Fred Doherty Joe Ashby Fiona Mc Ewan Calum

Fred Doherty, Joe Ashby, Fiona Mc. Ewan & Calum Gray Infrastructure & issues and solutions for ATLAS Upgrade Hybrid/Module strip bonding ATLAS Upgrade Module Bonding Trials Using Glass Asics

130 nm ATLAS Upgrade Module BJ 820 Bonding Setup 19 th January 2016 GU Meeting 2

130 nm ATLAS Upgrade Module with Glass Readout Asics 19 th January 2016 GU Meeting 3

130 nm ATLAS Upgrade Module Glass Readout Asics with 4 rows of metal bond pads 19 th January 2016 GU Meeting 4

1 st 130 nm glass asic module bonding • The Front End Bonding (FEB) software program was loaded and the XY positional rereferencing, pattern recognition and Z height adjustment for all 10 glass asics on the hybrid and silicon strip sensor were performed. • All bond pads checked for correct alignment, metallisation, dirt, debris (bond position moved if possible from problem areas). All 10 asics checked with (64) 256 bond wires requiring (128) 512 bonds per asic (Source - asic & Destination – sensor) and giving a total of (640) 2560 bond wires requiring (128) 5120 bonds in total for LH hybrid. • All glass asic bond pad welding parameters had to be modified from the original program settings. The existing parameters would not take to the glass asic pads in trials - deformation error (too small) - so changed (increased weld parameter settings). • Prepare wire for immediate bonding using trial bonds if problem stop during 1280 bonds have to understand fix error, sort and prepare wire to continue bonding. • All 10 glass asics bonded well with new parameters and after successful completion of the LH hybrid to sensor bonds the module sensor was rotated to continue RH hybrid to sensor bonds. • Checks on all bonding info including BJ 820 individual wire bond graphs & detailed visual inspections confirmed successful completion. • This procedure is only 1/8 of the FE Bonding for 1 module. 19 th January 2016 GU Meeting 5

1 st 130 nm glass asic module bonding Wire bond positions on ‘destination pads’ of silicon strip sensor 19 th January 2016 GU Meeting 6

1 st 130 nm glass asic module bonding The many parameters for wire bonds (source / destination) at any (Asic / sensor) referenced position 19 th January 2016 GU Meeting 7

1 st 130 nm glass asic module bonding • There are 3 options for a bonding program procedure: • Load 1 st bond wire layer program then bond both sensor hybrids (markings on sensor when rotating with suction tool 4 times for 4 wire layers). • Fully bond the one sensor hybrid loading 4 layer programs and then 1 rotation and load 4 programs to complete other side. • Create 4 program for 4 bond layer to complete both sides. Hybrid/Sensor Positional accuracy and repeatability may cause problems and added time setting up. Referencing and Pattern Recognition also for 20 rather than 10 in current smaller logical program. • Ideally we do not want to be rotating / moving the sensor at any stage in the bonding process so the 4 programs covering each layer for both sides will I’m sure be implemented. 19 th January 2016 GU Meeting 8

1 st 130 nm glass asic module bonding First LH Hybrid wire layer 1 fully bonded 19 th January 2016 GU Meeting 9

1 st 130 nm glass asic module bonding 19 th January 2016 GU Meeting 10

1 st 130 nm glass asic module bonding 19 th January 2016 GU Meeting 11

1 st 130 nm glass asic module bonding 19 th January 2016 GU Meeting 12

1 st 130 nm glass asic module bonding This is a recorded video of the actual wire bonding after preparatory stage. 19 th January 2016 GU Meeting 13

1 st 130 nm glass asic module bonding Preliminary conclusions • Good learning curve and experience gained for 130 nm Module bond phase. • A few issues identified but no show stoppers. • CNM glass asics and silicon sensor bond very well at set welding parameters. • Timing of all procedures cannot be tied down. • Program expansion to 2 hybrids likely. • Loop heights for wire layers look ok but electrical module required to confirm. • Bond pull test not required till proper asics available for trials. • Will look for a camera / microscope lense screw adaptor for videoing setup. • Will feedback to UK bonders group on experience so far. 19 th January 2016 GU Meeting 14