University of California, Los Angeles Endcap Muon meeting @ Purdue TMB-RAT Status Report CSC/ME1/1 Meeting Oct. 6 2006 Yangheng Zheng University of California, Los Angeles Outline: Mezzanine cards, TMBs, RATs production details Firmware/Software updates Endcap Muon meeting @ Purdue 6-Oct-2006
TMB-RAT pair (version 2006) RAT Board Receives ALCT And RPC data CFEB Input Connectors Xilinx XC2V4000 Mezzanine board Endcap Muon meeting @ Purdue 6-Oct-2006
Bench Testing Software ~ 15K lines of code Mezzanine card test Hard resets test Connection test: CFEB connectors on TMB ALCT-RPC connectors on RAT DMB, MPC, CCB and VME connections on backplane (done by loopback backplane) TMB-RAT connections through backplane Delay chips test Other test (LEDs, EPROMs etc…) Endcap Muon meeting @ Purdue 6-Oct-2006
Mezzanine Cards Status TMB Virtex-2 mezzanine cards: In 2004 – 2005, 542 mezz. cards assembled. 520/542 work fine without extensive rework 22 sent out for re-balling, only 7 recovered. New: 40 additional mezz. cards had been assembled in July, 2006. All passed the test. Plan: assemble 10 more mezz. cards with reballed Virtex-2 chips. Endcap Muon meeting @ Purdue 6-Oct-2006
Endcap Muon meeting @ Purdue TMB Production Status 468 TMB’s in final system plan 15% spares: 540 in total Deliveries 520 shipped to CERN (495 tested: 3 found with problems) 10 shipped to Rice 3 shipped to Florida In UCLA electronics lab: 2 ready for shipping 19 need to be repaired Endcap Muon meeting @ Purdue 6-Oct-2006
Endcap Muon meeting @ Purdue RAT Production Status 468 RAT’s in final system Plan 15% spare: 540 in total Deliveries: Currently 540 shipped to CERN (495 tested: no problems found so far) 3 shipped to Florida In UCLA: 7 RAT boards need to be repaired Endcap Muon meeting @ Purdue 6-Oct-2006
TMB-RAT testing experience TMB: ~280 boards pass the test without extensive debugging. (error rate: ~50%) RAT: ~440 boards pass the test without extensive debugging. (error rate: ~20%) Mezzanines: Error rate 4% (compare to ALCT mezz. boards : ~30% error rate) Good assembly house made big difference for BGA project. Typical errors: soldering bridge/cold soldering (>90%) Bad chips/components (~10%) Mechanical errors (~1%) Endcap Muon meeting @ Purdue 6-Oct-2006
TMB Firmware/Software Updates Done by Greg with the help of Martin and others at CERN Software for programming eproms via VME controller/backplane TMB, CCB, DMB, MPC XML configuration for variable parameters RAT, ALCT via TMB with JTAG Firmware/Software update for small “Configuration” Eproms on TMB (done) Read in after hard reset reloads the firmware. TMB downloads its own registers using FPGA control of on-board VME circuitry. TMB downloads ALCT registers using JTAG Endcap Muon meeting @ Purdue 6-Oct-2006
Firmware/Software Updates (cont.) TMB+RAT+ALCT configuration (Greg): All (relevant) parameters now in xml file… Corresponding registers read from TMB, decoded into “user-friendly” format Configuration for TMB, RAT, and ALCT now takes following path: Parameters in xml file compared with configuration data in 2 proms on TMB prom 0 = TMB configuration parameters prom 1 = ALCT+RAT configuration parameters If different, proms programmed and verified Upon hard reset, TMB firmware loads TMB, ALCT, and RAT configuration parameters from user proms New software/firmware are being checked on slice test ME+1a To be implemented during MTCC phase 2 Endcap Muon meeting @ Purdue 6-Oct-2006
Board-maintaining/repairing Crew Engineer: Y. Shi (UCLA physics department) Students Special thanks to the following students who worked for the UCLA TMB-RAT project. Daniel Maronde, Paul Cubre, Jennifer Goodman, Alfonso Vergara, Mike Lee, Aaron Senter, Donald Chang Amber Bechtel Rina Kakimi Endcap Muon meeting @ Purdue 6-Oct-2006