Interlock minutes of a meeting held on 7th of December 2009

Present: B. Puccio, K. Hanke, J-L. Sanchez, J-C. Bau, R. Scrivens, M. Paoluzzi, B. Mikulec.


  1. How can the beam pulse or parts of it be removed at source level? (R. Scrivens)
  2. Additional discussions (all)

1. How can the beam pulse or parts of it be removed at source level?

R. Scrivens presented a few slides to give some answers to the question above. Measurements on the source have been carried out at the 3 MeV Test Stand. The ~400 us long source pulse will have a head and a tail that has to be removed. The average beam current is about 17 mA at the moment with a slightly decreasing pulse shape that is very reproducible.
  • Remove head: The pre-chopper (electrostatic deviation of beam at 45 keV) has to be started well before the source (order of magnitude: ~ 1 ms). It will remove the part of the source pulse corresponding to the rise-time and some space charge compensation; the fall time of the pre-chopper is <2 us (to 99%). A timing (NX.STOPPCHOP) switches off the pre-chopper (reaction time: ~3 us); this timing can be started relative to a PSB injection warning. Additional time of the order of 30 us is needed to re-establish space charge conditions and will be removed with the chopper. A screen shot of the timings currently used for the pre-chopper can be seen on slide 3.
  • Cutting the beam within pulse: This could be achieved switching off the RF of the source (decay time of ~10 us; too long). The pre-chopper could be used in addition (chopping to about 2 us), but it is not clear if the source could be easily re-ignited and moreover the time would be too long to return to normal beam (~30 us of space charge compensation). In summary, this option has to be excluded.
  • Remove tail: The pre-chopper should be used. A timing, NX.STARTCHOP, is used to switch off the beam (fall time of about 2 us). The extinction ratio with beam is still unknown. It will be possible to measure the delay time to switch off the RF.
  • Remove the beam completely: There is an interlock from the source to the 2 MHz RF. The reaction time of this system should be determined (see below). Tests have been made on the extinction ratio of the source with this interlock, and less of 2 uA of beam were measured. The pre-chopper should be switched on during >the duration of this operation in addition.

M. Paoluzzi adds that the reaction time of the RF switch is expected to be of the order of 100 us, but this still needs to be tested. This time cannot be much reduced as the LL control loop runs at kHz frequency. A second option to cut the source would be to set the grid screen voltage to 0 V; one would have to count with a delay of ~50 us to go from 1500 V down to 0 V. It was therefore agreed that the BIS should act on the screen grid (a fast interlock could therefore see a reaction of the order of ~50 us). The interlock system should send in parallel to the signal to the screen grid a logic pulse to be integrated in the low-level RF system for the source.

Interlock on single PSB rings: from the above it is clear that in such a case the source pulse has to be shortened and the distributor timing adjusted accordingly. R. Scrivens mentioned that tests have been undertaken changing the source pulse length. This is possible, but it affects the pulse intensity: a difference of ~1 mA for ~16 mA average beam current has been measured. It should also be discussed with P. Baudrenghien how the Linac4 RF will react if a reduced number of PSB rings will be filled.

Assigned to Start date Description State Result
B.Mikulec 2009-12-07 Contact P. Baudrenghien to discuss ring-specific interlock situations with respect to the Linac4 RF and the delay of the LL control to the chopper.   edit

Summary pre-chopper characteristics (to be confirmed with measurements):

  • Start ~1 ms before source
  • Reaction time: ~3 us
  • Fall time: ~2 us
  • Extinction ratio to be measured

Summary chopper characteristics (to be confirmed with measurements):

  • Delay of LL control: probably a few ns (to be discussed with Ph. Baudrenghien)
  • Rise and fall time: a few ns
  • Max. duration of chopper pulse: 50 us

2. Additional discussions

J-C. Bau explained that the max. decision time for the calculation of the external conditions can be up to 3 basic periods. As this is already the case and cannot be reduced, we propose to accept this situation also for Linac4.

The tail clipper signal should be included in the master BIC as input and has to result in cutting the full beam.

An improvement would be if the interlock of the slow bumpers could be separated from the fast extraction kickers (per ring) and if possible should be a direct input into the combined PSB/BT BIC.

Assigned to Start date Description State Result
B.Mikulec 2009-12-07 Follow up a separate interlock can be created for the BSW and connected to the BIC. This will be possible. edit

All ring-specific interlocks should be treated as now with the external conditions.

-- BettinaMikulec - 07-Dec-2009

  • 2009-12-07-Beamkiller-Scrivens.pdf: What can be done at the source level to cut parts or the whole Linac4 beam pulse?%META:FILEATTACHMENT{name="2009-12-07-Beamkiller-Scrivens.pdf" attachment="2009-12-07-Beamkiller-Scrivens.pdf" attr="" comment="What can be done at the source level to cut parts or the whole Linac4 beam pulse?" date="1260206106" path="2009-12-07-Beamkiller-Scrivens.pdf" size="192579" stream="2009-12-07-Beamkiller-Scrivens.pdf" tmpFilename="/usr/tmp/CGItemp17105" user="bettina_2emikulec_40cern_2ech" version="1"}%
Edit | Attach | Watch | Print version | History: r7 < r6 < r5 < r4 < r3 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r7 - 2010-11-11 - BettinaMikulec
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    SPL All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright