2D CHARGE READOUT

1
2D CHARGE READOUT

• Chamkaur Ghag
• School of Physics
• University of Edinburgh

2
DRIFT I
DRIFT Collaboration
BDMC

• Directional Recoil Identification From Tracks
  detector at Boulby Mine
• 1 m3 cubic Negative Ion Drift TPC
• Central cathode at 20 kV, 270 V/cm drift field
• Electro-negative gas reduces diffusion
• MWPC 2D readout planes with an anode-pitch of 2mm
  and 1m2 area

3
DRIFT II

• DRIFT based on long range of expected WIMP
  induced nuclear recoils in low P gas
• Tracks reconstructed to give E loss recoil
  direction
• One of DRIFT II objectives to get better
  reconstruction by increasing spatial resolution
  of charge readout plane
• Tracks require sampling at narrower pitch than
  DRIFT I - not possible using MWPCs. Need
  different system

DRIFT Collaboration
4
GEMs
Avalanche occurs in holes etched in Kapton sheet
CERN Group
5
MICROMEGAS
Charpak Giomataris 1992

• Cu/Ni micromesh with holes 40microns square
  etched at a pitch of 50microns held above
  readout plane by Kapton spacers
• HV to mesh creates a large E field in the
  mesh/anode gap
• Avalanche occurs with high gains 104 between
  micromesh and readout plane. Acts as a
  parallel-plate avalanche chamber by inducing
  charge in strips.

6
Why MICROMEGAS?

• Can get high gain and spatial resolution
• Less likely to generate radiological background
  (made from Kapton, Cu, Ni)
• Robust, cheap, simple to make, resistant to spark
  damage (compared to GEMs)
• Can be tiled for larger areas - possible design
  for DRIFT II is 4 separate Micromegas detector
  readout planes each 25cm2 tiled together for
  total active area of 0.25m2

DRIFT Collaboration
7
CHARGE READOUT

• Cross strip grid on a Kapton PCB with all strips
  at 300micron pitch
• X direction 820 strips 100micron wide
  Y direction 820 strips 150micron
  wide
• Maintains equal exposed areas resulting in
  well-correlated charge sharing ? important for
  efficient detection on both coordinates
• PCB 25micron thick (Pyrolax flexible AP8525),
  standard copper wires 17.5microns thick can get
  lower

Universidade de Santiago
8
CODED GROUPS

• Large number of tracks and each track requires
  electronics (preamps, amps, digitizers, etc)
  ? EXPENSIVE!
• First and last 10 of 820 tracks have individual
  outputs to group into veto, calibration, testing,
  etc
• Remaining 800 coded grouped to minimize the
  number of output channels

9
SIMPLE CODED GROUP

• e.g., 9 wires read out with 6 outputs
• 9 wires grouped into three subsets for group
  signal on left
• Corresponding wires in each group connected for
  code signal on right
• Which of the group and code signals are triggered
  uniquely identifies wire
• Signal cannot propagate to outputs of other wires
  due to op-amps arranged as inverting amplifiers
• Grouping index is 2 each wire must have two
  connections to it (one group plus one code)

If c is no. of channels, N is no. of wires, then
T. Lawson
10
Better Coded Groups

• More efficient technique is to group the
  groups
• Means higher grouping index each wire is member
  of more groups hence connected to more outputs
• e.g., 32 wires are read out with 10 channels with
  grouping index i5
• 1-32 are wire numbers and A-J the 10 output
  channels. Each wire must be connected to 5
  separate output channels

c i N(1/ i)
11
CODED GROUPS OVERVIEW

• 800 wires split into three batches of 32, 256 and
  512 and grouped as separate detectors for testing
• N 32, i 2, c 12
• N 256, i 4, c 16
• N 512, i 3, c 24
• i of 512 lower than 256 so can remove effects of
  scaling up N to test efficiency of grouping
  scheme as function of i as well as N c
• Total number of output channels for the 800 wires
  is 52. Thus for 820 wires, 72 readout channels
• Can reduce total number of output channels for
  800 wires to 20 with a grouping index of 5 if
  grouped in one batch useful?

12
Connecting channels to wires

• Majority of 52 output channels must be connected
  to 64 separate tracks
• To physically do this, can have wire from output
  running across tracks extended out from the main
  active area of the grid
• Such extensions would add 3cm to the 25cm
• Output wire is connected to tracks with vias
  (shown as yellow points) to link output channels
  to their associated tracks as dictated by
  grouping scheme.

13
VIA Connections

• The track is widened into a circle where the
  connection with the output channel wire is to be
  made
• A hole is made through this and lined with
  5micron thick Tin-Lead solder
• layout design depends on size of such vias
  (including required clearance), i.e, if via
  diameter is greater than the 300micron pitch,
  they will not fit in the extended regions

PhotoMachining Inc
14
Non-Adjacent Grouping

• By coding and grouping all 800 tracks such that
  no 2 consecutive tracks are connected to the same
  output channel we can increase the available
  pitch to 500microns (20 thou)

15
LAYOUT DESIGN

• Standard vias 600microns wide. To accommodate
  these two options are
• (1)
• All four sides of detector utilized alternate
  tracks are read out on each side of the active
  area giving 600 micron pitch
• -Extended regions add 3cm to all sides
• -May cause complications with tiling
• -Output up from 72 to 104 due to repeated
  channels
• (2)
• In conjunction with non-adjacent grouping, the
  extended regions are fanned out for tracks to
  reach required pitch
• -Fanned region adds further 4cm

16
TILED READOUT
If fan out required
Extended region of flexible circuits can be
folded back
4 detectors can now be tiled for a 0.25m2 area
Can now halve number of output channels over 4
detectors by coupling outputs of adjacent sides
of separate detectors
17
NEXT
Vacuum Chamber with drift field
Opening front flange MICROMEGAS mounted on inside
connections out from rear flange
48cm
Drift
micromesh
120cm
strip readout

• 25cm2 readout grid as Micromegas to be
  manufactured and mounted in vacuum chamber at
  University of Sheffield with much of DAQ
  requirements at Edinburgh
• Readout mounted on inside of opening flange with
  connections out through rear flange cage fitted
  to define drift field

18
Tests

• Initial tests possibly to fire alpha source
  across the chamber for easily detectable
  charge-density tracks to establish
• ability to observe and reconstruct track shape
  dimensions
• effect on reconstruction efficiency and spatial
  resolution by changing lateral diffusion of
  charge cloud by altering drift field, drift
  distance, or the gas such as comparing CS2 with
  Arisobutane which has no diffusion suppression
• how well signal from individual wire is
  prevented spreading to outputs of other wires and
  feasibility of reduction of output channels from
  72 to 42 for 820 tracks
• efficiency of micromeshes including homemade
  woven wire with less action at points than
  manufactured mesh due to rounded edges

19
SUMMARY

• Charge readout strip development for MICROMEGAS
  with high spatial resolution
• 25cm2 readout grids with 820 tracks at 300micron
  pitch and tiling capability
• Tracks coded and grouped to reduce number of
  output channels to 72 for 820 tracks
• Grouped tracks connected with vias through
  extended regions to output channel wires
  running across
• Final design dependent on via size non-adjacent
  grouping scheme and fanning-out practical
  solutions
• Tests to be conducted in vacuum chamber to check
  feasibility of coded groups and setup for further
  development such as increased reduction in output
  numbers

20
Thanks

• Dr Tom Davinson1
• Dr Tim Lawson, University of Sheffield
• Dr Alex Murphy1
• Mr Colin Thomson1
• 1 University of Edinburgh