TOB System Test Status Report DS ROD System
TOB System Test Status Report • • • DS ROD System Test Setup DAQ Software DS ROD Characterization • • PLL, FED Timing Scans OPTO Scans • • • Juan Valls CERN Noise Characterization (ROD vs OTRI) Effect of Decoupling Caps on TOB Modules Conclusions 1 Juan Valls
SS ROD System Tests Conclusions from past TOB system tests SS ROD electrical design validated n n Bartalini et al. Chierici et al. optical readout (6 modules) electrical controls (electrical FEC, no DOHM) Grounding scheme found ok and validated Cooling performance and thermal behavior studied and verified at room temperature Noise performance studied Overall system performance validated 2 Juan Valls
DS ROD New Vienna AOH (LLD 2 ICs, 3 laser drivers) New final CCUM module (CCU 25 IC) Redistribution of resets and back-plane pulses on ROD ICB CCU 6 reset out top reset bottom reset 6 back-plane pulse lines PIO SS ROD ICB DCU CCU 25 2 back-plane pulse lines reset out top BP pulse bottom BP pulse PIO 6 reset lines DS ROD ICB New FEC 2 CCU PCB to mimic DOHM controls functionality (present during the testing of cabled RODs in production) (G. Magazzu, F. Ahmed) 3 Juan Valls
DS ROD New (prototype) LV PS (Sandor’s design) n n n DELTA switching power supply (8 V, 50 Amp) (old DELPHI HPC) Linear regulators (fixed 2. 5 V and 1. 25 V) Fast reacting PS (V 2. 5 overvoltage < 0. 2 V, long cables, up to 10 A) Sense voltages on the regulators for fast feedback Current limitation on both lines Interlock controls + V/I monitoring next version The CCUM voltages are provided through the FEC 2 CCUM board DS ROD 12 modules 48 APVs I 2. 5 ~ 6. 4 A I 1. 25 ~ 2. 6 A ICCU ~ 0. 17 A I 2. 5 (max) ~ 9. 2 A I 1. 25 (max) ~ 3. 2 A 4 Juan Valls
DS ROD Assembly (Controls) HV adapter card and connector SC out (and LV out) SC in (and LV in) CCUM (with CCU 25) Juan Valls LV adapter card and connector ICB Ground 5
DS ROD Assembly (Readout) 24 fibers New Vienna AOHs (LLD 2 ICs) from Jan Troska (Tracker Optical Links Web Page) 6 Juan Valls
DS ROD Setup (Building 598) Optical (~3 m) Readout HV LV TOB DS ROD Layer 1 FEC 2 CCUM board C 6 F 14 Cooling Plant Electrical Controls 1 k. W +5 C/+32 C 7 Juan Valls
DAQ Software XDAQ n n n System Tests Test-Beam Controls integration Introduces a non-flat CMN picked-up by some of the modules in the ROD (see past talks on SS ROD) XROD n n System Tests Electrical and Functionality Tests of RODS Simultaneous readout of FED buffers while arrival of input frames Needed optical control Separate location of BE boards (FEC card) Software throttle if FED overflow inhibit TSC triggers Subestructure Burnin Test Station (W. Beaumont) 8 Juan Valls
XROD TSC FED ROD FAST debugging tool CMS-like DAQ hardware Access to BE boards APV n TSC, FED, CCUM n Handles CCU 6 and CCU 25 Access to FE registers n PLL, MUX, APV, DCU, AOH n Handles DCU 1 and DCU 2 n Handles LLD 1 and LLD 2 Internal/external TSC triggers (and FED internal) Single GUI Interface 9 Juan Valls
XROD Frames Noise XROD handles up to 3 PMCFED cards n n n Gain Scan n Pulse Shape Scan 8 modules (4 or 8 APVs) 1 SS ROD (6 modules, 4 or 8 APVs) 2 SS RODs (4 modules, 4 APVs) 1 DS ROD (12 modules, 4 APVs) The use of K-MUX will enhance this capability http: //cern. ch/valls/CMS_SST/xrod. htm 10 Juan Valls
PLL Time Alignment Scan through PLL fine delays (1. 04 ns) and with a fixed FED digitization delay Reconstruct APV tick marks The DS ROD introduces shift delays of ~2 ns on the trigger arrival time to APVs. FED 0 FED 1 FED 2 XROD 11 Juan Valls
FED Timing Scan Find the FED optimal digitization point FED 0 Reconstruct APV tick marks by varying FED skew clock delay wrt data (PLL settings fixed) FED 1 Choose sampling point close to the back edge of the tick mark FED 2 XROD 12 Juan Valls
Optical Scan Characterization Inverted ticks into AOH ! (connector mismatch between ICC and AOH PCBs) Based upon Mirabito’s code Run FEDs in Scope Mode Fix AOH settings. Get distribution of ticks and baselines (over events and samples) fixed by patching OEC output connectors Ticks still arriving inverted into the AOH Juan Valls 13
Optical Scan Characterization Plot ticks and baselines as a function of bias (for a fixed gain) Get the tick amplitude from the difference between these distributions AOH Gain = 1 (24 fibers) baselines tick amplitudes ticks AOH bias Juan Valls AOH bias 14
Optical Scan Characterization Find optimal settings (gain and bias) for an 800 m. V AOH input tick amplitude What does this correspond to at the FED (in ADC counts)? Need to calibrate FED cards: FED gain ~3. 5 m. V/count, Optolink gain ~0. 8 V/V Bias 150 -210 counts Gain 0 Gain 1 Gain 2 15 Juan Valls
Measurements All measurements taken with: n Optimized timing (PLL, FED) and opto settings (gain and bias) n RMUX = 100 (to match termination with AOHs) n APV bias generator registers (as from “Procedures for Module Test”, Draft 2) All results given in terms of: n Total noise ( tot) n CMN substracted noise ( CMN-substracted) n Differential noise ( diff) RMS of ½(ADCi-ADCi+1) 16 Juan Valls
DS ROD Noise Position 2 Position 3 ROD ICB Position 4 Position 6 Deconvolution Non-Inverting (Doracil) (200 V) Position 1 Position 5 tot diff CMN Cicorel CCUM 17 Juan Valls
DS ROD CMN (flat) Calculation (running average pedestals) Non-Inverting ~40% 18 Juan Valls
DS ROD HV Scans HV Bias Scan on DS ROD 6 HV channels for 12 modules (CAEN SY-127, A 343 boards) Total noise (ADC) = f (Vbias) Full depletion at ~150 Volts Similar behavior for all modules FNAL M 658 (Cicorel) placed on top side of ROD (near to CCUM) 19 Juan Valls
FNAL M 658 (Cicorel/Hybrid. SA) OTRI Setup Peak Mode Inverting Peak Mode Non Inverting Total Noise ( tot) Deconvolution Non Inverting Differential Noise ( diff) CMN substracted Noise ( CMN-substracted) 20 Juan Valls
Noise (DS ROD vs OTRI) Peak Mode (Non-Inverting) OTRI tot diff Cdec tot diff DS ROD tot diff Deconvolution (Non-Inverting) DS ROD noisier than OTRI Slighter higher differential noise than total noise (uncorrelated CMN) 21 Juan Valls
Full Gain Scans (DS ROD) Ical=29 ~ 25000 elec Fit Range: Ical=18 to Ical=70 0. 6 – 2. 7 MIPs DS ROD Gains/APV Offsets/APV 22 Juan Valls
Full Gain (DS ROD vs OTRI) OTRI ROD Peak Mode Non-Inverting Gains (M 658) DS ROD vs OTRI (electrons/ADC count) OTRI ~850 elec/ADC (OTRI) ~650 elec/ADC (ROD) ROD Deconvolution Non-Inverting 23 Juan Valls
Noise (DS ROD vs OTRI) Peak Mode Non-Inverting OTRI Setup Peak: 1600 elec. Dec: 2600 elec. DS ROD Setup Peak: 1600 elec. Dec: 2700 elec. Deconvolution Non-Inverting tot diff APV 25 bare chip on PCB (Cinp=18 p. F) Peak: 900 elec. Dec: 1500 elec. 24 Juan Valls
Effect of Decoupling Cap TOB Cycorel Hybrid Detector Return Decoupling Cdec = 100 n. F Edge effect improvement on TIB modules (see Civinini talk) 25 Juan Valls
Edge Strip Correlation OTRI DS ROD OTRI Cdec=100 n. F No improvement No edge effect on ROD (w/o Cdec) 26 Juan Valls
TOB/TEC and TIB TOB TIB Vbias NAIS HV Connector on Kapton Cable GND Vbias Bias Connector on Kapton Cable GND (wirebond to bias ring) 27 Juan Valls
CMN (DS ROD vs OTRI) Peak Mode (Non-Inv) Common Mode Noise DS ROD vs OTRI CMN (Cdec=100 n. F) Deconvolution (Non-Inv) DS ROD CMN 28 Juan Valls
Conclusions (I) Flat noise, flat CMN in both setups Similar noise results for both setups (OTRI/ ROD) after full gain values applied) Slightly larger CMN for OTRI than for DS ROD No evidence of noise edge effects on ROD (optical readout) Edge effect seen in OTRI setup (FNAL modules M 658 and M 657, electrical readout), not cured with Cdec Most of the software tools and hardware designed for the system test setups will also be used during production for electrical and functionality tests of RODs 29 Juan Valls
Conclusions (II) Next. . . -source and cosmics studies (see next talk) Study the cooling performance (thermal behavior) of DS ROD in the cold (with final LV PS + interlocks) Integration of DOHM (or use of FEC 2 CCUM) Exercise >1 RODs in a control loop Exercise the back-plane pulse functionality Integration of K-multiplexer into the DAQ Integration of ROD objects into DB More at. . . http: //cern. ch/valls/CMS_SST/rod_system_tests. htm 30 Juan Valls
- Slides: 30