Investigation of a Cs137 and Ba133 runs. Michael Dugger and Robert Lee

Size: px

Start display at page:

Download "Investigation of a Cs137 and Ba133 runs. Michael Dugger and Robert Lee"

Jerome Boyd
5 years ago
Views:

1 Investigation of a Cs137 and Ba133 runs Michael Dugger and Robert Lee 1

2 Cs137 Using run 149 One million triggers Doing a quick analysis with fits: Not using Kei s noise corrections at the moment 2

3 ADC ADC ADC ADC ADC ADC First six fits time (ns) time (ns) time (ns) time (ns) Start time of event pulse time (ns) time (ns) 3

4 max adc baseline (from fit) ADC height versus baseline Baseline adc (from fit) Can t calculate the correct baseline for these events (start time < 0) 4

5 max ADC baseline (from fit) ADC height versus baseline with timing cut Baseline ADC (from fit) Baseline ADC (from fit) No timing cut Required start time of event pulse to be greater than 20 ns 5

6 ADC height versus baseline with timing cut Accidentals below the threshold Baseline ADC (from fit) 6

7 Cs137 spectrum Data Monte Carlo with 10 kev energy smear Accidentals ADC The shapes do not agree Energy (MeV) Need to figure out how to modify Monte Carlo to look like real data 7

8 Lower energy modification to Monte Carlo The Monte Carlo does not take the fluctuating baseline into account Need to find the threshold and baseline parameters 8

9 ADC Threshold time (ns) Threshold looks to be at about

10 max ADC baseline (from fit) Baseline Took x-projection for y > 150 and fit to Gaussian Baseline ADC (from fit) Baseline: Center = σ = 11.3 Baseline ADC (from fit) 10

11 Cs137 spectrum with baseline and threshold Data Monte Carlo Accidentals ADC Energy (MeV) The shapes still do not agree, but low energy is looking a bit better 11

12 Higher energy modification to Monte Carlo Tried putting in thin layers of aluminum between source and detector but the agreement between Monte Carlo and data did not get much better Decided to look at varying the depth of the depletion region 12

13 Depletion region (slide 1) The depletion region is where we can convert energy deposition to voltage. Outside the depletion region the detector is blind (no potential difference to sweep the charges) Applying a reverse bias potential to the detector causes the depletion region to grow We are fully depleted with a bias potential of 165 V (but run the detector at the manufacturer s suggested 200V) 13

14 Depletion region (slide 2) Our detector has a thin highly-doped p-type semi conductor on the ring side with the bulk of the detector being lightly doped n-type The dependence of the voltage to the depletion depth D is D = 0.53 [ρ n V] 1/2 μm, where is the resistivity in Ωcm and V is in volts Assuming that we are fully depleted at V f = 165 (D f = 1000 μm) then D = [V/165] 1/2 mm 14

15 Monte Carlo with different depths of depletion 15

16 Cs137 spectrum with 500 μm depletion Data Monte Carlo Accidentals ADC Energy (MeV) Not a perfect agreement on the shapes but very suggestive It is as though we have only half the depletion depth which would be equivalent to having only about 40 volts of bias applied 16

17 Rates Assume: 1 μci Cs137 source (2.22x10 6 dpm) Source 3.5 cm from detector Rate with baseline fluctuation, threshold and full depletion: 249k/minute Rate with baseline fluctuation, threshold and D = 500 μm : 19.0k/minute Time to one million events T 1M : With baseline fluctuation, threshold and full depletion: T 1M = 4.02 minutes With baseline fluctuation, threshold and D = 500 μm : T 1M = 52.5 minutes 17

18 Ba133 Used run 153 Only total events taken over ~30 hours! 18

max adc baseline (from fit) ADC height versus baseline with timing cut for Ba133 Baseline adc (from fit) Baseline adc (from fit) No timing cut Required

19 max adc baseline (from fit) ADC height versus baseline with timing cut for Ba133 Baseline adc (from fit) Baseline adc (from fit) No timing cut Required start time to be greater than 20 ns Threshold too high Timing cut kills too many events of interest Will not use timing cut for the Ba133 source data 19

20 Threshold and baseline fluctuations for Ba133 Threshold for run 153 was set at a higher level than for the Cs137 run analyzed thus far (run 149) Did same sort of analysis for run 153 as for run 149 to get the threshold and baseline parameters 20

21 Ba133 spectrum with 500 μm depletion Data? Monte Carlo Accidentals Energy (MeV) Energy (MeV) Used old energy calibrations Plots look similar but the data has much more background 21

22 Rates Assume: 1 μci Ba133 source (2.22x10 6 dpm) Source 3.5 cm from detector Rate with baseline fluctuation, threshold and full depletion: 1.45k/minute Rate with baseline fluctuation, threshold and D = 500 μm : 11.7/minute Time to events T : With baseline fluctuation, threshold and full depletion: T = 26.8 minutes With baseline fluctuation, threshold and D = 500 μm : T = 55.5 hours Longer time than what Kei saw but there was a bunch of unidentified background in the energy spectrum 22

23 New or old effect? If this is an old effect, shouldn t we have seen this sort of effect in the Spring run? 23

24 Spring run energy deposition and rates Needed to shift the energy deposition of the data by 6% to get good agreement between data and Monte Carlo The rates between the pair spectrometer and TPOL where consistent to within 1.1 standard deviations 24

25 Monte Carlo of triplets E dep Blue: Full depletion Red: D = 500 μm E dep (MeV) If D = 500 μm I would have had to shift the energy deposition of the data by about a factor of 2 to get the peaks to line up and I would not have been able to get consistency in rates between TPOL and the pair spectrometer The Spring data appears to have TPOL with nearly full depletion 25

26 Diagnostics We need to figure out what is going wrong Potential diagnostics: HV with signal on o-scope Rates The smoking-gun test (source on ring-side) 26

27 HV, capacitance and noise level Our detector acts as a parallel plate capacitor (C = ε area/separation), where the dielectric is the depletion region The separation between the plates is the depth of depletion (D) The preamps noise is highly dependent upon the input capacitance (mostly the capacitance of the detector) As the bias is increased, the depletion depth grows and the capacitance diminishes, which causes the noise level to decrease The noise level can be seen as a proxy for the depletion depth 27

28 Ripple and other noise as function of HV (slide 1) 10 mv/div 10 ms/div HV = 0V Using Po210 source HV = 20V HV = 40V HV = 60V 28

29 Ripple and other noise as function of HV (slide 2) 10 mv/div 10 ms/div HV = 80V HV = 100V HV = 120V HV = 140V 29

30 Ripple and other noise as function of HV (slide 3) 10 mv/div 10 ms/div HV = 160V HV = 180V α HV = 200V HV = 200V 200 mv/div & 1 µs/div 30

31 HV check The noise level at 40 volts is much higher than at the 200 volts we run at Make sure that as you go up in bias voltage that the noise becomes more suppressed (until about 120 volts) 31

32 Rates Check the distance of source to detector and the activity of the source For a 1 μci Ba133 source placed 3.5 cm from the detector, the rate should be something like 4k/minute For a 1 μci Cs137 source placed 3.5 cm from the detector, the rate should be something like 112k/minute The rates give us information that is valuable. So even if the rates are small, they should be recorded 32

33 The smoking-gun test (source on ring side) The depletion region starts at the interface of the p- and n-type material The p-type material is on the ring side and is highly doped and very thin. The depletion depth extends into the bulk n-type material from the ring side. Thus, the behavior of the detector, when not fully depleted, is very different for particles incident on the ring side than if the incoming particles were incident upon the sector side 33

34 Comparison of results on source placement Source on sector side Source on ring side For this study I changed the smearing to be 12 kev A problem of having only a partial depletion will show up clearly when comparing results of sector/ring side placement of source 34

35 Title 35

36 Title 36

Triplet polarimeter update

Triplet polarimeter update M. Dugger, February 2015 1 Plan Set up a polarimeter test bench in the Experimental Equipment Laboratory (EEL) Further test the silicon strip detector using fadc and sources