for this application. (Greater timing accuracy could be achieved with external gating pulses.) The VI control panel is shown in Fig. 2 and the VI block diagram in Fig. 3. You should be familiar with most of the elements in the block diagram at this point. The Express DAQ Assistant VI framework is used for input using counter input 0. Controls are provided for the length of the count interval, the number of samples, the histogram binning and whether or not to write the histogram contents to a le (you must specify this before the run). Indicators are provided for the total number of counts, the sample mean and the sample standard deviation.

Figure 1: Circuit block diagram. The salient features of the Geiger counter controller are shown in Fig. 4. It has a pair of terminals on the back marked Oscilloscope (bottom one is ground). The output is a fast-rising pulse with height of approx. 2.5 V and a decay time constant 200 s superimposed on a DC pedestal 5 V. To get the required TTL logic pulse for the computer interface, we need to trigger on Geiger counter pulses exceeding a certain amplitude above the pedestal ( 0.6 V). That is accomplished with the AC-coupled Schmitt circuit in Fig. 5. The 0.01 F capacitor at the input blocks DC while passing the rapidly rising leading edge of the Geiger counter pulse to the comparator input. The open collector comparator output is connected to a 1 k pull-up resistor, the Schmitt trigger feedback resistor and a TTL inverter. The inverter drives the base of a 2N3904 BJT capable of providing a TTL-compatible signal in the properly terminated 50 coaxial cable. The termination of the cable in its characteristic impedance prevents reected pulses which might otherwise cause spurious counts. The topics of transmission lines and terminations will be covered later in the quarter.

Figure 2: Geiger counter statistics front panel (during timing accuracy test).

Procedure
Precautions with Geiger tube Leave the cable from the Geiger counter controller to the Geiger tube in place at all times. The connector on the controller can carry a voltage of order 1000 V. If you leave the cable from the tube connected, nothing else can get connected to it by mistake (including ngers!). Leave the Geiger tube in its holder. It has a thin front window which could be broken. Precautions with radioactive source Dont touch the source. Leave the source in its tray at all times. The TA or lab assistant will provide the sources and deal with moving them from place to place. Review the material in Appendix D of Melissinos and Napolitano [1] on radiation safety. Remember that distance between you and the source is your friend. 1. Record your observations and make preliminary curves in a loose leaf notebook (preferably quadrille-ruled engineering paper). 2. Check and turn on Geiger counter Remember the warning above not to disturb the connection to the Geiger counter itself, but verify that it is connected to the control box HV. Turn on the Geiger counter control box, set the HV knob fully CCW (low voltage) and turn on the Geiger counter HV. We turn on this equipment at this point to let its temperature stabilize. 3

Figure 3: Geiger counter statistics VI block diagram.

Geiger counter tube 5 nF 2 kV
Amplifier

Geiger counter control box

To internal counter circuits Oscilloscope output

thin mica window

LND, Inc Model 7232 filled with Ne + halogen gas

470 k!
High Voltage min " 400 V max = 1000 V

Figure 4: Geiger counter and control box. 3. Discriminator circuit A discriminator circuit produces an output logic pulse for a (suitable amplied and shaped) detector pulse exceeding a given threshold. Wire the discriminator circuit shown in Fig. 5. Use the lower right corner of the prototype board array. Be reasonably neat and use 5 V and ground busses appropriately. The input should be from the BNC connector on the left and the output to the BNC connector on the right. Use twisted pairs of wires for the connections from the Geiger counter output to the left BNC block. The Geiger counter wires connect to the scope output screw terminals on the back of the Geiger control box (bottom terminal is ground). Also use a twisted pair for the connection to the circuit input. The output goes to the PFI 8 input of the small connector box coupled to the DAQ card in the computer via a terminated 50 coaxial cable. 4. LabVIEW VI Load the DAQ program (called Make count samples write for.vi) and gure out what it does. Note how the loop delay timer waits until a xed multiple of a system clock (in ms) before reading the continuously running counter at the end of each iteration of the loop. 4

+5V

100 k" +15 V +

2 3 8

+5V +5V

10 k" 1.8 k" "Oscilloscope" output from Geiger counter control box Vin .01 F

1.0 k" 1/6 7404

1 2

10 " 2N3904 Vout to DAQ interconnect box PFI 8 input (must be terminated in 50 " at DAQ card input)

LM311 C
1

*7
OC

4.7 k"

1.0 k" -15 V

Figure 5: AC-coupled Schmitt trigger discriminator circuit and line driver for Geiger counter. The Geiger count sample for the current interval is found by subtracting the counter reading from the previous iteration. This is put in an array. The zeroth and rst element are deleted: the zeroth element winds up being the initial value of the counter (0) and the rst element the value from the incomplete period of the rst loop iteration (the loop is only synchronized after the rst loop ends). This procedure will include counts outside the timer interval between loop iterations so the actual interval will be slightly longer than the specied value and somewhat variable. The effect of this on the data will be evaluated, although it is expected to be negligible for this application. More elaborate procedures exist for critical work. Also determine how counts are assigned at the histogram bin boundaries. 5. Test the count accuracy and repeatability for 100 ms and 1000 ms counting intervals (these are the intervals you need for the Poisson and Gaussian distributions, respectively). Use the signal generator to produce a 1 kHz square wave of 2 V amplitude (4 v peak-topeak). Disconnect the twisted pair from the Geiger counter control box and instead connect the signal generator to the input of your trigger circuit on the breadboard. Now set your VI to count 10 samples of 100 ms. When nished, set the histogram accordingly for a longer run. With a longer run (50 samples, say) you can accumulate statistics on the counting accuracy. Record the histogram, measured number of counts with error and signal generator frequency. Repeat for a 1000 ms counting interval. When you are nished, disconnect the signal generator from your trigger circuit input and reconnect the twisted pair from the Geiger counter control box output. 6. Set Geiger counter operating voltage The 137 Cs source should be in the second slot from the top of the Geiger tube holder. Be sure the Geiger counter control box is set to count. We will leave it in this mode at all times. Also, dont push the reset button. It can generate spurious pulses in the output to the discriminator. This source emits energetic particles and rays. Sufciently energetic s enter the Geiger tube through the thin window and ionize the gas to cause output pulses. The s can eject energetic electrons from atoms in the gas lling the Geiger tube to trigger the counter. (a) Finding the threshold 5

!
4

Note: connect 0.01 F capacitor between 5V (pin 14) and ground (pin 7) at chip

Use the LabVIEW VI (not the counter control box itselfleave it counting continuously; it has a somewhat lower counting threshold). Be careful not to exceed 1000 V. If you hear a ticking sound from the counter (HV breakdown), reduce the voltage immediately and call the TA. Set the VI to count a small number of samples (10, say) with a short count interval (100 ms). Start the VI in repetitive mode (recirculating arrow) and watch the total number of counts per run as you slowly increase the Geiger counter HV from its minimum setting. Find the minimum HV setting for which you get reliable counting. Record this threshold voltage. (This procedure may need to be modied.) (b) Making a plateau curve Start at the next 50 V HV multiple above the threshold (e.g., if the threshold was 730 V, start at 750). Set the VI to acquire 10 samples with a 1000 ms count interval and start the program. After 12 s or so, you will get a count measurement for this voltage (total counts reading). Record this and the voltage. Now increase the voltage in 50 V increments up to 1000 V (or 300 V above threshold, whichever is lessdont exceed 1000 V) and repeat the count measurements. You should get samples, ni , of order 1000 counts each on the plateau. Make a (hand-drawn) graph of your result with error bars. We assume Poisson statistics here, so we can estimate i = ni . The threshold voltage in our discriminator is too high to observe the linear or proportional region of the counter, but there should be a gently rising linear plateau in the Geiger region above the threshold and to the left of any sharp rise at large voltages. Set the voltage near the center of this plateau and record this operating voltage (it will probably be 100 V above your rst measured point above the counting threshold). Fig. 6 shows a typical plateau curve and the chosen operating point. Your results may vary somewhat.

Figure 6: Geiger counter plateau curve. 6

7. Observe Geiger pulses and dead time Sketch the pulses at four points: trigger circuit input, comparator input, comparator output and trigger circuit output. Note: check the output pulse at a high sweep speed to be sure there are no spurious extra pulses at the end (glitches) which could be counted by the 20 MHz counter. Be careful not to interfere with the proper termination of the output. Now set the oscilloscope time base to 50 s/division. The source should now be in the top slot, nearest the tube. You should see additional output pulses on the same trace as the triggering pulses. These appear at random, following a uniform distribution except near the beginning of the trace. Estimate how long this gap lasts where no additional pulses are seen. This is the overall dead time of the circuit. Record this information. Actually, the counter is recovering from delivering its previous pulse toward the end of this period. If you hook the scope probe to the comparator input, you should see small pulses occasionally near the end of the deadtime region below the 0.6 V threshold (scope trigger level needs to be set low enough to capture them). These pulses increase in size until they reach the maximum. This is the Geiger counter recovery region. See if you can observe this. 8. VI le output Optionally, an output text le of the histogram contents may be written in a spreadsheetcompatible format for further analysis. 9. Observe Poisson statistics with mean near 0 This should be done with the source in the next-to-bottom slot (away from the tube) and with a relatively short count interval (100 ms, say). The mean number of counts should be in the range of 1-3 counts; be sure it is low enough that you get a reasonable number of zero count samples. The histogram should be set to have bin widths of one count and cover the range 0-10 counts. Do a run of 10 samples and check that the histogram of the samples ni is correcti.e., make the histogram by hand and see that it agrees with the computer. Include the check histogram in your lab notes (this is to verify that the software is working as expected). Check that the count rate measured by the Geiger counter control box roughly agrees with the VI. Do this by comparing counts for a 10 s interval (total number of counts for 100 100 ms trials with the VI vs. a separate 10 s measurement with the Geiger counter scaler using its controls). The results should agree within statistics. If not, discuss with the instructor. Perform and record to le an extended run under these conditions (100 samples, say) to get a histogram with good statistics. Be sure to record the conditions of the run, particularly the histogram setup. Be sure your upper limit is such that there are no overows. Be sure the histogram looks reasonable before proceeding. 10. Observe Poisson statistics with mean 100 and compare with normal distribution The source should now be in the slot third from the top. A counting interval of 1000 ms will probably be appropriate for this section. Make a histogram which includes the entire distribution on ni (within at least 2) and has boundaries that are easy to understand. For 7

example, if the expected number of counts were 100, you could make a histogram with lower limit 80, upper limit 120 and 20 bins. Do a run of 100 samples, save the output data le for further analysis and record the histogram in your lab notes. Make sure the histogram looks reasonable. 11. Suggested analysis and writeup The idea is to plot the histograms and compare with Poisson (low rate) or Gaussian (high rate) statistics. For the high rate data, we want to compare with the Gaussian limit by calculating the number of counts expected in each bin using differences of the Gaussian cumulative distribution function, given the measured mean value. Use LabVIEW to produce comparison plots. Calculate and compare 2 to the extent possible. The writeup should be brief but complete. It should include a brief discussion of the 137 Cs radioactive decay sequence and Geiger tube operation (see Melissinos and Napolitano, Ch. 8).[1] It should describe your procedure and the equipment used, give your results and the degree of agreement with statistical theory.

References
[1] Adrian C. Melissinos and Jim Napolitano, Experiments in Modern Physics, 2nd. Ed., Academic Press (2003). [2] Bevington and Robinson, Data Reduction and Error Analysis for the Physical Sciences, 3rd Ed., McGraw-Hill (2003). [3] LND 7232 Geiger counter tube, LND, Inc., http://www.lndinc.com