Pixel Pilot system Hardware Status Simon Kwan August 28, 2012 CMS Pixel Upgrade Workshop, Grindelwald1

Parts and hardware Needed New 2x8 modules using PSI46dig ROC (PSI), new TBM (Rutgers) New sensor Bump bonding New Al cable prototype HDI POH New port card Other chips (CCU, gatekeeper, TPLL, ALT/DLT…); Module production & assembly (Purdue, Nebraska, …) Module test-stand, Full chain readout system test-stand, and DAQ test stand Mechanical support and cooling channel (exist) New FED (Vienna, Vanderbilt) ;  TCA FED (UK) Power supply (modified A4603), DC-DC converter Cables, fibers (exist) DCS system CMS Pixel Upgrade Workshop, Grindelwald2

Hardware Status ROC, TBM (other reports) Sensor (other report) Al flex cable Pixel Optohybrid Port Card DC-DC converter (Katja will cover this?) Mechanical support and cooling DCS system CMS Pixel Upgrade Workshop, Grindelwald3

Schedule With the run extended till ~ Feb 2013, we have a little bit more time to build the pilot system Our plan is to have the system ready for testing at CERN summer of 2013 The main schedule driver is the bump-bonded modules. We must have the ROCs and sensor available for bump-bonding no later than April 2013 We plan to place the order for bump-bonding and have the vendor prepared everything (masks, assembly fixture and tooling) in advance CMS Pixel Upgrade Workshop, Grindelwald4

Module Initially, we were thinking of using TBM07 and then have the Datakeeper chip mounted on the new port card But as of late, our thinking is to use TBM08 which means that the output from the module will be running at 400MHz. No need to multiplex two modules with the Datakeeper This is in line with our current thinking of “one module, one fiber” (Sergey’s slogan) CMS Pixel Upgrade Workshop, Grindelwald5

FPix Readout for Pilot Disk (using TBM8, and aluminum readout cables) Pilot Run system will use: 2x8 detector module TBM8 with digital readout Aluminum readout cable A new Port card with up to 16 readout channels (4 POHs) (only 4 channels will be needed) Data are sent from detector modules to Port card via aluminum flex readout cables, and converted to optical signals there (4-channel POH hybrid) ROC registers download used DOH/GateKeeper/TBM/ROC communication channel TPLL is used to recover commands from the Clock/Control channel Delay25 is used to align clock, commands, and data I2C slow control uses existing design of a CCU board (including resets) Port Card uses the same architecture, as the current one, plus it incorporates FAN-IN/OUT ics, that used to be on the Adapter Board. Not shown are TPLL, and Delay25 ASICs CMS Pixel Upgrade Workshop, Grindelwald 6

HDI Mike Matulik (engineer who designed the current HDI for FPIX): – HDI layout is 80% complete and will be finished middle of this week. – I will review the design carefully and if there aren’t many issues, I may send it out for a quote for prototype quantities about the beginning of the following week. Expect to have the HDI available this Fall CMS Pixel Upgrade Workshop, Grindelwald7

Aluminum flex cable 1 st prototype made by CERN plastic shop (Rui Oliveira) more than a year ago While test results were satisfactory, we pushed ahead with a full length (75cm) prototype using the right material This full-size prototype was sent out for fabrication to KCP (thanks for Alice Bean) and CERN Last Friday, received a batch of 10 cables from KCP. Visually, they looked OK (thickness 0.24mm) CERN: early August, Rui informed us that he has now all material on hand to start fabrication We have enough connectors for the cables. CMS Pixel Upgrade Workshop, Grindelwald8

Port Card During the last couple of months, we discussed various options: three POHs, four POHs, different versions of port card for inner and outer modules (with and without Datakeeper) We hope that at this workshop, we could settle this question Expect real work to get started once the TBM08 is submitted CMS Pixel Upgrade Workshop, Grindelwald9

FNAL POH V2 For Pilot System FNAL has designed and tested the Version 2 Pixel OptoHybrid (POH V2) This version uses commercial TOSAs and fibers secured with Diamond DMI latches POH V2 has been tested with: Analog Level Translator (ALT) Digital Level Translator (DLT) at 400 Mbps using PRBS7 patterns Test results indicate that the DLT needs further optimization (see next slide). The designer is working on a solution for next submission. Test results indicate that the POH with the current ALT could be used as a “fall back” solution if pressed for time.

Optical Modulation = 138  W POH V2 (400 Mbps, PRBS7) Optical Modulation = 248  W FNAL POH V2 Port Card Signal Insertion Card DMI Clip

DCS System Haven’t thought about this at all until yesterday What came off top of my head – One RTD placed on HDI of each module – May also place one RTD on the Al brazed channel – One RTD each on inlet and outlet of cooling loop for disk 3 – Total = 8 RTDs; can be readout by single PLC module – Do we still have the slot for a control cable (or spare channels on existing control cable) : to be looked into – For modules powered normally, monitoring of voltage and current is straight forward – For modules powered by DC-DC conversion, CCU will be used for configuring/control of the Converters. We will use DCU channels to monitor the Vaa and Vdd. CMS Pixel Upgrade Workshop, Grindelwald12

Mechanical Support Our constraint is that this has to fit in the existing half cylinder with C6F14 cooling channels We proposed to simply glue using conductive epoxy the 2x8 modules directly onto the Al brazed channel Heat sink will be directly underneath all the ROCs so we expect the thermal performance to be very good We plan to insert a “fiducial” or alignment target on the Al channel to guide the placement of the module We will build a simple CAD model to determine the location of the target and the module This target will be placed accurately on the Al channel before hand under CMM at Fermilab. CMS Pixel Upgrade Workshop, Grindelwald13

Al Brazed channel CMS Pixel Upgrade Workshop, Grindelwald14

Pilot System Model 15

Summary This is now a high priority project in the US It has to be a success Expect to have all the needed hardware parts by early summer 2013 We should be able to ship the all the tested modules and readout electronics to CERN and ready for testing and installation by Summer 2013 CMS Pixel Upgrade Workshop, Grindelwald16