Minutes October 3rd, 2008
Present: Gianluigi Arduini, Mike Barnes, Helmut Burkhardt, Rama Calaga, Riccardo de Maria, Elias Métral, Federico Roncarolo, Giovanni Rumolo, Benoit Salvant, Elena Shaposhnikova, Simon White.
Agenda
- Riccardo de Maria and Benoit Salvant
"First MD results on the fast TMCI at SPS injection" (slides, minutes)
- Mike Barnes
"MKDV heating/outgassing issues in the SPS" (slides, minutes)
- Elias Metral
"Impedance measurements and/or simulations needed to make more precise analyses of the beam dynamics " (slides, minutes)
- Elias Metral"SPS transition crossing with SFTPRO and CNGS" (slides, minutes)
- Federico Roncarolo
"Impedance measurement and simulations for FP420" (slides, minutes)
Minutes (thanks to Elias)
Presentation by Riccardo de Maria and Benoit Salvant:
- Results of 2007 MDs as well as preliminary results of the 2008 MDs on SPS fast instability at injection were presented. 2007 results agreed very well with HEADTAIL simulations. This year's MDs however were not so successful. Current understanding is that the chromaticity was set too low (i.e. negative). The Faraday cage exponential pickup was also used to reveal the frequency content of the signals. The bunch rotation in the PS and the SPS RF were turned off to inject longer bunches in the SPS, so that the frequency resolution can be increased. Analysis work is still ongoing.
- Comment from GA: Use the fast analog signals (other SPS wideband longitudinal and transverse pickups) recently installed on OASIS for next time.
- Comment from Mike => The 3 MKDH are not made of ferrite but laminated steel => They measured it and they found many resonances!!! => To be checked with Mike.
- 4 striplines (exponential PU) => Best suited for frequency domain whereas the head-tail monitor is better suited for time domain - 2 for getting H and 2 for V.
- Rise-time is much longer than both wall-current monitor and head-tail monitor. Wolfgang Hoeffle is working on it.- High-frequency activity near 1.6-2 GHz could be noise, as the signal lies far from the bunch sum signal in time domain. GA meas. this also in the past with exactly this wide-band PU and the spectrum analyzer. The point is that the signal appears after some cycles, i.e. it is anyhow triggered by the beam. But the fact that the lines stay straight seems to indicate that it is artificial. Elena said she saw this in the past.
- RF OFF : 12 ns bunch length from the PS. In the past they had 25 ns for the longitudinal impedance meas. => May be one should ask for that in the next MDs but Heiko had some difficulties to produce this 12 ns beam (cavity voltage could not be decreased). - Next step: See if we have activities between 0 and 1.5 GHz and redo measurement with 25 ns bunch length (ask Heiko). Gianluigi said it could be interesting to compare with RF off data from 200 to see what has changed in the machine.
- Suggestion of Elena => Analyze % losses vs. time for Benoit's measurement and simulations could be useful.
Presentation by Mike Barnes:
- MKDV2 => We don’t have a spare for it.
- Reminder: Installed in the SPS is 1 MKDV1 and MKDV2, and the spare is a MKDV1. It was said in the past that some meas. were done on MKDV2 but it is not possible so it was a MKDV1.
- Question1 and action GA: estimate from the optics if one can replace a MKDV2 by the spare MKDV1 as it has a smaller aperture.
- Question2 if pb with Question1: What about preparing a spare? Money needed etc.
- Meas.: - 25 ns => Normal case: T of MKDV1 is bigger than MKDV2.
- 75 ns => T of MKDV1 stays a reasonable T and the T of MKDV2 increased a lot. Mike attributes this to the denser spectrum at 75 nsec that is then more likely to hit one or more MKDV2 resonances. Elena said a computation of power loss due to such an overlap leads to about 100W, which can explain the temperature increase.
- It seems that it was the MKDV1 whose ferrite was not treated and installed like this last year (and kept this year).
- Suggestion from Elena => Install transition pieces in all kickers.
- Question: Do we understand the huge pressure rise with only the 50 ns beam? It seems to be much more violent than 25 ns.
- Question: did we reach nominal intensity with 4 batches and 25 nsec spacing this year?
- Gianluigi suggested it may be interesting to look on the BCT or Fast BCT at when the pressure increase happen in the cycle.
- Question: What is the batch spacing for 50 ns beams? Certainly the same as for 25 ns and 75 ns, i.e. 8 empty buckets = 225 ns.
- Next MD:
1) Correct the batch spacing.
2) Play with the batch spacing.
3) BCT and fast BCT all the time.
4) Action EM: Ask vacuum people to log the pressure (ion pumps current) => For MKDV and MKE6 (all in fact).
Presentation by Elias Metral on which FEA simulations and bench measurements are needed to obtain the relevant wake field(s) for HeadTail macroparticle simulations:
- In order to calculate the force acting on macroparticles, the Headtail code needs to import the dipolar wake field and the quadrupolar wake field separately for each plane (i.e. 4 wake fields). Elias presented computations of the generalised impedance
- Question from GA: Don’t we have already all the information in the time-domain MAFIA simulations? => To be followed up. MAFIA should be able to give the wake fields at any transverse location, so that the test particle can be at locations from which it is easier to obtain the driving and detuning terms.
- Anyhow the meas. are made with wires (1 or 2 or more?) in the frequency domain, but it is true that for the simulations it is important to understand what we can deduce from the time-domain simulations.
Presentation by Elias Metral on transition crossing in the SPS:
- Elias showed rapidly meas. by Karel et al. on CNGS and SFTPRO near transition, revealing oscillations, which can be studied to have an estimate of the longitudinal impedance.
- Elena said that we have to be careful with the interpretation of the oscillations as we can have oscillations already without impedance due to the different shape of the bucket before and after transition => nonlinear effects of the RF bucket. This was not included in the analysis but could be studied in detail with HEADTAIL.
Presentation by Federico Roncarolo:
- Made simulations for FP420 with both 1 wire (longitudinal impedance) and 1 wire displaced to try and observe the parabola.
- He did not find a nice parabola, which can be qualitatively understood by the fact that there is NO top/bottom AND left/right symmetry. Therefore one does not expect a parabola!
- however, nice agreement between simulations and measurements!
Author: Benoit Salvant CERN
AB/ABP-LIS