Choose timezone
Your profile timezone:
LIU Beam Parameter WG meeting #8
→
Europe/Zurich
874/R-018 - MEETING ROOM PRIORITY BE OP (CERN)
Follow up from last meeting
Transverse feedback in PS
- There is already a plan to follow up on this subject at the next meeting on 22 July with two presentations: Guido will provide the specifications for the transverse feedback in the PS, and Wolfgang B. will present the injection studies motivating these specifications. The injection of longer bunches (220 ns) for the standard LHC beams will also be part of the discussion.
Longitudinal emittance blow up in PSB with injection of phase noise (Danilo Quartullo)
- The blow up reviewed at the last meeting (and followed up in Simon's talk) makes use of the phase modulation of C16. The method of injecting band limited noise in the RF phase does not need a second RF system (and is presently used in the SPS and LHC), and could be applied also to the PSB
- BLonD simulations have been run for a post-LS2 scenario
- They include an accurate impedance model (Finemet cavieties, resistive wall, ejection kickers, etc), space charge, multi-turn wake, periodicity, phase and radial loops, multiple RFs, RF phase noise injection.
- The accelerating cycle from 160 MeV to 2 GeV designed by A. Blas has been used, however it would be good also to review it in the future and check whether it could be optimised in terms of longitudinal acceptance and Bdot variations.
- The beam parameters considered are the HL-LHC target parameters from the existing LIU tables with a slightly higher intensity to allow for margin. The injected beam is at the moment just a rectangle in the longitudinal phase space, it would be good to review the injection studies done by Elena Benedetto and Vincenzo Forte (future brightness line with Linac4) to be sure that the assumptions in terms of pulse length and energy spread (determined by the chopper and debuncher cavity) are consistent.
- The case of acceleration with 8 kV does not leave much space for the longitudinal emittance blow up to 3 eVs as this value of bucket area would be only reached less than 100ms before extraction. Besides, including intensity effects, this point moves even closer to extraction and there are losses along the cycle just due to limited bucket acceptance. However, when accelerating with 8 kV, 24 gaps of the Finemet cavities can be short-circuited, which would actually zero the losses when impedance and space charge are not included, but still leave very little margin for emittance blow up.
- The case of acceleration with 16 kV is more promising. Without intensity effects, the bucket area becomes 3 eVs before c500, with intensity effects slightly later but still leaving few hundreds of ms to apply the longitudinal emittance blow up.
- By applying band limited noise between 0.8 and 1.5 times fs between c450 and c600, the emittance can be efficiently blown up to the desired value of 3 eVs.
- The noise needs to be regenerated every few thousands of turns to follow the changes in frev and fs (the former changes by a factor 2 along the cycle, the latter by a factor 5)
- The final bunch length is 175 ns instead of the targeted 205 ns, so the bunch should still be lengthened before extraction (voltage reduction, second harmonic?)
- The LLRF needed to make this possible in actual machine studies is not yet available and would need some time to be deployed. The LLRF team will be mainly busy on LEIR and ELENA in the next few months, so that it becomes unlikely that this will be done before 2017 (large longitudinal emittances and longer bunches should be anyway available from the PSB using other blow up techniques, which could be injected already in 2016 into the PS for space charge studies at 1.4 GeV). On the positive side, however, the origin of the synchro problem at PSB extraction seems to have been found.
- This method based on injection of phase noise is used in the SPS (although lately not much emittance blow up is applied in the SPS thanks to the use of the 800 MHz RF system for beam stabilisation) and an idea would be to have simulations benchmarked against SPS data.
Longitudinal emittance blow up in PSB with parametric excitation (Simon Albright)
- The method was illustrated at the last meeting and depends on several RF parameters.
- A three step procedure with the present accelerating cycle (to 1.4 GeV) has been simulated including space charge. The voltage is trimmed to create three plateaus of constant bucket area during which the phase modulation of C16 is applied with constant parameters.
- The target parameters of 2.8 eVs and 220 ns bunch length with bunch shortening mode h1 + h2 = 8 kV + 8 kV can be obtained, but the loss of landau damping causes a lingering longitudinal structure on the bunch, which can be smoothed out by use of C16 voltage + modulation.
- For the future, assuming an available voltage of 24 kV, the blow up is demonstrated to work with a single step procedure (at constant bucket area of 5 eVs). Space charge is presently not included in these simulations.
- Machine tests have been made using the three step procedure starting from a cycle with 8 kV on C02, low voltage on C04 and 1 kV (ramped to 2 kV) on C16
- First, it is verified that the introduction of the C02 voltage dips (and trims on C16 voltage) to create the constant bucket area plateaus does not cause changes in the beam quality at extraction
- A program for the amplitude (Am) and frequency (fm) modulation for C16 is established.
- Settings are found to produce 2 eVs with 220 ns at extraction for all four rings. These settings are found to strongly differ ring by ring. This is mainly due to the present lack of control of the C02 and C16 phase.
- Reproducibility studies show that these parameters can be preserved on the relative cycle at least with the same repeatability of those for the standard 25 ns beam.
- To improve this method and make it more easily operational, the LLRF requirements are:
- Am (modulation depth as a function of time) + ωm (modulation frequency as a function of time) → already available
- θ2 (phase of C04 relative to synchronous phase)
- θ16 (phase of C16 relative to synchronous phase)
- Beam loading compensation on θ16
- In a meeting with the LLRF experts it was decided that all these requirements are technically feasible, as well as the noise injection to enable the blow up method described by Danilo. C16 phasing and noise injection will be activated (one this year, the other next year to allow both to be studied before LS2), with order and timeline defined by maturity of studies as well as progress on LEIR and ELENA. However, it is currently intended that both will be possible after LS2 during the same cycle if required.
- A. Findlay remarked that large longitudinal emittance LHC-type beams (up to 2.8 eVs) can also be produced with the standard blow up procedure based on pure C16 voltage trim over a certain part of the cycle. These can also be produced quite reproducibly and can be sent to the PS for MDs. It is clear that the final procedure used for the longitudinal emittance blow up in the PSB will have to be both mastered and understood, and simple enough that it can be used operationally without any expert assistance. Since we already have large longitudinal emittance + long bunches available at the PSB exit for PS studies, the implementation of the above described features in the LLRF control of the PSB is certainly desirable over the timescale outlined above and there is no need to request anything earlier than presently planned.
There are minutes attached to this event.
Show them.