First look at a hybrid for a stave
First look at a hybrid for a stave The University of Liverpool Ashley Greenall ATLAS Upgrade Week, CERN 24/02/2009 1
UK Stave Hybrid: Considerations • Current hybrid was designed as test vehicle to verify that: 1. Bare ABCns on a hybrid operate as expected. 2. Then confirm that ABCns bonded to a sensor meet specification. • Design more suited to a bridged hybrid than gluing onto sensor. • Only upon satisfactory completion of the above do we proceed to an integrated stave hybrid. • Now believe we are in a position to do this! What needs to be considered… • Flex circuit will be glued directly to sensor. • Because of this it becomes substrate-less. • Will need to understand how to stuff passives, asics and wire-bond etc. • Already touched upon in hybrid talk, so far no showstoppers identified. • • Need to take into account powering. Ideally integrated onto hybrid - currently 20 x ABCns draw >3. 5 A! • Are wire-bonds ok for the power connection? • Off module connections (to bus cable) made using wire-bonds. • Need to integrate BCC asic – required for module digital I/O. • Present sensor geometry results in hybrid width remaining ‘fixed’ at 24 mm. • But length will reduce to ~110 mm. ATLAS Upgrade Week, CERN 24/02/2009 2
UK Stave Hybrid: first look 7. 6 mm Still need to take into account real estate for powering… Dashed line shows sensor profile 97. 5 mm 110. 6 mm HV Filter detail with bond pads Integrated BCC detail Locations as presented by A. Weidberg (NIKHEF). COM L 1 5. 5 mm Bond pads for connection to bus cable. BCC BCO Data. Out Passives for AC-coupled LVDS (mix of 0402 and 0603 devices) ATLAS Upgrade Week, CERN 24/02/2009 3
UK Stave Hybrid: first look • • • First attempt at layout looks fine. Integrating BCC onto hybrid is achievable. Powering is still an unknown. • Need to understand what circuitry is required. But (my opinion)… • Are we taking a risk integrating the BCC directly onto the hybrid? • What if it doesn’t function as expected? • Then have to wait for another submission, incurring a 3 -4 month delay? • Powering the hybrid – why not keep ALL options open? • Could we consider both serial and parallel powering. An alternative take… • DON’T integrate the powering and BCC onto the hybrid, instead make use of add-on boards (see following slide). • Doesn’t fix us to a single solution – increases the flexibility. • BCC board could be replaced by a bare die LVDS variant (if so needed). • Likewise powering – could have both serial and DC-DC powered add-ons. • Would allow us to test DC-DC powering in a realistic environment. ATLAS Upgrade Week, CERN 24/02/2009 4
UK Stave Hybrid: add-ons (BCC example) • • Fabricated on FR 4 (~500µm thickness, matches to sensor+flex thickness). Geometrically identical to integrated BCC. • Should still connect to bus cable – as per integrated solution. Width stays fixed at 24 mm BUT length L is not fixed. Fixation of add-on needs to make use of a glue that is not permanent. • Could provide a similar solution for powering… Bond pad field for inter-hybrid to BCC add-on Bond pads for connection to bus cable L 24 mm ATLAS Upgrade Week, CERN 24/02/2009 5
UK Stave Hybrid: Summary • Hybrid layout, in principle looks ok. • Use of ‘add-ons’ increases flexibility of hybrid + stave. • Doesn’t tie us into a fixed solution for Digital I/O or powering. • Need ‘green light’ to proceed with this idea. • Otherwise roll back to fixing components onto hybrid. • Layout could be fixed very quickly (2 -3 weeks) for next submission. • Hybrids available early summer…. ATLAS Upgrade Week, CERN 24/02/2009 6
- Slides: 6