Overview, remarks, lamentations, hope and despair

Overview, remarks, lamentations,

hope and despair

M. Sozzi TDAQ WG meeting 14 December 2011

M. Sozzi

NA62 TDAQ Working Group meetingFebruary 8th 2012 - CERN

Introduction and stuff

Agenda

TDCB• 12 boards + cables distributed in November for

validation (sustainable rate, time resolution, noise, etc.): CEDAR, CHANTI, LAV, RICH, MUV, IRC/SAC, STRAWS

• Performance/rate tests: report by S. Venditti• Tests in Perugia: report by M. Piccini• Firmware rewrite completed, testing to be started.• Test bench hardware not started yet.• Final version PCB started yesterday: no major

changes foreseen.

Pisa

Waiting for validation of the boards by sub-detector groups

(none so far)

IF MORE BOARDS ARE NEEDED FOR 2012 THEY SHOULD BE REQUESTED BY

FEBRUARY 2012

Otherwise NO boards will be produced until 2013 on top of those which you have

Price will depend on the number of boardsBase: 1000 €/board + 500 €/4 cables

TEL62Status report by F. Spinella

• Most parts of the board tested on prototype• Second prototype to be mounted• Control software for programming (JAM): not yet working,

Perugia working with LHCb to fix this • Final firmware, documentation, control software started

GUI for multiple-board control to be started

• Pre-production (10-12 boards for 2012 runs) to be started in 1 month.

• Full production planned for 2013 (Pisa+Roma TV). • Test procedure (Roma TV) not started yet• Need to decide on STRAWS readout needs

Pisa

From NA62 TWiki pages:

TEL62 firmware - II• Work can start now (on TELL1 if hardware testing needed)

• The developing TEL62 firmware is stored at CERN under SVN (instructions on TWiki). Concurrent development works fine.Anybody can start to contribute even on TEL62.

• Need specifications of sub-detector requirements for processing blocks (data, packing, registers, memory space for look-up tables, etc.)

• Firmware writing guidelines on TWiki

TEL62 firmwareWhich (reduced) functionality for 2012? Prioritization Manpower issues still unresolvedResponsible people (1 per sub-detector) (almost) identified:

RICH: C. Santoni (Perugia)LAV: M. Raggi (Frascati)CEDAR: K. Massri (Birmingham)LKr/L0: A. Salamon (Roma TV)CHANTI: D. Di Filippo (Napoli)MUV: R. Wanke (Mainz) ?SAC/IRC: V. Kozhunarov (Sofia) ?

Work can start now (on TELL1 if hardware testing needed)Need specifications of sub-detector requirements for processing blocks (data, packing, registers, memory space for look-up tables, etc.)

STARTEDSTARTED

(FOR RICH)

STARTED

TEL62 firmware - IIISTARTED

STARTED

STARTED

ADVANCEDSTARTED

(FOR RICH)

ADVANCED

TEL62 firmware - III

STARTED

STARTED

NEW SINCE LAST MEETING

Crates, racks

Racks positions finalized

8 “Front-end TEL62” crates ordered to Wiener + all (36) special backplanes: CEDAR: 1STRAWS: 1LAV: 3RICH: 1MUV: 1Spare: 1

Clock modules - I

LTU (NA62 version, Birmingham): paid for, ready, available, first firmware and software versions ready: talk by V. Cerny Modules available to sub-detectorsNew documentation page available on TWiki

TTCex (new version, CERN):paid for, built, PLL problem (random clock offset) in new version. Modules required by sub-detectors for 2012 available1 high-quality module required by RICH rented from EPOOL

Optical splittersTested, sub-detector requirements collected, common order placed

BirminghamBratislava

LTUs are at Marian’s

TTCex are at Riccardo’s

Marian’s proposal:

out of the LTU/TTCex required and available for 2012, each sub-detector takes 1 pair to their lab for testing and (in case they ordered more than 1), leave the second one at CERN, in order to setup the central clock crate for the 2012 runs.

Clock modules - II

Clock infrastructure• Fibers: Specification of optical fibers (incl. termination)

and definition of optical network for TTC: doneInstallation will be in March 2012

• Choke/error: cable (& connectors) specification (CAT7 + RJ45): doneTest of longest cable (150 m, EB20 to CEDAR) with LVDS tester (special firmware on LTU): February 2012Installation: April/May 2012

• Central VME crate: prepared with 7 LTU, test with TALK board (L0TP for 2012) at CERN: February 2012Installation in ECN3: May/Jun 2012

• Master clock generator: tested, installed, no spare• Burst emulator: installed, to be adjusted

BirminghamCERN

L0TP No progress since last meeting.PC-based solution with Terasic FPGA development board not yet demonstrated to be valid. Two boards available (in Ferrara and Torino).

• Ferrara is developing a software library to access data coming from FPGA. This would allow to measure time performances of primitive-matching algorithms

• Torino started some preliminary measurements on ethernet transmission latency

• No candidate for software (primitive-matching algorithm).

2012: use talk board for run (any prototype to be tested?).

2014: critical issue

FerraraTorino

E. Rebuffello, E. Menichetti (Torino): tests on TCP/IP latency using FPGA development board and a PC

Latency is way too large, but TCP is not what we want

FPGA clock frequency is not a relevant parameter

Farm and higher-level triggers

Work in Mainz on L1/L2 farm software infrastructure Requirements for 2012 not yet clearData format change with respect to NA62 note (use UDP)

Work in Pisa on: (1) fast driver (for farm and all PCs) and (2) microsecond-level time-synchronization protocol between hardware and PC (for PC-based L0TP and possible GPU-based L0 trigger): talk by L. Deri

No LKr L0 trigger simulation yetNo L1 recipe so far to reach 100 kHzBandwidth allocation proposal needed (also true for L0)

TDAQ for sub-detectors DAQ Trigger

CEDAR Standard. Rates? Front-end? Radiation? Something in 2012

-

GTK Non-standard. Ongoing. Not in 2012 -

CHANTI Standard. Available in 2012 -

STRAWS Non-standard? Not in 2012 -

RICH Standard. Not needed in 2012 Ideas, some work

CHOD Standard. Front-end? Some in 2012? Ideas, no work

LKr Non-standard. Ongoing. Not in 2012 (but NA48 readout available)

Ongoing. Missing details and MC

LAV Standard. Front-end OK. In 2012 Ideas, work starting

MUV Standard. Front-end? Unclear for 2012 Ideas, no work

SAC/IRC Unknown. Not in 2012. -

Non-standard readout systems

Gigatracker: not relevant for 2012, to be reviewed

LKr: not relevant for 2012, work ongoing, to be reviewed

Straws: work ongoing, review today: talk by P. Lichard Decision expected at this meeting between proposed system and fallback solution.

Documentation TDAQ TWiki developing: https://twiki.cern.ch/twiki/bin/view/NA62/TdaqSystemInstructions to setup TEL62 firmware environment OKFeedback needed !

OnlineDiscussions on Control Program and online environment started: talk by R. FantechiHelp from EN/ICE available (lots of LHCb software can be reused)Some manpower identified: N. Lurkin (Louvain) to start real work soonOnline responsibility (group) to be clarified later

Dry run / test run

We have a TDAQ run coordinator! See G. Lamanna talk (and react to it)2012 will be a limited (but still important) TDAQ test

Coming soon…(Expected for next meeting)

• Review of TEL62 firmware

• First detailed plan for Control Program

• Detailed discussion of trigger farm software

• Review of Gigatracker readout integration

• Review of LKr readout integration

Critical issues1. L0TP: no proof of viable solution, no progress

2. Online software: no work done, no group responsibility

3. TEL62 firmware: lack of manpower: not ready in 2012

4. Radiation tests: no progress

5. Crates: only very basic tests done before order

6. Simulation: missing L0 for LKr, L1 and L2

7. Missing specifications: e.g. fine time L0 trig info

Critical issues1. L0TP: no proof of viable solution, no progress

2. Online software: no work done, no group responsibility

3. TEL62 firmware: lack of manpower: not ready in 2012

4. Radiation tests: no progress

5. Crates: only very basic tests done before order

6. Simulation: missing L0 for LKr, L1 and L2

7. Missing specifications: e.g. fine time L0 trig info

a working solution for 2012 exists (TALK board)

some manpower, work starting soon, help from EN/ICE

minimal DAQ firmware should be available

minimal mechanical/electrical tests were done

specifications will freeze automatically