LIU-SPS Coordination Meeting
Zoom
LIU-SPS Machine Coordination Meeting 18/11/2020
Invited (from): Alexander Huschauer, Andy Butterworth, Anne Funken, Anthony Rey, Antonio Grande Mallavia, Arthur Spierer, Brennan Goddard,Chiara Pasquino, David Mcfarlane, David Nisbet, Elena Shaposhnikova, Eric Montesinos, Eric Veyrunes, Etienne Carlier ,Fabio Follin ,Francesco Maria Velotti ,François-Xavier Nuiry, Frank Gerigk, Gerd Kotzian, Giovanni Rumolo, Giulia Papotti ,Gregoire Hagmann ,Grzegorz Kruk, Hannes Bartosik, Ivan Karpov, Ivan Romera Ramirez, James Ridewood ,Johan Dalla-Costa, Jonathan Meignan, Julie Coupard, Kevin Shing Bruce Li ,Lars Jensen, Loic De Oliveira, Louis Pereira, Luciana Kolbeck, Luis Medina, Malika Meddahi, Marine Gourber-Pace, Mauro Taborelli, Paul Cruikshank, Philippe Baudrenghien, Reyes Alemany Fernandez, Sonia Bartolome Jimenez, Stephane Cettour Cave, Sylvie Prodon, Verena Kain, Wolfgang Hofle, Yannick Le Borgne
Matters arising and actions
- Dynamic injection pattern to synchronise with timing system and BQM, being followed-up.
- VSC storage of some samples under vacuum not yet followed up; to be checked with team
- High intensity scrubbing coating strategy for LS3; still open and needs to be folded into the schedule for 2023
- Can't have both long and short ion cycle in the SC, as too long, OP will look at longer term removal of this limit (SPS-OP)
- Update on ions commissioning planning: Discussions took place within RF; The sequence of steps needed for ion commissioning being prepared together with Reyes to come up with a solid plan, based on the time required for the different steps.
- TIDVG interlocking has been sorted out.
- RF power interlocking implemented for Thales. Still to be clarified how this will be implemented for the Siemens and Philips plants.
News on dry-run and test progress
Bunch-by-bunch distribution at SPS flat top for LHC injection losses (Luis Medina)
- BLonD simulations have been performed to model the bunch-by-bunch characteristics at SPS extraction. A model of the One Turn Feedback (OFTB) and the Feed Forward (FF) loops has been implemented. It has been benchmarked with measurements from 2010 with 72 bunches, and with measurements from 2018 with BCMS beams. These measurements have been used to calibrate the LLRF parameters in BLonD.
- A model for bunch intensity, bunch length and binomial exponent of the distribution has been established. Using this model, a distribution of 72 bunches can be generated based on a measurement of 48 bunches.
- To understand the residual difference between the model and measurements, the sensitivity to several parameters has been checked. The biggest uncertainty is probably still coming from the parameters used in the LLRF model.
- The bunch distributions generated with the model agree well with the measured distribution at LHC injection, in particular concerning the beam core. The tails cannot be easily measured and thus remain somehow a free parameter. Further studies are ongoing to assess the impact of the tails.
- Elena proposed to perform additional measurements in the SPS to assess the population of tails in the distribution, as this is an important parameter for the injection losses. This will be followed-up offline to see what types of measurements can be done (eg. reduce voltage until particles spill out of the bucket).
Bunch-by-bunch distribution expected from the rigid bunch model (Philippe Baudrenghien)
- A rigid-bunch model has been developed to assess the bunch-by-bunch phase shift due to residual uncompensated beam loading when including the LLRF loops and a model of the transmitters including their bandwidth limitations and the clamping. It considers only steady-state distributions. Each bunch stays at the stable phase of its perturbed bucket.
- The bunch-by-bunch phase deviation has been calculated for 4x72 bunches at 2.3e11 p/b, resulting in 2 degrees rms and 15.6 degrees peak-to-peak phase deviation within the available RF power. These numbers seem compatible with past measurements when taking into account the difference in intensity. However, the phase shift at the head and the tail of the batches seems larger compared to the multiparticle simulations with BLonD (presented by Luis before), which could be due to the fact that the latter do not take into account the clamping of the RF power amplifiers (requires extension of LLRF model in BLonD). This will be followed-up.
Preparation of controlled longitudinal emittance blow-up (Giulia Papotti)
- To achieve longitudinal stability of the high-intensity LHC proton beams, they require controlled longitudinal emittance blow-up. This can be achieved by adding band-limited noise to the phase loop.
- In the pre-LS2 implementation, the noise was generated in the FESA class. There were a few trimmable functions from which the frequency high and low were derived from make rules.
- The specifications for the implementation in the new LLRF are in EDMS 2114577. The noise generation algorithms will remain in the FESA class due the flexibility provided. The new beam control will include a function generator for the noise. A phase-loop response compensation will be available as an option (to avoid that the phase loop suppresses the injected noise). The new system will work for both proton and ion beams.
- A comparison of the available noise generation algorithms shows good agreement, however, there is some small residual noise outside the desired band observed with the VariNoise algorithm used pre-LS2. Even though this should be of no concern, it will be further investigated.
- An example of the adjustments for controlled blow-up of an LHC pilot bunch from an MD in 2018 was presented, illustrating the process of parameter optimization which is rather time consuming (even without strong intensity effects).
- In operational conditions for LHC bunch trains, the setting up of the blow-up relied on small shot-to-shot variations and also on small bunch-to-bunch differences. To ease the setting-up procedure in the future, ideally the optimum frequency band should be predicted by a combination of calculations and simulations. Furthermore, the proposed procedure for setting up is to start first with a reasonable value of the gain, then optimize the frequency band to achieve similar blow-up on all bunches, and finally optimize the gain to achieve the target emittance. This approach will be studied in analytical calculations and simulations including the impedance model and the LLRF loops. Later on this will be complemented with beam studies.
- First results of calculations of the bunch length evolution and the required synchrotron frequency bands for different intensities have been shown. In double RF, there are flat parts in the synchrotron frequency distribution that need to be avoided to limit the blow-up to the beam core. The stable phase offset and the bunch length along a train of 72 bunches has been calculated, showing that the bunch-by-bunch phase offset is dominated by TWCs+OTFB while other impedance sources give a constant phase shift. A calculation for a full train of 4 batches shows that it will be challenging to find proper excitation band to blow-up all bunches consistently due to the bunch-by-bunch 800 MHz phase variation resulting from beam loading. This will be subject of detailed studies.
- Hardware and firmware are well advanced, studies with intensity effects have started.
- A simple mechanism should be prepared to allow verifying that there is no overlap between blow-up and scraping. To be followed up by Giulia.
AOB
- Status of fast valves in BA2: Two valves are not responding properly (VVFA and VVFB). VSC is trying to resolve the issue, but propose to block them open for the moment to avoid issues in operation (e.g. due to valves not responding). There are no spares for the VVFB available, thus needs a discussion how to proceed for spares in the future, and if the layout could be simplified. To be followed up by Chiara and Verena.