Digital Signal Processing for HPGe Detectors July 28, 2012 David Radford ORNL Physics Division
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Digital Signal Processing
for
HPGe Detectors
July 28, 2012
David Radford ORNL Physics Division
2 Managed by UT-Battelle for the U.S. Department of Energy
• Hyper-Pure Ge (HPGe) detectors are the “gold standard” for gamma-ray spectroscopy − Unsurpassed energy resolution
− Indispensible to in-beam nuclear structure studies for many decades
• Made from a single large crystal pulled from molten hyper-pure Ge • Operated as a large reverse-biased diode; up to 5 kV bias
• No current flows until a gamma ray interacts with an electron in the Ge, lifting its energy above the band gap of the Ge semiconductor
• This electron scatters off other electrons, creating many electron-hole pairs; each pair takes ~ 3 eV in energy
• The electrons and holes separate in the strong electric field and are collected at the electrodes
• The resulting charge pulse is proportional to the deposited gamma-ray energy, and is amplified and digitized
• Operated at cryogenic temperatures to prevent thermal generation of electron-hole pairs
HPGe Detectors
3 Managed by UT-Battelle for the U.S. Department of Energy
Closed-end coaxial
n+
p+ p
n+
p
Planar
HPGe Detectors Historically, there have been two designs, both cylindrical:
Coaxial and Planar.
Later we will discuss a new design, “Point Contact” detectors
p+
4 Managed by UT-Battelle for the U.S. Department of Energy
Closed-end coaxial
n+
p+ p
n+
p
Planar
HPGe Detectors Historically, there have been two designs, both cylindrical:
Coaxial and Planar.
p+
~ 3 kV
−
+
5 Managed by UT-Battelle for the U.S. Department of Energy
HPGe Detectors
Electric potential
Drift paths of electrons and holes
Closed-end coaxial
Fast ADC Charge-integrating pre-amp
Signal Processing
Once the preamplifier signal/waveform has been digitized, need to process the data to extract interesting numbers, e.g. • Amplitude
• Gives energy, so we want it to be as accurate as possible
• Timing • Pulse-shape analysis to select specific types of
events • Noise filtering and/or reduction • Pile-up rejection and/or correction
Extracting Signal Amplitude
Height of signal is proportional to
detected energy
Preamp signal • Decays exponentially with
~ 50µs time constant
Time
Volta
ge
Extracting Signal Amplitude
But the signal has noise… • Need to integrate signal for several microseconds
to determine amplitude more precisely • Use “Trapezoid Filter”
Extracting Signal Amplitude
Output of Trapezoid Filter
Undershoot due to exponential decay of signal
Extracting Signal Amplitude
• Need to correct for signal decay time • “Pole-zero” correction
Trapezoid and P/Z Correction • Need to know the baseline (asymptote)
and signal decay time constant tau • E ~ S2’ – S1’
Baseline
S1’
S2 S2’
S1
14 Managed by UT-Battelle for the U.S. Department of Energy
P-type Coaxial P-type Point Contact
p p+
n+ n+
p+
p
Point Contact Detectors • No deep hole; small point-like central contact • Length is shorter than standard coaxial detector • Excellent discrimination between single-interaction and multiple-
interaction events • Excellent resolution at low energies
Pulse-Shape Response Point Contact detectors are ideal for discrimination between single-site and multi-site events (or determining the number of interactions)
0
100
200
300C
harg
e Si
gnal
0
20
40
60
Cur
rent
Sig
nal
200 600 1000 1400Time (ns)
0
100
200
300
400
0
20
40
60
80
200 600 1000 1400Time (ns)
a) Single-site b) Multi-site
16 Managed by UT-Battelle for the U.S. Department of Energy
Pulse-Shape Response
9/30/09 Radford, RedTeam Review
Red: all events Blue: PSA-selected events
Th source
95% efficiency for double-escape peak = single-site events 99% rejection of single-escape peak = multi-site events
17 Managed by UT-Battelle for the U.S. Department of Energy
“Inverted-Coaxial” Point-Contact Detector • Designed and developed here at ORNL • Drift of charges is radically different from a normal coaxial detector • Very long drift times, ~ 2 µs • A segmented prototype has recently been produced by Canberra France
19 Managed by UT-Battelle for the U.S. Department of Energy
Am “SuperPulses” • Finely collimated Am source,
directed at known location on the detector surface
• Select events with 60 keV in a single hit segment
• Use PSA to select only single-site events
• Time-align events to a common time
• Take average signal to reduce noise to negligible level
RadWaresig2sp.ps, created 08-Apr-13 03:18:08
Cha
rge
1 1 1 10 2 2 2 2Time ( s)
Superpulse
Singleevent
PC Seg1 Seg2 Seg3
Related Documents
