Attendees: Patrick, Steph, Kuhkee, Debbie, Till, Hamid
Topics:
Requirements / Requests from Joe Frisch (physicists) and Howard Smith/Mike Stanek (operations)
1) pulse stealing - how can feedbacks deal with this?
We concluded that the feedbacks need to learn when pulses are missing from the pulse pattern. This happens with the kicker magnets - there is a plan to add the kicker control to the EVG, but not scheduled. This work must be done by Steph before the downstream feedbacks can respond to the kicker activity. We will have to add this effort to the fast feedback schedule if this becomes a requirement. Stephanie estimates it will take ~1 month of her time.
2) Timeslot control and FACET - it is known that when FACET is running there will be a ~30Hz disturbance on the beam.
In order to correct for this in feedback, the feedbacks will need to know which timeslot FACET is running on. If FACET and LCLS share an EVG, or if we can get that FACET timeslot information to LCLS, then the feedback algorithm can work with it. Additionally, the 120Hz feedback algorithms may need to change significantly (to a base + difference type of algorithm) when FACET is on. It is projected that FACET will begin operating sometime in 2011 - It was decided we did not have to deal with these issues immediately, but plan for it with FACET. We should make it a goal to develop a good test environment so that we can test new algorithms after January 2010, to prepare for FACET.
3) RF control and longitudinal feedback
we need to develop a transition path for upgrading Joe's RF control so that it works cleanly with higher level applications like the longitudinal feedback, phase scans, energy_change, etc.
Other Items:
Till to purchase 2 units AMPRO Readyboard 820 EPIC SBCs
BLENs are already using the private network interface between VME and coldfires