Yujong Kim

1
May 15, 2007, Fermi National Accelerator
Laboratory, USA
Experimental Optimization of TTF2 RF
Photoinjector Bunch Compressors
Yujong Kim Free Electron Laser Laboratory Duke
University
TESLA-S2E-2007-71
yjkim_at_fel.duke.edu, http//www.fel.duke.edu/yjkim
2
Contents

• Introduction on DESY TTF2/FLASH
• Layout, Commissioning History, Main Components,
  and Upgrades in 2007
• Optimization of TTF2 RF Photoinjector to get a
  lowest Emittance
• L-band RF Gun and Booster Module (ACC1)
• Invariant Envelop and Difficulties
• 1st Experimental Demonstration of Emittance
  Damping in Booster Module
• Optimization of TTF2 Bunch Compressors to get a
  fs long Electron Bunch
• Principle of Bunch Compressors and CSR effects
• Optimization of TTF2 Bunch Compressors against
  CSR effects
• Setting of RF Phases in Modules
• 1st FEL Lasings with about 10 20 femtosecond
  Long FEL Photon Pulse
• Summary and Acknowledgement

3
Introduction on DESY TTF2/FLASH

• Layout of TTF1
• Layout of TTF2/FLASH for Nominal Mode Operation
  in 2008 (?)
• History of TTF2/FLASH Commissioning
• Layout of TTF2/FLASH for Femtosecond Mode
  Operation (2004 - 2007)
• TTF2 Main Components
• Major Upgrades in 2007

4
TESLA Test Facility Phase 1 (TTF1)
In November 2002, we removed all components of
TTF1 and started TTF2 project !
e- beam diagnostics
e- beam diagnostics
bunch compressor
laser driven electron gun
15 m, undulator
photon beam diagnostics
booster
TESLA superconducting accelerator modules
E 230 260 MeV I 1.0 1.5 kA enx 4 7 mm
Photon beam properties
Saturation Wavelength 80 -120
nm Spectral width (FWHM) 1 Pulse
energy at saturation 30 -100 mJ Pulse
duration (FWHM) 30 -100 fs Peak
power 1
GW Average power 5
mW with 70 pulses Radiation spot size FWHM
250 mm Angular divirgence FWHM 260
mrad Peak brilliance up to
41028
To develop superconducting accelerator technology
with a higher gradient, To develop needed
technologies for European XFEL and a future
linear collider, To supply femtosecond long FEL
photon at soft X-ray region to users, Conversion
of TTF1 to TTF2/FLASH was started in November
2002.
5
TTF2/FLASH Layout - Nominal Operation
Layout of TESLA Test Facility Phase 2 (TTF2) /
FLASH
Beam energy 1.0 GeV (2007) Peak current 2.5
kA rms bunch length 50 µm 166 fs Bunch
separation 111 ns Number of bunch per train
7200 Repetition rate 10 Hz Slice transverse
normalized rms emittance 2 µm rms beam size
68 µm rms energy spread 1 MeV _at_ 1.0
GeV Undulator period 27.3 mm Undulator gap 12
mm K-parameter 1.17 Undulator length 30
m Peak magnetic field 0.495 T (NdFeB
permanent) SASE source wavelength 6.4 nm 120
nm Saturation length 27 m Peak power 2.8
GW Average power 40 W SASE source pulse length
(FWHM) ca. 100 fs Peak (average) spectral
brightness 2.41030 (3.51022)
photons/sec/mrad2/mm2/0.1BW
2003 2004 2005 2006 2007 2008
Linac undulator installation Gun Test at
PITZ Gun (PITZ II) installation Injector
commissioning Complete vacuum installation 1st
beam at beam dump Beam through undulator 1st
lasing _at_ 32 nm 1st FEL user service _at_ 32 nm 1st
lasing _at_ 13 nm 1st FEL saturation _at_ 13 nm 1st FEL
harmonics _at_ soft X-ray Lasing with a long bunch
train Shutdown for upgrade - Now Lasing at 6 nm
with 1 GeV Installation of 3rd harmonic
Installation of seeding option Seeding
demonstration
ACC39, ACC6, and SEEDING are not installed yet !
6
FLASH Operation Conditions for fs FEL Mode
Delayed Installation of ACC39 and users' request
on femtosecond range photon pulse
Original Plan
Modified Plan (2004 - 2007)
For 30 nm lasing Q 0.5 nC ACC1 phase -8 deg
off crest ACC39 off BC2 energy 125 MeV BC2
bending angle 18 deg ACC23 phase on crest BC3
bending angle 5.4 deg (close to edge) BC3
energy 380 MeV ACC45 phase on crest Beam
energy at the undulator 450 MeV Peak current
1.3 kA Slice emittance 3 µm at spike Slice
energy spread 300 keV at spike Saturation
length 20 m
For 32 - 13 nm lasing Q 1.0 nC (to get a good
stability) ACC1 phase -9 deg off crest against
CSR ACC39 off BC2 energy 127 MeV BC2 bending
angle 18 deg ACC23 phase on crest (or
slightly off crest) BC3 bending angle 3.577 deg
(center) BC3 energy 380 MeV ACC45 phase on
crest for a low E spread Beam energy at the
undulator 445 - 700 MeV Peak current 1.1
kA Slice emittance 2 µm at spike Slice energy
spread 400 keV at spike Saturation length 27 m
7
TTF2 Layout for the fs FEL Mode Operation
One example layout - during FEL operation on
April 9th, 2005
low energy spread, low emittance, high peak
current for FEL operation
Q 1.0 nC
sz 1.821 mm 794 µm
728 µm (FWHM 200 fs 10 fs)
BYPASS
BC2
BC3
RF-GUN
ACC1
ACC2
ACC3
ACC4
ACC5
LOLA
COLLIMATOR
6 UNDULATORs
DUMP
E 127 MeV R56 181.3 mm sd 0.60 T
18.0355 deg (15 deg 21 deg)
E 380 MeV R56 43.1 mm sd 0.18 T
3.577 deg (1.7 deg 5.4 deg)
E 445 MeV sd 0.16
Under no compression and after BC2 ACC12345
phase on crest projected emittance 1.1 µm
for 90 intensity projected emittance 1.8
µm for 100 intensity Under compression with
this layout and at undulator ACC1 phase
-9.0 degree ACC2345 on crest peak current
1.0 1.5 kA bunch length 155 (FWHM) fs
for -10 deg off crest slice normalized rms
emittance 2 µm at a spike slice energy
spread 400 keV rms horizontal beam size
160 µm rms vertical beam size 100 µm
12.95 MV/m 16.84 MV/m -9.0 degree
17.29 MV/m 13.53 MV/m 0.0 degree
3.85 MV/m 4.03 MV/m 0.0 degree
40.25 MV/m (peak) 38 degree from zero crossing
Final bunch length (200 fs 10 fs) depends on
phases of ACC1, ACC2, ACC3 Here all module
gradients are the average accelerating gradient
instead of the peak ones
8
TTF2 Main Components RF GUN
Q 1.0 nC
E 127 MeV R56 181.3 mm sd 0.60 T
18.0355 deg (15 deg 21 deg)
E 380 MeV R56 43.1 mm sd 0.18 T
3.577 deg (1.7 deg 5.4 deg)
E 445 MeV sd 0.16
1.3 GHz 1.5 cell OFHC copper cavity Cs2Te
(Cesium Telluride cathode), Q.E a few Fully
symmetric coupler with an axial coupling Water
cooling for up to 50 kW Average RF power for 0.9
ms 27 kW Long RF pulses (up to 0.9 ms) 10 Hz
rep rate Duty cycle 0.9 Klystron power 5
MW Max gradient 40 MV/m on the cathode ( 3
MW) Best main solenoid current 277 A 0.163
T Bucking current 20 A
9
TTF2 Main Components GUN Driving Laser
PITZ already installed. 21
ps flat-top pulse
TTF2 not installed yet.
4.4 ps RMS Gaussian Beam
1047 nm
2
10
TTF2 Main Components - Gun Driving Laser
Laser image on virtual cathode
11
TTF2 Main Components TESLA SC Modules
E 127 MeV R56 181.3 mm sd 0.60 T
18.0355 deg (15 deg 21 deg)
E 380 MeV R56 43.1 mm sd 0.18 T
3.577 deg (1.7 deg 5.4 deg)
E 445 MeV sd 0.16
12.95 MV/m 16.84 MV/m -9.0 degree
17.29 MV/m 13.53 MV/m 0.0 degree
3.85 MV/m 4.03 MV/m 0.0 degree
Laser image on virtual cathode
40.25 MV/m (peak) 38 degree from zero crossing
12
TTF2 Main Components Bunch Compressors
13
TTF2 Main Components Two 3 FODO Cells
14
TTF2 Main Components Two 3 FODO Cells
15
TTF2 Main Components Two 3 FODO Cells
Q4DBC2H Q5DBC2 Q6DBC2
Q7DBC2 Q8DBC2 Q9DBC2
Q10DBC2H
DBC2 section

• 4 OTR in 3 FODO cells
• OTR4DBC2
• OTR6DBC2
• OTR8DBC2
• -OTR10DBC2
• Cell phase advance 45 deg
• One cell length 1.9 m
• 7 QMs in three FODO Cells
• Q4DBC2 - defocusing
• Q5DBC2 - focusing
• Q6DBC2 - defocusing
• Q7DBC2 - focusing
• Q8DBC2 - defocusing
• Q9DBC2 - focusing
• Q10DBC2 - defocusing

0.46 m
1.9 m
1.9 m
1.9 m
OTR4DBC2 OTR6DBC2 OTR8DBC2
OTR10DBC2
16
TTF2 Main Components Two 3 FODO Cells
Q5SUND1H Q4SUND2 Q4SUND3
Q3SEED Q7SEED Q12SEED
Q16SEEDH
SEEDING section

• 4 OTR in 3 FODO cells
• OTR3SUND1
• OTR2SUND3
• OTR5SEED
• -OTR14SEED
• Cell phase advance 45 deg
• One cell length 8.9802 m
• 7 QMs in three FODO Cells
• Q5SUND1 - defocusing
• Q4SUND2 - focusing
• Q4SUND3 - defocusing
• Q3SEED - focusing
• Q7SEED - defocusing
• Q12SEED - focusing
• Q16SEED - defocusing

8.9802 m
8.9802 m
8.9802 m
2.19205 m
2.29805 m
OTR3SUND1 OTR2SUND3 OTR5SEED
OTR14SEED
17
TTF2 Main Components Pyroelectric Detector
Interferometer beamline at BC2 last dipole
Courtesy of O. Grimm
If time-dependent synchrotron radiation hits the
pyroelectric crystal, it generates time-dependent
temperature change in the crystal. That
temperature change in the crystal induces
time-dependent current or voltage (our measuring
quantity). Therefore it is possible to check
status of bunch length compression.
18
TTF2 Main Components LOLA Cavity
TM11 mode
BYPASS
19
TTF2 Main Components Collimators
We need collimators to reduce radiation dose
accumulation in undulator !

• Three Collimators at TTF2
• GUN BC2 collimators to chop dark current
• Collimators at Dog-Leg region to chop followings
• energy spread gt 3
• large betatron amplitude (beta function gt 10 m)

One of 4 collimators at Dog-Leg with 4 mm IRIS
diameter
titanium absorber included
water cooling
20
TTF2 Main Components 6 Undulators
BYPASS
BC2
BC3
RF-GUN
ACC1
ACC2
ACC3
ACC4
ACC5
LOLA
6 UNDULATORs
COLLIMATOR
DUMP
undulator period 27.3 mm undulator gap 12 mm,
fixed peak field 0.470 T K parameter
1.17 average beta function 5 m magnet material
NdFeB QMs on µMovers
21
TTF2 Other Components OTR - Wirescanner
OTR
TTF2 OTR system is designed and constructed in
collaboration between DESY and INFN-LNF and
INFN-Roma2 (resolution down to about 11 µm) Wire
scanner (WS) is modification of CERNs wire
scanner (resolution down to about 20
µm) Undulator wire scanner (resolution down to 5
µm)
WS
TTF2 OTR and Wire Scanner
Courtesy of K. Honkavaara
22
TTF2 Other Components Stripline BPM
FODO channel at BC2 downstream
TTF2 standard BPM
Stripline BPMs (resolution lt 30 µm, 34 mm
Pipe) This BPM is installed in QMs at FODO
channels.
Courtesy of D. Noelle
23
TTF2 Main Components Users' Beamlines
24
TTF2/FLASH Upgrades in 2007
Long Shutdown (March 26 - July 1) for Major
Upgrades
Module No 2
1 3
4 5

• change of module 3 (ACC3) by module 7
• repair of module 5 (ACC5)
• installation of the new module 6 (ACC6) with 28
  MV/m for 1 GeV operation
• change and redesign of the Gun diagnostic
  section
• Installation of Infrared undulator for THz
  source
• infrared beamline
• optical replica (seeding section)
• feedback system (XFEL)
• new MCP detectors
• new BPMs incl. BC3
• new OTRWS stations (seeding section)
• new phase monitors
• new magnet power supplies
• new synchrotron radiation beamline (BC3)
• modification dogleg vacuum chamber

BYPASS
BC2
BC3
RF-GUN
ACC1
ACC2
ACC3
ACC4
ACC5
LOLA
ACC6
COLLIMATOR
6 UNDULATORs 1 EM UNDULATOR
Module No 2
1 7
4 5 6

DUMP
440 MeV 28 MV/m
1 GeV
25
Optimization TTF2/FLASH RF Photoinjector

• Layout of TTF2/FLASH RF Photoinjector
• Space Charge Induced Emittance Growth and
  Compensation
• Invariant Envelope in Booster for the Emittance
  Damping in Booster
• Various Difficulties to get the Emittance
  Damping in Booster
• Our Experimental Results and Comparison with
  Simulation
• 1st Experimental Demonstration of Emittance
  Damping in Booster

26
Experiences Excellent Emittance at Injector
By optimizing TTF2 injector properly, we could
get an excellent emittance at the injector.
Without any bunch length compression, projected
normalized emittance is about 1.1 µm for 90 beam
intensity in 1.0 nC and 4.4 ps (rms) long bunch.
Let's see following pages how we could optimize
TTF2 injector !
27
Projected Emittance Growth and Compensation
initial zero emittance just after emission from
cathode
increased emittance due to space chare force
Solenoid Compensation
Before BOOSTER
rotated phase space by an external focusing
solenoid
compensation by reaction of space charge force
after some drift
28
Invariant Envelope and Emittance Damping
Booster Compensation
Invariant Envelope
BOOSTER LINAC
Before BOOSTER
After BOOSTER
Space charge force induces oscillations in
envelope and emittance. The emittace and envelope
oscillation can be damped by acceleration in
booster. (L. Serafini and J. Rosenzweig PRE Vol
55, Page 7565) Invariant envelope is an
ideal case which makes a constant slope (-?'/2)
for all different slices in the phase space by
the acceleration of booster. In this case, beam
spot size as well as transverse momentum are
reduced together due to reduced space charge
force in booster. Emittance damping in booster !!!
29
Booster Matching for Invariant Envelope
30
Experiences Matching Conditions at Booster
31
Experiences Matching Conditions at Booster
What are difficulties to get emittance damping at
booster linac ?
32
Difficulties Misalignment at GUN Region
Horizontal misalignment is about 9 mm, and
vertical misalignment is about 5 mm
This was fixed by K. Flöttmann and H. Weise in
December, 2004
33
Difficulties Misalignment at GUN Region
Emittance after BC2 is significantly changed for
small change in steerer current
Q 0.35 nC H1GUN -1.267 A BC2 on but no
compression
34
Difficulties Too Small Aperture at Dipole
Full height of the vacuum chamber for the dipole
12 mm Beam was slightly chopped there for a low
solenoid current (lt 277 A 0.163 T) Low charge
operation and optimization of steerers (H1GUN and
V2GUN)
35
Difficulties Too Small Aperture at Dipole
IDUMP Dipole chamber with 12 mm diameter cuts
beams for 277 A. But higher solenoid current
gives mismatching in booster, needed readjusting
of ACC1 gradients We improved this chopping by
optimizing steerers at GUN region.
36
Difficulties Small Vacuum Chamber in BC2
BC2 Chamber Height 8 mm Large vertical
emittance due to wakefiled in BC2 region This
was cured by adjusting vertical steeers (V10ACC1
and V1UBC2) at UBC2 region.
1.380 µm
2.116 µm
37
Difficulties - Steering Dispersion by ACC1
By changing dV/V of ACC1 and while monitoring
beam positon on the screens at DBC2, we measured
steering and dispersion effects of ACC1.
Actually, this is one of disadvantage of the
ponderomotive RF focusing in SW linac. This was
cured by adjusting steeers at GUN region.
8DBC2 screen
H1GUN 0.90H3GUN 1.20V2GUN -2.30V3GUN
0.95
best H1GUN steerer setting to reduce dispersion
by ACC1
38
Difficulties Jitter or Error
0.4 MV/m lower GUN gradient gives over focusing
at the entrance of ACC1 0.4 MV/m higher GUN
gradient gives under focusing at the entrance of
ACC1 Both cases give a large emittance growth
(well matched case 2.0 µm)
-1.0 GUN gradient error
1.0 GUN gradient error
emittance
emittance
beamsize
beamsize
39
Difficulties Jitter or Error
0.00163 T lower solenoid field gives under
focusing at the entrance of ACC1 0.00163 T higher
solenoid field gives over focusing at the
entrance of ACC1 Both cases give large emittance
growth (well matched case 2.0 µm)
-1.0 Solenoid Field error
1.0 Solenoid Field error
40
Experiences - Matching Conditions at Booster
How about emittance damping after solving those
difficulties ? And in this case, what is
measured emittance at DBC2 ?
41
Measured Emittance at DBC2 Section
Minimum emittance is obtained at around 277 A
0.163 T Other machine parameters are the nominal
ones
42
Measured Beam Images along Injector
April 7th, 2005, V. Ayvazyan, N.-I. Baboi, Y. Kim
43
Measured Beam Images along Injector
April 7th, 2005, V. Ayvazyan, N.-I. Baboi, Y. Kim
44
Measured Beam Images at 2GUN
April 12th, 2005, S. Schreiber, Y. Kim
Minimum beam size is generated around 342 A
45
Measured Beam Images at 2GUN
Sources of beam size difference between
simulation and measurement at 2 GUN screen -
difficulty to get beam image due to a high dark
current at 2 GUN screen - generally more wider
beam image of Cerium doped YAG (Yittrium Aluminum
Garnet) - nonlinearity and nonuniformity in gun
driving laser beam
0.5
0.5 nC
A. Murokh, PAC2001
46
Measured Beam Images at 3GUN
Please note that simulation gives about 5 µm
emittance at 3GUN for 277 A. Dark current was
chopped by IDUMP chamber, which is located at the
upstream of 3GUN YAG screen.
April 12th, 2005, S. Schreiber, Y. Kim
47
Measured Beam Images at 3GUN
April 7th, 2005, V. Ayvazyan, N.-I. Baboi, Y. Kim
main solenoid 277 A 3.07 mm for x plane 2.84
mm for y plane
right side vertical up
48
Measured Beam Images at 4DBC2 OTR
April 7th, 2005, V. Ayvazyan, N.-I. Baboi, Y. Kim
130 µm for both planes
49
Measured Beam Images at 6DBC2 OTR
April 7th, 2005, V. Ayvazyan, N.-I. Baboi, Y. Kim
120 µm for x plane 140 µm for y plane
50
Measured Beam Images at 8DBC2 OTR
April 7th, 2005, V. Ayvazyan, N.-I. Baboi, Y. Kim
150 µm for x plane 160 µm for y plane
51
Measured Beam Images at 10DBC2 OTR
April 7th, 2005, V. Ayvazyan, N.-I. Baboi, Y. Kim
140 µm for x plane 120 µm for y plane
52
Measured Beam Emittance at DBC2 Region
April 7th, 2005, V. Ayvazyan, N.-I. Baboi, Y. Kim
2.331 µm
2.395 µm Q 0.91 nC Intensity
100 Strong damped emittance in ACC1 (5 µm ? 2.3
µm)
53
ASTRA Simulation up to the ACC1 End
Note that we measured about 2.3 µm Measured
result is well agreed with simulation one !
April 7th, 2005, Y. Kim
ACC1 Module
normalized emittance 2.1 µm
rms beamsize at the end of ACC1 500 µm rms
beamsize at DBC2 140 µm
54
One Example of our Best Emittance
February 23rd, 2005 by K. Klose, F. Stulle, Y.
Kim
After fine scannings for all parameters !
55
One Example of our Best Emittance
February 23rd, 2005 by K. Klose, F. Stulle, Y.
Kim
1.892 µm
1.826 µm Q 1.04 nC Intensity
100 Strong damped emittance in ACC1 (5 µm ? 1.9
µm)
56
ELEGANT Simulation on our Best Emittance
Excellent agreement with measurements
horizontal and vertical beam sizes at DBC2
screens 140 µm

500 µm
140 µm
57
ELEGANT Simulation on our Best Emittance
Excellent agreement with measurements
horizontal and vertical beam sizes at DBC2
screens 140 µm

58
One Examples of our Best Emittance
February 23rd, 2005 by K. Klose, F. Stulle, Y.
Kim
1.160 µm
1.181 µm Q 1.04 nC Intensity
90 Well agreed with ASTRA simulation (1.0 µm for
90 intensity)
59
ASTRA Simulation on our Best Emittance
excellent agreement with measurements 90 core
emittance 1.1 µm, 100 emittance 1.9 µm
Excellent agreements between simulations and
measurements !!! 1st experimental demonstration
of emittance damping in booster !
60
Optimization of TTF2 Bunch Compresors
Let's see how we optimize bunch compressors to
generate femtosecond long bunch (lt 200 fs) ?
61
Working Principle of Bunch Compressor
Bunch Compressor Layout for SCSS Project - Y. Kim
et al, NIMA 528 (2004) 421
dE
dt
62
Incoherent Synchrotron Radiation
Incoherent Synchrotron Radiation (ISR) is
generated when relativistic long beam goes
through dipole magnet. Since ISR is a random
quantum process, it generates incoherent (
slice, or uncorrelated) energy spread, hence
slice emittance growth in the bending plane. For
one dipole magnet, the relative uncorrelated
energy spread due to ISR is given by Slice
emittance growth due to ISR is given by
- lower energy - longer dipole - smaller bending
angle are good against ISR effects
Electron path
Dipole region
63
Coherent Synchrotron Radiation
In BC where dispersion is nonzero, bunch length
becomes smaller. Short electron bunches in dipole
can radiate coherently (CSR) at wavelength CSR
from tail electrons can overtakes head electrons
after the overtaking length. CSR
generates correlated energy spread along whole
bunch Electrons are transversely kicked at
the nonzero dispersion region or in BC Hence,
projected emittance is increased in BC due to CSR.
LOT
Tail
Head
CSR from tail
Electron path
Dipole region
64
Coherent Synchrotron Radiation
Courtesy of M. Dohlus
Without CSR self-interaction
Head with lower energy
DM3
DM4
DM2
DM1
With CSR self-interaction
Head energy gain by CSR
x
Tail energy loss due to CSR
z
65
Against Coherent Synchrotron Radiation

• Strong focusing lattice around BCs to reduce CSR
  induced emittance growth

Strong focusing against CSR ?-functions 0
?-functions 3
Before BC1
FODO cells
After BC1
66
Coherent Synchrotron Radiation Wake
67
Nonlinearity in Longitudinal Phase Space

• Linearization of longitudinal phase space with a
  higher harmonic cavity

- means deceleration !
Linear range 60 degree 60 deg in 2856 MHz 60
ps (18 mm) 60 deg in 1300 MHz 126 ps (37.8
mm) 60 deg in 650 MHz 252 ps (75.6 mm) Good
for 9.0 mm (rms) ILC bunch
-70 MeV deceleration
Only C-band Linac
C-band Linac X-band Correction Cavity
SCSS BC
Nonlinearities in the longitudinal phase space
due to RF curvature, short-range wakefields,
T566, and space charge force can be compensated
by harmonic cavity.
Y. Kim et al, NIMA 528 (2004) 421
68
Operation Point of TTF2 Bunch Compressors
Blue After BC for different compression Green
Before BC
Without higher harmonic compensation cavity,
nonlinearity in dz-dE chirping becomes stronger
after BC, and there is some local charge
concentration in very small local region. This
nonlinearity by RF curvature helps in making
femtosecond spike.
69
ELEGANT Simulation - Generation of fs Spike
Nonlinearity in longitudinal phase space due to
RF curvature makes the spike in BC Advantage of
Nonlinearity !!!
3BC2 OTR D1BC2 61.6 A ? 18.0355 deg
Machine Status on April 9th, 2005
Starting charge concentration or spike
Q -1.0 nC ACC1 phase -9.0 deg off crest to
get needed dz-dE chirping for compression. No
space charge consideration from BC2
70
ELEGANT Simulation - Generation of fs Spike
Nonlinearity in longitudinal phase space due to
RF curvature makes the spike in BC Advantage of
Nonlinearity !!!
3BC2 OTR D1BC2 61.6 A ? 18.0355 deg
Machine Status on April 9th, 2005
Reduced energy spread by ACC23
Compressed bunch length and stronger spike by BC3
Q -1.0 nC ACC1 phase -9.0 deg off
crest ACC2345 on crest because there is nonzero
chirping after BC2 No space charge consideration
from BC2
71
Spike Status on April 9th, 2005
ELEGANT simulation
9DUMP OTR
188 fs (FWHM)
ELEGANT simulation
2 µm
peak gain 100
SASE source on CeYaG crystal behind a gold wire
mesh (0.25 mm) which is located at 18.5 m
downstrem of undulators.
72
Experiences Projected Emittance after BC2
Note that during these measurements, injector was
not optimized perfectly. Hence emittance is
somewhat higher than our nominal case.
February 6th by F. Loehl and Y. Kim
Big horizontal emittance growth from around -10
deg off crest due to CSR
No strong space charge effects
We chose around -9.0 deg for a good operational
phase
73
ELEGANT Simulation - Best ACC1 Phase
Maximum bunch length compression at BC2 is
happened for ACC1 phase -13.0 deg If ACC1 phase
is lower than -11 deg, peak current is not
increased further at BC3.
After BC2
After BC3
max compression at -13 deg
around from -10 deg to -9.0 deg give a good slice
emittance and a high peak current. In this case,
compression done at core region
The best ACC1 phase is from-10 deg to -9 deg
74
Experiences Bunch Length from LOLA
On May 11th, measured bunch length with LOLA is
about 155 fs (FWHM) for ACC1 phase -10 deg,
which is close to estimated by ELEGANT
simulation, 160 fs On April 14th, measured bunch
length with LOLA is 11 16.5 fs (FWHM) for the
maximum compression phase, around -13.0 deg
16.5 fs Spike with LOLA
155 fs Spike with LOLA
155 fs is close to our simulation one !!!
75
What should we set up for BC Operation ?

• Energy at BCs against Space Charge and Wakefield
  Effect
• Energy spread or off crest phase in
  precompressor linacs
• (ACC1 phase for BC2 and ACC23 phase for BC3)
• Chicane Strength R56 constant by using the
  fixed bending angle
• (by tuning phase of precompressor linacs, we
  adjust compression factor)

76
Experiences - Fixed Bending Angle in BCs
3BC2 OTR D1BC2 61.6 A ? 18.0355 deg
E1 gt 127 MeV gt E2
By the geometry of vacuum chamber, beams go
through center of the vacuum chamber when bending
angle is 18.0355 deg. We fixed bending angle
for easy energy setting at BC2 even though the
vacuum chamber geometry can accept 15 deg to 21
deg bending angle. If beam energy is changed,
then we adjust DIBC2 dipole current to send beam
in the middle of the vacuum chamber (D1BC2 61.6
A for 127 MeV, D1BC260.5 for 125 MeV).
77
Experiences - Max Compression Phase at ACC1
There is a constant phase difference between on
crest phase of module and the maximum compression
phase. This is very useful to find on crest phase
and BC operational phase.
Max. Pyro signal at ACC1 phase 118.34 deg
If time-dependent synchrotron radiat ion hits
the pyroelectric crystal, it generates
time-dependent temperature change in the crystal.
That temperature change in the crystal induces
time-dependent current or voltage (our measuring
quantity). Therefore it is possible to measure
bunch length and CSR strength
Phase difference between two phase monitors
around BC
Maximum of pyro signal was at 118.34 deg. phase
monitor signal was its minimum on 130.90 deg. On
crest of module is located about 13 deg higher
from the maximum compression phase.
78
Experiences - Search of On Crest Phase
On screen in BCs, by monitoring branch in the
tail region, we can determine on crest phase
exactly. Source of this branch is unknown yet !
ASTRA Simulation Result on longitudinal phase
space after ACC1 Module, Phase -3 deg off crest
It seems there is vertical dispersion (Short-range
wakefield, or roll error in diopole, etc.)
3BC2 OTR for on crest phase
3BC2 OTR
one longer branch at top ACC SP Phase
about -3.0 deg off crest
Tail in bunch (dzlt0) Head in bunch (dzgt0)
Tail in bunch (dzlt0) Head in bunch (dzgt0)
Branch from tail
Energy spread 215.9 keV
79
Experiences - Search of On Crest Phase
On screen in BCs, by monitoring branches in the
tail region, we can determine on crest phase
exactly. Source of these branches is unknown yet !
minimum energy spread or minimum beamsize at
-2.0 deg
ASTRA Simulation Result on longitudinal phase
space after ACC1 Module, Phase -2 deg off crest
3BC2 OTR
two symmetric branches ACC SP Phase about
-2.0 deg off crest
Tail in bunch (dzlt0) Head in bunch (dzgt0)
Tail in bunch (dzlt0) Head in bunch (dzgt0)
Branch from tail
Symmetry in head and tail Energy spread 170.5
keV (Minimum phase)
80
Experiences - Search of On Crest Phase
On screen in BCs, by monitoring two branches in
the tail region, we can determine on crest phase
exactly. Source of these branches is unknown yet !
energy spread or beamsize is not its minimum at
on crest
ASTRA Simulation Result on longitudinal phase
space after ACC1 Module, Phase on crest
3BC2 OTR ACC SP
Phase on crest phase (13 deg from max
compression phase)
Tail in bunch (dzlt0) Head in bunch (dzgt0)
one longer branch at bottom
Asymmetry in head and tail Energy spread 253.6
keV (Not minimum phase)
81
Experiences - Max Compression Phase at ACC23
After setting on crest phase in ACC1, we
performed pyro scanning of ACC23 phase to find
the maximum compression phase at BC3. The maximum
compression phase at BC3 is about 4243 deg from
the on crest phase.
Max compression phase of ACC23 -180.29 deg. On
crest phase of ACC23 -137.29 deg
82
Experiences BC Fine Tuning
From SASE Gain and Beam Image at 9DUMP, we
optimized further !
Good compression for SASE
Weak compression close to on crest No SASE
83
Experiences 1st Lasing at 32 nm
First beam through the collimator
03.12.2004 Beam transport through the undulator
13.12.2004 First lasing around 32 nm,
14.01.2005 FEL pulse energy up 42 µJ (16
µJ ave) 30.06.2005
Good compression for SASE
one of the first single shot diffraction images
produced behind a double slit - 0.15 mm slit
distance - 30 µm slit width
Weak compression close to on crest No SASE
FEL told that our machine was optimized properly
as designed.
about 20 fs (FWHM) pulse length
84
Summary
After curing various difficulties (misalignment,
small aperture, dispersion, steering effects, and
wakefields), we could get an excellent emittance
( 1.1 µm for 1.0 nC and 90 intensity) even
though we used Gaussian laser beam profile. Our
measured emittance can be reproducible ! We
have firstly demonstrated the Ferrario's matching
principle at the booster linac with the Gaussian
laser beam profile. By the help of nonlinearity
in the longitudinal phase space due to RF
curvature (mainly), higher order momentum
compaction factor (T566), and space charge, etc,
we could generate femtosecond spike at BCs. To
set up needed dz-dE chirping for compression, we
should choose proper off crest phase. Setup of
proper operational phase in precompressor linacs
(ACC1, ACC23) was done from information on
emittance measurements, pyro phase scanning, and
beam branch in the tail region.
85
Summary Acknowledgments
We can continuously generate lasing at 32 - 13 nm
with about 10 20 fs photon pulse length since
January 2005. This means our injector is well
optimized and bunch length is well compressed
within 200 fs range to give high peak
current. Y. Kim sincerely thank all TTF2 shift
operators and run coordinators. Specially to K.
Floettmann, S. Schreiber, and Prof. J. Rossbach
for useful discussions and encouragements, and to
Dr. M. V. Yurkov, B. Faatz, R. Treusch, and S.
Schreiber for their kind supply of their
presentation materials. Y. Kim also sincerely
thank FERMILAB's invitation to open this seminar.
86
TTF2 Commissioning Experiences

• Stability of SASE Source and Safety Issues
• Charge and RF stability
• Low feedback to keep constant bunch length and
  charge
• Orbit feedback system to keep constant orbit in
  collimator and undulator
• Scanning tool to get more stable SASE source
• PETRA ramping
• Gun dark current

87
Charge Stability
88
GUN RF Phase and Laser Synchronization
89
Energy Stability and Drift at BC2
GUN and ACC1 with RF feedback variation of energy
0.054
variation of energy spread 1
90
Timing Jitter due to Energy Drift after BC2
91
Feedback for Stable Charge Bunch Length
ACC1 phase to control bunch length after
BC2 Laser flashlamp to control charge ACC1
voltage to control energy and arrival timing
92
ACC7 Orbit Feedback for Stable Orbit
BYPASS
BC2
BC3
RF-GUN
ACC1
ACC2
ACC3
ACC4
ACC5
LOLA
6 UNDULATORs
COLLIMATOR
DUMP
to get stable orbit in collimator and undulator
!!!
93
Scanning Tool to get Golden Orbit in Undulator
BYPASS
BC2
BC3
RF-GUN
ACC1
ACC2
ACC3
ACC4
ACC5
LOLA
6 UNDULATORs
COLLIMATOR
DUMP
steerer fine scanning to find golden orbit in
undulator !
94
Effect of PETRA Ramping
BYPASS
BC2
BC3
RF-GUN
ACC1
ACC2
ACC3
ACC4
ACC5
LOLA
6 UNDULATORs
COLLIMATOR
DUMP
steerer fine scanning to find golden orbit in
undulator !
95
GUN Dark Current in Undulator
BYPASS
BC2
BC3
RF-GUN
ACC1
ACC2
ACC3
ACC4
ACC5
LOLA
6 UNDULATORs
COLLIMATOR
DUMP
Loss due to GUN dark current and Loss due to beam