Slide 1L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder UCSB Testing Status Anthony Affolder (for the UCSB module testing group) Current testing infrastructure Test results Module test equipment needs Software Needs Current testing through-put
Slide 2L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder Testing Facilities Readiness Clean room adjacent to production area à Room layout finished à Currently in high bay with intentionally same layout –Missing only hybrid thermal cycling test à Clean room will be finished in April Storage cabinet with dry air flow in place for testing phase Long term module storage cabinets under construction at KSU
Slide 3L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder UCSB Recent Infrastructure Projects Hybrid holding plates à Matches hybrid thermal cycling test stand à Can bond and test with low noise in same holder Burn-in low voltage distribution crate à LED indicators and fuse protection Crowbar à HV protection circuit à 12 each for FNAL/UCSB Module cold plate à Attaches to current module holder Hybrid clamshell Module clamshell
Slide 4L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder Hybrid Visual Inspection Once FNAL indicated lift off problem, closely visually inspected our 5 hybrids à Found 1 wire bond lift off à Bonds have redundancy so no failure seen –Removed broken bond to eliminate chance of short à Will closely monitor for lift off/ wire breaks throughout production process
Slide 5L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder Current UCSB ARC Setup 3 ARC Controller + 5 new ARC FE + 1 old ARC FE 1 Hybrid testing (2 Hybrid clamshells) 2 Module testing (Module clamshell) à 2 LED systems à DEPP HV supply –Missing hybrid-to-utri adapters to fully use capacity Hybrid Clamshell Module Clamshell DEPP LED Controller ARC Controller ARC crate
Slide 6L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder ARC Hybrid Test Results (1) 5 hybrids tested on arrival à All good, but fail current requirement given in the Procedures for Module Testing à Pedestal test reproducible, but noise measured very sensitive to digitization effects –Measured noise varying the pedestal by changing VPSP value –Digitization effects cause variations larger than ±20% Hybrid noise requirements should be changed à N i within 50% of average à 0.3 < N i < 1.5 –Removes chance of chips with uniformly low/high noise from passing requirements See hep.ucsb.edu/cms/cmsUCSB.html for more complete description
Slide 7L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder ARC Hybrid Test Results (2) 4 hybrids have had APV bonded to pitch adapters à All channels are good Hybrid clamshells enable testing without large pickup effects at chip edges à Hybrid isolated with base plate and lid grounded à Same requirements can be made for hybrids with or w/o PA bonded
Slide 8L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder ARC Gain Measurements (1) Multi-point gain added to ARCS 6.0 beta software à Thanks to the Aachen group Tests are very quick and have advantages over pulse shape measurements à More uniform à More stable à Gain linearities à Gain variations Strongly suggest that test be added to all M800 production. Will be included TOB testing Make requirements on gain slope, offset and fit 2
Slide 9L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder ARC Gain Measurements (2) 2 of 24 chips so far have no response with low injection values à Only occurs in peak mode inverter on Caused by initialization defaults in ARCS tests à VPSP set too low –Thanks to T. Franke and M. Raymond for help VPSP setting will be modified in new release à Removes all seen irregularities
Slide 10L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder ARC Module Test Results (1) 1 Prototype Module Tested 1 Production Module Tested à Module testing in UCSB clamshell decreases common-mode noise to the point where location of opens become detectable by their noise levels –PA-sensor –Sensor-sensor à Pinholes act as if saturated à High current channels can have higher noise (Bad IStrip à Slight noise increase on chip edges à Two neighboring channels have high noise only in peak inverter on Pinhole PA-sensor open Sensor-sensor open Bad CAC Bad Istrip Unknown Problem PEAK ON
Slide 11L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder ARC Module Test Results (2) LED pinhole test à Bonded one inherent pinhole on purpose à Test acted as expected –120 A needed to unsaturate pre-amp Not clear if maximum current (400 A) supplied sufficient to find “high current” pinholes Pinhole Unknown Problem PA-sensor open Sensor-sensor open PEAK ON
Slide 12L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder ARC Module Test Results (3) Gain test à Pinhole clear by lack of gain à Opens clear by higher gain Pinhole Sensor-sensor open PA-sensor open Unknown Problem Bad CAC Bad Istrip PEAK ON
Slide 13L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder ARC Test X-calibrations Minimal X-calibration has been performed à Same faulty channels found at FNAL/UCSB on prototype model à CERN FHIT/UCSB ARC test compared –Pedestals very similar –Noise is slightly higher than CERN FHIT measurement Not understood if it is FHIT/ARC difference, a software difference, an environmental difference, or a digitization effect Once model production begins, plan to exchange modules with FNAL for X-calibration
Slide 14L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder ARCS Database No entries yet put into official database à Waiting for ARCS interface for database upload and database definition to stablize All testing results have been made available to silicon community on local website à hep.ucsb.edu/people/affolder/module_test_UCSB.html
Slide 15L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder DAQ status 1 full DAQ stand available at UCSB à Planned only for Vienna box tests with modules Limited number of tests performed à See P. Gartung’s talk à Focus on bringing up hardware/software for Vienna box à Xdaq, LT, and Xrod software all compiled and run –Still working to understand controls/output No LED stand planned for DAQ No database work done as of yet
Slide 16L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder Vienna Box Status Vienna Box recently arrived at UCSB à No modification necessary to use with TOB module carrier plates à Full compliment of plates currently in machine shop à LV distribution ready à Just received HV controllers (both the A128HS and A1303) –Integrating as fast as possible à Demonstrated each slot works with hybrids à Thermal cycled box between –25 to 30° C with chiller à Need hardware to test more than one channel at a time –See P. Gartung’s talk for more details
Slide 17L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder Current Hardware Needs ARC Needs à 1 DEPP à 1 LED system –Will be picked up this week Torino Interlock Box à In-route 3 Electrometers à 2 in-route Flat cable-Lemo adapter à Under construction at UCSB 13 hybrid-to-utri ver. D à 1 in-route –Used for both ARC module and burn-in test –Limiting factor in testing through-put Hybrid thermal cycling box à Lack hardware/software à Tests will be performed by A. Honma’s group for M800 à We are aiding in the manufacturing of test stand DAQ components à 8 VUTRI –10 in-route à 9 PAACB –6 in-route à 1 TPO –1 in-route à 1 CCU –1 in-route à 2 PAACB to TPO interface –Needed for backplane pulsing
Slide 18L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder Current Testing Software Needs Hybrid Characterization à Database input Module Testing à IV scan, backplane pulsing test, database input Module Burn-in Rods à Beginning to assemble necessary infrastructure for rod assembly/ burn-in à Xrod and LT software package already compiled and running
Slide 19L3 Testing Meeting March 7, 2003UCSB Status-Anthony Affolder Current Estimated Testing Through-put Hybrid Arrival Testing à 24 hybrids per day à Matches peak production (+ contingency) à Requires dedicated tech Module Basic Test à 2 ARC LED stands but only 1 hybrid-to-utri version D à 12 modules per day à Requires dedicated tech Module Burn-in (Vienna Box) à See P. Gartung’s talk for details à Lack of hybrid-to-utri adapters limits testing through-put to 1 –While burn-in test underway, no module testing possible à Module characterization test will be done at beginning/end of test cycles ARC After Burn-in Test à Not clear how much ARC testing necessary after burn-in –Depends on grounding/pickup issues –LED is minimal test Time estimates do not include database entry. After finishing test equipment acquisition/integration, database will be first priority.