Recent Results in Analysis of Errors in Fringe Scans by Shintake Monitor (IPBSM) ATF2 Topical Meeting July 8, 2013 KEK ATF2 Topical Meeting Jacqueline Yan, S. Komamiya （ University of Tokyo, Graduate School of Science ） Y. Kamiya （ University of Tokyo, ICEPP ） T.Okugi, T.Terunuma, T.Tauchi, K.Kubo, J. Urakawa (KEK) 113/07/08

Outline Beam Time Status of 2013 Spring Run IPBSM Performance & Error studies Statistical Fluctuation Statistical Fluctuation M reduction Factors M reduction Factors Summary&Plans 2 Recent Updates ….. Relative Position Jitter (Phase Jitter) Relative Position Jitter (Phase Jitter) Polarization related issues Polarization related issues 13/07/08 ATF2 Topical Meeting

Stable IPBSM performance in beam tuning reiteration of linear /nonlinear knobs in aim of small σy Recent Beamtime Status (174 deg mode) ATF2 Topical Meeting ex ） consecutive 10 fringe scans preliminar y Time passed preliminary 10 x  x *, 1 x  y * 13/07/ ° mode ”consistency scan” M 〜 ± (RMS) correspond to σy 〜 65 nm Best record from Okugi-san’s Fri operation meeting slides

deg (2-8 deg) mode IPBSM phase drift study Dedicated data for IPBSM error study H and V relative position jitter slow drifts (see next page and Okugi-san ‘s slides) Study of wake-field effects Ex) intensity dependence /ref cavity & bellows scan consistency between Cherenkov and CsI results Cherenkov S/N ～ 15, CsI S/N ～ 3 high M at 30 deg Demonstration of Stability Demonstration of Stability: Consistency scan ( x 26) fitted M = / ICT: E9 preliminary 6/20 – 21 13/07/08 ATF2 Topical Meeting 4

13/07/08 5 ATFII Review Phase Drift (initial phase vs time) 5 final set of scans on deg final set of scans on deg (initial phase) vs (time) 6 deg (6/14: 1:00 am) drift ～ 93 mrad / min (73 nm / min at IP) 26 consecutive scans over 1 30 deg (6/21) generally stable overall : init. phase = / rad (~ 2.5%) 33 mrad / min (5.5 nm / long range scans typical heavy Drift varies for different periods stable ATF2 Topical Meeting

Comp Signal Jitter Jitter BG jitter 13/07/08 *scaled by S/N iCT monitor fluctuation Relative beam –laser position * Intrinsic CsI detector energy resolution detector energy resolution Signal Jitter Sources < 10% under investigation < 1 % ~ 3 % % ( PIN-PD signal) < 5 % ICT monitor accuracy measured Comp sig energy normalized by beam intensity varies with beam condition and S/N Spring, 2013: 174 deg mode contribution to Sig Jitter ΔE sig / E sig, avg Signal Fluctuation < 1.5 % (photo-diode) offline veto for timing, power Maybe few % from each of Δy and Δx signal jitter derived directly from actual fringe scans (peaks) ： typically ～ 25% signal jitter derived directly from actual fringe scans (peaks) ： typically ～ 25% ATF2 Topical Meeting 6 jitter （ RMS ） 〜 1.3 ns Relative timing cut (beam – laser) (beam – laser) e.g. 1-sigma

hard to separate from other fluctuation sources (laser pointing jitters, drifts, ect….) hard to separate from other fluctuation sources (laser pointing jitters, drifts, ect….) Vary over time Vary over time Phase Jitter / Relative Position Jitter Can’t push all fluctuation to phase jitters Can’t push all fluctuation to phase jitters fitted energy jitters with contributions from statistics, timing, BG, and Δx preliminary take high statistics scans (Nav ~ 100) under optimized conditions for dedicated analysis derive horizontal rel position jitter Δx using high statistic laserwire scan if Δy < 0.3 * σy (ATF2 beamline design) (ATF2 beamline design) CΔy > 90 % for σy* = 65 nm CΔy > 90 % for σy* = 65 nm Issue 1: Δy  M reduction residual M reduction factors must be assessed in order to derive the true beamsize !!! Issue 2 : fluctuation source during fringe scan If Δx 〜 2.5 μm cause 〜 4 % signal jitters (assume Gaussian profile σ laser = 10 μm) 13/07/08 ATF2 Topical Meeting 7

（１） “Direct Method” compare M measured at different modes under same beam conditions  observe upper limit on M meas  observe upper limit on M meas Goals: （２） “Indirect Method” : evaluate each individual factor offline and “sum up” represents the typical conditions of a particular period however …… hard to derive overall M reduction (only “worst limit “ ) ?? over- evaluate σ y 13/07/08 How to evaluate M reduction? 1 st : suppress M reduction  aim for Ctotal 〜 1 2 nd : precisely evaluate any residual errors  derive the “true beam size” priorities 1) Obtain “overall” M reduction only apply to a particular data (2) investigate a certain “unknown factor” (2) investigate a certain “unknown factor” : (e.g. rel pos jitter, spatial coherence) however,we must first ……. however,we must first ……. clear up all other factors !! clear up all other factors !! Prove these factors are θ mode independent Prove these factors are θ mode independent ATF2 Topical Meeting 8

Error sourceM reduction factor Fringe tilt (z, t) Laser polarization Optimized to “S state” using λ / 2 plate Relative position jitter main theme of today’s talk !! Phase drift under investigation Spatial coherence under investigation, monitor by CCD (??) Major bias if unattended to Minor factors profile imbalance : negligible after focal lens scanned well power imbalance: power measured directly for each path Laser path alignment : Resolution of mirror actuators aligning laser to beam alignment precision is important Could be major bias ATF2 Topical Meeting Spring 2013, 174 deg drift : typically mrad / min Still quantitatively uncertain Beamtime  optimization by “tilt scan” Individual M Reduction Factors typical conditions of a particular period 13/07/089 Cleared up through measurements of laser polarization and half mirror reflective properties

Deriving Relative Position Jitter Δy (phase Jitter Δφ) 10 Test validity of method using simulation fix M from fringe scan to jitter plot Plot signal jitter vs phase (1) generate fringe scan Aim for “realistic” ATF-like assumptions e.g. nominal σ y = 70 nm, 174 deg S/N = 5, BG fluc = 20%, ICT = 1E9/bunch timing jitter, power jitter, ect….. fringe scan vertical jitter σ pos : horizontal jitter from Δy (Δφ) (2) obtain Δy from fitting Model 13/07/08 ATF2 Topical Meeting

Deriving Δy (Δφ) : Test of Method using Simulation 11 various Δy inputs ---  10% * σy first 5 seeds extracted ideal Compare Nav = 100 and Nav = 10 extracted Δy agree with input within errors 35 % * σy 20% * σy assume nominal σ y = 70 nm Nav = 10 Nav = 100 test with 14 nm Δy input first 5 seeds also observed effect of input vertical jitters  light jitter ( 〜 10%), S/N = 5  heavy jitter ( 〜 20%), S/N = 2 signal jitters disturb precision in Δy extraction ( i.e. large errors) less vulnerable for large statistics (e.g. Nav = 100) M reduction from Δy Δy inputs Fitted M 13/07/08 ATF2 Topical Meeting

Deriving Δy (Δφ) : apply to actual data ATF2 Topical Meeting 12 Note: errors are only from fitting (we may have other systematic effects Not enough (large Nav) data yet for drawing any conclusions fix M from fringe scan to jitter plot Δy = / nm Δφ = / mrad Fitted σ y : – 9.0 nm Corrected σ y : – 10.1 nm Fitted σ y : – 2.5 nm Corrected σ y : – 2.7 nm Δy = /- 9.6 nm Δφ = / mrad M reduction due to Δy about 85 – 95% (?) Nav = 100, 30 deg, 5/19 Nav = 10, 174 deg, 3/8 preliminary 13/07/08 very preliminary attempts !! no Nav = 100 data available for 174 deg

Deriving Δy (Δφ) : Current Status possibly valid model for deriving Δy (Δφ) (???) BUT !!!!! be aware of limitations on reliability When applied to actual data……. Slow drifts, other large jitters, conditions always changing how realistic were assumptions in test simulation ?? how to apply this method reliably ?? resolve slow drifts (laser & e beam) and large jitters take large Nav data occasionally under appropriate conditions further development of method optimize data taking scheme Proposal / Plan 13/07/08 ATF2 Topical Meeting 13

laser polarization related issue are cleared up Just after injection onto vertical table very close to linearly S polarization very little polarization related M reduction polarization measurement λ/ 2 plate setting 13/07/08 ATF2 Topical Meeting 90 deg cycle “P contamination”: P p /P s = (1.46± 0.06) % Set-up IPBSM laser optics is designed for pure linear S polarization even more precise evaluation planned in autumn (hardware prepared) individual measurements for upper and lower paths near IP also measured reflective properties of “half mirror” Rs = 50.3 %, Rp = 20.1 % Match catalog specifications !! half mirror 14 power ratio

ATF2 Topical Meeting 15 lower “S peaks” (maximum M) also yield best power balance  Minimize M reduction 15 S peak P peak 45 deg between S and P During Beamtime “λ/2 plate scan “ to maximize M laser polarization and power balance and power balance Rotate λ/2 plate angle lens upper power meter investigate power balance: U vs L path 90 deg 180 deg Rotate λ/2 plate and measure high power Immediately in front of final focus lenses 13/07/08 M reduction factor due to power imbalance

Mismatch in axis between fringe and beam transverse longitudinal laser path observed on lens: precision ~ 0.5 mm (few mrad) Fringe Tilt issues: Position drifts e beam rotation in transv Current method ： “tilt scan” fringe pitch / roll adjustment: C tilt : % if uncorrected directly use e beam as reference for tilt adjustment 16 (study of fringe tilt by Okugi-san) important adjustment to eliminate M reduction 13/07/08 ATF2 Topical Meeting ex fringe pitch M 0.07  0.32 Mirrors for adjusting tilt M174L Y (8.9 mm  9.01 mm )

as an indispensible device for achieving ATF2 ‘s Goals  contribute with stable operation to continuous beam size tuning  Consistent measurement of M 〜 0.3 （ 174 ° mode) at low beam intensity correspond to σ y ~ 65 nm (assuming no M reduction) after correction for relative position jitter : maybe 〜 60 nm ??  Clearing up of residual M reduction factors  First attempts at evaluation of relative position jitter the most unknown and possibly most dominant error source ?? Summary ATF2 Topical Meeting  Maintain / improve beamtime performance : e.g. stability, precision  continue to assess residual systematic errors. Especially focus on Δy  derive the “true beam size”  aim for stable measurements of σy < 50 nm within this run Goals Shintake Monitor (IPBSM) Towards confirming σ y = 37 nm 13/07/0817

ATF2 Topical Meeting Backup 13/07/0818

Deriving Δy (Δφ) : apply to actual data very preliminary examples M reduction due to Δy about 85 – 95% Current Status: possibly valid model for deriving Δy (Δφ) (???) BUT !!!!! be aware of limitations on reliability When applied to actual data……. Slow drifts, other large jitters, conditions always changing how realistic were assumptions in test simulation ?? how to apply this method reliably ?? resolve slow drifts (laser & e beam) and large jitters take large Nav data occasionally under appropriate conditions further development of method optimize data taking scheme Proposal / Plan 13/07/08 ATF2 Topical Meeting 19

Crossing angle θ 174°30°8°2° Fringe pitch 266 nm1.03 μm3.81 μm15.2 μm Lower limit20 nm80 nm350 nm1.2 μm Upper limit110 nm400 nm1.4 μm6 μm Measures σy* = 20 nm 〜 few μm with < 10% resolution Expected Performance select appropriate mode according to beam focusing ATF2 Topical Meeting σ y and M σ y and M for each θ mode 13/07/0820

Laser interference scheme Time averages magnetic field causes inverse Compton scattering ・ phase shift at IP  α ・ wave number component along y-axis 2k y = 2k sin φ ・ modulation depends on cosθ S-polarized laser Wave number vector of two laser paths Fringe pitch 13/07/08 ATF2 Topical Meeting 21

Calculation of beam size Total signal energy measured by γ-detector Convolution of ・ Laser magnetic field ： Sine curve ・ Electron beam profile ： Gaussian M : Modulation depth Laser magnetic field Electron Beam profile with beam size σ y along y-direction S ± : Max / Min of Signal energy 13/07/08 ATF2 Topical Meeting 22

13/07/08 ATF2 Topical Meeting 23

Phase control by optical delay line Optical delay line (~10 cm) Controlled by piezo stage Movement by piezo stage : Δ stage Phase shift 13/07/08 ATF2 Topical Meeting 24

Other studies using IPBSM 13/07/08 ATF2 Topical Meeting 25 Beam intensity scan others: Test various linear / nonlinear tuning knobs IPBSM systematic error studies “Reference Cavity scan” in high β region (ex: 30 deg mode) wakefield studies Check linearity of BG levels in IPBSM detector  Observe “steepness” of intensity dependence compare with other periods to test effects of orbit tuning and / or hardware improvement for wake suppression (ex: 30 deg mode) beam intensity 5E9 / bunch BG level

IPBSM phase drift study 6/20-21 by plotting initial phase of the last Bellows y pos scan (ICT : E9/bunch) slight drift about 25 +/- 5 mrad  convert to 3.3 – 5.0 nm / min at IP 13/07/08 ATF2 Topical Meeting 26

13/07/08 ATF2 Topical Meeting 27

13/07/08 ATF2 Topical Meeting 28