LCLS Beam Diagnostics

Transcription

1 LCLS Beam Diagnostics International Beam Instrumentation Conference 2014 Henrik Loos September 17, 2014

2 Outline Overview LCLS accelerator diagnostics LCLS-II Charge and beam position Beam profile measurement Bunch length diagnostics Summary 2

3 LCLS and LCLS-II Beam Parameters LCLS LCLS-II Unit Baseline Operation SC Unit RF frequency MHz Repetition rate Hz Electron energy GeV Bunch charge 200 & pc Bunch length 20 < µm (rms) Emittance norm µm X-ray energy kev X-ray pulse energy < 2 < 4.7 < 2.2 mj X-ray pulse length 230 < fs (FWHM) 3

4 LCLS Diagnostics Development Machine tuning optimization - Fast wire scanner High brightness beam issues - YAG screen (PSI) New capabilities - SXRSS, overlap diagnostics Extended machine parameters - Low charge mode - Mid IR spectrometer - XTCAV LCLS-II - BPM µtca receiver - RF-BPM (PAL) gun L0 TCAV0 L1S heater 3 wires 3 OTR DL1 135 MeV L1X BC1 220 MeV 3 wires 2 OTR σ z1 L2-linac BC2 5 GeV 3 OTR σ z2 TCAV3 5.6 GeV 4 wire scanners old screen µ wall L3-linac BSY 14 GeV 5 wire XTCAV scanners vert. FWS dump DL2 YAG SPARO undulator 14 GeV 4

5 LCLS-II Diagnostic Challenges Single bunch properties like LCLS-I High beam power - Put (almost) nothing into the full rate beam - Low rate (diagnostic) or variable rate (postspreader) lines for most invasive diagnostics High beam rate - Fast DAQ needed, FPGA processing, low latency networks for feedback, MPS, etc. - Single shot detector signals Low charge mode - BPM & R-BLM sensitivity Courtesy J. Frisch 5

6 Charge Measurement Upgrade Existing toroid electronics issues - only local calibration - up to 15% read-back variation - Up to 4% noise at 150 pc from up to 500 cable run (BPMs give 0.05%) New electronics and DAQ - Add remote calibrator to in-tunnel amp - Use differential cable TORO:LTU0:195:TMIT Charge (Nel) x 10 8 Correlation Plot 05-May :47:09 TORO:LTU0:195:TMIT TORO:LTU1:605:TMIT y = + A x 1 + B x 0 A = B = e e+07 χ 2 /NDF = 2.16e+13 rms fit error = 4.65e+06 Nel y = + A x 1 + B x 0 A = B = e e+06 χ 2 /NDF = 5.1e+12 rms fit error = 2.25e+06 Nel BPM Charge Estimate (Nel) x Gated charge amplifier CAEN QDCV965A already used for PMTs Get ~1% agreement for absolute charge between two toroids Noise of 0.2% to 0.5% LCLS-II requires high dynamic range average current measurement, use commercial solution (Bergoz Turbo-ICT) 6

7 µtca Strip-line BPM Electronics Part of SLAC µtca development - Initiated for NC LCLS-II project New AFE, use 300 MHz rather than 140 MHz, also higher BW Uses SIS bit ADC at 109 MHz Up to 8 BPMs per crate possible Courtesy S. Hoobler Poster C. Xu, WEPD17 7

8 LCLS Test Installation Implemented for 4 strip-line BPMs in L3 linac Operational for almost 2 years without issues Achieve same resolution as from existing electronics Resolution (µm) 10 1 BPM Resolution L3 21-Dec :43:46 SVD Linear Predictor µtca BPM Resolution ( µm) BPM Resolution vs. Charge BPM26901 BPM27201 BPM27301 BPM27401 BPM pc z (m) Charge (pc) 8

9 X-Band RF Cavity BPM PAL Prototype SLAC Receiver Schematic Dipole Cavity Monopole Cavity PAL-SLAC collaboration GHz (4x S-band) for flexible bunch pattern New receiver with coax input and µtca Test BPM next to ANL undulator RF-BPM Preliminary results already meet <1 µm LCLS-II requirement Noise issue with power supply resolved, awaiting beam test Undulator RF-BPM Resolution 9

10 LCLS-II BPMs Most critical performance at 10 pc low charge limit Strip-line BPMs (30 µm) for most of beam transport RF cavity BPMs in special locations, X, S, or L band - Energy measurement - Orbit for fast feedback - Wire scanner jitter correction - Undulators Cold button BPMs inside cryo-modules (100 µm) 10

11 Fast Wire Scanner COTR makes WS critical for LCLS beam tuning LCLS uses existing SLC design - Stepper motor driven, mm/s speed - Vibrations from wire card support on single side and stepper motor - 45 actuator with 3 wires for x, y, u plane Fast wire scanner development - Linear motor, up to m/s speed possible - 2 bellows to cancel vacuum forces - Also 45 scan orientation, 2 stroke for 3 wires - Encoder with sub-µm resolution Beam µ-metal Wire card Linear motor 11

12 FWS Motion Profile Motion profile to minimize scan time Beam synchronous data acquisition of encoder position and beam loss signal Makes motion stability not critical SLC-style 4 location emittance measurement for x, y, and coupling takes 8 min Expect < 30 sec with FWS Upgrade project started for LCLS Position (mm) Position X wire Y wire U wire FWS Motion Profile Scan Time (s) 12

13 FWS Magnetic Shielding First prototype installed upstream of undulators Observed significant drop in FEL during scan Related to ~20 µtm magnetic field from linear motor FEL Energy (mj) Motor Position (mm) FWS moves FEL Energy GeV Wire hits beam FWS Position Pulses BPM Pos y (µm) Undulator Beam Orbit Pulses µ-metal shielding was added Now reduced to ~1 µtm Tolerance limit for LCLS & LCLS-II 13

14 LCLS-II FWS MW beam power Carbon wire for least beam loss Scan simulation with typical beam parameters of wire heating Stays below safe fluence studies established for SLC 60 Relative Temperature Increase Wire Heating Simulation α = 17 α = 87 α = 1084 No cooling Position (m/s) Velocity (m/s) Accel. (m/s 2 ) Time (s) Time (µs) Speed of 400 mm/s already demonstrated with FWS Higher speed requires longer stage May add thick wire for beam halo measurement 14

15 Injector OTR Measurements Straight beam path, no COTR affect on beam size - OTR and wire scanner emittance agree Laser heater chicane - introduces small R56, see 2x COTR enhancement, emittance 25% too small - Energy modulation from laser interaction in undulator 4000 Enhancement reduced to 20% See laser 2. harmonic - Emittance still underestimated Even small enhancements of COTR can affect emittance measurements Counts gun L0 TCAV0 L1S heater 3 wires 3 OTR DL1 135 MeV (C)OTR Spectra L1X B 220 LH off LH on Chicane off Wavelength (nm) see also F. Zhou et al., FEL14, THP031 15

16 SwissFEL Profile Monitor PSI development Installed at SLAC for GeV beam test at factor 10 5 COTR location YAG viewing geometry - Smallest spot size - COTR reflected away from CCD - Tilted focal plane needs tilted CCD LTU Installation YAG Fluorescence Mirror COTR Beam Screen Holder Talk R. Ischebeck, TUCYB3 16

17 Commissioning Results Saturation of YAG tested - None at 20 pc, indication at 180 pc Test for coherent enhancement - Scan RF phase to change bunch length - COTR enhancement reduced from 10^5 to small factor at full compression or 10 s of percent in normal setup YAGS:LTU1:743:TMIT charge estimate (Nel) x Correlation Plot 12-Mar :09:05 YAGS:LTU1:743:TMIT COTR Test BLEN:LI24:886:BIMAX (A) YAGS:LTU1:743:TMIT charge estimate (Nel) x CODR? Correlation Plot 12-Mar :53:53 YAGS:LTU1:743:TMIT y = + A x 0 A = χ 2 /NDF = 17.9 rms fit error = 944 Nel YAGS:LTU1:743:YRMS Y rms (um) Beam 20 pc 13.2 GeV Saturation Test 17

18 SXRSS Beam Overlap Diagnostics Soft X-Ray Self Seeding ( ev) Both beams diverted by chicanes Need diagnostics to measure both Combine wire scanner and YAG screen Wire 40 µm carbon, YAG 20 µm thick View both with CCD camera ~10 µm position measurement needed Wires YAG X-rays e-beam 18

19 Overlap Diagnostics Performance Move supporting girder to scan wire and find e-beam position Move x-ray mirror to steer x-rays onto YAG, find position Use mirror response matrix to overlap beam, get seeding CR effects are serious issue PMT:DMP1:430 Signal () BOD:UND1:1005 Position µm) ( Wirescan on BOD:UND1: May :53:23 xarea = 0.036± 0.00 Mcts xmean = -0.39± 0.00 mm xrms = 31.1± 0.55 µm xskew = 0.00± 0.00 xkurt = 0.00± 0.00 BOD10 Wire Scan y (mm) Profile Monitor YAGS:UND1: Feb :06: BOD10 CCD Image BOD13 CCD Image 10 shot average BG subtracted 4000 cts X-rays x (mm)

20 Mid-IR Spectrometer C*R based bunch length measurement of LCLS um and sub-um beams needs 1-20 µm Single shot preferred KRS-5 prism based spectrometer developed Images OTR from foil onto 128 Dry air input port I J λ = 632 nm edge mirror H AR-coated Si window F E Stray-light shield CVD Diamond G OTR33 C Iris and moveable filters A Tip/tilt adjustable D B e-beam HeNe Reverse-injection alignment HeNe (w/ assoc. optics) z y x element pyroelectric line array Transfer function determined by fitting spectra at different bunch lengths to simulated bunch spectra T. Maxwell et al., PRL 111, (2013)

21 MIR Spectrometer Results Form-factor extrapolation for λ > 20 µm necessary Bunches as short as 0.7 µm rms at 20 pc measured 150 pc T. Maxwell et al., PRL 111, (2013) Non-invasive version possible using CER from DL2 bends 21

22 LCLS-II Bunch Length Monitors Relative BLMs similar to LCLS-I detecting edge radiation R-BLMs at full beam rate for feedback system Average THz radiation power at few W level becomes issue - Required attenuation leaves insufficient single shot energy - Cooled pyroelectric detectors being investigated High dynamic range from wide charge and length range - Use of Schottky diodes at few 100 GHz with much higher sensitivity Fast detector response for MHz rate 22

23 X-Band Deflecting Cavity Existing S-band deflecting structure about 5 µm resolution X-band provide ~10x better - 4x higher frequency - 2.5x higher gradient Installation post-undulator in main dump beam line - Non-invasive for FEL users Courtesy P. Krejcik Direct observation of longitudinal phase space on dump YAG C. Behrens et al., Nat. Commun. 5 (2014)

24 XTCAV Bunch Length Measurement Simple calibration with phase sweep Bunch length with fit to off and ±90 Achieved resolution - 1 fs (4 GeV) - 4 fs (14 GeV) Doubling plan using SLED Beam Size (µm) TCAV bunch length on OTRS:DMP1: Mar :03: σ x = 46.39± 0.80 µm σ z = 1.415±0.086 µm r 15 = 0.078±0.030 cal = ±3.183 µm/µm Bunch Length 4.2 GeV, 22 MV σ BL21 / σ XTCAV Correlation Plot 07-Oct :41:09 y = + A x 0 A = χ 2 /NDF = 24.2 rms fit error = R-BLM vs. XTCAV OTRS:DMP1:695:BLEN Bunch length (um) Checked R-BLM calibration 5% average deviation to XTCAV R-BLM not sensitive below 2 µm TCAV:DMP1:360:AACT (norm) 24

25 XTCAV as FEL and X-Ray Diagnostics X-ray pulse reconstruction - Compare FEL off and on - Measure time resolved energy loss - Energy spread increase also used FEL Off FEL On Electron Trapping X-Rays Electrons Longitudinal bunch manipulations - Slotted foil emittance spoiler - Double bunch setup 25

26 Summary Diagnostics was sufficient for first LCLS operation New developments driven by enhancements in beam parameter range and capability, and also operational needs LCLS-II diagnostics benefits greatly from existing projects, but still many challenges remain 26

27 Acknowledgements Thank you for your attention! Special thanks to the diagnostics teams for Toroids: D. Brown, S. Condamoor, R. Larsen PAL RF-BPM: S. Babel, C. Kim, S. Hoobler, P. Krejcik, A. Young, C. Xu Fast Wire Scanner: S. Anderson, M. Campell, A. Cedillos, M. D Ewart, P. Krejcik, R. Iverson, Z. Oven 27