updates on the large area picosecond photo-detector (lappd ...hanohano/post/temp/test.pdf · scale...

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

http://psec.uchicago.edu/�

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



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



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.



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



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



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


14 Seon Lee, Marcel Demarteau, Henry Frisch, Junqi Xie, Dean Walters, Zikri Yusof, Klaus Attenkofer

Theory and Applications of Transmission Mode Metal (Aluminum) Photocathode



47

Slide #



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


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


27 Kurtis Nishimura, et al. Communication of ongoing work at University of Hawaii


48

Slide #



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)


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


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

updates on the large area picosecond photo-detector (lappd ...hanohano/post/temp/test.pdf · scale...

Documents