Transcript Arno_Gadola_PhD_Seminar_270812_V4x

A new camera concept for Cherenkov
telescopes
Zurich PhD seminar, 27. - 28. August 2012
27.08.2012
Arno Gadola
Motivation
Ground-based very high energy gamma-ray
astronomy
Flux
IR
• Non-thermal processes generate gamma-rays (keV – PeV).
R
• Investigation of galactic and extragalactic objects:
SNR, pulsars, x-binaries, Active Galactic Nuclei etc.
-4
O UV
X
γ
Inverse Compton
Synchrotron
1
6
11 log(E) [eV]
• Improve understanding of processes of high energy gamma-ray production.
• New physics: indirect dark matter searches, test of quantum gravity models
• Ground-based high energy gamma-ray astronomy uses atmosphere as calorimeter and thus profits
of a very large collection area.
• Imaging Atmospheric Cherenkov Telescopes cover energy range of 10 GeV – 100 TeV.
Current and future ground-based telescopes:
MAGIC, H.E.S.S., VERITAS, FACT
Cherenkov Telescope Array (under development)
2
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Courtesy NASA/JPL-Caltech
γ
Working principle
Eγ: 10 GeV – 100 TeV
Gamma energy: 2.193 TeV
Impact parameter: 231 m
Telescope: 12 m dish
Cherenkov light
1 to >5000 photons/pixel
3
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Courtesy NASA/JPL-Caltech
Working principle
6
Eγ: 10 GeV – 100 TeV
Pair production
e+
γ
γ γ
Bremsstrahlung
γ
Cherenkov light
Sea level
Mountain level
10
2
10
~100 photons/m2
for Eγ = 1 TeV
0
10
-2
10
e-
Data taken from T.C. Weekes, Very High Energy Gamma-Ray Astronomy
4
[photons m -2]
γ
Cherenkov photon density
10
0
10
2
4
6
10
10
10
Primary gamma energy [GeV]
Gamma energy: 2.193 TeV
Impact parameter: 231 m
Telescope: 12 m dish
<100 ns
Ø 250 m at ground level
Cherenkov light
1 to >5000 photons/pixel
4
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Courtesy NASA/JPL-Caltech
Working principle
6
Eγ: 10 GeV – 100 TeV
Pair production
e+
γ
γ γ
Bremsstrahlung
γ
Cherenkov light
Sea level
Mountain level
10
2
10
~100 photons/m2
for Eγ = 1 TeV
0
10
-2
10
e-
Data taken from T.C. Weekes, Very High Energy Gamma-Ray Astronomy
4
[photons m -2]
γ
Cherenkov photon density
10
0
10
2
4
6
10
10
10
Primary gamma energy [GeV]
Gamma energy: 2.193 TeV
Impact parameter: 231 m
Telescope: 12 m dish
<100 ns
Ø 250 m at ground level
Cherenkov light
1 to >5000 photons/pixel
Mirror with Ø 12 m has
100 m2 collection area
© CTA
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
5
Working principle
Primary gamma particle
Top of atmosphere
n=1
~ 120 km a.s.l.
Pair production
Cherenkov light emission with θ(n)
𝑐
cos 𝜃 =
𝑛𝑣
~ 20 km a.s.l.
~ 20 m radius
~ 5 km a.s.l.
θ(n)
~ 0.5 -2 km a.s.l.
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
n = 1.00029 a.s.l.
θ ≈ 1.3°
6
Handling large dynamic signals
Large dynamic amplitude range can be treated in two ways:
• dual signal path with
• a high gain covering the low amplitude range (e.g. 0.2 – 200 pe) and
• a low gain covering the high amplitude range (e.g. 20 – 5000 pe)
High gain
PMT
RT
FADC
A
D
Low gain
A
Storage and
offline signal
reconstruction
D
7
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Handling large dynamic signals
Large dynamic amplitude range can be treated in two ways:
• dual signal path with
• a high gain covering the low amplitude range (e.g. 0.2 – 200 pe) and
• a low gain covering the high amplitude range (e.g. 20 – 5000 pe)
• single signal path with non-linear signal treatment
High gain
PMT
RT
FADC
A
D
Low gain
A
Storage and
offline signal
reconstruction
D
Non-linear
A
PMT
RT
D
Storage and
offline signal
reconstruction
8
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Signal shaping
Simulation
RT = 50 Ω
RT = 2000 Ω
PMT
RT
CPMT
RT
Signal bandwidth: 3 times lower
Termination resistor RT influences signal bandwidth:
50 Ω versus 2000 Ω gives 3 times lower bandwidth
 Larger termination resistor allows to digitize signals with a lower sampling rate (~ 3 times)
 Use of reconstruction algorithm on digitized data still produces good timing and amplitude
resolution
9
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
3. Differentiation and smoothing
 Center of gravity above zero = photon arrival time
4. Base line restoration: pole-zero cancellation (0 < p < 1)
and smoothing
 Peak maximum = pulse amplitude
 Peak area proportional to pulse amplitude
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Amplitude [ADC counts]
Amplitude [ADC counts]
2. Up-sampling: linear interpolation (1 ns sample interval)
and smoothing (moving average filter)
Amplitude [ADC counts]
1. Measured and digitized PMT signal (4 ns sample interval)
Amplitude [ADC counts]
Signal reconstruction
Time [ns]
60
40
20
0
-20
150
60
200
Samples [ns]
250
𝑓(𝑡)
40
20
0
-20
60
150
40
200
Samples [ns]
𝑔 𝑡 =
20
250
𝑑𝑓(𝑡)
𝑑𝑡
0
-20
150
60
200
Samples [ns]
250
ℎ 𝑡 = 𝑔 𝑡 + 𝑝 ∙ 𝑓(𝑡)
40
20
0
-20
150
200
250
10
Amplitude [ADC counts]
Signal reconstruction
Amplitude [ADC counts]
1. Measured and digitized PMT signal (4 ns sample interval)
3. Differentiation and smoothing
 Center of gravity above zero = photon arrival time
4. Amplitude determined by integration of up-sampled signal
over fixed window size (typically 200 ns)
𝑖+200
 Peak area proportional to pulse amplitude:
𝐴=
𝑓 𝑡 𝑑𝑡
𝑘=𝑖
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Amplitude [ADC counts]
Amplitude [ADC counts]
2. Up-sampling: linear interpolation (1 ns sample interval)
and smoothing (moving average filter)
Time [ns]
4000
3000
2000
1000
0
150
4000
200
Samples [ns]
250
𝑓(𝑡)
3000
2000
1000
0
150
4000
2000
200
Samples [ns]
𝑔 𝑡 =
250
𝑑𝑓(𝑡)
𝑑𝑡
0
-2000
4000
150
200
Samples [ns]
250
2000
0
NOT USED
-2000
150
200
250
11
Towards a real camera…
12
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
The Photo-Detector-Plane
Photo-Detector-Plane (PDP) consisting of:
- Photomultiplier
- Signal amplifier: analogue signal transmission over CAT5/6 cables
- HV generators: 500 and 1500 kV
- CAN bus for slow control
PMT
RT
Developed and built at University of Zurich
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
13
The Photo-Detector-Plane
Lab test setup: PDP with 12 PMTs
Possible PDP assemblies for different telescope sizes
900 pixel, 1550 mm flat to flat
1296 pixel, 1860 mm flat to flat
1764 pixel, 2170 mm flat to flat
2304 pixel, 2480 mm flat to flat
14
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Mechanical structure
A possible design of the mechanical structure
- ~2.7 m flat to flat
- ~2 m depth
- 1.7 t weight
Back view
Design study by S. Steiner, UZH
15
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Test camera setup
Front
1 Laser pulses fed in with optical fiber
1
8
2 Board with 12 PMTs
3 CAN bus (slow control + 24V)
4 Analogue pixel signal transmission via
CAT 5 cables (1 cable per 4 pixels)
Back
2
7
3
5 FADC drivers for 8 FADC
6 FPGA board
6
5
4
7 Event transmission via LAN to
computer disk
8 Dark box
Not shown:
- PC with offline analysis software
- Slow control PC
- 24 V power supply
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
16
Single photoelectron response
Measured with:
• High voltage of (1000 ± 20) V
• Amplifier gain of 12
Noise
240
Signal
4
Example of single
photoelectron
230
3
220
Counts
Amplitude [ADC counts]
10
10
210
2
10
PMT gain ~50’000
200
190
80
1
90
100
110
120
130
Channels [4 ns]
140
150
160
10
-100
0
100
200
Peak area [ADC counts * ns]
300
400
17
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Full range resolution
Measurement of linearity with electric pulser
Output of preamplifier saturatates while
input · gain is larger than output swing.
Eventually, the output recovers as saturation
conditions fail.
Pulse amplitude reconstructed
by pulse area
Pulse amplitude reconstructed
by pulse maximum
Input signal:
~10 PE
~30 PE
~100 PE
~300 PE
~1000 PE
𝑖+200
𝐴=
𝑓 𝑡 𝑑𝑡
𝑘=𝑖
ℎ 𝑡 = 𝑔 𝑡 + 𝑝 ∙ 𝑓(𝑡)
Transition from linear
to saturation regime
Measured signals, resolution 1 ns / sample
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
18
Summary
• Investigation of a single signal path concept for signals with large dynamic range
• Development of signal reconstruction algorithms
• Advantages
• Lower power consumption
• Lower costs (everything is needed only once)
• Digitized signals don’t produce huge data streams
• Disadvantages
• Amplitude and time resolution slightly worse than dual signal path concept
• Concept may be installed in first Cherenkov camera for the Cherenkov Telescope Array
19
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Full range resolution
Measurement of linearity with pulsed laser, light bulb as noise source (0.24 events / ns) and PMT
Amplitude resolution
Time resolution
ns
Measured ●
Simulated ───
Simulated + 250 ps trigger RMS ───
Measured, peak ●
Measured, area ●
Simulated, peak ───
Simulated, area ───
Smallest possible resolution √x ───
Transition from linear
to saturation regime
Simulation results kindly provided by T. Kihm
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
20
‘Non-linear’ amplifier
Voltage-feedback-operational-amplifier
Current-feedback-operational-amplifier
Emitter-follower
Differential input stage
Bandwidth independent
of gain
gain * bandwidth = const.
Gain rolls off earlier
𝑈𝑜𝑢𝑡
~
𝑈𝑖𝑛 1 +
1
𝑅 +𝑅
1
∙ 𝐹𝑅 𝐺
𝐴𝑂𝐿 (𝑗𝜔)
𝐺
27.08.2012, A new camera concept for Cherenkov telescopes, A. Gadola, UZH
Gain rolls off same
𝑈𝑜𝑢𝑡
~
𝑈𝑖𝑛 1 +
1
1
∙𝑅
𝑍𝑇𝑅 (𝑗𝜔) 𝐹
21