Online Major Milestones

Introduction

The complete DAQ (approximately 80 full crates) must be ready for testing the complete detector with cosmic or pulser triggers by 1Q FY 2014. First beam to the hall is expected in 4Q FY 2014, with low-rate production data taking expected to start in 4Q FY 2015. High-rate production data taking may start as early as 4Q FY 2016, assuming funds are available to upgrade the detector. Beneficial occupancy of Hall D is expected in summer 2010, counting house in fall 2011.

Note that the 12 GeV project specifies a beam rate of 10**7 photons/sec, but that we plan an eventual upgrade to 10**8. All components must be designed to be capable of, or upgradable or expandable to, 10**8 running.

Needed for Subsystem Review May 2008 and for CD-3 Review Summer 2008

Understand event size and resulting backplane and network bandwidth requirements.
Revisit and verify energy-sum trigger algorithm using latest hardware design.
Interface document from Electronics and/or DAQ group defining timing distribution system.
Interface document from Electronics group defining energy-sum and master trigger systems.
Interface document from Electronics and/or DAQ group defining trigger distribution system.
Interface document from DAQ group defining CODA 3 major components and functionality.
Develop conceptual design of DAQ system specifying strategy to integrate devices and software described in the above four documents into a working system.
Preliminary libraries for reading out existing custom boards (F1TDC, FADC).
Demonstrate 200 kHz triggering and high-speed readout (> 100 MB/s) of multiple FADC's in a single crate.
Choice of controls framework and basic implementation plan.

Trigger/DAQ Milestones

2006

Production event transfer system - done Aug 2006
Production event I/O package - done Aug 2006

2007

DAQ test farm by 28-Sep-2007 (PED deliverable) - done Jan 2007
Prototype software component framework - done Jun 2007
Prototype run control system - done Jun 2007
Production inter-process communication package - done Nov 2007

2008

Subsystem review May 2008 and CD-3 review July 2008, see milestones above
Develop CODA-based electronics test stand by Jul 2008
Preliminary energy-sum and master trigger design by Sep 2008 (PED deliverable)
Prototype staged/parallel event builder by Sep 2008 (PED deliverable)
Prototype single crate timing and trigger distribution by Nov 2008
Simulation of full system at low rate (run control, ET, multiple EB's, ER, etc) by Nov 2008
Prototype TI board by Nov 2008
Prototype energy-sum and master trigger boards by Nov 2008
Demonstrate 200 kHz staged parallel event building by Nov 2008
Demonstrate 200 kHz block event readout and transfer by Nov 2008
Prototypes of major CODA 3 components, including readout controller, staged/parallel event builder, event recorder, and run control system by Nov 2008.
Document overall strategy for hall/accelerator inter-operation with Accelerator division by Nov 2008.

2009

Prototype L3 farm management software by Sep 2009
Master trigger design by Aug 2009 (PED deliverable)
Prototype multi-crate timing and trigger distribution by Oct 2009
Overall trigger design by Oct 2009
Prototype TS board by Oct 2009

2010

Demonstrate multi-crate high-speed readout
Simulation of full system at high rate
Production front-end driver software
Production timing distribution system
Production trigger distribution system
Production experiment control system
Production component framework
Production software components
Production TI board
Production TS board
Production energy-sum and master trigger boards

2011

Integration and test of all final Trigger/DAQ system components by Sep 2011
Begin to build trigger and DAQ in Hall D as electronics and detector components begin to arrive

Trigger/DAQ Detector Needs

DAQ systems of various sorts will be needed during detector development, construction, and checkout. Below, "parallel" means that parallel or staged event building is needed. "Small" means four crates or less, and "large" means more than four crates. In general, parallel readout is only needed at very high rates, where a single network connection cannot handle all the data.