OWG Meeting 18-Mar-2009

From GlueXWiki
Jump to: navigation, search

Agenda


Time/Location

1:30 PM Wed 18-Mar-2009 CC F326


To connect by telephone:

1.) dial:

  • 800-377-8846 : US
  • 888-276-7715 : Canada
  • 302-709-8424 : International

2.) enter participant code: 39527048# (remember the "#")


Announcements

Next Meeting

1-Apr-2009 1:30pm CC F326


New Action Items from this Meeting

Minutes

Attendees: Elliott W, Beni Z, David L, Alex S, Vardan G, Ed J, Carl T, Graham H, Ben R, Dave D, Chris C, Elton S, Mark I, Eugene C.


DAQ Update

Graham gave a short report on recent happenings in the JLab DAQ group.

  • In process of hiring new person to help Dave A with VxWorks and embedded computer issues. Have interviewed a few people.
  • Lost additional hire during freeze, but may get it back with new budget. This person would help Ed with electronics projects.
  • Software moving along at expected pace.


EPICS Demo and Discussion

Elliott showed a demo of the EPICS StripTool, Alarm Handler, and the JLab Accelerator archiver. PV's were served from a Linux softIOC running on a desktop. The data was just software-generated sawtooth functions. Hall D likely will not use these particular software packages as better ones are under development at the SNS and other places. Purpose here was just to demonstrate a proof-of-principle of using Linux softIOC's and EPICS-aware utilities.

  • Drivers must be written to access underlying hardware in real world.
  • If you are lucky someone in the EPICS community has already written the driver.
  • Most drivers are written for VxWorks, which we do not plan to use.
  • The softIOC makes the data available via the Channel Access (CA) protocol. Data can be accessed by CA clients that use a very simple client library.
  • The client library implements two-way access (read/write) so clients can both read hardware data and set parameters in the hardware.
  • Important question is efficiency of EPICS for bulk downloads and readbacks.
  • Chris mentioned we need end-to-end drawings for all electronics and controls systems. The IRMIS system can help, but it is not a substitute.


We also discussed the potential role of Labview in the Hall D online.

  • JLab has a Labview site license, and Labview runs on Linux.
  • Linux Labview plugins are available that can read EPICS CA data. I.e. a Labview application can be a CA client. Not clear whether a Labview application can be a CA server.
  • Labview applications can be compiled and no license is needed for deployment. Licenses only needed for development.
  • Simplest use is as an GUI development tool, i.e. as a replacement for MEDM, EDM, CSS (Java/Eclipse based package), etc. All these packages do the same thing, wrap graphics around the basic CA client library. Labview is just another variant on the same theme.
  • Big advantage is that Labview is way more advanced than any of the others concerning graphical display (e.g. it has a much bigger widget library).
  • Further very big advantage is that Labview supplies analysis tools that are completely absent in MEDM and EDM, or that have to come from outside packages if CSS is used. Examples are endless: Fourier transform analysis, statistical analysis, multi-dimensional analysis, etc.
  • An important question concerns memory and cpu resources needed by Labview applications.


JLab Hall B Discriminator

Ben discussed the Hall B discriminator board. It is being redesigned for the 12 GeV upgrade, and may be interesting to Hall D. See his presentation for numerous details.

  • 16 channel, lemo inputs.
  • 20 boards deployed in Hall B in 2001, more needed, old ones probably will be replaced (obsolete parts).
  • 12V DAC per channel, dual threshold, 100 MHz.
  • Trigger outputs, scalers, test inputs, FPGA based.
  • Custom logic on board for Hall B. Board works if these circuits are not populated.
  • Also works if one of the two-per-channel threshold circuits is not populated.
  • Uses a very low dispersion comparator, one of the cost drivers.


Electronics Update

Chris gave a brief update, Fernando was at the doctor's office.

  • Looking to hire an additional engineer. Likely to get approval soon.
  • Trigger test stand running with two crates, CTP in each.
  • SD testing soon in 20-slot crates.
  • Full two-crate tests (FADC, CTP, SD, timing, trigger generation) in about two weeks.
  • Ed said he got dual-edge source-synchronous transfers working.
Retrieved from "https://halldweb1.jlab.org/wiki/index.php?title=OWG_Meeting_18-Mar-2009&oldid=12538"