Jason McPhate, John Vallerga, Anton Tremsin and Oswald Siegmund

1
A noiseless, kHz frame rate, imaging detector
base on MCPs readout with a Medipix2

• Jason McPhate, John Vallerga, Anton Tremsin and
  Oswald Siegmund
• Space Sciences Laboratory, University of
  California, Berkeley
• Bettina Mikulec and Allan Clark
• University of Geneva

2
WFS detector for future AO systems

• kHz frame rates
• Match atmospheric timescales
• Many pixels - eventually 512 x 512
• More subapertures and more pixels per subaperture
• Very low readout noise (lt 3 e-)
• Lower penalty for more pixels per subaperture
• High (80) optical QE
• Use dimmer guide stars or higher frame rates

Angel et al., A Road Map for the Development of
Astronomical AO
3
Imaging, Photon Counting Detectors
4
Why would you want one?

• No readout noise penalty
• Use as many pixels as you wish
• Continuous temporal sampling to nsecs
• Choose integration period(s) after the fact or on
  the fly
• Other advantages
• Large area, curved focal planes
• Cosmic ray 1 count
• LN2 not required
• Low dark current (0.16 attoamps cm-2)

5
Whats the Catch?

• Global Counting Rates
• 1000 Shack-Hartmann spots per WFS
• Kilohertz feedback rates
• 1000 counts per spot for sub-pixel centroids

• 1 Gigahertz counting rate!

• Requires integrating detector

• Quantum Efficiency
• Historically Optical Photocathodes lt 15
• Silicon devices (CCDs) can get 90
• Noiseless helps, but not that much

• Requires GaAs Photocathode

6
Our AO detector concept

• An optical imaging tube using

• GaAs photocathode

• MCPs to amplify to 104

• Medipix2 ASIC readout

7
Medipix2 ASIC Readout

• Pixelated readout for x and gamma ray
  semiconductor sensors (Si, GaAs, CdTe etc)
• Developed at CERN for Medipix collaboration
• 55 µm pixel _at_ 256 x 256 (abutable to 512 x n x
  256).
• Pixel level amp, discriminator, gate counter.
• Counts integrated at pixel
• No charge transfer!

14mm
16mm
Applications Mammography, dental radiography,
dynamic autoradiography, gamma imaging, neutron
imaging, angiography, x-ray diffraction, dynamic
defectoscopy, etc.
8
Readout Architecture

• Pixel values are digital (14 bit)
• Bits are shifted into fast shift register
• Choice of serial or 32 bit parallel output
• Maximum designed bandwidth is 100MHz
• Corresponds to 286µs frame readout in parallel

3584 bit Pixel Column 0
3584 bit Pixel Column 255
3584 bit Pixel Column 1
256 bit fast shift register
32 bit CMOS output
LVDS out
9
First test detector

• Demountable detector
• Simple lab vacuum (10-7 Torr)
• UV sensitive, no photocathode

10
Lab Detector Lessons

• Medipix ASIC works well as MCP readout
• Sub-pixel centroiding of Shack-Hartmann like
  spots was achieved
• Optimized parameters for use in optical tube
• Chevron stack of 10 µm pore MCPs (protect cathode
  from ion feedback)
• MCP gain of about 104 (longer tube life and
  higher counting rates)
• MCP to Medipix gap of 300 to 500 µm (Medipix
  wirebond clearance)
• Approximately 1600 V rear field (minimize MCP
  charge cloud spread)

11
Vacuum Tube Design
No GaAs capability at UCB So GaAs photocathode
by industrial vendor Means using standard size
tube Only marginally larger than the Medipix2
device
12
Thick Film Ceramic Header

• Internal mounting/GND surface for Medipix
• Route 60 Medipix signals out of vacuum
• Multi-layered to better match Medipix pitch
• Maintain hermetic seal of tube to 10-9 Torr
• Provide land pads for external I/F connectors

13
Vacuum Tube Design
14
Vacuum Tube Design
15
Vacuum Tube Design
16
Medipix on a Header
17
MCP/Medipix Serial I/F Board
18
Vacuum Tube Design
19
Vacuum Tube Design
20
Vacuum Tube Design
21
Vacuum Tube Design
22
Vacuum Tube Design
23
Parallel Readout Design

• Development by ESRF
• 1 to 5 Medipix2 chips
• FIFO for each chip
• Flat field, deadtime corrections
• Optional centroid calculation
• High speed serial out

24
Future Work (3 yr. NOAO grant)

• Seal a MCP/Medipix tube with a GaAs photocathode
• Perhaps a multi-alkali photocathode tube (_at_UCB)
• Finalize and build parallel readout
• Test at AO laboratory at CFAO, U.C. Santa Cruz
• Test at telescope

25
Acknowledgements
This work was funded by an AODP grant managed by
NOAO and funded by NSF
Thanks to the Medipix Collaboration

• Univ. of Barcelona
• University of Cagliari
• CEA
• CERN
• University of Freiburg
• University of Glasgow
• Czech Academy of Sciences
• Mid-Sweden University

• University of Napoli
• NIKHEF
• University of Pisa
• University of Auvergne
• Medical Research Council
• Czech Technical University
• ESRF
• University of Erlangen-Nurnberg