Controls Meeting 8-Sep-2011

From GlueXWiki
Jump to: navigation, search

Agenda

  • Announcements
    • Power engineer being hired
    • GlueX collaboration meeting 6-8 Oct
  • Minutes from previous meeting 11-Aug-2011
  • Equipment orders - Mark, Hovanes
    • test equipment
    • rack hardware, Ethernet power strips
  • Test Lab status
    • what has Sarin come up with concerning the Danfysik power supply?
    • what needs to be moved out and when?
  • Upgrade to solenoid control system - All
    • needed about one year from now
    • new hardware design based on experience in Test Lab
    • need to collect requirements for hardware and software
    • do we want to upgrade to Ethernet/IP control on Danfysik power supply (from RS232)?
    • need subgroup to work on this
  • A-B licenses and license server - Josh
    • add licenses from other groups? dedicated licenses? priorities?
  • Interim networking in the hall - Elliott
  • Big UPS ordered - Elliott
    • should we run circuits to the hall? If so, to what?
  • Solenoid controls racks in the hall - Elliott
  • ControlNet vs Ethernet/IP - Hovanes
  • Controls test stand, BOY, etc. - Hovanes
  • Nerses Gevorgyan coming - Hovanes
    • what should he work on? device drivers vs EPICS utilities vs Java stuff vs ...?
  • Target controls needed by next summer
  • IRMIS - Elliott
  • Mantis task tracker



Time/Location

9:30 AM Thurs 8-Sep-2011 CC L207


Announcements

Next Meeting

Thurs 22-Sep 9:30am CC L207


Minutes

Present: Elliott W, Josh B, Hovanes E, David L, Scot S, Mark S.


Equipment

  • Mark ordered two meters and an oscilloscope with FY11 money.
  • We have additional money in FY12 for test equipment.
  • Mark and Scot will propose an order for rack hardware, Ethernet power strips, etc, needs to be in by 1-Oct.


Test Lab

  • We need to be out by 1-Oct according to the nominal schedule.
  • Sarin and Scot worked on the Danfysik power supply. Sarin understands the slow dump mechanism and he and Scot verified it is working. The problem was when we went to a smaller dump resistor another part of the circuit needed modification. They found it worked with the nominal dump resistor.


Solenoid Control System

  • A subgroup (Elliott, Mark, Josh, Scot) will meet regularly to design the new system based on experience in the Test Lab.
  • This needs to be operational about one year from now.
  • Ethernet control of the Danfysik needs to be investigated, perhaps in cooperation with Hall C (Steve Lassiter).
  • Upgrades to Danfysik recommended by review committee need to be evaluated.
  • Fast and slow dump chain strategy needs to be revisited.
  • New quench detector needs design.


Allen-Bradley Licenses

  • Josh received another RSLogix license and a ViewPoint license (one seat).
  • We are setting up an A-B license server on a CUE virtual machine (visible everywhere on site).


Interim Networking in Hall

  • Elliott worked out an interim networking plan with Bryan H.
  • We may purchase our big switch sooner rather than later since no price breaks are expected with time (unlike computers).
  • We will initially rely on wireless in the hall, but can run temporary copper if needed.
  • Main fiber run to the hall will come after platform is complete.
  • Elliott will investigate running fiber to the trailer (to eliminate $250/month cell-phone modem cost).


UPS

  • Walt ordered our big counting house UPS.
  • Delivery by 1-Jan, should be able to install immediately in the counting house.
  • Discussion on what systems in the hall, if any, should be connected to this UPS.
    • solenoid controls racks seemed to be a good candidate.
    • some network switches in the hall, esp. those supporting controls.
    • target control system?
    • other systems?
    • how much power needed and how long?
    • note that most components connected to the UPS (computers) will be shut down during power loss within 10 mins.


ControlNet vs Ethernet/IP

  • Current design uses Ethernet/IP connections between all controls chassis.
  • Hovanes wondered whether the solenoid control system should have a local PLC and run ControlNet between chassis.
    • advantage is that it will still run if network is down.
    • disadvantage is you need local PLC, or must run copper to the counting house.
    • Bryan Hess reports that modern networking gear is extremely reliable.


Controls Test Stand

  • Hovanes reported continuing progress in test stand.
  • Some equipement in the Test Lab will be moved to the test stand.
  • Hovanes is testing some new EPICS utilities, more need to be tested.
  • Ongoing/upcoming projects include FCAL bases and CAEN/Wiener crate control.


Nerses G

  • Nerses could arrive as early as Oct, but he needs a visa first.
  • He will likely work on both low-level EPICS drivers, crate control, as well as high-level EPICS utilities.
  • Elliott noted that the new EPICS Java framework might be interesting, he will find out the latest on this.


Target

  • Control system needed by Summer 2012.
  • Should be easy, all control loops will likely reside in manufacturer hardware.
  • All we need to do is download setpoints and monitor health.


IRMIS

  • Schema mostly stable, user interface in a state of flux.
  • New interface should be out this fall.
  • Current interface somewhat cumbersome for bulk entry.
  • Does IRMIS have a good long-term future, or might it fade away? Fewer developers working on it now.