E' F' Eikenberry - PowerPoint PPT Presentation

1 / 24
About This Presentation
Title:

E' F' Eikenberry

Description:

Pixel: Comparator and trimming. Paul Scherrer Institut. Ch. Br nnimann, SLS ... Threshold discriminator with 4-bit trim. 15-bit counter 1 MBq x-ray rate. CHIP: ... – PowerPoint PPT presentation

Number of Views:34
Avg rating:3.0/5.0
Slides: 25
Provided by: eiken
Category:

less

Transcript and Presenter's Notes

Title: E' F' Eikenberry


1
PILATUS A 2-D Pixel DetectorFor Protein
Crystallography
E. F. Eikenberry1, Ch. Brönnimann1, S.
Kohout1, B. Schmitt1, G. Huelsen1, H. Toyokawa2,
R. Baur3 and R. Horisberger3 1 Swiss Light
Source, Paul Scherrer Institut, CH-5232
Villigen-PSI 2 SPring8, Japan 3 CMS-Project, Paul
Scherrer Institut, CH-5232 Villigen-PSI
E. F. Eikenberry
2
SLS Layout
Green first phase beamlines Red proposed
beamlines
E. F. Eikenberry
3
X-Ray Diffraction From A ?X-174 Virus Crystal
E. F. Eikenberry
4
Macromolecular Crystallography
Determine the structure of highly complex
molecules to atomic resolution
Molecular Weight gt 1 MDa (gt 50,000 atoms)
Applications
Enzymes Viruses Catalytic mechanisms Electron
transport Energy transduction Motility Drug
Design Structural genomics
E. F. Eikenberry
5
Braggs Law
Diffracted rays
Incident rays
?
?
d
Condition for constructive interference 2d
sin(?) n?
Experimental Setup
Diffracted rays
X-ray beam
Detector plane
2?
Rotating crystal
E. F. Eikenberry
6
Using X-ray Diffraction to Determine Molecular
Structure
Structure factors, F(hkl), are given by
where x, y, z coordinates in unit cell,
?(xyz) electron density distribution, h, k,
l coordinates in reciprocal space, and the
integral extends over 1 unit cell
The electron density distribution in the unit
cell, which represents the desired structure, is
obtained by
The observed intensities in the diffraction
pattern are given by
I(hkl) F(hkl)2
E. F. Eikenberry
7
Interpretation
Convolution makes a crystal
X-rays take the Fourier transform of the crystal
The Phase Problem
The pattern is actually x-ray intensity, not the
complex amplitude required to calculate the
desired unit cell structure. Many methods are
used to overcome this.
E. F. Eikenberry
8
  • SLS Pixel Detector Goals
  • Protein crystallography
  • Small Crystals
  • Large unit cells
  • Energy range 5 - 17.5 keV
  • Intensity 1013 x-ray/s on sample
  • Spot 15 ?m x 25 ?m
  • Divergence 150 ?rad x 30 ?rad
  • Rapid tunability
  • Fine slicing data collection
  • Detector Requirements
  • Resolve gt 500 orders
  • Large detector area
  • Good PSF
  • Large dynamic range
  • High accuracy
  • Fast readout
  • Radiation tolerant

E. F. Eikenberry
9
The Perfect X-Ray Detector
  • Quantum efficiency (at given energy)
  • Event rate
  • Signal-to-noise
  • Linearity
  • Saturation
  • Dynamic Range
  • Detector size
  • Spatial resolution
  • Energy resolution
  • Stability of calibration
  • Radiation damage susceptibility
  • Ease of use
  • Speed Throughput
  • Cost

It all depends on the experiment!
E. F. Eikenberry
10
Paul Scherrer Institut
The SLS Pixel Detector
  • Size 40 x 40 cm2 (0.16m2)
  • 2000 x 2000 pixels
  • Pixel size 200 x 200 mm2
  • Modular detector -gt dead area 6
  • High frame rate gt10Hz
  • High duty cycle gt94 (Tro6ms)

Base plate (water cooled)
Bank with 5 modules
Tilt angle variable
Modules with 16 chips
E.F. Eikenberry
11
Paul Scherrer Institut
SLS Pixel Module
Ch. Brönnimann, SLS
12
Paul Scherrer Institut
Pixel Preamp-Shaper
Ch. Brönnimann, SLS
13
Paul Scherrer Institut
Pixel Comparator and trimming
  • AC-coupled comparator with diode feedback
  • simple inverter as comparator
  • comparator biasing done via pn-junction of the
    diode -gt
  • expect lower threshold variations on the chip
  • radiation insensitive
  • low power consumption

Ch. Brönnimann, SLS
14
Bump-bonded module See landscape supplement
15
SLS Pixel Chip
217 x 217 ?m2 pixels 44 x 78 3432
pixels/chip Chip 10 x 20 mm2 Fabricated in
DMILL 0.8 ?m process 6 wafer yields 64 chips
First batch of wafers had processing defects gt
we dont have to pay gt thats all we have to
work with 0 goodchips 10 to 50 interesting
chips/wafer Left half of wafer is dead in most
cases 2 bad pixels
Second batch of wafers is in hand and being
preprocessed
16
Bump-Bonded Module
Indium bump-bonded 8 x 2 chips per sensor Sensor
366 x 157 57,462 pixels Sensor is 8 x 3.6
cm2 Double-size pixels between readout chips
give the sensor continuous coverage
90Sr test - shows almost all pixels good 55Fe
test - shows a range of defects 0 defects (1
chip) 4 defects (1 chip) many defects (most
chips)
17
366 x 157 Pixel Module
E. F. Eikenberry
18
Protein Diffraction Pattern See landscape
supplement
19
Trimming a Single Chip See landscape supplement
20
Trimming A Module With A Global Vcmp
E. F. Eikenberry
21
Threshold vs. Vcmp
E. F. Eikenberry
22
Retrimming A Module
E. F. Eikenberry
23
SLS Detector Architecture See landscape supplement
24
SLS 2k ? 2k Pixel Detector
PIXEL 217 ? 217 ?m2 Preamplifier and
shaper Threshold discriminator with 4-bit
trim 15-bit counter gt1 MBq x-ray
rate CHIP CMOS 0.8 ?m DMILL rad-hard Designed
at PSI 44 ? 78 pixels, 9.9 ? 18.3 mm2 Read out
at 10 MHz MODULE Fully depleted diode array 81
? 36.6 mm2 ? 300 ?m, 57,462 pixels Indium
bump-bonded to 16 chips Chips read out in
parallel (5 ms) Module control board BANK 5
modules Bank control board caches
data DETECTOR 12 banks Adjustable
V-shape With fine-slicing, a 180? rotation may
produce 2000 8 MB images (16 GB) in 3 min.
E. F. Eikenberry
Write a Comment
User Comments (0)
About PowerShow.com