Upload
truongkhuong
View
212
Download
0
Embed Size (px)
Citation preview
Updates on the Large Area Picosecond Photo-Detector (LAPPD)
Project and Integrated Readout Electronics
Gary S. VarnerUniversity of Hawai’i at Manoa
5-OCT-2012 AAP Meeting in Hawaii
Old ways die hard…
2
“Flat panel display” vs. CRT
Photocathode
Input photons
MCP1MCP2
Anode
Readout Electronics
3
Starting point is a Micro-Channel
Plate Photo-Multiplier Tube
4
Pushing limits on multiple frontiers: timing, volume, & cost.
The LAPPD Collaboration
A diverse group, consisting of:• National laboratories• Universities• Private companies
• Combining the strengths and resources of many
More information at:http://psec.uchicago.edu
5
∑>>i
iPartsTotal
As of April 2009 (original Proposal)Table of attribution provided at end
Elements of an MCP-PMT
• Photocathode• Micro-channel plates• Collection anode• Readout electronics• Mechanical design / tile
assembly.
Active research is ongoing for all elements… there are a lot of impressive achievements, will try to hit highlights (and hopefully not get lost in technical detail)
Photocathode
Input photons
MCP1MCP2
Anode
Readout Electronics
6
Elements of an MCP-PMT
• Photocathode• Micro-channel plates• Collection anode• Readout electronics• Mechanical design / tile
assembly.
Photocathode
Input photons
MCP1MCP2
Anode
Readout Electronics
7
Photocathodes
12/16/2010
Move forward on multiple fronts…
8
Attrib: Klaus Attenkofer
Alchemy??
9
Even for well established vendors, difficult to always make the “tried and true” recipe work
Why not apply recently available tools to make a science out of photocathode fabrication ??
Measurements courtesy Nagoya U (Belle II PID group)
A Portfolio of Risk
Multiple, parallel R&D approaches:1. Pursue the fundamental science behind what makes photocathodes tick 2. Work to scale conventional bi/multi-alkali technology to large sizes3. At the same time, investigate novel photocathode concepts (III-V)
• …while simultaneously needing to remain cognizant the need to integrate & move to mass production 10
Modern Porfolio Theory(cf: wikipedia)
The science: characterization during growth @ ANL/BNL
11
Movable optical station
X-ray, AFM Facility- Visualization of growth and activation process (APS, BNL)
Photocathode Characterization
LED light source and fiber optics
“chalice”
See references for recent publications
Perfecting the Known: Scaling up to 8” @ SSL
12
Good Q.E., uniformity
Scale up to larger chamber
Exploring the Future: III-V materials and Field Guiding
Extend to longer wavelengths –utilize commercial semiconductor fabrication infrastructure?
13
Electric Field
Delivering Today: so that there is a future
A penultimate dream 24” x 16” integrated module:1. Within the portfolio, engage industrial partners2. If there is a market, they will come3. Collaboration members concentrate on what they do best
14
Elements of an MCP-PMT
• Photocathode• Micro-channel plates• Collection anode• Readout electronics• Mechanical design / tile
assembly.
Photocathode
Input photons
MCP1MCP2
Anode
Readout Electronics
15
Micro-channel Plates
• Conventional MCPs:– Drawn/sliced lead-glass fiber bundles.– Chemical etching & heating in hydrogen to improve emissivity.– Expensive!
• LAPPD approach:– Use atomic layer deposition (ALD) on low-cost substrates.
• Allows precision control over thicknesses, down to single atomic layers.• Can be used with a large variety of materials.• Potentially significant cost savings.
16
Atomic Layer Deposition (ALD)
• A conformal, self-limiting process
• Allows atomic level thickness control
• Applicable for a large variety of materials
J. Elam, A. Mane, Q. Peng, T. Prolier (ANL:ESD/HEP), N. Sullivan (Arradiance), A. Tremsin (Arradiance, SSL)
17
ALD Activation of Glass Capillaries
12/16/2010
Side view
Top view
1) Begin with insulating glass capillaries
2) Use ALD to apply a resistive coating.3) Use ALD to apply an emissive coating.
18
Performance of ALD MCPs• Successful functionalization of glass capillaries,
optimizing SEY materials
Gain characterization
190 200 400 600 8000
1
2
3
4
5
Elec
tron
Gain
(sec
onda
ries/p
rimar
y)
Primary Electron Energy (eV)
MgO 40Å 30Å 20Å
Al2O3
40Å 30Å 20Å
Aging effects
20
SL-10 (w.o/ & with Al protection layer)
K. Inami et al (Belle II PID group)
MCP1MCP2
Ion bombardment
photocathode
20um poreALD coated
This is huge – lifetime is a major MCP-PMT limitation
Not temperature controlled
Matching gain with theory/simulation
21
• Working to develop a first-principles model to predict MCP behavior, at device-level, based on microscopic parameters
• Will use these models to understand and optimize our MCP designs
Valentin Ivanov, Zeke Insepov, Sergey Antipov, Nucl. Instr. Meth. A639 (2011) 158-161.
Elements of an MCP-PMT
• Photocathode• Micro-channel plates• Collection anode• Readout electronics• Mechanical design / tile
assembly.
Photocathode
Input photons
MCP1MCP2
Anode
Readout Electronics
22
Stripline Anodes (Prototype)
• Photonis-Planacon on transmission line PCB:
• Striplines allow coverage of a large area with a manageable number of channels.
23Courtesy Fukun Tang, Greg Sellberg
Stripline Anodes Measurement• Average time ((t1+t2)/2) along strip gives arrival time
• Time difference (t1-t2) gives the position
σt of order ps feasible for large Npe 24
2525
StriplineReadout
-- silkscreen on glass
2 GHz
1 GHz
• 30 anode strips• 40 anode strips
Ana
log
band
wid
th
0.5 GHz
Anode strip length [cm]
10 20 30 70
Elements of an MCP-PMT
• Photocathode• Micro-channel plates• Collection anode• Readout electronics• Mechanical design / tile
assembly.
Photocathode
Input photons
MCP1MCP2
Anode
Readout Electronics
26
Readout Electronics• Building on experience from existing devices &
readouts.– Readout based on waveform sampling
– Requirements of the readout vary significantly by application
27
6.4 psRMS
CH1
CH2
NIM A602 (2009) 438
σ ~ 38 ps
Single p.e. resolution
4x4 anode “1 inch”MCP-PMT (HPK SL-10)
28
The single threshold is the
least precise time extraction
measurement. It has the advantage
of simplicity.
Single threshold
The multiple threshold method takes into account the finite slope of
the signals. It is still easy to implement.
Multiple threshold
The constant fraction algorithm is very often used
due to its relatively good performance and its simplicity.
Constant fraction
The waveform sampling above the Nyquist frequency
is the best algorithm since it is
preserves the signal integrity.
Waveform sampling
In principle, sampling above the Nyquist-Shannon frequency and fully reconstructing the signal attains the best timing information.
28
Timing Extraction Methods
Attrib. Jean-Francois Genat
• The four algorithm models have been simulated.
• In principle, pulse sampling gives the best results.
• To realize this performance, sampling frequency is taken to be 2× the fastest harmonic in the signal: 10Gs/s.
Timing Extraction SimulationsHow to get to picosecond timing
From Jean-François Genat*Nucl.Instrum.Meth.A607:387-393,2009
29
Sampling Rate 2.5 GSa/s-17GS/s
# Channels 6 (or 2)
Sampling Depth 256 (or 768) points
Sampling Window Depth*(Sampling Rate)-1
Input Noise <1 mV RMS
Analog Bandwidth 1.5 GHz
ADC conversion Up to 12 bit @ 2GHz
Latency 2 µs (min) – 16 µs (max)
Internal Trigger yes
PSEC-3 + PSEC-4 ASICs 4.3mm
4.0mm
30
31
Surface charge induced on the strip as a function of time and position
31
PSEC-4 measurements with an 8” MCP
+6mm pos.
50% Const. Frac.Discriminator
PSEC-4
Dual-ends of 8” MCP w/ PSEC-4 @ 10 Gsa/s
-- left anode strip-- right anode strip
Modular Readout System architectureRespin Analog card only for different ASIC
A variety of data collection configurations possible
Demonstrate timing distribution @ pslevel between modules
32
5/20/2011
ASIC Amplification? # chan Depth/chan Sampling [GSa/s] Vendor Size [nm] Ext ADC?
DRS4 no. 8 1024 1-5 IBM 250 yes.
SAM no. 2 1024 1-3 AMS 350 yes.
IRS2 no. 8 32536 1-4 TSMC 250 no.
BLAB3A yes. 8 32536 1-4 TSMC 250 no.
TARGET no. 16 4192 1-2.5 TSMC 250 no.
TARGET2 yes. 16 16384 1-2.5 TSMC 250 no.
TARGET3 no. 16 16384 1-2.5 TSMC 250 no.
PSEC3 no. 4 256 1-16 IBM 130 no.
PSEC4 no. 6 256 1-16 IBM 130 no.
Success of PSEC3: proof-of-concept of moving toward smaller feature sizes.• Next DRS plans to use 110nm; next SAM plans to use 180 nm.
Portfolio of options, leading the way
33
Elements of an MCP-PMT
• Photocathode• Micro-channel plates• Collection anode• Readout electronics• Mechanical design / tile
assembly.
Photocathode
Input photons
MCP1MCP2
Anode
Readout Electronics
34
An ideal (integrated) Photodetector
• DC power in, fiber optic out• Integrated photon electrons T,Q data out
35
36Flat panel sealed glass is the ultimate solution
Elements of an MCP-PMT• Photocathode• Micro-channel plates• Collection anode• Readout electronics• Mechanical design / tile assembly.Photocathode
Input photons
MCP1MCP2
Anode
Readout Electronics
37
•High speed laser•Diagnostic UV beam•laser power monitoring•Position scanning
•MCP testing setup at ANL
(EFI/ANL)-fs laser at APS38
Next steps – outfit various test benches
Argonne 8” test chamber
Argonne 33mm test bench
Hawaii ps-Laser &Cross-talk characterization bench
SSL 8” XDL measurement + many others…
39
Electronic test benches as well
Mock tile Configurable strips
Coupling tests with readout boardsAnd ASICs, as well as SMA tests
40
First article Status (2-OCT-2012)
Photocathode
Input photons
MCP1MCP2
Anode
Readout Electronics
41
>20% QE on 8” B33 glassMost pins <10^-9Almost there!
Innovations/Achievements
42
After 3 very intense, productive years of R&D, the LAPPD Collaboration is poised to deliver a first 8” MCP-PMT• A major technological achievement
• Compared with a 2” tube (16 times area!)
• Improved MCP lifetime (no need to “scrub”)
• Compact readout with mm spatial and TTS-limited timing
• Lessons learned• Promising directions for future photocathode work
• Simplified mechanics/construction toward mass assembly
43
Summary
Back-up Slides
44
Personal Observations• LAPPD is an example of the innovation and
R&D breakthroughs can arise when a well-posed, targeted problem is undertaken by a diverse group of scientists who are willing to work together and learn each other’s language
• First 8” MCP-PMT is a culmination of this structured program, where continual collaboration dialog and internal reviews and external collaboration was extremely helpful
• A model for detector R&D in the future45
Where credit is due (part 1)
46
Slide #
Primary Contributor(s)[necessarily incomplete]
Source Cited Publication/ Comment
3 Matt Wetstein RICH 2010 Conf talk
8 Henry Frisch Myriad presentations, as guiding principle for Collaboration activities
14 Junqi Xie, Marcel Demarteau, Henry Frisch, Seon Lee,Alexander Paramonov, Robert Wagner, Zikri Yusof, Klaus Attenkofer
Instrumentation for Theory-Inspired Photocathode Dev. Within the LAPPD Project
To appear TIPP 2011Proceedings (ID 473)
14 Seon Lee, Marcel Demarteau, Henry Frisch, Junqi Xie, John Smedley, Zikri Yusof, Klaus Attenkofer
Revealing the Correlations between Growth Recipe and Microscopic Structure of Multi-alkali Photocathodes
To appear TIPP 2011Proceedings (ID 443)
14 Seon Lee, Marcel Demarteau, Henry Frisch, Junqi Xie, Dean Walters, Zikri Yusof, Klaus Attenkofer
Theory and Applications of Transmission Mode Metal (Aluminum) Photocathode
To appear TIPP 2011Proceedings (ID 438)
Where credit is due (part 2)
47
Slide #
Primary Contributor(s)[necessarily incomplete]
Source Cited Publication/ Comment
15,22,23, 41
Ossy Siegmund, Jason McPhate Communication of ongoing work at Berkeley Space Science Laboratory
16 Ryan Dowdy, Seon Lee, Klaus Attenkofer
Large area transmission mode NEA GaAs photocathodes for sub-400nm wavelength operation
To appear TIPP 2011Proceedings (ID 423)
16 Jim Buckley, Dan Leopold Communication of ongoing work at Washington Univ. St. Louis
J. Appl. Phys. 98, 043525 (2005)
22 Slade Jokela, Jeff Elam, Anil Mane, Qing Peng, Igor Veryokvin, Alexander Zinovev
Composition and thickness dependence of electron-induced secondary electron yield for MgOand Al2O3 from atomic layer deposition
To appear TIPP 2011Proceedings (ID 256)
27 Kurtis Nishimura, et al. Communication of ongoing work at University of Hawaii
Where credit is due (part 3)
48
Slide #
Primary Contributor(s)[necessarily incomplete]
Source Cited Publication/ Comment
28 Herve Grabas, Razib Obaid, et al.
Communication of ongoing work at the University of Chicago
33,34 Eric Oberla, Herve Grabas A 4-Channel Waveform Sampling ASIC using 130nm CMOS technology (PSEC-4 updates based on continued development/test)
To appear TIPP 2011Proceedings (ID 219)
35 Mircea Bogdan, Henry Frisch, Jean-Francois Genat, HerveGrabas, Ed May, KurtisNishimura, Richard Northrop, Eric Oberla, Gary Varner
Design of a Data Acquisition System for Large Area, Picosecond-Level Photodetectors
TIPP 2011ID 51
40-42 Matt Wetstein, Slade Jokela, Bernhard Adams, MatthieuChollet, Valentin Ivanov, Zeke Insepov
Integration-Level Testing of Sub-Nanosecond Microchannel Plate Detectors for Use in Time-Of-Flight HEP Applications
To appear TIPP 2011Proceedings (ID 253)
40-42 And a cast of many ANL, U Chicago, U Hawaii
Work I couldn’t squeeze in (LAPPD TIPP2011 papers)to be published
49
Topic Primary Contributor(s) TIPP11 ref
A novel atomic layer deposition method to fabricate economical and robust large area microchannel plates for photodetectors
Anil Mane ID 457
Steps Towards 8”x8” Photocathode For The Large Area Picosecond PhotodetectorProject At Argonne
Zikri Yusof ID 60
Multipurpose Test Structures and Process Characterization using 0.13µm CMOS: The CHAMP ASIC
Michael Cooney, Larry Ruckman, Kurtis Nishimura, Eric Oberla, Matt Andrew, Jean-Francois Genat, Gary Varner
ID 214
A correlation-based timing calibration and diagnostic technique for fast digitizing ASICs
Kurtis Nishimura, Andres Romero-Wolf
ID 193
Large Area MCPs
Large Area MCP structure:- Custom photo-cathode
- 2-plate chevron (high gain)
- Transmission line 2D readout
(limit the number of electronics channels)
Electronics:- GSa/s Waveform Sampling +
Signal processing
Jean-Francois Genat, LAPPD Electronic Review, Chicago, May 20th 2011 50
Examples of Timing Algorithm Studies…
*for details, see SLAC-PUB-14048 (also to appear in NIM A)
Comparison of two algorithms*:1. “Reference”(~template fitting)
2. Constant fraction
• Performance demonstrated on two different waveform sampling ASICs.
On both ASICs, the reference method performed slightly better than CFD.
The difference was small… for many applications CFD may be enough.
51
MCP signal development
MCP signal rising edge: qE = ma l = 1mm, E=100V/mm, tr=250ps
Jean-Francois Genat, LAPPD Electronic Review, Chicago, May 20th 2011
- First gap: 10ps
- Assume 2-stage pores TTS of 20ps
- Anode gap: 10ps
- Noise: 20ps
Total is 32ps
(Measured Burle-Photonis 2’’ x 2’’ is 30ps)
52
S = G Npe
The Factors that Limit Timing Resolution, University of Chicago, April 28-29, 2011
Transit Time NoisedetectorRise time Gain (Signal/noise)
Electronics:NoiseelecSample rateAnalog bandwidth
σn = σn det + σn elec
Understanding what matters
53
Picosecond timing and 2D position for large area detectors: delay lines and Waveform Sampling
Delay line readout and pulse sampling provide fast timing (2-10ps). Delay lines should have a signal bandwidth matched to the detector
Fewer electronics channels for large area sensors
t1, a1t2, a2
½ (t1+ t2) = timev(t1-t2) = longitudinal positionΣ αi ai / Σ αi = transverse position
Pico-second electronicsPico-second electronics
3.8 ps translates in 190 µm position resolution with 50 photo-electrons
The electronics contribution to spead is small:3.8ps / sqrt(2) = 2.7ps,
For 50PEs, MCP is 30ps / sqrt(50) = 4.2ps, equal at 100 PEs
3.8ps
54
•Stefan RittApril 28th, 2011 Timing Workshop, Chicago
How is timing resolution affected?
dBs ffUut
331⋅
⋅∆
=∆
U ∆u fs f3db ∆t100 mV 1 mV 2 GSPS 300 MHz ∼10 ps
1 V 1 mV 2 GSPS 300 MHz 1 ps
100 mV 1 mV 20 GSPS 3 GHz 0.7 ps
1V 1 mV 10 GSPS 3 GHz 0.1 ps
•today:
•optimized SNR:
•next generation:
•includes detector noise in the frequency region of the rise time
•and aperture jitter
•next generation•optimized SNR:
•How to achieve this?
•Assumes zeroaperture jitter
Stefan Ritt slide
UC workshop 4/11
Copper strip testbench
Godparent review - Hervé Grabas
Simulation
Measurement – good agreement
56
Components of a Waveform Sampling ASIC
Chicago Hawaii ASIC, Multi-Purpose (CHAMP)• Develop new circuit elements, train additional students in the
IBM 130nm process
5/20/2011 Nishimura - CHAMP & ASIC Options
12-bit DAC
PSEC3 VCDL
CHAMP VCDL
CHAMP low Vt VCDL25GSa/s
Many useful building blocks, will allow increased functionality, more compact implementation
58
5/20/2011PSEC-3 Noise, linearity, bandwidth• Input noise
ranges ~1-1.5 mV RMS
• Dominant source from analog buffers Issue addressed
in PSEC-4 design mV
Due to charge injection, correctable
Eric Overla LAPPD Electronics + Integration GPC review
Input bus layout problem –high R, parasitic C
PSEC-3 already very usable, PSEC-4 will be even better
59
Oscilloscope on a chip? Calibration
~= ??
60