APD Module Development @ IU

1

APD Module Development _at_ IU B. Adams, F. Busch,
P. Childress, W. Fox, S. Mufson, J. Urheim, G.
Visser with much input from T. Chase, L.
Mualem August 2006 Collaboration Meeting, FNAL, -
Electronics breakout

• System Overview
• Summary of Earlier Design / Thermal Testing
  Efforts
• Developments Since Intl Falls Mtg.
• Open Questions
• Plans for the Coming Months

2
Requirements Goals

• Reqs for Photodetector Interface / APD Module
  Enclosure
• Optical Alignment
• Fibers must be aligned w/ APD pixels (both
  transversely longitudinally)
• I.e., module must serve as an alignment jig for
  Cookie holding the fiber-ends
• Cooling
• APDs must be cooled to -15 C
• Environmental
• Must be impervious to leakage of light
  humidity
• Electronic Interface
• APD signal, bias lines, thermal sensor, TEC
  power connections to FEB.
• Our current goals
• Finalize design so that production can commence
  on 500 enclosures for IPND
• in short term, fabricate 10 prototypes matched
  to prototype APD/carrier board assemblies.
• work on interface with scintillator module
  manifold snout (w/ T. Chase)

3
Additional Requirements

• Flip-chip mounting of APD on 0.060-in circuit
  board.
• We call this the carrier board. It has
  cutouts aligned w/ APD pixels.
• APD is cooled directly by a Peltier Effect TEC.
• The TEC must not apply significant mechanical
  stress to the APD array
• For this reason we deploy a deformable crush
  pad between TEC APD
• The hot side of the TEC is thermally connected
  to a heat sink
• 15 C water-cooling is the current baseline.
• TEC mean time to failure 23 yrs, so failure
  rate 3 / day ?
• The TECs will need to be swappable.

4
Preliminary Design

• Clam-shell design for APD Module Enclosure
• Injection-molded pieces surround (most of)
  carrier board,
• creates gas volume to keep APD dry and dark,
• and limits heat conduction paths.
• Our optical connector (cookie) is actually an
  entire 1/2 clam-shell.
• It will be permanently fastened to the
  scintillator module manifold snout
• The clamshell design simplifies alignment and
  establishment of a good gas/light seal for
  environmental isolation
• The other half of the clamshell consists of a
  heat sink holder plus heat sink /epoxy /TEC
  /crushpad stack assembly.
• We have investigated both air- water-cooled
  heat sinks
• Focusing on 15 C water, since thats the
  current baseline

5
APD Module TEC side

• Heatsink holder
• Heatsink

TEC glued here
Water inlet/outlet
Hole (1 of 2) in wall of heatsink for TEC leads
6
Prototype heatsinks

• Three of the prototype heatsinks used for testing.

Water inlet/outlet will be plastic - i.e., with
cap will be a single injection molded piece
Bored out aluminum stock - cheap to fabricate
7
APD Module Optical Connector
8
APD Module Optical Connector
Hole for x-y alignment pin, matching holes in
carrier board
Each line is actually two fiber ends, side by side
Carrier board sits against this surface Note
o-ring.
This design meets alignment specifications
minimizes heat conduction paths but is not well
suited for easy threading of fibers
These surfaces are faced off together. Provides
registration in z to APD side of carrier board
2nd half of clamshell abuts this surface
9
APD Housing/Heatsink testing

• cookie side with insulation removed
• APD housing assembly with insulation removed

10
Understanding TEC Heat Load

• Keep Cold Side at 0o C
• Let heat sink temp float
• See how much current needed to attain given DT.
  Open circles are our data.
• From this can infer Heat Load from TEC specs for
  this Tcold
• This turns out to be 0.55 W.
• Also see how much power is consumed by TEC
• For DT 30 C, this is about a watt
• Of course heat load is different at -15 C
• But TEC operates as advertized, at least at 0 C
  

11
Results w/ water-cooled heatsink
RTD Heatsink Diff Amps Volts power IV
-15 0 15 0.76 0.88 0.67
-15 0.5 15.5 0.79 0.91 0.72
-15 3.5 18.5 0.88 1.03 0.91
-15 4.8 19.8 0.94 1.1 1.03
-15 6 21 0.98 1.15 1.12
-15 9.2 24.2 1.14 1.35 1.53
-15 9.3 24.3 1.14 1.37 1.57
-15 9.5 24.5 1.16 1.38 1.60
-15 9.7 24.7 1.16 1.38 1.60
-15 12 27 1.28 1.54 1.97
-15 15 30 1.60 1.93 3.09
-15 15 30 1.60 1.89 3.03
Aim to run APD at -15 C TEC is linear in
current required up to DT 30 C (see previous
slide) We measure 1.60 A for this configuration,
thus heat load is 2.5 W (see X on
previous slide) Must supply 3.0 W to TEC This
means total heat to remove is 3.0 2.5 5.5
W This can be done w/ water _at_ near 15 C at 2
cc/s
12
APD Housing/Heatsink testing

• wet end of insulated APD test setup
• cookie end of insulated APD test setup.

13
APD Housing/Heatsink testing

• Insulated APD housing test unit wrapped with foil.

14
Impact of Insulation
Power into TEC
3.4 Watts
3.2 Watts
2.7 Watts
Water flow 2cc/sec
Full Insuln
No Insuln
Half Insuln
Time (minutes)
15
Developments Since Intl Falls

• Turn attention toward interface w/ manifold
  snout
• Meeting w/ Tom Chase at IU, June 27
• o Holes versus Slots for fibers ? General
  threading considerations ?
• o How to glue fibers ? (note can use
  wicking glue if slots)
• o How to affix cookie to manifold snout ?
• o Insulation issues ?
• o Clearance issues ?
• Suggestion that APD module should not be wider
  than depth of scint mdl.
• o This means the carrier board width should
  shrink 50.8mm --gt 40mm
• also reduces warping of board, beneficial
  for Hamamatsu for APD mounting
• Aug. 2 receive suggestion from Tom for glue
  slot in cookie.
• o Requires additional injection molded piece
  w/ hole pattern (see drawing).
• Other developments
• Concerns for stresses to APD
• o Leads to notion of mylar tape shim applied
  to cookie after flycutting
• o Also redesign of heat sink holder addn of
  IM spacer frame modifcn
• heat sink (holder) to set height of TEC.

16
New Optical Connector (Walt)
Note cookie is no longer deep, for ease of
threading. Potential thermal impact of this is
unknown
Shims would go here
17
Manifold Snout / Cookie xface
Manifold snout bottom piece end
Groove for insertion of cookie
Tab on cookie
18
Glue Slot in Cookie (Tom C.)
19
New Design for TEC Side
Hole for alignment pin
Heat sink assembly
Cookie
Spacer Frame
20
New Heatsink Holder
Heatsink assembly
Holder / Spacer frame (glued to carrier board)
21
Plans for the coming months

• Prototype fabrication
• Aim to build one or two prototypes now for
  thermal studies
• Once we get APD/carrier board assemblies from
  Hamamatsu, plan is to machine 10 to be
  distributed.
• A new technician, Fritz Busch, has joined our
  group will help with this effort.
• Outstanding questions
• Desiccant
• Must keep relative humidity to 3.4 to avoid
  condensation!!
• But where to put desiccant and associated
  installn issues