Sector Collector Electronics Consolidation of DT Sector Collector Electronics C. Fernández Bedoya on behalf of DT Upgrade group CMS Upgrade Week October 26th, 2010

DT Electronics Sector Collector Second level of DT read-out and trigger C. Fernández Bedoya October 26th, 2010 2 DTTF

DDU ROS ROB DT CHAMBERS DT Read-Out System UXC55 USC55 DDU (FED) 40 m copper 240 Mbps ~16 Mbps throughput 60 m optical 800 Mbps ~80 Mbps throughput S-LINK64 320 MB/s ~ 200 MB/s throughput DDU (FED) 10 DDU => half a wheel 12 ch/DDU (only half needed) Data merging Data quality monitoring TTS interface ~ 0.7 kB muon event size/DDU Chambers 5 wheels 60 sectors 250 chambers 660 super-layers 1640 layers ~172200 channels Minicrates 1500 ROB 128 ch/ROB Time digitalization (0.7 ns resolution) 1 µs time window Sector Collector 60 ROS 25 ch/ROS => 1 sector Data merging Data quality monitoring ~ 260 bytes muon event size/ROS C. Fernández Bedoya October 26th, 2010 3

DT Trigger System TSC 40 m copper 480 Mbps 60 m fiber 1.6 Gbps OptoRX C. Fernández Bedoya October 26th, 2010 4

DT Sector Collector - 2 SC crates per wheel - Located in tower racks in UXC level 2 Near - 60 ROS and 60 TSC boards in the system (1 per sector) - Complex electronic system - Main elements: LINCO, TIM, ROS, TSC LINCO C. Fernández Bedoya October 26th, 2010 5

DT SC Consolidation: Motivation Proposed upgrade is not motivated by the physics performance but by the fact that aging and other risks may jeopardize detector operation and contribute to an accelerated degradation Magnetic fields (40 mT) => tangential turbines SC crate power consumption (aprox 1 kW) is already marginal for CMS cooling system in the cavern. Temperature of some boards reach 45ºC. Turbines aging will lead to accelerated aging of the system and increased failure rate. Radiation doses 0.2 Gy per year (charged particle fluxes 20 cm-2 s-1) SC boards include large amount of logic, sensitive to SEU. Higher L => higher SEU rate. Important constrain for future upgrades. Limited access tighten to LHC technical stops and radiation protection issues In case of failure, it can take easily one week until access to the cavern is granted. Impact of failures in the detector can be VERY LARGE: Part failing Affected region %DT affected LINCO Half a wheel in trigger and read-out 10% TIM ROS 1 Sector in the readout 1.7% TSC 1 Sector in the trigger C. Fernández Bedoya October 26th, 2010 6

Proposed Upgrade -Relocation of DT SC electronics in the USC counting room -Make a “simple” copper to OF conversion at SC level -Modify input mezzanines of ROS and TSC Low impact modifications: compatible with present system and with possible future upgrades Present system Proposed upgrade C. Fernández Bedoya October 26th, 2010 7

Cu to OF conversion Present proposal is to make a 1 to 1 channel Cu-OF (Present links are copper based which length cannot be increased without compromising its reliability) Optical fiber 25 @ 240Mbps 25 @ 240 Mbps 32 @ 480 Mbps 32 @ 480Mbps Copper In the tower racks (substituting present SC) Plus few components for bias setting (DAC) and monitoring. OF could be extracted from the back of the SC crate VME interface at tower rack may not be needed Power can be extracted from present power supplies C. Fernández Bedoya October 26th, 2010 8

Installation of fibers Number of links between Minicrates and SC electronics Per Sector Per Wheel Total ROB to ROS 25 300 1500 SB to TSC 32/40* 400 2000 57/65 700 3500 *(S4 and S10 have 5 DT chambers each instead of 4) Large number of fibers 4 of these cables per SC crate (40 cables in total, large number of spare fibers) Total cross section needed around 100 cm2 144 fibers cable C. Fernández Bedoya October 26th, 2010 9

Routing of the fibers should be done through present “trigger” tunnels

-CERN EN/EL/EF section can provide and manage optical fiber installation. -We are in contact with them. -http://en-dep.web.cern.ch/en-dep/groups/el/Sections/cf.htm

OF input stage In principle, there is enough space below the false floor in S1 USC to recover extra cable lengths (though it depends on the exact racks to be used). Main problem is to allocate the SC crates in S1: -10 SC crates 11U each -To minimize L1A latency, they should be close to DTTF racks (S1D01 and S2D02) -In DT racks at present there is only space to allocate 6 SC crates (and not very close to DTTF) (Relocation can be done in batches of half a wheel) C. Fernández Bedoya October 26th, 2010 14

OF to Cu conversion -TSC and ROS input stages are mounted on mezzanines. -Modification to receive optical signals instead of copper can be made reusing most of the present modules -Interface with DDU and DTTF can remain unchanged C. Fernández Bedoya October 26th, 2010 15

First tests -Two prototype boards to perform Cu-to-OF and OF-to-Cu conversion have been developed at Ciemat. -With these boards we intend to qualify different devices at the market. -First tests are positive both in standalone and in the integrated chain with ROB and ROS -More studies are ongoing mainly in the reliability of the link Cu -> OF Cu -> OF ROB ROS C. Fernández Bedoya October 26th, 2010 16

DT SC Consolidation: Motivation Additional issues that may make the effort worth: ROS: -very big noise events that sporadically fill up the buffers Present FIFOs are external, cannot be easily enlarged Faster processing time requires higher performance FPGAs (Very expensive solution in radiation environment) OptoRX: -Problems with JTAG interface for configuration and monitoring -Upgrade connected with possible DTTF redesign Plus, we are talking of running CMS until 2030 (20 years from now), system will evolve (at some point by force), OF from the cavern to USC is very likely what we may want at some point (it is not a wasted effort). Any increase in performance cannot be accompanied by an increase of power consumption if present location remains Any improvement in the present location has a higher cost and longer design time (due to radiation constrains) C. Fernández Bedoya October 26th, 2010 17

Compatibility with future upgrades Step 1 Future upgrades? Present Cu-OF (trg) Minicrate Minicrate Minicrate Cu-OF (RO) OF Patch panel ROS TSC Cu-OF (RO) Cu-OF (trg) UXC UXC UXC USC USC USC OptoRx ROS TSC ROS TSC + OptoRx DDU DTTF OptoRx DTTF DTTF DDU DDU C. Fernández Bedoya October 26th, 2010 18

DT SC consolidation: Plan PRESENT PLAN 1. Fibers installation (2012): -Requires long access time -Needs to be done simultaneously for all the fibers 2. Scaled installation at any time (winter shutdowns). Minimal unit is half a wheel. -Modification of TSC and ROS input mezzanines -Relocation of SC electronics in USC Estimated cost 800 k€ (+600k€ if ROS and TSC are totally redesigned) C. Fernández Bedoya October 26th, 2010 19

Back up

DTTF BPTX 33 U 16 U 10 U* DTTF bottom 10 U 13 U 13 U empty 13*+13 U empty 15 U 0 U 16 U 10 U* 16 U 10 U* 14 U* 33 U 0 U 15 U Since 26 U Since 26 U 0 U In row D we could put more racks to the left, in the corridor, there is space 0 U

SC to USC -Move SC crates to USC (VME access in UXC may not be needed, compatible with future RS485 board) -Modify input ROS and TSC mezzanines for OF reception Fully compatible with present system Drawbacks: -trigger latency may be slightly increased -Find space in USC for 10 SC crates -Route large number of fibers (with present 48 fibers => 73 cables)

DT Chambers YB+2 YB+1 YB0 YB-1 YB-2 250 chambers 5 wheels 26 May 29th, 2009 C. F. Bedoya DT Chambers Sector 4 YB+2 YB+1 YB0 YB-1 YB-2 Sector 5 Sector 3 Sector 6 Sector 2 Sector 7 Sector 1 MB2 MB1 MB3 MB4 Sector 8 Sector 12 Sector 9 Sector 11 Sector 10 250 chambers 5 wheels 12 sectors/wheel 4 layers/sector: MB1, MB2, MB3, MB4 172,200 drift cells Covers 0<η<1.2. Provide muon identification Precise momentum measurement. Good pT resolution at high transverse momenta: σ(pT)/pT~ 10% @ 1TeV/c Reliable and robust trigger: pT standalone measurement @ Level-1 and High Level Trigger and precise Bunch Crossing (BX) assignment.

DT Chamber DT Cell  DT Chambers 27 May 29th, 2009 C. F. Bedoya Tmax < 400 ns Efficiency ~ 99% Drift velocity ~ 55 μm/ns  DT Cell Anode wire 3 Superlayers per chamber, 2 for Ф coordinate and 1 for  coordinate. Almost linear space-time relationship. Single wire resolution ~ 250 µm Local reconstruction (rФ) ~ 100 µm Gas: Ar/CO2 (85/15)% High Voltage: wires 3.6 kV strips 1.8 kV cathods -1.2 kV

28 May 29th, 2009 C. F. Bedoya Minicrate Attached to the DT Chambers it contains the first level read-out, trigger and full chamber control electronics. CCB: Full chamber control and monitoring: configures, sets thresholds, reads temperatures, etc. CCBlink: Connects the CCB to the external DT DCS system. TRB: Searches track segments and performs bunch identification. SB: Performs track selection and transmits to TSC. ROB: Time digitalization of signals coming from the chambers. ROLINK: Collects outputs from ROBs and sends it to the ROS. New Minicrate firmware means new CCB firmware.

MC secondary link upgrade for 2012 shutdown From Franco G. Replacement of 485 boards (10) housed in SC crates Improvements in secondary link system done 2 years ago have solved the many RS485 IC ruptures on MC linkboard But: Improvements were realized with many ‘handmade’ patches added to boards The UXC-USC link for half wheel is slow, 38.4Kbps Often 485 boards lost communication with DCS (last week 2 of 10, sometimes more) Recovering requires cycling on/off the SC crate Enhancement of MC communications reliability Integration of all patches on PCB new 485 board Maximization of USC-UXC link speed Boards remain compatible with present hardware Required the modification of part of DCS server software Cost: about 15Keuro. Man power by INFN PD

present system after 2008 improvements MC SECONDARY LINK present system after 2008 improvements From Franco G. Primary serial link -> optical fiber MC communication Secondary serial link -> RS485 copper chain Half wheel RS485 board 38.4Kbps Sector 1/7 driver485 Sector 2/8 driver485 38.4Kbps UXC Upper/bottom SC 9U crate Sector 3/9 controller Sector 4/10 driver485 38.4Kbps Sector 5/11 driver485 38.4Kbps Sector 6/12 UXC-USC optical link 38.4Kbps 485 chain termination & overvoltage protection

From Franco G. Proposed new 485 BOARD 485 chain termination & overvoltage protection RS485 38.4Kbps RS422 link to ADLINK PMC8681 (PMC board already mounted on VME SC crate controller) 230.4Kbits full duplex S1/S7 S2/S8 485 link controllers 1/sector S3/S9 USC Interface controllers S4/S10 S5/S11 S6/S12 Backup optical link 38.4Kbps for present system compatibility (upgradable to 230.4 full duplex) Power from LV caen module

Consequences of LHC plan for DT: Reassess the robustness of our stock of spares, in particular for on detector electronics (with only 2 openings of CMS, the number of hot spares needs to be larger than planned): Meticulous failure rate measurements during 2010 On detector electronics: Not sensible redoing the Minicrates before 2020, however we are for sure short of BTIM. Present approach is remaking the theta trigger boards and use old boards to retrieve spare BTIMs UXC racks electronics: CAEN Low Voltage connectors: weak point in particular for A3100. Plan is to move to A3100B (bolts) this year. For A3050 the plan is to test lubricant in 10% of the detector this year and see how it behaves. Sector Collector electronics: Currently we need to wait until LHC stops to get UXC access in order to fix a SC problem. A SC problem may handicap a large fraction of the detector (min a sector in trigger or readout). Present approach under study is to make a “simple” copper to OF conversion at SC level and move SC electronics to USC Expectations: 3 ROBus/year , 5 TRB/year, 1 CCB Link/year Failure rate= 46 in 2009, 4 in 2010 Failure rate low at present (~1/year)

Equalizer Laser driver 25/ROS 32/TSC … Equalizer