Transcript MEDIPIX - University of Houston

Preparing For a Medipix-Based
Dosimeter Demonstration on the
International Space Station
Lawrence Pinsky
Physics Department
University of Houston
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
1
Lawrence Pinsky1*, Nicholas Stoffle1,2, Son Hoang1,
Ricardo Vilalta1, Kerry Lee2, Neal Zapp2, Edward Semones2,
Daniel Turecek3, Jan Jakubek3, Zdenek Vykydal3,
Stanislav Pospisil3, Hisashi Kitamura4,, Yukio Uchihori4,
Nakahiro Yasuda4 and Ondrej Ploc4
1Physics
2NASA
Department, University of Houston, Houston, Texas, USA
Johnson Space Center, Houston, Texas, USA
3Institute
for Experimental and Applied Physics, Czech Technical University
in Prague, Czech Republic
4Radiation
Measurement Research Section, National Institute of Radiological
Sciences, Chiba, Japan
Medipix2 and Medipix3 are CERN-Based Collaborations…
(IEAP-CTU-Prague and the University of Houston are Members
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
2
Hybrid Pixel Detector
Detector and electronics readout are optimized separately
3
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
Hybrid Pixel Detector - Cross Section
UH is currently working on epitaxial deposition techniques that
will facilitate the creation of high efficiency Embedded-Neutron-Converter detectors
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
4
65,536 Bump-Bonds on each unit…
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
5
TimePix Cell Schematic
• Charge sensitive Preamp/Shaper w/ individual leakage current compensation
• Discriminator with globally adjustable thresholds & individual 4-bit fine tuning offset
• Individually settable test and mask bits for each pixel
• External shutter activates the counter (can be as short as 10 ns,
•
but 100 ns is the practical limit)
• 14-bit output register (11,810 decimal)
• …with Overflow indication
• Each pixel can have its mode
(ADC or TDC) set independently
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
6
TimePix Threshold Calibration

Calibrates the 4-bit (16 level) Threshold Offsets for each
pixel. The RED histogram is the distribution of noise turn-on
points with all bits set to high. The BLUE histogram is the
corresponding low setting. The BLACK histogram is the
corrected result. Each channel is ~20 e-…
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
7
Time-Over-Threshold (“ADC”) Mode
(P0=1,P1=0)
“Shutter” Window
33
50
12
Clock Out
Discriminator Output
Analog Signals
2Ke-
4Ke-
2Ke-
10MHz Clock
Threshold
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
100MHz Clock
8
TimePix and its TOT mode
TOT
Counter in each pixel can be used as
Timer to measure detection time => TOF experiments, TPC detectors, …
Wilkinson type ADC to measure energy of each particle detected.
Calibration
curve
Energy
Energy
Threshold
Time

If the pulse shape is triangular then
Time over Threshold is proportional to
collected charge i.e. to energy.

Due to limited bandwidth the pulse can
be NEVER perfectly triangular.

Non-linear TOT to energy dependence
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
9
Charge Clusters
The electron-hole pairs liberated by
traversing moving charges drifts in the
bias voltage and also diffuses during the
process, creating a multi-pixel cluster.
TimePix
Incident
Particle
“PIXELMAN”
Image
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
10
Dosimetry in Space

Our approach is to try and Characterize the Radiation Field as
precisely as possible as a function of time.

To do that, we need to assess the radiation environment in terms
of the Charge AND Energy of the individual particles that are
present.

…HOWEVER, because of the “Z2 effect” and the shape of the
energy-loss curves, it is possible for different ions to have the
same dE/dx in a thin detector…

Slow lower-Z particles seen in the dosimeter will not penetrate
deeply into the body, and can be mimicked by higher-Z faster
particles, which CAN penetrate deeply…

SO, again, “Our approach is to try and Characterize the Radiation
Field as precisely as possible as a function of time.”
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
11
Measuring Charge AND Energy



If you know b then measuring
the charge is reasonably
simple because of the Z2
dependence.
In accelerator experiments
the interaction fragments
from the projectile particle is
generally moving at close to
the projectile’s original
velocity, at least for forward
fragments.
However, if you have no a
priori velocity information, the
problem is the BETHE-BLOCH
Equation… (One can observe
the behavior over a longer
distance, or look for track
structure differences…)
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
All charge-1
particles.
Higher charge
particles are raised
by the square of the
charge at the same
momenta
12
Also, dE/dx ≠ LET
(High Energy d-Rays)



Because the number of d-rays
produced per unit track
length in the Air prior to
entering the Si is much less
than in the the Si…
…The highest energy d-rays
carry away more energy from
d-Rays
the Si than enters from the
air.
HOWEVER—It is the High
Energy d-rays that offer the
prospect of telling the
difference between the
different particle velocities
with the same dE/dx…
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
TimePix
Air
Incident
Particle
Si
Si
13
FLUKA Medipix (TimePix)Heavy Ion
Simulations
Note the ~10% differences in the
peak location, with the
measurements being 10% lower
than the simulations due to peak
saturation effects… However,
the widths are the same  The
Saturation effects are common…
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
14
TimePix in the HIMAC Beams



Data have been taken at
HIMAC in some dedicated
and many parasitic runs…
Runs at a range of
incident angles and with
different detector settings
are typically taken…
The primary beams were
(MeV/A):
 1H (p) 160
 4He 180 & 230
 12C
230
 14N
180 & 290
 16O
100 & 230
 20Ne 180, 430 & 600
 28Si 400, 600 & 800
 56Fe 500
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
15
Examples of Individual (Normal Incidence)
“Tracks”
100 MeV/A O
180 MeV/A Ne
600 MeV/A Si
800 MeV/A Si
The high energy d-rays are clear in the higher
energy tracks. These are not yet calibrated,
and the study to be preformed is to explore the
detailed resolution possible when all the
information is included…
The next goal is to be able to model these
tracks in the FLUKA Monte Carlo code…
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
16
800 MeV/A Si 85 degree Tracks



The tracks are from
particles diving
downward from left to
right.
As they pass though
the solder-bumps and
into the underlying chip
after leaving the Si
detector layer.
Some of the high
energy d-rays from the
chip enter the overlying
detector layer…
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
17
16O
@ 100 MeV/A—0° & 60° Averaged Cluster
Shapes
(Azimuthal & Polar Angle resolution ~ 1 Degree)
0°
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
60°
18
TimePix as an in- Flight
Dosimeter
84°N Latitude
~ 180 ° Longitude
@ Local Midnight
at 35,000 ft
Australia
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
100 Sec Integrated Image
The Dose RATE is ~ 50 times
the surface background rate
(in Houston) w/o Neutron Corr.
19
M.D. Anderson Proton Therapy
Center Scanning Beam Frame
This 10 ms frame was taken in the
high intensity scanning beam at
the M.D. Anderson Proton Therapy
Center in Houston, Texas. The
fluence is > 108 protons/cm2s. The
beam is centered near the lower
right edge of the frame and is
nominally 1 cm in diameter. The
frame is ~1.4 cm across. At this
fluence, the charge sensitive preamp shaping return feedback had
to be minimized to reduce the total
current draw on the Medipix chip
to avoid a voltage sag that would
have affected the chip’s overall
functioning. Individual p tracks
are visible in the core of the beam
and in the beam’s halo…
20
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
NASA HyPRD—SDTO
Hybrid Pixel Radiation Dosimeter—
Station Demonstration Test Objective



At the request of SRAG, We are
currently preparing to fly
several Timepix-based USB
interfaced units (HyPRD’s) as
an SDTO on the ISS possibly
later this year…
The SDTO units are the size
and weight of a typical
“Memory Stick” Flash Drive.
There is also a Phase I SBIR
that is producing a prototype
stand-alone interface (BatteryPowered, Wireless, with local
Analysis Capability and
Readout with Alarms)—Phase II
will hopefully begin this Fall
Earlier Medipix2 USB Interface
First ISS Flight Unit Prototype
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
21
SDTO ISS Flight
Software
(Via USB
Interface)




The flight software will run on the existing ISS Lenovo laptops
attached to the HyPRD unit the via a standard USB Type-A connector
It will have automatic frame-length and rate adjustment capability…
It will display cumulative (dose) and current (dose-rates) [NCRP153] Dose-Equivalent locally on the laptop.
Full data frames will be downloaded to the ground for analysis in a
Root-based software environment…
22
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
HyPRD Hardware Processing at CERN
HyPRD
Wire-bonds
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
23
LUCID – Langton Ultimate Cosmic ray Intensity Detector
TimePix’s First Space Mission—Educational Outreach
(UK Satellite to be Launched Q1—2012)
Velocity



5 TimePix Detectors (0.7 mm Al dome)
Sun-Synchronous Polar Orbit (98° @
11:00)
Programmable Signal Processor for
on-orbit data analysis…
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
Down
24
LUCID

Educational Outreach





PI is Becky Parker from the
Langton School in
Canterbury, UK.
Data will be available in
Daily downloads to High
School groups worldwide
via the Web…
Online analysis tools will
be provided…
Correlations with surface
Cosmic Ray Detectors
deployed in schools
worldwide is possible…
CERN@School

TimePix kits are being
provided for laboratory
use.
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
25
Medipix3 is Coming


Even more Rad-Hard!
Dual Circuit Capability in EACH Pixel.




Linear and Log Pre-Amps Possible…



Eliminate Saturation Effects
Smaller Pixel Sizes Possible


Either and ADC and a TDC simultaneously in each pixel…
Or, two ADCs with sequential use for 0 dead-time (< 100 ns)
Or, two ADCs with different Pre-Amp scales or responses…
256 x 256 - 55mm  512 x 512 - 25 mm
Lower Noise
On-Chip Output Clustering & Charge Centering

On Chip charge-sharing for sub-pixel position resolution
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
26
ATLAS Area Neutron Monitors
Detector Configuration (IEAP-Prague)
Medipix2 ASIC with 300µm Si
sensor + USB interface
Neutron conversion structures:
1)LiF+50µm Al foil area
2)100µm Al foil area
3)PE area
4)PE+50µm Al foil area
5)Uncovered area
X-ray image of conversion layers
LiF
PE
27
PE + Al
Al
Uncovered
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
Neutron Efficiency Calibration
Calibrated efficiency:
X-ray image of conversion layers
Courtesy of
Jan Jakubek
& Zdenek Vykydal
Thermal neutrons – 500s, 25meV
Thermal: 1.41E-2 ± 7.11E-4 cm-2s-1
252Cf: 1.19E-3 ± 1.89E-5 cm-2s-1
LiF
AmBe: 2.86E-3 ± 5.46E-5 cm-2s-1
PE
VDG: 7.23E-3 ± 5.81E-4 cm-2s-1
PE + Al
PE / PE+Al cluster count ratio:
252Cf: 10.70 ± 0.04
Al
AmBe: 5.18 ± 0.03
VDG: 2.51 ± 0.03
252Cf – 2000s, 2MeV (mean)
Uncovered
AmBe – 2000s, 4MeV (mean)
Van de Graaff – 1000s, 14MeV
28
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
Thank You for Your Attention
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
29
Backup Slides








Daya Bay Application (4)
Electron Measurements (1)
Drift Cloud Structure (1)
Using the Timepix Mode for Fission Fragments (3)
The “Volcano Effect” (4)
Solder Bumps (High Magnification) (1)
Timepix Mode Schematic (1)
60 Degree Fe Tracks (1)
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
30
Background Radiation @ Daya Bay(An
Underground Neutrino Oscillation Experiment
Being Constructed Near Hong Kong
This is an Integral of the
sum of all pixels for the
May 20, 2010 Hall 5 run:
(83280 sec = 23.13 hours)
The plot has a high
relative threshold to
suppress the MIPs
The heavily ionizing
tracks are from RadonChain Alphas…
The Pattern Recognition analysis takes into
account the shape and energy per pixel as well as
the integral energy in the total track cluster…
31
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
3 Prevalent Event Types in Daya Bay Hall 5
Penetrating Muon
Total Cluster Energy
< 120 Raw Counts
(angle dependent)
Alpha (Radon Chain)
Total Cluster Energy
> 1000 Raw Counts
Electron (g Compton)
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
32
Hall 5 – May 20 – 23 hour Pixel Sum
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
33
Radon

Radon is produced in the 238U decay chain:









…226Ra (1602 y 4.87 MeV a)  222Rn (3.8d 5.59 MeV a) 
218Po (3.1m 6.12 MeV a) 
These atoms are
typically left in an
214Pb (26.8m 1.02 MeV b) 
ionized state, and
214Bi (19.9m 3.27 MeV b) 
quickly adhere to the
nearest surface or dust
214Po (0.16ms 7.88 MeV a) 
particle
210Pb…
Summary  5.59 + 6.12 + 7.88 MeV = 19.59 MeV a
✚ 1.02 + 3.27 MeV = 4.29 MeV b
“Normal” Radon Backgrounds…



The sea-level atmosphere contains on the average of
150 222Rn atoms/ml. This causes ~100 Bq/m3 indoors, and
~10-20 Bq/m3 outdoors, with health limits ~ 230 Bq/m3
1 pCi/l = 37 Bq/m3 (1 Bq = 1 decay/s)
Given the ranges of a’s in air, the Medipix device in open
air samples a volume of about 8-12 cm3 or ~ 10-5 m3
34
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
23 MeV e- Measurements @ IAC
with the TimePix v. FLUKA
FLUKA ePrimary eAll Clusters
KeV
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
35
Charge Drift Cloud Image (241Am 5.5 MeV a)
Time of Arrival
image from a
5.5 MeV a from
an 241Am decay.
Common global
threshold can
be adjusted to
get time (i.e.
charge)
contours
through the drift
cloud…
(Single Event)

ISSRMP - Shanghai - L. Pinsky - 6/2/2011
36
Nuclear Fission Measurements
Using TimePix (TDC) Mode
Radioactive fragments from 239Pu selected by mass
separator LOHENGRIN (Grenoble) and deposited
onto the Timepix surface.
Reactor (ILL)
Timepix
Separator: E, B
Fragments37
Neutrons
Courtesy of
Jan Jakubek
Target
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
8He
Test with
8He
b
n
120 ms
b
810 ms
a
1. Decay is slow => Timepix
clock set to 2.5 kHz
2. To suppress the initial
cluster from ion the delay
of 25 us inserted
< ns
delay
Trigger
25 us
120 ms
8He
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
b
Courtesy of
Jan Jakubek
120+810 ms
b+a
38
Measured events: 8He decays (TDC Mode)
Primary
ion
2ndbeta
decay
1st beta
decay
Alpha
decay
The other decay mode
39
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
Courtesy of
Jan Jakubek
The “Volcano” Effect




We see a dip in
response for the
highest charge
deposition rates…
This may be due to
detector saturation
effects…
…Or to a plasma
effect that causes
high recombination
rates…
So far we see this only
in the Fe tracks…
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
40
Close Up of the “Partial” Event



The right-hand event is a
“normal” iron event, which
does show a clear “Volcano”
Effect. The scale is so high
that the d-rays are not visible.
The left-hand event is a
“Partial-Event.” One that was
partially cutoff by the
“Shutter.”
Because the central hole
essentially goes to zero, it
would appear that this event
occurred at the end of the
Shutter window and was only
the early part of the drift
image…
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
41
56Fe
@ 500 MeV/A
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
42
Fe “Volcano” Event
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
43
Bumps on the readout side – close up
44
ISSRMP - Shanghai - L. Pinsky 6/2/2011
Timepix (“TDC”) Mode
(P0=1,P1=1)
open
close
“Shutter”
17
164
Clock Start
Stop
Discriminator Output
Analog Signals
10MHz
100MHz
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
45
56Fe
60 degree Runs…
Cluster volume distribution
Vol All
Vol OK
12
Count
10
8
6
4
2
0
0
20000
40000
60000
80000
1e+005 1.2e+005
Volume [keV]
ISSRMP - Shanghai - L. Pinsky - 6/2/2011
46