Beam Stability Issues at Light Sources
- An Overview R. Hettel, SSRL
 Stability criteria
 Noise sources
 Orbit stabilizing technology
 Future directions
 Comments from SRI 2001
Stability Workshop
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Acknowledgments
Aladdin
J. Bisognano, M. Green, J. Stott
ALS/LBL
J. Byrd, D. Robin, C. Steier
APS
J. Carwardine, G. Decker, L. Emery, O. Singh
BESSY II
K. Holldack, R. Mueller
Daresbury
C. Nave
ESRF
L. Farvacque (“Beam Centre of Mass Stability”,1996)
NSLS
R. Biscardi, S. Krinsky, L.H. Yu
SLS
M. Boge, S.Zelenika, V. Schlott, A. Wrulich
SSRL
J. Corbett, J. Galayda, T. Rabedeau, J. Safranek, H. Wiedemann
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Beam Stability Requirements
Users want stability of:
• intensity after apertures
< 0.1%
apertures in phase space
• steering accuracy on small samples
< few % photon beam dimensions
pointing accuracy
• e- trajectory in IDs
< few % electron beam dimensions
emission pattern, off-axis energy pattern,
switched polarization, etc.
• photon energy
• timing
< 10-4 resolution
< 10% of critical time scale
pump-probe, etc.
• beam lifetime
many hours
Stability requirements depend on:
• time scale
• machine properties
• photon source phase space and experiment phase space acceptance
including optics
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Electron Beam Properties and Phase Space
Electron beam characterized by conjugate variable pairs in 6-D phase space:
x, x
E, t (or )
y, y
-------- transverse ----------
longitudinal
For each conjugate pair, beam occupies phase space ellipse of constant area - or
emittance
transverse:
  (s) x 2  2(s) xx   (s)x  2  cons tan t
 y  k k  coupling , k ~ 0.1
e- beam size:
 x (s)   x  x (s)  (s)E / E 
e- divergence:
 x (s)   x  x (s)  (s)E / E 
longitudinal:
 (rad) 
 y (s)   y  y (s)
2
2
h c dE
s E
( ~ 40

    / 2

 y (s)   y  y (s)

1 2




x
x
dE
for SPEAR 3)
E
Have coupling between phase space planes (6-D emittance preserved)

H-V by skew quads, orbit in sextupoles, resonances

transverse-longitudinal (Touschek scattering, x = E/E)
Beam Stability Issues at Light Sources
R. Hettel

etc.
SSILS, Shanghai, China
Sept 25,
Relationship Between Photon and Electron Parameters
(approximate, assuming constant lattice Twiss parameters; e- = Ee-/me-c2; s is Twiss parameter)
Photon Param
Relationship to Electron Parameters
1. Size at L
ph (L) = [ph(0) + (L'ph ) ]
2
2
1/2
ph (0) = [e- + diff ()]
2. Divergence
at L
2
2 1/2
2 1/2
'ph (L)='ph (0)=['e- +' ] (unfoc) 'ph(L)= -'ph(0) (1:1 foc) 'e-=[s+(' E/E) ]
-1
-1
-1/2
2
'  e- (dip, wigg)
'  e- (nNu) (und harm n, Nu periods)
E
3. Position at L
yph (L)=ye-+L y'e- (unfoc)
yph (L) = ye- (1:1 focus)
ye- (Ee-) = Ee-/Ee-
4. Angle at L
y'ph (L) = y'e- (unfocused)
y'ph (L) = -y'e- (1:1 focus)
y'e- (Ee-) = 'Ee-/Ee-
5. Critical freq/
undulator harm
2
c  Ee- (dipole)
2
-2
2
n  n Ee- /[1/2+ K + (/K) ]
6. Energy/freq res
Eph/Eph= y'ph /B (xtal mono; B=~90-800 mrad)
7. Spectral flux
density
dF()/d  Ee-Ie- S(/c)
(dip, wigg on-axis; vert angle-integrated; S(/c) in Fig.3)
2
dF(n)/dd  Ie-/'ph (n) (und, on-axis)
F()= photons/s/unit freq BW; Ie-= e- curr
8. Bunch length
t = (/s) E/E
2
2 1/2
(unfocused)
diff () = /[4' ()]
2
2
c  Ee- [1- (/K) ] (wigg.)
 = horiz view angle
/K = max ID deflect angle
(und harm n)
n/n = 1/nNu (und harm n, Nu per)
 = momentum compaction factor, s = synchrotron frequency
9. Bunch time osc tb = /rf = (/s)Ee-/Ee-
Beam Stability Issues at Light Sources
ph (L) = ph(0) (1:1 focus)
2 1/2
e- = [(s) + ((s) E/E) ]
rf = RF frequency
R. Hettel
SSILS, Shanghai, China
Sept 25,
Photon Beam Properties
Photon beam dimensions (unfocused):
 ph (L)  e  2  diff 2  (Lph ) 2
diff () 
ph  e  2 

4ph2 ()
1
nN u  2
Nu = number of undulator periods (= 1 for dipoles and high k wigglers); n = undulator harmonic
Notes:
e- = ~100-500 m horizontal
e- = ~20-100 rad horizontal
~20 - 50 m
vertical
order 10 rad vertical
1/ = order 100 rad  dominates vertical divergence except for large Nu
ph- (horiz) = order mrads for dipole and wiggler sources
Translation of photon phase space between source and experiment:
y
y
-L
ya/L
-ya
+
ya
y
-ya/L
yexp = ysp + Lysp
yexp = ysp
L
+
ya
y
aperture
experiment
source point
Beam Stability Issues at Light Sources
-ya
R. Hettel
SSILS, Shanghai, China
Sept 25,
Beam Position Instability and Emittance Growth from Orbit Motion
y
cm
o
y
eff
vertical
aperture
For disturbance frequencies >> experiment integration time:
 = o + cm
/ = cm/o
For disturbance frequencies < experiment integration time:
 (envelope) = o +2 o cm + cm
/  2  cm/o
(cm<< o; L. Farvacque, ESRF)
Note: can apply similar analysis to other phase space dimensions
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Experiment Photon - Electron Parameter Sensitivity
[Pexp(i)] = [M(i,j)] [Pe-(j)]
Iph
Ie-
Eph
Eph/Eph (rms)
x
y
EeEe-/Ee- (rms)
x
y
[Pexp(i)] = z
[Pe-(i)] = z
x
y
tbunch
polarization
rot
x
y
t bunch
coherence
polarization
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Beam Stability - Time Scales
Disturbance frequencies >> experiment integration time:
Orbit disturbances blow up effective beam  and , reduce intensity at experiment,
but do not add noise.
For / = cm/o < ~10%:
ycm(rms) < ~0.3 y
ycm(rms) < ~0.3 y'
Note: can have frequency aliasing if don't obey Nyquist….
Disturbance frequencies  experiment integration time:
Orbit disturbances add noise to experiment.
For / = ~2 cm/o <~10%:
ycm(rms) < 0.05 y
ycm(rms) < 0.05 y'
Disturbance frequencies << experiment time (day(s) or more):
Realigning experiment apparatus is a possibility.
Sudden beam jumps or spikes can be bad (low rms)
Note: peak amplitudes can be > x5 rms level
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Intensity Stability
Want high level of flux (I) constancy through aperture or steering
accuracy to hit small sample (sample size on order of beam size ).
I/I < 10-3 (typical)
Note: some experiments require < 10-4 flux constancy
Photoemission electron spectroscopy combined with dichroism spectroscopy
(subtractive processing of switched polarized beam signals)
(H. Padmore)
Flux variations caused by
• orbit instability
• beam size instability
• energy instability
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Beam Stability at Apertures
Photon Intensity Noise after Aperture
Beam Position Change
Photon Intensity Noise after Aperture
Beam Size Change
10
d = 0.5
1
d = 1
noise factor (%)
noise factor (%)
10
d = 0.1
d = 2
0.1
d = 3
0.01
d = 0.5
1
d = 1
d = 0.1
d = 2
0.1
d = 3
0.01
0.001
0.001
1
10
d =half aperture
dy/ %
100
dy =displacement from center
0.1
1
d =half aperture
10
d/ %
Intensity variations for beam with Gaussian size  due to position motion dy from the
center and beam size change d for various sized apertures.
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Intensity Stability - Orbit
For 0.1% intensity stability, orbit stability should be
xcm, ycm < .05 x,y
at source point for focused beams
xcm, ycm< .05 y'
at source point for unfocused beams
~140 rad @ 3 GeV, N = 1
Vertical:
y ~ 20-50 m
y = ~17 rad @ 3 GeV, N = 100;
~7 rad @ 7 GeV, N = 100
 ycm < ~1-3 m,
ycm< ~0.5-10 rad
Horizontal:
for 3rd gen sources
~mrads for dipoles, wigglers
x ~ 100-500 m
x = ~35 rad @ 3 GeV, N = 100;
~20 rad @ 7 GeV, N = 100
But for wigglers: dF()/d  Ee- Ie-S(/c)
1
S(/c)
c  E2[1 - (/K)2]
x = /c
1.333 x1/3
 xcm < ~5 - 25 m,
-x
0.777 x1/2e
0.1
0.001
0.01
/c
0.1
1
10
Beam Stability Issues at Light Sources
xcm< rads
(for undulators and off-axis wigglers
at /c >> 1)
0.01
0.0001
 = horiz view angle
R. Hettel
SSILS, Shanghai, China
Sept 25,
0.1% Intensity Stability - Energy
Orbit perturbation:
Coherent electron energy variations in lattice dispersion sections at frequencies < data
integration time (synchrotron oscillations, RF)
x =   E/E < .05x = ~10-20 m

E/E (coherent) < ~10-4
(x= ~0.1 m at dipole sources; y= ~0.02 m)
( < 0.2o for SPEAR 3)

fRF/f RF = c E/E < ~10-7
( c = ~10-3)
(fRF < 50 Hz for f RF = 500 MH); imposes limit on phase noise for RF source ~10 kHz BW)
Photon emission:
dF()/d  Ee- Ie- S(/c)
1
c 
E2
(dipole, wiggler)
S(/c)
x = /c
1.333 x1/3
-x
0.777 x1/2e
0.1
dF(n)/d d  Ie-/ph2  Ee-2 Ie- n Nu
(undulator)
0.01
0.0001
 E/E (coherent) < ~10-4
Beam Stability Issues at Light Sources
R. Hettel
0.001
0.01
/c
0.1
SSILS, Shanghai, China
1
10
Sept 25,
Intensity Stability - Beam Size
For 0.1% intensity stability, beam size stability should be
/ < ~10-3
Beam size-perturbing mechanisms:
• changes in horizontal-vertical electron beam coupling
ID gap change, orbit in sextupoles, energy ramp without coupling correction
• collective effects
coupling resonances, single- and multibunch instabilities in transverse and
longitudinal planes, intrabeam scattering
• gas bursts, ions, dust particles
• electron energy variations in lattice dispersion sections at frequencies > data
integration time (synchrotron oscillations, Landau damping mechanisms, etc.)
2 =  + (E/E)2 + (E/E)2
( = ~350 m;  E/E = ~100 m for  = 0.1 m and
E/E (natural energy spread) = ~ 0.1%; 0 = ~360 m)
 E/E (rms) < ~ 10-4
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Photon Energy Resolution
Photon energy resolution <10-4 after monochromator:
Bragg:
E ph
E ph

where B = ~5o-45o (~90-800 mrad)
B
 ycm < ~10 rad

Note: some monochromators reaching 10-5 resolution  ycm < ~1 rad
Undulator line energy and width not degraded:
line wavelength = n = n u(1+k2/2)/22
 d n/ n = -2 E/E
natural width =  n/n = 1/Nn (N = # periods, n = harmonic
• for <10% line width increase (N = 100, n = 10)  E/E (rms) < ~2 x 10-4
• for <10-4 coherent energy shift (N = 100, n = 10)  E/E (coherent) < ~5 x 10-5
Note: for 10-5 resolution  E/E (coherent) < ~5 x 10-6
(  < 0.01o for SPEAR 3
Beam Stability Issues at Light Sources
fRF < 2.5 Hz for f RF = 500 MHz)
R. Hettel
SSILS, Shanghai, China
Sept 25,
Timing/Bunch Length Stability
Bunch time-of-arrival stability (tbunch):
tbunch < ~0.1 of critical time scale in experiment (pump-probe sync, etc.)
- or < ~0.1 bunch
( bunch = ~5-50 ps)
whichever is larger
Time-of-arrival variations caused by energy oscillations:
t bunch 
 (rad )
2f rf
 tbunch = ~1.3 ps for dE/E (coherent) = 10-4 in SPEAR 3
( = ~0.08 bunch)
Bunch length variations associated with changes in energy spread causes
beam size variation:
E/E (rms) < 10-4  bunch < 0.5% bunch
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Lifetime
Lifetime contributors:

quantum lifetime

gas scattering lifetime (Coulomb, bremsstrahlung)

Touschek lifetime

ions and dust particles
Touschek usually dominant lifetime factor for low E rings (< 3 GeV):
 x  y  s  2  p  m
 p
 Touschek 
N
p/p = ring momentum acceptance
N = number of particles in bunch
 control and stabilize bunch volume
Ion trapping prevented by having gap in bunch fill pattern
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Sources of Electron Beam Instability
Long term (weeks-years)

ground settlement (mm)

seasonal ground motion (< mm)
Medium term (minutes-days)

diurnal temperature (1-100 m)

river, dam activity (1-100 m)

crane motion (1-100 m)

machine fills (component heating, BPM intensity dependence)

filling patterns (heating, BPM processing) (1-100 m)

RF drift (microns)

gravitational earth tides (sun and moon) (C = 10-30 m)

coupling changes
Short term (milliseconds-seconds)

ground vibration, traffic, trains, etc. (< microns, <50 Hz typ)
ground motion amplified by girder + magnet resonances (x~20 if not damped) and by lattice (x10-x40)
 nm level ground motion can be amplified close to m level

cooling water vibration (microns)

rotating machinery (air conditioners, pumps) (microns)

booster operation (microns)

insertion device motion (1-100 m)

power supplies (microns)

vac chamber vibration from BL shutters, etc. (microns)
High frequency (sub-millisecond)

high frequency PWM and pulsed power sources (microns)

synchrotron oscillations (1-100 m)

single- and multibunch instabilities (1-100 m)
Note: relative component motion more critical than common mode motion
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Ground Motion - slow
Ground settling (ALS)
Gravitational earth tides
due to sun and moon
(ESRF)
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Ground Vibration
APS
worldwide
G. Decker
V.D. Shiltsev
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Vacuum chamber, BPM stability
INVAR
INVAR
INVAR
NSLS chamber motion
SPEAR 3 chamber/BPM supports
J. Safranek
3 m/oC vert, 15 m/oC hor
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Beam Stabilizing Technology
Accelerator component motion
• short girders, rigid supports (high frequency modes)
• damping materials (visco-elastic double-sided sticky tape, Anocast)
• reduced vibration from water pipes, machinery, pumps, booster, etc
• minimal chamber temp gradients, chamber water cooling, discrete photon stops
• temp-stable BPM supports (Invar, carbon fiber - <3 m/oC @ 1 m)
• tunnel air and cooling water temperature stabilization
(<0.2oC).
• component position sensors (HLS, micron-res encoders, etc.)
• calibration of temperature-induced motion
• at-energy and top-off injection
Corrector magnets
• high frequency, low inductance (kHz)
strong DC correctors (iron core) + low inductance tweeters (e.g. air core)
• high transverse linearity vs. frequency
• high resistance vac chamber sites (low eddy currents)
• uniform magnetic environment at all sites (field clamps)
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Beam Stabilizing Technology - cont.
Power supplies
• low-ripple PWM chopper technology (20 kHz for mains, 40-60 kHz for correctors)
• high stability (transductors, even for correctors)
• high frequency, low quantization noise corrector supplies (>18-bit DACs)
• digital control and regulation
Orbit monitoring systems
• > kHz orbit acquisition
• 1st turn, turn-turn and narrowband (high resolution)orbit processing modes
• non-multiplexed button processing (calibration tones, hybrid combiners)
• compensation for BPM motion, intensity dependence, etc.
• beam-based offset correction (quadrupole modulation)
• digital processing (including digital receivers)
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Beam Stabilizing Technology - cont.
Feedback and other active systems
• unified global/local orbit feedback with electron and photon BPMs (up to ~200 Hz BW)
• automatic magnet girder alignment (ESRF, SLS, etc.)
• mode-damped RF cavities
• mode-0 longitudinal feedback through RF system
• multibunch feedback systems (longitudinal and transverse)
• harmonic cavities for bunch length control and Landau damping (NSLS, ALS, etc.)
• tune feedback (Aladdin)
• beam line component feedback (mirror, monochromator, etc.)
More sophisticated orbit feedback algorithms include:
• compensation for ID gap changes
• compensation for accelerator component motion (measured or calibrated)
• real-time digital processing and high speed data links
• RF feedback
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
ESRF Girder Damping
L. Zhang
beam line intensity noise
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
SLS Girder Mover System
V. Schlott and S. Zelenika et al.
System
System
cam mover and hydrostatic level detector (micron res.)
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Local 3- and 4-Magnet Feedback Systems
SSRL, 1982-86
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Global Orbit Feedback Systems
NSLS Harmonic Feedback System
APS SVD Global Feedback System - 1990s
1988
J. Carwardine, Y. Chung, F. Lenkszus, et al.
L.H. Yu, R. Biscardi, J. Bittner, E. Bozoki, J. Galayda,
S. Krinsky, R. Nawrocky, O. Singh, G. Vignola
• 360 BPMs
• 80 correctors/plane
• BPM de-spiking
“DSP Scope”
1) Poor regulation of
sextupole supply
2) Steering supply
oscillating at 248 Hz
3) Bad BPM with
broadband noise
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Local vs. Global Feedback
yID y’ID
y1
Local correction:
+
+
L
BPM 1
yID = (y1 + y2 )/2
 y2ID = y2/2
y2
yID = (y1 - y2 )/2L
BPM 2
 y2ID = y2/2L2
y = measurement error
e.g., y = 10 m, L = 3 m  7 m rms position error, 2.4 rad rms angle error
 want L to be large
Multi-loop crosstalk  reduced performance
Global correction:
• reduced set of correction eigenvectors filters
response to BPM noise
lower spatial BW, more BPMs in average
ESRF
• correction matched to most likely disturbances
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
SVD Orbit Correction
SPEAR (J. Corbett)
NSLS X-Ray ring inverse response matrix
with different SVD thresholds (J. Safranek)
SVD threshold = 0.01
SVD threshold = 0.0001
Beam Stability Issues at Light Sources
R. Hettel
SVD threshold = 0.001
SVD threshold = 0
SSILS, Shanghai, China
Sept 25,
RF Feedback
NSLS
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Orbit Feedback - Performance Limitations
Electron BPMs (electron)

BPM mechanical motion (thermal)

intensity or fill pattern dependence

TE10 mode in antechamber

local position/angle calculation errors

orbit acquisition rate (rate ~10x fdbk BW  1 kHz for 100 Hz BW; rate ~2 kHz to avoid 720 Hz alias)

BPM muxing (switch rate > ~10x orbit acq BW to avoid H-V coupling; limited by switch settling time)
Photon BPMs
• corruption of undulator photon monitor signal by dipole radiation
• dependence on ID gap (emission pattern) and beam intensity
• blade surface contamination
Orbit corrector system

transverse field uniformity: orbit-dependent response matrix

field uniformity vs. frequency: freq-dependent response matrix

vacuum chamber BW

hysteresis

quantization noise and step resolution
Response matrix
• inadequate BPM and/or corrector quantities (>4 per  period desirable)
• inefficient BPM and/or corrector placement (low orthogonality)
• non-optimized eigenvector cut-off for R-1 (trade-off between azimuthal correction resolution and
corrector strengths, sensitivity to BPM noise)
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Undulator Photon Monitor Limitations
• ID photon emission pattern (flux and
energy) changes with ID gap
• Causes apparent beam motion of
~1-100 m, which can propagate
around ring if photon monitors used
in feedback.
• Can compensate for gap changes with
calibrated feedforward orbit and tune
corrections + feedback to correct
residual perturbations.
ESRF
Beam Stability Issues at Light Sources
• Blade monitors work better for low-E
rings (photon spectrum better
matched to blade photoemission
properties) - K. Holldack et al., BESSY II)
R. Hettel
SSILS, Shanghai, China
Sept 25,
Photon Monitors - Blade Geometries
APS
O. Singh
BESSY II
K. Holldack
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
APS Undulator Soft-Bend Chicane
G. Decker
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Orbit Feedback - Corrector Resolution
R. Mueller et al., BESSY II
Noise from 16-bit DACs solved with 24-bit
DACs (~20-bit ENOB)
(feedback cycle rate = 0.2 Hz)
16-bit DAC noise
beam line intensity noise
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
SPEAR 3 Cu Chamber Eddy Current Break
Horizontal Corrector - Chamber Frequency Response (on-axis)
(from Mafia)
0.00
-3.00
0.1
1
10
100
1000
attenuation (dB)
-6.00
-9.00
CuNi
-12.00
Cu
-15.00
-18.00
-21.00
-24.00
-27.00
frequency (Hz)
Vertical Corrector - Chamber Frequency Response (on-axis)
(from Mafia)
0.00
0.1
1
10
100
1000
attenuation (dB)
-3.00
CuNi
-6.00
Cu
-9.00
-12.00
CuNi
Cu
f = 200 Hz
-15.00
frequency (Hz)
Cu
CuNi
CuNi
f = 200 Hz
Beam Stability Issues at Light Sources
R. Hettel
Cu
insulation
SSILS, Shanghai, China
Sept 25,
Multibunch Feedback
Longitudinal
J. Fox et al., SLAC
Transverse
W. Barry et al., ALS
Harmonic Cavities
• increase bunch length and Touschek lifetime
• induce tune spread to damp multibunch instabilities
ALS 1.5 GHz cavity
J. Byrd, R. Rimmer et al.
(Landau damping)
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Mirror Feedback
• error signal obtained from
position sensitive detector
near beam focus
T. Rabedeau, SSRL
• error signal used to control
piezo high voltage
• piezo provides mirror fine
pitch control with typical full
range of motion +/- 30 rad
or +/- 0.6mm or more focus
motion.
focus 1.4 m rms
source 17.3 m rms
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Future Light Sources
Energy Recovery Linac
MARS, G. Kulipanov (Budker Inst.); ERL, S. Gruner et al.(Cornell, JLab)
PERL, J. Murphy et al. (NSLS); Ultrafast Light Source, A. Zholents et al. (LBNL)
x = ~0.15 nm-rad @ 5-7 GeV x,y = 3-40 m x,y = 3-10 rad
 ~ 0.1 m, 0.1 rad, <100 fs stability
Challenges:
 energy
stability
 bunch
charge stability
 pump-probe
s = 100 fs - 10 ps
timing sync
 other
LCLS (& SPPS)
 = ~0.05 nm-rad @ 15 GeV
x,y = ~30 m
x,y = ~0.4 rad
s = 85 fs
 micron, 10s nrad, <100 fs stability
Challenges:
<
<
0.9 ps gun timing jitter
 <0.05%
klystron voltage stability

0.07 S-band klystron phase stability
pump-probe timing sync
 other
Diffraction-limited storage rings (J.-L. Laclare)
x = ~0.1 nm-rad
y = ~8 pm-rad (1Å diffraction limit)
x = ~100 m
y = ~10 m
y = <10 rad
s = ~10 ps
 ~ 0.1 m, 0.1 rad stability
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Beam Stability Requirements - Summary
Parameter
Present
Future trend
intensity stability
<0.1%
<0.01%
steering accuracy
<5-10% e-, ph
<2% e-, ph
beam size stability
<0.1% ph
10-4
<10% bunch length
<0.01% ph
10-5
<10% bunch length
~10-2 - 105 Hz
~10-2 - 106 Hz
hrs
days
emittance
~5-20 nm-rad
~0.05-0.2 nm-rad
e- beam size (vert)
ph beam divergence
e- bunch length
~30-300 m
~10-200 rad
~3-30 m
~0.5-10 rad
~10-100 ps
~100 fs (FEL)
~1-5 m
~1-10 rad
~1-10 ps
<~5 x 10-5 ( < 0.1o)
~0.1-1 m
~0.05- 0.5 rad
~10-100 fs (FEL)
<~5 x 10-6 ( < 0.01o)
energy resolution
timing stability
data acquisition rate
stability period
e- position stability (vert)
e- angle stability
e- bunch length stability
e- energy stability
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Beam Stability Requirements
DRAFT rev. 8/27/01
e- stability need
beam param
user stability need
Parameter
Present sources (rings)
Future sources (ring, ERL, FEL)
intensity (shot-shot with normalization)
<0.1%
<0.01%
steering accuracy
<5% e-, ph
<2% e-, ph
beam size stability
<0.1% ph
<0.01% ph
-5
-4
energy resolution/accuracy
10 -10
10-5
timing stability
<10% bunch length
<10% bunch length
-3
polarization stability (switched)
10 (?)
10-4 (?)
expt data acquisition rate
~10-2-105 Hz
~10-2-106 Hz
stability period (w/o realignment)
secs-hours
secs-days
beam availability
>90%
>90%
emittance
~5-20 nm-rad
~0.05-0.2 nm-rad
ph beam size (vert)
~30-300 m
~3-30 m
ph beam divergence
~0.5-10 rad
~10-200 rad
H-V emittance coupling
<0.1-10%
<0.01-100%
e- beam rotation
<?? mrad
<?? mrad
e- bunch length
~10-100 ps
~10 ps (ring); ~100 fs (ERL,FEL)
position stability (vert)
~1-5 m
~0.1-1 m
angle
~1-10 rad
~0.05- 0.5 rad
coupling
<0.1%
<0.01%
rotation
<?? mrad
<?? mrad
-2
-11
-2
-11
bunch-bunch / turn-turn charge
~10 / 10
~10 /10 (ring); ~10-2/10-2 (FEL, ERL)
bunch length
~1-10 ps
~1 ps (ring); ~10-100 fs (FEL)
-5
o
energy
<~2-5 x 10 ( < 0.04-0.1 )
<~5 x 10-6 ( < 0.01o)
stability bandwidth
1/hrs - 105 Hz
1/hrs - 106 Hz
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
SRI 2001 Beam Stability Workshop - Comments
Madison, WI
Aug. 24, 2001
• ~50% users, 50% accelerator people at Workshop
• More communication between users and accelerator groups needed to understand
instability requirements
• Users could specify stability needs for experiments; could specify generic needs for
different classes of experiments via specialized committees
• Characterize photon-electron sensitivity for each beam line - joint effort between
accelerator and beam line staff
• More robust experiment designs might be possible (i.e. optimization of experiment
phase space acceptance with respect to photon phase space)
• Accelerator group could be more actively involved working on beam line problems
• More cooperative accelerator and beam line diagnostics
• “Look at the light” -- i.e., use optical monitors in beam lines to determine beam
stability, don’t rely solely on electron monitors
• Facility user meetings could include stability discussion session
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
User observable - electron observable sensitivity matrix
M. Greene, Aladdin
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Summary
Stability requirements are stringent:

intensity stability < 0.1%

pointing accuracy < 5% beam dimensions

photon energy resolution < 10-4

timing stability < 10% bunch length
 orbit < 1-5 m, <1-10 rad
beam size < 0.1 %
e- energy < 5 x 10-5
Requirements are becoming more stringent:

x 5-10 more stringent stability with beam source and beam line development
 orbit < .1-1 m, <.05-.5 rad
beam size < 0.01 % e- energy < 5 x 10-6

faster data acquisition time-scales


short bunch machines present pump-probe timing sync challenge: <100 fs
fast-switched polarization, ID changes
Stabilizing technology development:

control noise sources

temp-stable materials

vibration-damping materials

motion sensors, calib

automatic alignment

better BPMs (e- and photon)

faster orbit feedback

adaptive optics

improved signal processing

robust expt design

expt component feedback (integrate with accel feedback?)
may relieve need for absolute accelerator stability
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,
1 m
Beam Stability Issues at Light Sources
R. Hettel
SSILS, Shanghai, China
Sept 25,