1 DAQ for FVTX detector Implementation Mark Prokop Los Alamos National Laboratory.

Post on 01-Jan-2016

217 views 0 download

transcript

Mark Prokop, FVTX Review Nov 2008 1

DAQ for FVTX detector Implementation

Mark Prokop

Los Alamos National Laboratory

Mark Prokop, FVTX Review Nov 2008 2

Talk Outline

• Overall Readout System

• ROC Pre-Production Prototype

• FEM Pre-Production Prototype

• Risk Factors

• Summary

Mark Prokop, FVTX Review Nov 2008

Technical Challenges for Readout

• Challenges– Radiation environment

around the detector• 10 year Total Integrated

Dose - <200 kRad• Acute Radiation Effects

on FPGA configuration SRAM memory

– Data Bandwidth• 3.38 Terabits/sec

– Number of Data I/O Lines

• ~17k LVDS Pairs

• Solutions– Use Radiation Tolerant

FPGAs close to the detector

• ACTEL FLASH based FPGA

– IR Data Compression Circuitry

• Sync word removal

– FO Data Transport to FEM• 16 - 2.5 Gigabit/sec FO

data links

3

Mark Prokop, FVTX Review Nov 2008

Overall Readout Strategy

• ½ of each detector arm is read out independently

• 6 ROC cards collect and compress the data from the detector

• Each ROC card sends fiber output to two FEM boards in the Counting House

• Slow Control fiber sends control data stream up/down the FEM↔ROC Slow Controls link

• Clock Distribution Board distributes Beam Clock and Start signals to individual ROC boards (the signals are sent over dedicated optical fibers)

Mark Prokop, FVTX Review Nov 2008

DAQ

• ROC board layout started• Clock Distribution Board layout started• Components, cost, power consumption

estimated• Electronics supply voltages regulated on ROC• Voltages Required:

– 3.3V: FO Tx/Rx– 2.5V: SERDES, ROC I/O, Wedge– 1.5V: FPGA Core

• Total power per ROC board ~ 20 W board + 18 W for V regulators = 38 W

• Total Power per ½ Arm = 228 W• Total Power per Arm = 456 W• Total Power for Detector = 912 W

5

12 ROC boards + 2 Clock Distribution boards in a “big wheel”

Mark Prokop, FVTX Review Nov 2008

ROC Board - WBS 1.5.2

6

Slow Controls

FPGA

Gas Seal Area

FPGA A

FPGA B

FPGA C

FPGA D

8x FO Tx

8x FO Tx

Slow Controls

FO Tx/Rx

SERDES

SERDES

SERDES

SERDES

Wed

ge C

on

necto

rsW

edg

e Co

nn

ectors

Wed

ge C

on

necto

rsW

edg

e Co

nn

ectors

Power Power

JTAGFPGA

SC SERDES

10.2”(W) x11.75”(L)30°

Mark Prokop, FVTX Review Nov 2008

FEM Board – WBS 1.5.3

7

8x FO Rx

Slow ControlsFO Tx/Rx

SlowControls

FPGA

MainFEM

FPGA

SERDES

SERDES

SERDES

EEPROM

RotarySwitch

Power

Power

Bu

s Interface

Slow ControlsSERDES

DCM II SERDES

DCM II FO Tx

VME 6U Board

Mark Prokop, FVTX Review Nov 2008

Risk Factors

• FVTX DAQ design utilizes well tested, commercially available parts, nothing “state-of-the-art”.

Risk – low

• Design of the ROC requires a significant amount of digital signal tracing and excellent ECAD skills. Risk – moderate

• FEM board is simple VME board with a few components. Moderate number of I/Os.

Risk – low

• None of the WBS 1.5 items is on the critical path

8

Mark Prokop, FVTX Review Nov 2008 9

SummaryWBS 1.5 – Technical

• Specification Document TDR, DAQ overview document• Prototype status ROC and FEM pre-production

prototypes – started• Electronics Components All available from distributor

stock• Production Quantities 24 + 4(spares) ROC boards

48 + 6(spares) FEM boards• Heat load 38 W per ROC board

456 W per arm• On-project Manpower 1 EE + 1 ECAD• Institutions Involved LANL• Infrastructure Defined draft• QA procedures in place Full QA plans for ROC and FEM

Mark Prokop, FVTX Review Nov 2008

DAQ Implementation

Backup Slides

Mark Prokop, FVTX Review Nov 2008 11

SummaryWBS 1.5 - Schedule

• PHENIX System Test: 1/09-3/09• ROC Pre-prod Proto: 10/08-1/09• FEM Pre-Prod Proto: 3/09-7/09• ROC Production: 10/09-1/10• FEM Production: 10/09-1/10

Mark Prokop, FVTX Review Nov 2008

SummaryWBS 1.5 - Cost

• ROC Pre-production: $74k

• FEM Pre-production: $85k

• ROC Production: $390k

• FEM Production: $350k

• Ancillary: $116k

• Fibers and Lab Equipment: $124k

12