Track Imaging Cherenkov Experiment TrICE

1
Track Imaging Cherenkov Experiment (TrICE)
Collaborators Argonne, University of Chicago,
University of Utah

• Ground-based Gamma-ray Astronomy Towards the
  Future
• May 12, 2006
• Karen Byrum

2
TrICE People

• Argonne Group
• Karen Byrum Group Leader
• Gary Drake Electronics Group Head
• Vic Guarino Mechanical Engineering
• Liz Hays Data Acquisition
• Eve Kovacs Computing Support
• Steve Magill Simulation
• Larry Nodulman Data Analysis
• Rich Talaga in situ Calibration
• Bob Wagner Phototubes
• Ken Wood Technical support

• Univ. of Chicago Group
• Simon Swordy Group Leader
• Rich Northrop Mechanical Engineer
• Scott Wakely Data Acquisition/Alignment
• Stephanie Wissels Mirrors / Alignment

• Univ. of Utah
• Dave Kieda Mirrors

3
What is TrICE?

• TrICE SCIENCE A high resolution method for
  directly measuring TeV-PeV cosmic-ray
  composition.
• Search for exotic (strange quark) charge states
  (Zgt92)
• Characterization of composition tests models of
  cosmic ray acceleration mechanism above knee.
• DETECTOR RD A high resolution camera and fast
  electronics for gamma-ray telescopes.
• Multi-channel PMTs give an angular pixel spacing
  (0.08 deg) better than any existing Cerenkov
  telescope (0.15 deg).
• Development of GHz photon counting ASIC based on
  HEP particle flow algorithm (ASIC Specs mostly
  defined limited chip prototype end of year).
• GAMMA-RAY SCIENCE Small pixels and fast timing
  will allow gamma-ray telescopes to lower Et
• Finer pixel spacing allows more detailed
  morphological mappings (such as the shell
  structure of SNR).
• Finer pixel spacing and fast timing allow better
  night sky background rejection (lower
  Ethreshold).

Fresnel lens to image EAS
MAPMT Camera Array
Spherical mirrors 4m focal length
Plane mirror images direct Cerenkov onto Camera
4
TrICE The Mechanism (small pixels fast timing)

• Use EAS Cherenkov signal as trigger Light Yield
  ? Energy (TeV to PeV)
• Direct Cherenkov signal ? Z2

15ns
0ns
0
0.5deg
Direct Cherenkov light all arrives within lt 1ns
dispersion ( 300psec).
EAS Cherenkov imaged through Fresnel lens.
Direct Cherenkov imaged
by spherical mirrors. Longer
path length gives 20ns time delay
5
Trice Testbed for Technology
100 TeV 56Fe
Digitizied at several thresholds
50 GeV gamma-ray
Single bit elect.
Top Simulated Cerenkov emission due to a 100
TeV 56Fe nucleus collected in bins of 1 ns
resolution in time delay and 0.01 resolution in
the emission angle. Photons falling within a
radius of 67-94 m from the shower core reveal a
well-defined peak at small angles from the DC
emission (left). If this signal is digitized for
several thresholds, the DC peak remains visible
(right). Bottom Simulated Cerenkov emission
from a 50 GeV gamma ray binned as above, but
using an effective area of 500 m2 (left). When
digitized for a single threshold, the the shower
structure and total number of counts remains
intact.
S. Wissel
6
Camera RD Multianode Phototubes (MAPMT)
Burle 85011-501

• Camera Requirements
• Linearity over full dynamic range
• Good single pe resolution
• Quantum efficiency as high as possible
• Good long term gain stability
• Pixel-to-Pixel gain uniformity
• Low crosstalk
• High dark current operation w/o significant gain
  loss
• Triggerable on dynode signal

Hamamatsu H8500
7
Camera RD

• Camera circuit board 16 PMTs

8

Camera RD Single pe Resolution H8500 vs R8900

H8500 single pe
R8900 single pe
R. Wagner
9
Optics
Stephanie Wissels

• 4 Spherical Mirrors installed
• Chopper drives Motors Actuators
• Alignment Software working
• Alignment procedure in place
• First Course Alignment - 3/1/06

10
Optics Aligning the Mirrors
4 mirrors before alignment
Focused rays from a point source at 12m
4 mirrors aligned
11
DAQ Trigger
Liz Hays

• Use Preproduction Electronics for Fermilab Main
  Injector Neutrino Oscillation (MINOS) experiment
  (readout by VMIC VME processor) - Gary.
• Digitize sample every 19ns (Tevatron RF clock
  frequency)
• Uses multirange ADC to digitize integrated charge
• Trigger on phototube summed dynode above set
  threshold
• DAQ developed by Liz Hays
• Records eight 19ns integrated samples beginning
  with one prior to trigger
• Need to determine pedestals and calibrate each of
  8 ADC ranges
• Measure data rates versus thresholds
• Implement dynode trigger

Single pe noise
Bias curve measured in lab dark box
Rate vs dynode trigger threshold -gt using LED
filter wheel (20 - 30 pes) onto pixel 16
12
First Ever Air Shower Candidate Event readout
with Multi-channel PMTs
PMT 1
PMT 2
ADC
PMT 3
PMT 4
TS
19ns Time Slices (TS)
Problems
HV not gain matched Mapping wrong Triggering on
or of 4 tubes
E. Hays
13
Night Sky Peds Use Data to correct for gains.
(random pulser triggered pedestals)
Raw Data
Corrected Data
Raw
Gain corrected
ADC
64 Pixels
Two step gain correction Using Night Sky
Pedestals
Corrected spectrum looks more reasonable, steeper
at higher counts
- GainMap1 correction uses Ncountsgt40 -
GainMap2 trimmed with ltph 40200gt (which makes
counts gt80 look good!)
L. Nodulman
14
Selection of Best Events
Triggered data
Cosmic Rays?
Map Gain Corrected Events with 3 or more tubes
having ADC gt 4000 counts
4 PMT ADC Sum
Single pe Spectrum
L. Nodulman
15
Future Plans for TrICE

• Immediate Future
• Complete Stage 1
• 16 R8900 PMTs with MINOS pre-production
  Electronics
• 8 spherical mirrors installed/aligned
• Dynode Trigger installed/ Current monitoring
  protections systems working
• Install in-situ calibration system
• Through the summer Commissioning, Data taking,
  MC simulations
• Continue with RD of photon counting ASIC -gt
  Digital TELescope (DTEL)
• Conceptual Design ANL U of C
• Design is based on architecture of the ASIC being
  built for the ILC hadron detector (Big
  difference is speed 1GHz)
• Second Stage of TrICE (requires new
  proposal/funding)
• Implement ASIC into readout 256 ch -gt 1600 ch
• Increase angular acceptance 16 MAPMTs -gt 100
  MAPMTs
• Observation of direct Cherenkov may demand this
  increase.

16
Backup Slides
17
Camera RD Night Sky Background Simulation
Use incandescent bulb to provide various DC
currents
Measure response of a pixel using pulsed LED
signal for various levels of background
light. Response is relative to no background
light (zero NSB)
Conclude No effect on gain for expected night
sky background
R. Wagner
18
TrICE Camera Simulations

Simulated image of 100 TeV Fe with core at 72m
(from detector) The camera consists of a 40 40
array of 0.08pixels. The signals are given in
photoelectrons assuming an efficiency of 20 and
integrating over the duration of the shower.
The DC emission peak (also shown separately,
bottom) appears at the leading edge of the image
separated from the bulk of the shower Cerenkov
emission. Cuts on timing properties further
enhance separation.
S. Wissel
19
Looking at Triggered Data
Night Sky
Triggered PH
Triggered PH
Night Sky
19ns Time Slice
Triggered ph is in 1 ts mostly, and 2 ts, night
sky everywhere
Each PMT see triggers and night sky, late
timeslices show only lower peak
6 files, 7k events
L. Nodulman
20
Direct Cerenkov Window

• Plot shows window for direct Cherenkov detection
  as function of Z of primary.
• Upper limit due to EAS overwhelming DC signal
• Lower limit from threshold for Cherenkov emission

21
First Telescope Data with Multi-channel PMTs
Typical triggered single pixel noise event
22
Typical Cherenkov Spectrum
Cherenkov Spectrum
Classical PMTs
23
100 TeV Iron
24
Software Trigger
Everyone has night sky, triggers mainly 1 guy,
cut harder than trigger gets interesting, gets
interesting faster if you ask for two
I took good guys to be gt2 gt4000, all 7 of them!