Download
1 / 15
150 likes | 344 Views
Main Beam / Drive Beam Phase alignment. B. Jeanneret CERN/BE CLIC Worshop , October 2009. Tolerance on Phase and Gradient. Luminosity and understanding/adjustment of the Beam Delivery and of the Final Focus. D.Schulte & R.Tomas , PAC09. Overall cycle.
E N D
Main Beam / Drive BeamPhase alignment B. Jeanneret CERN/BE CLIC Worshop , October 2009 CLIC_WSHOP_2009,BJ,WG5
Tolerance on Phase and Gradient • Luminosity and understanding/adjustment of the BeamDelivery and of the Final Focus D.Schulte & R.Tomas, PAC09 Overall cycle DB intertrain (or sector to sector) Not muchmargin on these CLIC_WSHOP_2009,BJ,WG5
326 klystrons 33 MW, 139 ms combiner rings Circumferences delay loop 72.4 m CR1 144.8 m CR2 434.3 m drive beam accelerator 2.38 GeV, 1.0 GHz CR1 CR1 1 km delay loop CR2 326 klystrons 33 MW, 139 ms drive beam accelerator 2.38 GeV, 1.0 GHz 1 km delay loop CR2 decelerator, 24 sectors of 868 m BDS 2.75 km BDS 2.75 km BC2 BC2 245m 245m IP1 e- main linac , 12 GHz, 100 MV/m, 20.85 km e+ main linac TA R=120m TA R=120m 47.9 km 83ps DB CLIC 3 TeV booster linac, 9 GeV, 2 GHz or 4 GHz ? BC1 MB e- injector 2.4 GeV e+ injector, 2.4 GeV 45fs e+ DR 365m e- DR 365m CLIC_WSHOP_2009,BJ,WG5
Implications for differentsystems • RF : • DB Linac : 1 GHz Unless feed-back in between : δφ = 0.015o • Timing system : • transmission stable over ~km down to 5μm • Instrumentation : • Detectionaccuracy down to 15fs • (Quasi-) Staticpath-lengtherror in rings • Groundstability/vibrations in rings Tolerance to rms : divide by 3 σz = 5 μm , σt = 15 fs Not discussed today CLIC_WSHOP_2009,BJ,WG5
Phase control / RF CLIC_WSHOP_2009,BJ,WG5
Timing and feed-back strategy in the tunnel • To relax on δφ : feed-forward& back beforeenteringeach DECEL Notional, as for now 9 GeV MB as a fastclock Timing line A B DB ∆t to: injectors DB & MB Damp. Rings MB TA DB TAs @50 Hz TA C E D MB to IP Phase from B, EnergyjitterδE from A&B, Phase correction for δE C&D, Phase error MB/DB , to synchronize theinjectorlinacs Open issue: frequency band of δE correction CLIC_WSHOP_2009,BJ,WG5
Phase control - I At 12 GHz • measurement of the 9 GeV MB ( keep in memory for T = [0..23] × 6 μs] ) • error in the 9 GeV MB fromInjector and in the MB-TA • DB TA lengthstability ⇔ magnetstability to bespecified • measurement of the DB • error of the feed-forwardadjustement of the chicane • transmission error in the timing system • bandwidthlimit for correction (30 MHz the filling of the MB_ACS) DOABLE ? • periodicdE/E structure complicating the measurement (slide 13) • Path-lengths variations to beworked-out in all rings (MB and DB) Scheme : correct the phase withfastfeed-forward in the final chicane and slow feed-back to the injectors Ingredients See talk A. Andersson CLIC_WSHOP_2009,BJ,WG5
Phase control – II : data L0 Chicane after TA α • Tolerance on α : • 1/10 δztol • 4 magnets L0 = 8m α = 0.1 rad Use staticmagnets + variable deflectors for range δφr=10o All thisbeingre-worked (lessdemanding) F.Stulle, D. Schulte Range insidewhich correction applies relax for DB linac (Instead of 0.015o) CLIC_WSHOP_2009,BJ,WG5
Fast timing network • Little (or no) expertise in the project as of today • Presentationsatthis workshop : WG3 THURSDAY 11h00-12h30 • Long distance Optical Fiberswithfsresolution F. OemerIlday / BilkentUniv. Ankara • Overview of the Phase Measurement System at SLS/PSI Vladimir Arsov / PSI • Femtosecondopticalsynchronization system for FLASH Matthias Felber / DESY CLIC_WSHOP_2009,BJ,WG5
326 klystrons 33 MW, 139 ms combiner rings Circumferences delay loop 72.4 m CR1 144.8 m CR2 434.3 m drive beam accelerator 2.38 GeV, 1.0 GHz CR1 CR1 1 km delay loop CR2 326 klystrons 33 MW, 139 ms drive beam accelerator 2.38 GeV, 1.0 GHz 1 km delay loop CR2 decelerator, 24 sectors of 868 m BDS 2.75 km BDS 2.75 km BC2 BC2 245m 245m IP1 e- main linac , 12 GHz, 100 MV/m, 20.85 km e+ main linac TA R=120m TA R=120m 47.9 km CLIC 3 TeV booster linac, 9 GeV, 2 GHz or 4 GHz ? BC1 e- injector 2.4 GeV e+ injector, 2.4 GeV e+ DR 365m e- DR 365m CLIC_WSHOP_2009,BJ,WG5
Fast timing network DL CR1,2 DB Linac TA Source 50m 1km 50m All this for σt = 15fs 24 x DECEL 500m? … 9Gev LTL EXP IP MB LTL MB TA Fback box 20km Main Linac 5km 2km Hererelaxed, tol must be set for full capture and match to damping time MB linac DR e+ DR e- Pre-DR e- 100m 1km Here to synchronize the twocomplex Tolrance : σt = 7.5fs CLIC_WSHOP_2009,BJ,WG5
Fast longitudinal measurement CLIC_WSHOP_2009,BJ,WG5
A complication : Incoherent synchrotron radiation • X-ray @ 6 keV : L_absorption = 50 µm (Aluminium) • Impact angle : ~ 0.2 rad • ∆T = 80 oK @ 50 Hz • Risk of rapidageing ∆x = 10 µm TIME STRUCTURE • OveralldE/E of ISR varies between • Trains travelling : • MIN : No DL, 1×CR1, 1×CR2 • MAX : 1×DL, 3×CR1, 4×CR2 • MaxdE/E = 2 o/oo • Converts to dz after last chicane before DECEL • 1% longitudinal z-pitch modulation R56 = -0.33 m • Cure : • make the DL, CR1,CR2 longer accordingly • complicates the feed-forwardφ–correction • (seeslide 20) 24 bunches x 121 CLIC_WSHOP_2009,BJ,WG5
Detection of the bunchcentroid 10x tol 9x tol • Existing CTF3 data ? 2.5 x tol A gaussianbunch, with sσz = 1mm And tolσ(δz) = 0.005mm ( σ(δt) = 15fs ) CLIC_WSHOP_2009,BJ,WG5
Summary • Timing and feed-back in the femto-second range not trivial • Much work in many areas • Multi-group issue (RF,detectors,network?,beam) • Long distance (km) & precise (10fs,5μm) new for us CLIC_WSHOP_2009,BJ,WG5
