Cooling issues
Fred Hartjes
NIKHEF
Nikhef/Bonn LepCol meeting
May 8, 2017
Thermal properties
Coca glued on intended way to stump
1.3 ml Araldite 2020 (low viscosity) + 1 g
Boron-Nitride
=> paste like substance
Sufficient glue to fill all holes
Dummy flexes included
Fred Hartjes
Nikhef/Bonn LepCol meeting. Nikhef. May 8, 2017
2
Re-measurement using thermal glue
Stump thermally coupled to cold plate
Water cooled
4 heaters on coca
2.5 W each
1 heater on stump (2.5 W)
Simulating LV regulation
Fred Hartjes
Nikhef/Bonn LepCol meeting. Nikhef. May 8, 2017
3
Module isolated (on Styrofoam)
Coca temperature up to 105 °C
Stump follows coca
•
ΔT = 10 °C
Fred Hartjes
Nikhef/Bonn LepCol meeting. Nikhef. May 8, 2017
Thermistor list
• White => Coca
• Red => Coca
• Green => Stump
• Blue => Cold plate
4
Coca temperature reduced
105 => ~56 °C
Module on cold plate
•
Stump well cooled
ΔT = 5 °C
Biggest thermal resistance in
joint Coca – Stump
ΔT = 27 °C
Fred Hartjes
Nikhef/Bonn LepCol meeting. Nikhef. May 8, 2017
Thermistor list
• White => Coca
• Red => Coca
• Green => Stump
• Blue => Cold plate
5
Thermal resistance of glued TPX3 chip
Assume thermal resistivity of glue of 2500 K/W.mm (Araldite 2011)
Glue parameters
Surface: 200 mm2
Thickness 0.1 mm
=> thermal resistance 1.25 K/W
2.5 W => ΔT ~ 4 °C
We may use Gloptop
UV curing => only the edges
Thickness << 0.1 mm
=> ΔT << 4 °C
Hopefully the thermal resistance of chip to coca is not really an issue
Fred Hartjes
Nikhef/Bonn LepCol meeting. Nikhef. May 8, 2017
6
Conclusions thermal issues
Stump well cooled by cold plate
We may not need a cooling line in the
Stump
Weakest point: joint Coca to Stump
Presently ~ 60 °C chip temperature
Improvements expected by
Filling holes Coca – Stump
Enlarging Stump
Soon to be verified
Still for the present design we have to
live with 50 °C chip temperature
Not constant, greatly depending on RO
rate
Significant improvement only expected
by major changes in mechanical design
Fred Hartjes
Nikhef/Bonn LepCol meeting. Nikhef. May 8, 2017
7
Quad assembly
All PCBs / flexes due to arrive any
moment
Bonding board + flexes + LV regulator
HV board
Guard plate
concentrator
Modified alignment block produced,
tested and black anodized
Final drawings coca and stump
completed
In production
Includes filled hole between coca and
stump
Finished May 12?
Fred Hartjes
Nikhef/Bonn LepCol meeting. Nikhef. May 8, 2017
8
Pickup blocks
Pickup blocks (for chips with
InGrid) still bit problematic
Big contact area to grid
Surface structure bit rough
Modified design (aluminium) just
fabricated
Contact surface to grid greatly
reduced
Just been fabricated
Assembly jig for gluing coca onto
stump still to be designed
Waiting for PCBs
Fred Hartjes
Nikhef/Bonn LepCol meeting. Nikhef. May 8, 2017
9
SPARE
Fred Hartjes
Nikhef/Bonn LepCol meeting. Nikhef. May 8, 2017
10
Thermal conductivity glues
Two aluminum plates 20 x 20 mm
Equipped with NTCs and one heater
1 mm wide gap filled with glue
Heater power adjusted getting 15 C
temperature difference between the
plates
Measurements
Araldite 2011: 2500 K/W.mm
Araldite 2020: 2900 K/W.mm
1.3 ml Araldite 2020+ 0.69 g Boron
Nitride: 760 K/W.mm
Fred Hartjes
Nikhef/Bonn LepCol meeting. Nikhef. May 8, 2017
11