Recent results from CTF3 Piotr Skowroński for the CTF3 Collaboration 22 January 2016 CLIC Workshop

2015 CTF3 Programme Two Beam Acceleration with the Two Beam Module Deceleration studies in Test Beam Line Phase Feed Forward The Dogleg Experiment: AS Breakdown Rate in presence of beam Beam instrumentation tests  Wakefield Monitors  Cavity BPMs for the main linacs  Strip-line BPMs for the drive beams  Beam Loss Monitors  Optical Diffraction Radiation beam size monitors Setup of the beams and Machine Development  Emittance control  Dispersion control  Combined beam emittance reduction  Stability 22 January 2016 CLIC Workshop

Test Beam Line in CLEX 22 January CLIC Workshop 2016  Deceleration  Power production  Form factor  Emittance preservation  Energy spread  Beam transport  Overall consistency Steffen Doebert

RF power production 25A Drive Beam delivered Basically nominal power (135 MW) in last PETS with recirculation, > 90 MW in regular PETS Total peak power production: ~ 1.3 GW at 12 GHz 22 January CLIC Workshop 2016 Steffen Doebert

Deceleration results, new record 22 January CLIC Workshop 2016 Initial energy: 135 MeV Minimum energy (10% threshold): 65.8 MeV  51 % deceleration Steffen Doebert

Two Beam Module (TBM) 22 January CLIC Workshop 2016 PETS 2 ACS4 PETS 1 ACS3ACS2ACS1 Drive Probe The Lego block of CLIC  High precision assembly  Active alignment  Vacuum  Easy to repair  Not expensive Wilfrid Farabolini

Two Beam Module 22 January CLIC Workshop 2016 Wilfrid Farabolini

Two Beam Module Comparison with energy gain 22 January CLIC Workshop 2016 Direct DB power production with 23 Amps – 140 ns Power input on each ACS: P in = 10.4 MW (ACS1 rescaled) Nominal TBM energy gain: = 45.5 MeV TBM energy gain: 43 MeV Wilfrid Farabolini

Measured phases between structures 22 January CLIC Workshop 2016 Output phases from structures ACS4ACS3ACS2ACS1 PB  43  21  31 11 PB generated power DB generated power Correct if: Phase error without recirculation Phase error with recirculation Phase control  21 PB  21 DB Equal-6 o -4 o No control  43 PB  43 DB Equal-13 o -12 o No control  31 PB  31 DB Equal-31 o -6 o Priming control  1 PB  1 DB At 180 o 0 o CALIFES phase control cos(15.5 o ) x 45.5 = 43.8 MeV Wilfrid Farabolini

Two Beam Module 22 January CLIC Workshop 2016 PETS 2PETS Tank ACS4 PETS 1 ACS3ACS2ACS1 Drive Probe  A  A  31 11  PETS Generated powers Additional PETS to prime the module  The CTF3 drive beam is less intense then in CLIC  This way we can reach the CLIC power level and the accelerating gradient Wilfrid Farabolini

Highest energy gain obtained (so far) 22 January CLIC Workshop 2016 TBM energy gain: 58 MeV ACS2 input: 45 MWPETS 1 output: 85 MW With DB 15 A (factor 4), pulse length 140 ns Limited by break downs: the module is not fully conditioned yet! 58 MeV Wilfrid Farabolini

Unloaded Loaded (CLIC) Increasing current Gradient along the structure Average gradient 100 MV/m Dogleg Beam-Loading Experiment 22 January CLIC Workshop 2016 Drive beam 1-3A MeV 12 GHz RF from klystron 12 GHz accelerating structure J.L. Navarro Beam loading changes the field distribution for the same average gradient ⇨ how is the break-down rate affected? Reactivated an old beam line (dogleg) ~1.2 A DB current (like CLIC Main Beam) Measure BDR with/without beam for a direct comparison Frank Tecker

March-April run: a hot spot on the Dogleg 22 January CLIC Workshop 2016 Initially: breakdown distribution inside structure as expected Later: breakdowns mostly detected at beginning of structure  => hot spot for breakdowns had developed there structure unusable for beam-loading experiment => changed Aug 2014 Apr 2015 position along structure Frank Tecker

New structure conditioned and measurements started 22 January CLIC Workshop 2016 Initially BD more at downstream end Later at the beginning Now uniformly distributed In the structure Phase information + timing Aug Sep Oct Nov Dec Robin Rajamäki Frank Tecker

Beam instrumentation tests The CLIC requires top notch performance from the beam instrumentation, or above Very large number of devices needed in CLIC imposes cost effective solutions Non-destructive measurements wherever possible  The machine safety and the immense power density of the beams In almost all the cases it means detection of very tiny signals in direct vicinity of hundreds of Mega Watt RF Must be tested and studied in realistic conditions 22 January CLIC Workshop 2016

Wake Field Monitors 22 January CLIC Workshop 2016 Reidar Lunde Lillestol

Wake Field Monitors WFM was up to the spec in the first acceleration structures in TBTS Since then the design has changed  Were moved to the last cell 22 January CLIC Workshop data Reidar Lunde Lillestol

Resolutions January CLIC Workshop 2016 Reidar Lunde Lillestol

Noise from the Drive Beam 22 January CLIC Workshop 2016 Reidar Lunde Lillestol

Cavity BPMs (version 2) Resolution measured with 3 BPMs where the position in pickup is predicted by the signals measured in the other 2 Simple correlation between BPMs gives a resolution of ~1 μm. Due to the phase behaviour in the centre of the BPM and the angled trajectory of the beam, the position is not accurately measured around 0  A model is required to predict the behaviour in this region 22 January CLIC Workshop 2016 Jack Towler

Strip Line BPM for Drive Beam (version 2) The first generation was sensitive to the 12 GHz 22 January CLIC Workshop 2016 Parameter Shorted BPM Terminated BPM Stripline length 25 mm37.5 mm Angular coverage 12.5% (45°)5.55% (20°) Electrode thickness 3.1 mm1 mm Outer radius 17 mm13.54 mm Ch. Impedance 37 Ω50 Ω Duct aperture 23 mm Resolution 2 μm Accuracy 20 μm Time Resolution 10 ns Alfonso Benot Morell

Strip Line BPM 22 January CLIC Workshop 2016  Resolution test using Singular Value Decomposition (SVD): Separate systematic beam effects (i.e. betatron motion, cavity phase/energy errors, RF jitter…) from uncorrelated BPM noise floor  P=999 consecutive, synchronous shots (22 A beam) analysed for all M=9 BPMs in the Drive Beam of the Two- Beam Module  B = U·S·V T, where: B PxM  Position data for all BPMs U PxP  Temporal eigenvectors S PxM  Diagonal matrix (eigenvalues s ii ) V MxM  Spatial eigenvectors  s ii give correlation level between U and V  Set s ii = 0 in the high correlation region and recompute B  B’  σ (stdev) of columns of B’ : resolution for each BPM. For our prototypes: BPM0645 (#5): 1.4 µm (H), 0.5 µm (V) BPM0685 (#6): 2.4 µm (H), 2.5 µm (V) Resolution close to specified value of 2 µm Alfonso Benot Morell

Optical Fibre Beam Loss Monitors 22 January CLIC Workshop 2016 Signal subtraction to account for showers from TBL only Optical fibre at the TBL Background fibre Signal fibre Real signal fibre 28 m Connecting fibre 25 m Connecting fibre 75 m TBL TBM Maria Kastriotou

TBL: losses with long bunch trains 22 January CLIC Workshop 2016 Observing losses from a 1µs long pulse o Controlled losses generated by switching off quadrupoles o BPM signals to correlate Nominal Q500 off Q550 off Q600 off Determination of loss location from signal leading edge o Good qualitative agreement between oBLM and BPM profile loss measurements Localisation of loss down to (below) 2 m achieved! First measurement of beam loss crosstalk to BLMs at TBM Potential limitation of BLMs due to RF cavity dark current and RF breakdown Maria Kastriotou

Optical Transition Radiation Interference 22 January CLIC Workshop 2016 Beam pass 2 screens: OTRI Beam pass 1 screen: OTR Formation length: Max separation = 4 L 400nm) Screens config: Possibility to estimate the beam divergence Measured OTRI Vertical Polarization Robert Kieffer

OTRI Results 22 January CLIC Workshop 2016 Experimental measurement of the shadowing of the electromagnetic field. OTRI data in the far field (angular) confirm existing predictions. First experimental study of shadowing in imaging conditions. Results to be published soon. First measurements using RF wavelength Robert Kieffer

Phase Feed Forward Goal: Stabilize the phase of the drive beam down to 0.2 degree at 12 GHz (50fs) 22 January CLIC Workshop 2016 Improvements in 2015 Resolution of the phase monitor down to x lower uncorrected downstream phase jitter:  2 degrees  0.8 degrees 2x more powerful amplifier  +/- 350 V  +/- 700 V (20 kW  40kW) Corrected downstream phase jitter: 1.4 degrees  below 0.3 degrees Jack Roberts

Phase Feedforward: Current Jitter Record 22 January CLIC Workshop 2016 Interleaved data:  Even pulses have FF on and odd pulses have no correction applied 0.74 degrees phase jitter reduced to 0.28 degrees Simulated best possible (unlimited) FF correction given beam conditions 0.27 degrees Jack Roberts

2015 improvements 22 January CLIC Workshop 2016 High bandwidth (~30 MHz) correction: Corrected variations along the pulse not only jitter on the mean Phase variation along the pulse (between black lines) reduced from 1.68 to 0.26 degrees (mean deviation of samples along pulse) Jack Roberts

Drive Beam Performance 22 January CLIC Workshop 2016 Improved stability Improved transmission to CLEX TBTS CR TBL Piotr Skowronski, Tobias Persson, Davide Gamba

Drive Beam Performance Also improved long term stability, all thanks to  New feed-backs, improvements and tuning of the existing ones  Reduced dynamic aperture: dispersion control, Dispersion Target Steering, automatic orbit closure, improved alignment, removed vertical dispersion, … 22 January CLIC Workshop 2016 Drive Beam current in BPMs around Two Beam M odule rel. 5.3 ∙ Current behind TBM Piotr Skowronski, Tobias Persson, Davide Gamba

New Delay Loop Optics Optics in the Delay Loop had large nonlinearity in momentum (nonlinear dispersion was especially ugly) Davide Gamba designed and commissioned improved one 22 January CLIC Workshop 2016 New one Footprint in the horizontal phase space for factor 8 combined beam in CLEX (simulation) Old one Measured dispersion after DL Piotr Skowronski, Tobias Persson, Davide Gamba

Irradiation tests in CALIFES ESA tests electronics for the JUICE mission  JUpiter ICy moons Explorer The choice of the dark current offer the better beam characteristics and operational easiness  Low intensity, very stable, doesn’t need the laser 11 days (weekends or nights) have been dedicated to irradiation in 2015 Small modifications ongoing to improve the beam homogeneity 22 January CLIC Workshop 2016 During all of the runs single event effects were observed Wilfid Farabolini, Maris Tali

Conclusion When developing such challenging technologies testing in realistic conditions is a must 22 January CLIC Workshop 2016

Backup 22 January CLIC Workshop 2016

Conclusion (TBM) 22 January CLIC Workshop 2016 The CALIFES beam has bean extensively used for many experiments Concerning TBM tests and considering the complexity of the RF scheme more studies and DB time are necessary this year Improve the structures conditioning Improve some RF calibrations Validate the module performances Study the beam quality after acceleration by the TBM Test all the subsystem in integrated conditions (WFMs, girder alignment, BLM…)

22 January CLIC Workshop 2016

Cavity BPMs for the Main Linac Centered beam excites monopole mode (TM 010 )  Amplitude dependent on charge Away from the center, other modes are excited  First order dipole mode (TM 110 ) depends linearly on beam offset and charge. TM 110 splits in 2 orthogonal modes Beam excites other unwanted higher order modes.  Requires suppression of unwanted modes. 22 January CLIC Workshop 2016