ATTENDEES

Steph, Till, Kuhkee, Debbie, Diane, Patrick, Hamid

ACTION ITEMS

  • Everyone, task list and 'best estimates' of effort to Diane by Friday April 24
    • Kuhkee - RF Actuator (including 120Hz operation, network interface, 'pattern awareness')
    • Steph - timing system changes
    • Till - network and bpms
    • Debbie&Diane - controller,
    • Debbie - magnets
    • Diane - blens
  • Debbie and Diane , Controller Reqs & Conceptual Design on Friday 24th at 11am meeting
  • Patrick to plan for LCLS interface with FACET timing as part of Linac Upgrade project. Must be in place before Oct 2010.
  • Steph to look at SLC actuator code and see if/how actuator DES and ACT were compared.

TOPICS

Req. & Conceptual Design Review:

Diane and Debbie will present the Controller IOC Reqs and Conceptual Design to the Controls dept on Friday 24th.

Network API:

Till is working on a network API proposal for all of us to review. Till suggests that he create the higher level interface first, placing it over regular BSD sockets, so that the rest of us can begin integrating the network interface to our subsystems. He will add the lower levels as it is developed, but integrating these levels should be invisible to our subsystems.

NOTE:

There is a caveat regarding the use of the high-level API over regular BSD sockets: it might well be that IP multicast is not supported or blocked by the SLAC network so that testing software using regular sockets might not work.

Test Facility:

Ernest has been working on the Fast Feedback Test Lab. Gathered the requirements, hardware list, created a task list and estimated effort. The current plan is to put this lab into the small conference room downstairs, adding fiber from the EVG Steph has upstairs. Thanks Ernest.

Actuators:

We discussed the Actuators in more detail - specifically the 'pattern-aware' requirement and the error checking on the actuators that is required by the feedback loops.

  • Pattern-aware - Each actuator will be designed with N pattern-aware DES values, and configured with an associated pattern, where the value is applied to the actuator when the pattern appears. For the initial implementation N can be set an a nominal value of N=10.
  • Error checking in Actuators - readback value (ACT) can be compared to desired value (DES) and alarm if they are not within tolerance. This can highlight when a magnet power supply is not working properly; apparently a common occurance with SLC. Since we are using the old power supplies we may want to do this check. Steph to research this. RF internal feedback loops essentially make DES=ACT when there is a PAD...So we can get an alarm if there is a problem in the stations with a PAD.
  • No labels