1. SLHC SCT Hybrid (CERN 2nd July 2007)1 SLHC SCT Hybrid Concept Ashley Greenall The University of Liverpool

2. SLHC SCT Hybrid (CERN 2nd July 2007)2 What will be discussed What we would like to achieve, and how it could possibly be done. Hybrid layout and its interaction with other module components. What are the constraints? Initial thoughts on layout options But no complete solutions Still a lot of unanswered questions! What I will not discuss Services – a lot of unknowns Hybrid trace layout detail – ditto Progress is difficult without ‘real’ components

3. SLHC SCT Hybrid (CERN 2nd July 2007)3 What we would like Ideally, a 40 ASIC hybrid which will Use minimal material in its construction Match up to a square sensor geometry (99mm x 99mm) How could this be achieved? By reducing the hybrid width Reduced material Making use of direct ASIC-to-Sensor wire-bonding Eliminates the need for pitch adaptors But, this means that Readout ASIC (ABC-N) die size has to take into account ASIC-to-Sensor bonding Especially its width –Must be optimised to match the sensor geometry

4. SLHC SCT Hybrid (CERN 2nd July 2007)4 Direct ASIC-to-Sensor Bonding Is this possible? YES, if we match (closely) the ASIC Input and Sensor bond pads Setting ASIC width to 7.5mm could help – any wider becomes problematic (SMD placement for front-end decoupling etc.) ASIC input bond pad pitch becomes 55µm Sensor bond pad pitch is set to 75.6µm Mismatch can be taken up by the wire-bonds But… Wire bonding issues We use a H&K 710 wire-bonder, max. bonding angle <17° (see following diagram.) –Using 25µm bond wire Bond foot displacement is a problem at the ASIC input pads (see Tim’s talk) 17° bond angle constrains ASIC-to-Sensor bond pad distance to ≥4.1mm Results in ‘long’ bond wire lengths – maybe not problematic? These parameters will determine where sensor bond pads can be located Need to converge on a ‘standard’ set of parameters!

5. SLHC SCT Hybrid (CERN 2nd July 2007)5 ASIC – Sensor Detail 16.6° bond angle Front-end Decoupling Capacitor ASIC Cap Hybrid Substrate ASIC Capacitor on topCapacitor on underside “Well” in substrate Continued… 7.5mm Width ASIC Detail 55µm ASIC input bond pad pitch 75.6µm Sensor pitch <17° bonding angle ASIC-to-ASIC separation is 2.1mm –Miniature capacitor could be placed in gap –100nF, X7R, 0402 SMD (Murata…) –1mm x 0.5mm x 0.5mm (LxWxH) alternatively Place capacitor on the underside of hybrid

6. SLHC SCT Hybrid (CERN 2nd July 2007)6 Opportunity to also consider optimising ASIC bond pad placement Data & Token Passing Hybrid Traces ABCD ByPass Hybrid Traces ‘Old’ ABCD Routing“Revised” Routing Data & Token Pads moved to side Allows direct wire-bonding ByPass routing on chip Reduces hybrid trace routing Data & Token Passing wire bonds ByPass routing ASIC Bond Pad Placement

7. SLHC SCT Hybrid (CERN 2nd July 2007)7 SLHC ‘Narrow’ Hybrid (concept) Taking these ideas on board, conceptually a hybrid could look like below Digital I/O Power/Aux Bus 99mm MCC Power Regulation 24mm 74mm SENSOR Readout ASICs Detail 2 ‘fingers’ per hybrid populated with 2 columns of 10 ASICs (40 total/hybrid) Electrically one item 4 Cu Layers with Kapton dielectric (example from LHCb VELO) 100µm track and gap, 300µm vias (conservative design to maximise yield) 2 x Outer layers (routing) are 5µm Cu with additional ~8µm plating 2 x Inner planes (PWR/GND) are 12µm Cu 50µm Bond Ply (Espanex) between layers 200µm total build thickness But, Services detail unknown Connector detail not finalised Power Regulation & Module control not defined ??? HV MCC: Module Controller Chip

8. SLHC SCT Hybrid (CERN 2nd July 2007)8 Example of routing density But there are caveats… Reduced material could mean Power/Gnd Planes compromised Might impact upon module electrical stability Consider Present SCT Barrel hybrid has 2 planes of ~3178mm 2 Cu area/plane & 1.6A total current EndCap required 4 dedicated layers of ~2500mm 2 Cu area/plane for stability Narrow hybrid has ~2376mm 2 Cu area/plane per finger with 2.74A current (nominal) Based on 4 layer build (2 Power/Gnd Planes) ~70% increase in current with ~25% reduction in Cu (compared to Barrel). Increasing width from 24mm to 32.5mm, Cu area becomes comparable to Barrel hybrid But still 70% more current! Add more Cu (by increasing Cu thickness/more layers) Lowers Plane impedances BUT Need to consider skin affect? High frequency parasitics could still be problematic Increasing Cu thickness does not necessarily help – increased Cu area does SLHC ‘Narrow’ Hybrid (concept)

9. SLHC SCT Hybrid (CERN 2nd July 2007)9 Hybrid Width (impact of) Prudent to consider 2 flavours of hybrid: 1.Preferred ‘narrow’ (24mm width) hybrid, minimal material approach 2.Backup solution of a ‘wide’ (32.5mm width) hybrid Or do we just increase the material on the preferred solution? If we adopt the 2 flavour approach, then Sensor bond pads must be placed to accept the differing hybrid geometries

10. SLHC SCT Hybrid (CERN 2nd July 2007)10 Sensor Bond Pads (a first look) Hybrid “finger” is centred w.r.t. sensor strip start/end points

11. SLHC SCT Hybrid (CERN 2nd July 2007)11 Sensor Bond Pads (wide hybrid) Wide hybrid detail For either ‘narrow’ or ‘wide’ hybrids Sensor bond pads are asymmetrically aligned w.r.t. the sensor strip Sensor pad location is determined by Adhering to the <17º bond angle Maintenance of the 4.1mm separation from ASIC to sensor bond pads Peculiar to Liverpool

12. SLHC SCT Hybrid (CERN 2nd July 2007)12 Summary The step from the original 12 device SCT hybrids to 40 device SLHC hybrid is a bold decision to make! We need to now: Optimise the ASIC width to match that of the Sensor –Fix at 7.5mm? Revisit ASIC bond pad placement –Some bond pads can obviously be moved, some not so… Understand what are the constraints if we wish to bond directly from ASIC-to-Sensor –What is the max. bonding angle that is universally acceptable? –What other wire-bonding parameters do we need to unify? Will minimal material approach compromise electrical performance? –Is it really necessary to consider 2 flavours of hybrid? –Even if we don’t, would it not be prudent to engineer the sensor such that it is not ‘flavour’ exclusive? i.e. add bond pads on the sensor so that we can (maybe) try numerous flavours of hybrid ???

13. SLHC SCT Hybrid (CERN 2nd July 2007)13 Backup

14. SLHC SCT Hybrid (CERN 2nd July 2007)14