Laser-wireMeasurement
PrecisionGrahame Blair
Beijing- BILCW07, 6th February 2007
• Introduction
• Overview of errors
• Ongoing technical work in this area
• Plans for the future.
BESSY: T. KampsDESY : E. Elsen, H. C. Lewin, F. Poirier, S. Schreiber, K.
Wittenburg, K. BalewskiJAI@Oxford: B. Foster, N. Delerue, L. Corner, D. Howell, M.
Newman, A. Reichold, R. Senanayake, R. WalczakJAI@RHUL: G. Blair, S. Boogert, G. Boorman, A. Bosco, L.
Deacon, P. Karataev, S. Malton , M. Price I. Agapov (now at CERN)
CCLRC: I. RossKEK: A. Aryshev, H. Hayano, K. Kubo, N. Terunuma, J.
UrakawaSLAC: A. Brachmann, J. Frisch, M. WoodleyFNAL: M. Ross
Laser-wire People
Laser-wire Principle
PETRAII• 2d scanning system• DAQ development• Crystal calorimeter
→ PETRA III• Ultra-fast scanning• Diagnostic tool
The Goal: Beam Matrix Reconstruction
50% Reconstruction success <5% error on σy
NOTE: Rapid improvementwith better σy resolution
Reconstructed emittanceof one trainusing 1% error on σy
Conclude: Essential to measurethe spot-size at the few % level or better
I. Agapov, M. Woodley
x
y u
x
y u
Skew Correction
y
x
1
optimal tan
ILCat 8868
21
222
4
y
y
x
x
u
uyxxy
Error on coupling term:
ILC LW Locations Eb = 250 GeV
x(m) y (m) opt(°) u (m)
39.9 2.83 86 3.9917.0 1.66 84 2.34
17.0 2.83 81 3.95
39.2 1.69 88 2.39
7.90 3.14 68 4.1344.7 2.87 86 4.05
2σscan
2σe
2αJσe
Ntrain bunches
2σL=2cτL
2σe (1 + strain)
αtrainσe
Scan of an ILC Train of Bunches
Not to scale!
Dyx
e HfNN
L4
rep2
train 2train3
1111 s
Need for Intra-Train Scanning
For <0.5% effect, strain<0.12; otherwise, the effect must be subtracted
For 1m bunches, the error after subtracting for any systematic shift (assumed linear ±αtrainalong the train) is:
trainBPM3
nm 100109.1
e
e
For <0.5% effect, αtrain<2.6; otherwise, higher precision BPMs required
Machine Contributions to the Errors
Bunch Jitter
21222scan EeJe
Dispersion
nm1000.5105 BPM
22
J
e
e
2mm/3.2
e
eAssuming can be measured to 0.1%,then must be kept < ~ 1mm
BPM resolution of 20 nm may be required
Alternative Scan Mode• R&D currently investigating ultra-fast scanning (~100 kHz)
using Electro-optic techniques• Alternative: Keep laser beam fixed and use natural beam jitter
plus accurate BPM measurements bunch-by-bunch.Needs the assumption that bunches are pure-gaussian
• For one train, a statistical resolution of order 0.3% may be possible
m10 m,1 exey
Beam jitter fixed at 0.25σ BPM resolution fixed at 100 nm
Single-bunch fit errors for
σeyσexσℓ
xR
laser beam
electron bunch
√2σℓ
xfzy
xfIzyxI
RR2
22
20
2exp1
2,,
ℓℓℓ
For TM00 laser mode:
2
1
RR x
xxf
#2 fM ℓ
2#
2 4 fMxR
Laser Conventions
Compton Statistics
2
21
Detected exp2
11212m
y
m
N
Approximate – shoulduse full overlap integral(as done below…)
Where :
m2.0nm 532102MW 100.05 10
det
fNP eℓ
Laser peak power
Detector efficiency(assume Cherenkov system)
Compton xsec factor
Laser wavelength
e-bunch occupancy
TM00 Mode Overlap Integrals
m10
m,1
ex
ey
m100
m,1
ex
ey
Rayleigh Effects obviousMain Errors:• Statistical error from fit ~ -1/2
• Normalisation error (instantaneous value of ) – assume ~1% for now.• Fluctuations of laser M2 – assume M2 known to ~1%• Laser pointing jitter
2
22
2
#
MMMf
ee
e
%10/
rad10102.2
23
e
e
Y. Honda et al
Estat EM2
TM01
TM00
TM01
TM00
TM01
TM01 gives some advantage for larger spot-sizes
Laser RequirementsWavelength 532 nmMode Quality 1.3Peak Power 20 MWAverage power 0.6 WPulse length 2 psSynchronisation 0.3 psPointing stability 10
rad
ILC-spec laser is being developed at JAI@Oxfordbased on fiber amplification. L. Corner et al
ATF2 LW; aiming initiallyat f2; eventually f1?
TM00 mode
Error resulting from5% M2 change
Statistical ErrorFrom 19-point scan
• Optimal f-num1-1.5 for = 532nm• Then improve M2 determination• f-2 lens about to be installed at
ATF
Relative Errors
Towards a 1 m LW
Wavelength 266 nmMode Quality 1.3Peak Power 20 MWFF f-number 1.5Pointing stability 10
radM2 resolution 1%Normalisation ()
2%
Beam Jitter 0.25BPM Resolution 20 nmEnergy spec. res 10-4
Goals/assumptionsE 2.5
Epoint 2.2
Ejitter 5.0
Estat 4.5
EM2 2.8
Total Error
8.0
Resultant errors/10-3
Could be used for measurement→ E
Final fit, including dispersion
preliminary
Lens Design + Tests
Aspheric doublet
N. Delerue et al.
Vacuum window
• f-2 lens has been built and is currently under test.
• Installation at ATF planned for this year
M. Newman, D. Howell et al.
Designs for f-1 optics are currently being studied, including:
S. Boogert, L. Deacon
ATF Ext
ATF/ATF2 Laser-wire• At ATF2, we will aim to measure micron-scale electron spot-
sizes with green (532 nm) light. • Two locations identified for first stage (more stages later)
1) 0.75m upstream of QD18X magnet2) 1m downstream of QF19X magnet
LW-IP (1) LW-IP (2)
σx = 38.92 m σx = 142.77 m
σy = 7.74 m σy = 7.94 m
Nominal ATF2 optics
ATF2 LW-test optics
P. Karataev
LW-IP (1) LW-IP (2)
σx = 20.43 m σx = 20 m
σy = 0.9 m σy = 1.14 m
Ideal testing ground for ILC BDS Laser-wire system
ATF LW Plans• March 07: Start upgrading ATF LW hardware• April 07: aim to install f2 lens system• May/Jun 07: aim to take first micron-scale scans
Longer term• Upgrade laser system to reduce spot-size further• Install additional LW systems, building towards emittance
measurement system for ATF2.• Investigate running with UV light.• Implement ultra-fast scanning system (first to be tested at
PETRA, funding permitting)• Build f-1/1.5 optical system
Summary• Very active + international programme:
- Hardware- Optics design- Advanced lasers- Emittance extraction techniques- Data taking + analysis- Simulation
• All elements require R&D- Laser pointing- M2 monitoring- Low-f optics- Fast scanning- High precision BPMs
• Look forward to LW studies at PETRA and ATF• ATF2 ideally suited to ILC-relevant LW studies.