Electron bunch shape measurements at the TTF-FEL
ABSTRACT The TESLA Test Facility linac has been operated in the first half of 2002 with two bunch compressors to drive the TTF-FEL free electron laser. During this running period, SASE radiation with a wavelength around 100 nm has been routinely delivered to experiments. The longitudinal shape of the electron bunches is a crucial property of the electron beam: a peak current in the order of 1 kA is required to drive the SASE process with the given undulator design, transverse emittance, and energy spread. We report on measurements of the bunch length and shape for different operating conditions of the two bunch compressors.
- SourceAvailable from: Karsten Holldack
ELECTRON BUNCH SHAPE MEASUREMENTS AT THE TTF-FEL
K. Honkavaara, S. Schreiber∗, Ch. Gerth, Ph. Piot, DESY, 22603 Hamburg, Germany
The TESLA Test Facility linac has been operated in the
first half of 2002 with two bunch compressors to drive the
TTF-FEL free electron laser. During this running period,
SASE radiationwith a wavelengtharound100nm has been
routinely delivered to experiments. The longitudinal shape
of the electron bunches is a crucial property of the elec-
tron beam: a peak current in the order of 1kA is required
to drive the SASE process with the given undulator de-
sign, transverse emittance, and energy spread. We report
on measurements of the bunch length and shape for differ-
ent operating conditions of the two bunch compressors.
To drive the TTF-FEL free electron laser at DESY, ex-
cellent beam properties in the transverse and longitudinal
planes are essential: a small transverse emittance in the
order of a µm, a high peak current in the kA range, and
a small uncorrelated energy spread below 0.1%. Opti-
in Ref. . The saturation has been achieved in the VUV
wavelength region (80 to 100nm), and SASE radiation
has been routinely deliveredto the experimentsduring sev-
eral running periods in 2001 and 2002.
A sketch of the TTF linac is shown in Fig. 1. The elec-
tron source is a laser-driven RF gun with a Cs2Te cathode.
The RF gun section is followed by a booster, a standard
TESLA 9-cell superconducting accelerating cavity oper-
ated at 11.5MV/m. After the booster the beam energy
is 16.5MeV. The beam is accelerated by two 12m long
conducting accelerating structures each. After a collima-
tion section, the beam is injected into the undulator mod-
ules with an energy of up to 300MeV. Two magnetic chi-
cane bunch compressors are installed: BC1 is downstream
of the booster cavity, BC2 between the accelerating mod-
At the end of the last FEL run beginning of 2002, both
bunch compressors have been used to shape the longitudi-
nal charge distribution of the electron bunch. In the follow-
ing, measurements of the longitudinal bunch shape using a
streak camera for different settings of the two compressors
To measure the bunch distribution, we use synchrotron
radiation emitted by the horizontally deflecting spectrom-
eter dipole after the undulator (see Fig. 1). The optical
part of the synchrotron radiation is guided by four flat alu-
minum mirrors to a streak camera situated outside of the
accelerator tunnel. An achromat lens is used to focus the
light ontothe entranceslit of the camera. In orderto reduce
chromatic effects, a narrow-band wavelength filter is being
used.  The data presented here are obtainedusing a filter
of 500 ± 40nm.
The streakcamera (HamamatsuFESCA-200C6138)has
an intrinsic resolution of 208fs (FWHM) measured with a
fs probe laser. Only some data have been taken with the
fastest streakspeedof20ps/10.29mm. Most ofthethedata
presented here are obtained with the second fastest streak
speed of 50ps/10.29mm. For this speed, the resolution
is 200fs (sigma) only. In the case of long uncompressed
bunches, the streak speed of of 100ps/10.29mm was more
suitable. The entrance slit is adjusted to be as small as re-
quired to obtain the best resolution. In most of the mea-
surements the slit is between 20µm and 40µm, the latter
has a better photon yield. For these slit sizes the intrinsic
camera resolution is not significantly changed. The gain of
the streak camera multi-channel plate has been adjusted in
A bunch signal with a jitter of about 50ps served as a
trigger for the streak.
For a given machine setting, several streak images have
In order to reduce the noise due to photonstatistics, several
single profiles (typically from five to ten) are overlayed.
Due to the trigger jitter, the recorded images are not in the
same time position. The shifting of the profiles for overlay-
ing along the time axis is performed manually taking the
leading edge of the profile as a reference. The average of
the overlayedprofiles is calculated. Since the bunchcharge
was stable during the measurement, a charge scaling of the
profiles was not necessary. An example of nine overlayed
profiles with the average profile superimposed is shown in
All profiles shownin the followingare averagedprofiles,
and scaled according to their peak intensity. The profiles
have not been corrected for the resolution of the camera.
EXPERIMENTS AND RESULTS
The longitudinal electron bunch profile has been mea-
sured with different settings of the bunch compressors. Ta-
ble 1 summarizes some design parameters. In all the mea-
surements presented here, the beam was passing through
BC2 with a deflecting angle of 18◦. Nominally, the first
0-7803-7739-9 ©2003 IEEE 911
Proceedings of the 2003 Particle Accelerator Conference
electron rf gun
Cryomodules with superconducting cavities
Module # 1
Module # 2Collimator
Figure 1: Schematic overview of the TTF-FEL linac phase 1 (not to scale). Beam direction is from right to left, the total
length is 100m.
Figure 2: Several measurements of the same longitudinal
beam profile. The average of the profiles is superimposed.
accelerating module (module 1) RF is dephased by 12◦
to obtain maximum compression. BC1 is either bypassed
or used with two different deflecting angles: 24◦and 30◦.
When BC1 is bypassed, the booster cavity RF is operated
with a phase 10◦off-crest corresponding to a minimum in
energy spread. The RF phase of the RF-gun has always
been nominal 40◦from zero crossing. The second module
is operated with on-crest acceleration.
Table 1: Some design parameters of the bunch compres-
defl. angle (◦)
max. disp. (mm)
compression ratio 2
For all settings, the general shape of the bunch is dom-
inated by a sharp leading peak and a long tail. The small-
est measured width of the peak is 600±100fs (sigma) and
Figure 3 shows the effect of the booster phase and mod-
ule 1 phase on the bunch shape. The beam passes through
both compressors, BC1 is operated with a deflecting angle
of 30◦, BC2 with 18◦. The bunch charge is about 1nC.
In the upper plot, the module 1 phase is on-crest. The
profiles are for the case of the booster at minimum energy
spread (+10◦, red), and at compression (+22◦, blue). The
compression effect of BC1 is only visible in the tails and in
Figure 3: Average bunch profiles with both BC1 and BC2
in use (1nC). Upper: Module 1 on-crest. Booster cav-
ity phase corresponding to minimum energy spread, 10◦
off-crest (red), booster cavity phase at 22◦off-crest (blue).
Lower: Booster phase 22◦, module 1 phase on-crest (0◦)
(red), module 1 phase in full compression (12◦) (blue).
the calculated rms over the whole distribution.
The lower plot shows the effect of the module 1 phase:
the module is on-crest as before (red), and with a phase of
+12◦corresponding to maximum compression (blue).
The compressing effect of BC2 is very clear (Fig. 3,
lower plot). A similar behavior has been observed bypass-
ing BC1 , as well as in our earlier measurements with a
lower resolution camera, where the peak could not have
The effect of the booster cavity phase is small (Fig. 3,
upper plot). The aim of using BC1 is not to fully com-
press the bunch, but rather to shape the leading peak. We
expect the precompression with BC1 to shorten the bunch
before acceleration with module 1, and thus to reduce the
curvaturein thelongitudinalphasespaceduetotheRF. The
reduced curvature leads – after compression with BC2 – to
a larger and more gaussian peak and a shorter tail.
Figure 4 compares three bunch shapes, while the ma-
chine was lasing with SASE close to saturation. In one
Proceedings of the 2003 Particle Accelerator Conference
case, BC1 has been bypassed (red), in the other cases, BC1
has been set to compression (24◦deflection), and tuned to
keep the machine lasing. The booster phase is 25◦(green)
and 26.5◦(blue). For all cases, the bunch charge was be-
tween 2.5 and 3 nC.
From the comparison it is clearly visible, that the pre-
bunching with BC1 effectively widens the sharp peak and
shortens the tail. The effect of this tailoring of the lasing
bunch slice on the measured internal mode structure of the
FEL radiation has been reported in .
Figure 4: Average bunch profiles for settings, where the
FEL was lasing. (a) BC1 is bypassed (red), (b) BC1 set to
compression (24◦deflection), booster phase at 25◦(green)
(c) and 26.5◦(blue). Charge for all cases between 2.5 and
DISCUSSION AND CONCLUSION
We have measured longitudinal bunch shapes with dif-
ferent settings of the bunch compressors. The data show
clearly a shape with a dominant leading peak of a cuurent
of 1kA and a long tail. The first bunch compressor BC1
is used as a pre-buncher to tailor the leading peak of the
bunch profile. This peak shows up after compression with
the second chicane BC2. It is actually the part or slice of
the bunch which contributes to the SASE lasing process.
In this sense, TTF1 was able to influence the mode
structure of the SASE FEL radiation by carefully adjust-
ing phases and compression of BC1.
We like to thank E. Janata from HMI, Berlin for the syn-
chronization unit of the streak camera trigger and L. Plu-
cinski, DESY for his help in setting up the synchrotron ra-
diation beam line.
 J. Andruszkow et al., “First Observation of Self-Amplified
Spontaneous Emission in Free-Electron Laser at 109 nm
Wavelength”, Phys. Rev. Lett. 85 (2000) 3825-3829.
 S. Schreiber et al., “Performance of the TTF Photoinjector
for FEL Operation”, Proc. of the workshop ”The physics
and applications of high brightness electron beams”, Chia
Laguna, Sardinia, July 1-6, 2002
 V. Ayvazyan et al, “Generation of GW radiation pulses from
a VUV Free-Electron Laser operating in the femtosecond
regime”, Phys. Rev. Lett., 88 (2002) 104802.
 Ch. Gerth et al, “Bunch Length and Phase Stability Mea-
surements at the TESLA Test Facility”, FEL2002, Argonne,
Ill., USA, Sept 9-13, 2002, to be published
 Hamamatsu C6138 FESCA-200, Test report, September
 The synchronization of the trigger has been provided by E.
Janata, HMI, Berlin
 K. Honkavaara, Ph. Piot, S. Schreiber, D. Sertore, “Bunch
Length Measurements at the TESLA Test Facility using a
Streak Camera”, Proc. of the 2001 Partical Accelerator Con-
ference, Chicago, Ill., USA, June 18-22, 2001, p. 2341.
 V. Ayvazyan et al, “Study of the statistical properties of the
radiation from a VUV SASE FEL operating in the fem-
tosecond regime”, FEL2002, Argonne, Ill., USA, Sept 9-13,
2002, to be published.
Proceedings of the 2003 Particle Accelerator Conference