Reworking of Indium bump bonded pixel detectors

Size: px

Start display at page:

Download "Reworking of Indium bump bonded pixel detectors"

Caroline Nash
6 years ago
Views:

1 Reworking of Indium bump bonded pixel detectors A.Airoldi 1, G.Alimonti 2, M.Amati 1, A.Bulgheroni 1, M.Caccia 1, D.Giugni 2, W.Kucewicz 3, F.Tomasi 2 1 Universita ' dell ' Insubria, Como, Italy 2 INFN Milano, Italy 3 University of Mining and Metalurgy, Kracow, Poland Reworking:first tests The chip stripping machine Conclusions: first results Sept/11/2002 G.Alimonti-Pixel

2 Reworking? For pixel detectors, due to the contact density and the required single bump failure (<10-4 ), the only interconnection technique is bump bonding on flipped chips Moving from a detector prototype to a full scale system requires a production chain optimization where hybridization plays a crucial role In current HEP applications up to 16 chips are attached to the same sensor; replacing a faulty chip ( reworking ) could considerably improve the yield. For instance in the DELPHI pixel detector the module production yield after bonding of 16 known good chips was ~80% and a good part of the faults were on a single chip Sept/11/2002 G.Alimonti-Pixel

3 Feasibility study The first tests have been performed at Politecnico di Milano, using a dynamometer operating in a temperature controlled environment The sandglass can turn around this axis so that, after the stripping, the 2 chips can easily be picked-up The upper holder is connected to 3 slides to absorb any shear stress component Chip vacuum holders Sept/11/2002 G.Alimonti-Pixel

4 Results The dispersion of the tensile strength and the quality of the re-worked chips ( mechanics grade assessed) looks fair Visual inspection confirms the good quality of the stripped bumps Sept/11/2002 G.Alimonti-Pixel

with the same planarity it has been attached (<0.

5 Test conclusions Reworked assemblies are quite good in terms of mechanical strength and single bump failure Reworking with In bumps has been shown to be possible Care has to be taken in: Removing the chip with the same planarity it has been attached (<0.1 mrad) to prevent shear stress to cause asymmetric bump breaks Warming up (~100 C) the bonds to exploit In plastic characteristics Sept/11/2002 G.Alimonti-Pixel

6 Conceptual design basically a microscope with few added features! Micrometric positioning Vacuum module holders Module heater Everything PC controlled Sept/11/2002 G.Alimonti-Pixel

7 Actual machine Venturi vacuum pump Z axis Pick-up tool with load cell Hot plate with Peltier cell X, Y stage Peltier power supply Motion control power unit Everything PC controlled Let s take a look at the single components Sept/11/2002 G.Alimonti-Pixel

Poli Comet II Coordinate measuring machine (CMM) by Poli, Varallo, Italy that guarantees the needed planarity Pneumatic guides of the granite moving XY axis and pneumatic balancing

8 Poli Comet II Coordinate measuring machine (CMM) by Poli, Varallo, Italy that guarantees the needed planarity Pneumatic guides of the granite moving XY axis and pneumatic balancing of the Z axis Maximum travel of 400x300x250 mm (X,Y,Z). Optical encoders with a resolution of 1 µm About 1x1x1 m dimensions and 140 Kg total mass Sept/11/2002 G.Alimonti-Pixel

9 Motion control NI 7344 controller up to 4 axis (3 implemented) Mini Maestro servo motor Pulse Width Modulation power units Z axis motor PWM power units Magneto-thermic breakers on the motor circuit Sept/11/2002 G.Alimonti-Pixel

10 Motion control (2) Performances after PID parameters have been optimized Position Vs Time Starting position Target position Position Vs Time Position Vs Time Position [um] Time [ms] Position [um] ,2 6,4 6,6 6,8 7 7,2 7,4 7,6 7,8 8 Time [s] Sept/11/2002 G.Alimonti-Pixel

within few microns from the target position for at least 300 ms.

11 Motion control (3) A double PID has been introduced to better maintain the target position after it has been reached: the approaching PID is changed to the stationary PID after being within few microns from the target position for at least 300 ms. The error is smaller and it does not depend any more on the approaching direction Sept/11/2002 G.Alimonti-Pixel

12 Peltier cell Hot plate T 2 V T 1 Heat sink Macor thermal insulators Hot plate with vacuum grooves Sept/11/2002 G.Alimonti-Pixel

13 Peltier cell (2) Peltier cell is biased by a GPIB controlled power supply The hot plate temperature is read out by thermistors: a PID loop reaches and maintains the needed temperature Sept/11/2002 G.Alimonti-Pixel

14 Pick-up tool Vacuum from Venturi Connection to Z axis Spring on load cell Load cell Moving vacuum pipe Teflon touching tool Sept/11/2002 G.Alimonti-Pixel

15 Load cell Wheatstone bridge Vout Vs Load -49,00-50,00-51,00-52,00-53,00 Vout[mV] -54,00-55,00 y = 0,4575x - 58,447 R 2 = 1 Series1 Best fit -56,00-57,00-58,00-59,00 0,0 5,0 10,0 15,0 20,0 Load [N] Sept/11/2002 G.Alimonti-Pixel

16 Touch down Maximum load on chip is the pick-up tool weight Load Vs Position 1,2 Pick-up tool weight 1,0 2 Soft touch down Load [N] 0,8 0,6 0,4 1 3 No load on cell 0,2 0,0 0,000 1,000 2,000 3,000 4,000 Position[mm] do not worry: there is also a safety stop on the Z axis position! Sept/11/2002 G.Alimonti-Pixel

17 Software LabView code to control the integrated system CMM Load cell Thermistors Peltier Cell Motion Control DAQ - I/O GP-IB Sept/11/2002 G.Alimonti-Pixel

18 Software (2) First let s check and calibrate the axis motion control and the load cell now we can set the target chip, go there and. land on it keeping an eye on the force on the load cell Sept/11/2002 G.Alimonti-Pixel

19 Conclusions First chips have been stripped out of a dummy module that already suffered some injuries Sept/11/2002 G.Alimonti-Pixel

required The chip stripping machine is working and

20 Conclusions (2) Pictures show a plastic detachment but somehow it looks like the planarity was not as required The chip stripping machine is working and (almost) ready! Sept/11/2002 G.Alimonti-Pixel

Characterisation of Hybrid Pixel Detectors with capacitive charge division

Characterisation of Hybrid Pixel Detectors with capacitive charge division M. Caccia 1, S.Borghi, R. Campagnolo,M. Battaglia, W. Kucewicz, H.Palka, A. Zalewska, K.Domanski, J.Marczewski, D.Tomaszewski