120Hz BSA Time Slot Pulse Stealing

1
120HzBSATime SlotPulse Stealing
2
120Hz

• Currently, BPMs and toroids acquire at 120hz with
  no problem.
• Wire-scanner PMT acquisition at 120hz is not
  tested yet but no problems expected.
• Bunch length measurements need to be upgraded to
  faster PAD--gtVME communication.
• Some LLRF PADs in IN20/LI21/LI24 can now acquire
  at 120hz. However, the PAD--gtVME--gtPAC
  communication (LLRF control) is not yet ready for
  120hz Dayle is working on it.
• Dayle fixed diagnostics LLRF PAD IOCs that went
  CPU-bound at
• Network and clients not yet studied at 120hz.
  One of the potential problems of 120hz will be
  due to clients (EDM, etc) that monitor
  120hz-updating PVs (some of which could be
  waveforms). Clients that connect to already
  loaded PAD IOCs and need 120hz updates may cause
  problems.
• Displays, archiver, etc should connect to slower
  PVs when at all possible.
• Clients should connect to PVs on MVME6100 IOCs
  that "shadow" PVs on PAD IOCs.
• HLA and Matlab apps may need to use slower PVs.

3
BSA

• BSA users who care about timeslot will need to
  set either "TS1" or "TS4" in the inclusion mask
  when they setup the EDEF.
• Double-processing of the same AI in the same
  8.3msec pulse new version of event software
  released to throw out any data with unchanged
  timestamp.
• Remember devices which cannot be
  acquired/available in 7.3 msec after beam, cannot
  be used in BSA.
• If a device can make the 7.3msec time budget but
  is rate-limited for other reasons, BSA users must
  select a rate-limit option in the inclusion mask
  of the EDEF.

4
Time Slot

• RF feedbacks will need to do things differently
  based on timeslot. Need spec.
• Some Matlab (and future fast) feedbacks will need
  to do things differently based on time slot.
  Need spec.
• Time slot is currently available at 120hz
  (ltiocgtPATTERN.C) to all IOCs with EVRs.
• So far, no changes, based on time slot, are
  required for all beam-related non-RF monitors
  (BPMs, toroids, PMTs, bunch-length).
• So far, MPS will need no changes based on time
  slot (though Stephen is using time slot for other
  reasons).

5
Pulse Stealing Sector 25

• TCAV3 OTR - trigger enabled, event code 160
  (TCAV3 10hz). For background images, the TCAV3
  OTR event code needs to be set to 159.
• TCAV3 Modulator - if TCAV3 is to be activated on
  stolen pulse, its modulator must be activated on
  the proper pulse(s) via the old control system.
  The MPG tags the pulses that have activate TCAV3
  so the EVG can send out event code 156 used by
  the TCAV3 PAC and LI25 kicker and event code 160
  for TCAV3 OTR.
• LLRF TCAV3 PAD - triggers all the time but will
  not feedback on pulses that TCAV3 is activated.
• LLRF TCAV3 PAC - triggers off-beam except when
  TCAV3 is activated.
• LI25 kicker triggers off-beam except for kicked
  pulses. To kick, it must have its kicker (aka,
  abort) trigger enabled and event code set either
  a desired rate OR only when TCAV3 is activated.
  GUI and PVs to do this are available.
• Devices downstream of LI25 kicker will still
  trigger on stolen pulses. Problems?.
• BSA any BSA that only cares about data when
  TCAV3 is activated needs to set "TCAV3" in its
  inclusion mask when they set up the EDEF. Apps
  must check status of downstream devices to know
  if beam is there or not.

6
Pulse Stealing LTU

• 2 trigger inputs to the LTU kicker
• Standby (constant rate) EVR trigger with proper
  delay to kick off beam.
• Standby EVR trigger with different delay to kick
  on-beam. (?) MPS enables or disables this
  trigger based on fault conditions and user
  request. What happens when EVR stops triggering
  due to reboot or other problems?
• No special devices (no OTR, no TCAV, etc) in the
  dump line which gets the kicked pulses that
  require anything special from the event system.
• Right now, all devices downstream of the kicked
  pulses will trigger as if there is beam.
  Problems?
• BSA apps will need to check device status to know
  if device is on beam or not.
• Will the status (pulse kicked or not) be in the
  pattern for BSA to set in their masks?