Commissioning experience and beam physics measurements at the SwissFEL Injector Test Facility

Editors' Suggestion
Open Access

Commissioning experience and beam physics measurements at the SwissFEL Injector Test Facility

T. Schietinger et al.

Phys. Rev. Accel. Beams 19, 100702 – Published 26 October 2016

Abstract

The SwissFEL Injector Test Facility operated at the Paul Scherrer Institute between 2010 and 2014, serving as a pilot plant and test bed for the development and realization of SwissFEL, the x-ray Free-Electron Laser facility under construction at the same institute. The test facility consisted of a laser-driven rf electron gun followed by an S-band booster linac, a magnetic bunch compression chicane and a diagnostic section including a transverse deflecting rf cavity. It delivered electron bunches of up to 200 pC charge and up to 250 MeV beam energy at a repetition rate of 10 Hz. The measurements performed at the test facility not only demonstrated the beam parameters required to drive the first stage of an FEL facility, but also led to significant advances in instrumentation technologies, beam characterization methods and the generation, transport and compression of ultralow-emittance beams. We give a comprehensive overview of the commissioning experience of the principal subsystems and the beam physics measurements performed during the operation of the test facility, including the results of the test of an in-vacuum undulator prototype generating radiation in the vacuum ultraviolet and optical range.

63 More

Received 9 June 2016

DOI:https://doi.org/10.1103/PhysRevAccelBeams.19.100702

Published by the American Physical Society under the terms of the Creative Commons Attribution 3.0 License. Further distribution of this work must maintain attribution to the author(s) and the published article’s title, journal citation, and DOI.

Published by the American Physical Society

Physics Subject Headings (PhySH)

Beam dynamics Beam techniques

Accelerators & Beams

Authors & Affiliations

Click to Expand

Article Text

Click to Expand

References

Click to Expand

Issue

Vol. 19, Iss. 10 — October 2016

Reuse & Permissions

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
Schematic layout of the SwissFEL Injector Test Facility (original configuration, not to scale). The total length of the beam line is approximately 60 m.
Reuse & Permissions
Figure 2
Horizontal (blue) and vertical (red) $β$ -functions for a typical beam optics setup, covering the beam line from the S-band booster exit down to the profile monitor at the high-energy spectrometer (upper panel). The corresponding magnetic lattice is shown in the lower panel, where the larger rectangular shapes indicate the locations of the main dipole (bending) magnets and the smaller those of quadrupole (focusing) magnets. The wave-like curve at 2.5 m represents the position of the X-band cavity. The green circle at 20 m marks the location of the rf deflector.
Reuse & Permissions
Figure 3
Longitudinal pulse profile of the $Nd ∶ YLF$ laser, measured with cross-correlation in the UV.
Reuse & Permissions
Figure 4
Longitudinal pulse profile of the $Ti ∶ sapphire$ laser, measured with cross-correlation in the UV. The profile is obtained by pulse stacking with four crystals, resulting in 16 pulse replicas. The cross-correlations for s- and p-polarized pulses are done in sequence (black and red curves) and summed up numerically (blue curve).
Reuse & Permissions
Figure 5
Beam transport and transverse beam profiles along the beam line for the $Ti ∶ sapphire$ and $Nd ∶ YLF$ gun laser systems.
Reuse & Permissions
Figure 6
SwissFEL gun cathode plug (copper, no coating), with dimensions indicated (left). Two cathode plugs mounted in the carousel holder of the load-lock chamber (right, see also Fig. 7). The rf contacts can be recognized around the circular cathode surfaces.
Reuse & Permissions
Figure 7
3D drawing of the vacuum suitcase connected to the load-lock chamber showing the cathode transfer principle and the two storage carousels.
Reuse & Permissions
Figure 8
Quantum efficiency evolution of the cathodes used at the SITF between October 2009 and September 2014.
Reuse & Permissions
Figure 9
Schematic layout of the SITF rf system, showing the accelerating structures as cylinders with the corresponding power sources (klystrons).
Reuse & Permissions
Figure 10
Geometry of the CTF gun No. V, with magnitude of the electric field in the horizontal and vertical planes.
Reuse & Permissions
Figure 11
Amplitude modulation scheme for the SwissFEL gun: input rf power (solid blue line), reflected power (dash-dotted blue line), and resulting peak gradient (red line) as a function of time for the nominal modulated rf pulse of $1 μ s$ total length. For comparison, the peak gradient for an unmodulated $3 μ s$ pulse is also shown (magenta line) [28].
Reuse & Permissions
Figure 12
Cutaway 3D views of the SwissFEL gun. Details of the backplane and the cathode plug are visible in the upper right figure.
Reuse & Permissions
Figure 13
Resonant frequency of the $π$ -mode as a function of the force applied to the cathode plug (top). On-axis electric field profiles for different frequencies, obtained by varying the position of the cathode plug (bottom).
Reuse & Permissions
Figure 14
Measured reflection coefficient ( $| s_{11} |$ ) as a function of frequency revealing the frequency spectrum of the SwissFEL gun. The operating frequency in vacuum is 2 997.912 MHz at 53°C. The frequency separation to the adjacent mode is 16.2 MHz.
Reuse & Permissions
Figure 15
Progress in rf conditioning the SwissFEL gun toward high power. The pressure gauge is mounted on the rf waveguide at the input of the rf gun.
Reuse & Permissions
Figure 16
Example of a non-Gaussian horizontal beam profile in the spectrometer at 7.1 MeV (black points), with a modified super-Gaussian function fitted to the data (blue line). The inset shows the original screen image.
Reuse & Permissions
Figure 17
Bunch mean energy jitter (top) and bunch rms energy spread (bottom) at the gun exit determined from 49 nonconsecutive shots over 30 s with a laser pulse repetition rate of 10 Hz for rf running at 10 Hz (blue squares) and 100 Hz (red asterisks).
Reuse & Permissions
Figure 18
Example images of dark current at nominal conditions on a $YAG ∶ Ce$ screen near the gun exit for the CTF (left) and SwissFEL (right) guns.
Reuse & Permissions
Figure 19
Integrated dark current per rf pulse as a function of the peak electric field at the cathode at 10 Hz and 100 Hz repetition rate (SwissFEL gun).
Reuse & Permissions
Figure 20
Measured beam energy gains per accelerating structure as a function of the rf input power for the FINSB02 (red squares, one structure) and $FINSB 03 / 04$ (green circles, two structures) rf stations. The solid lines of the corresponding colors represent square-root fits to the data, the dashed lines show the respective design expectations.
Reuse & Permissions
Figure 21
Field amplitude (top) and phase advance (bottom) measurements of the X-band structure using the bead-pull method, after tuning. Measurements and tuning were carried out at CERN [60].
Reuse & Permissions
Figure 22
Cutaway drawing of the X-band structure including waveguide system and motorized support.
Reuse & Permissions
Figure 23
Measured beam energy gains (blue squares) for the X-band structure as a function of operating temperature (top) and rf input power (bottom). The dashed lines correspond to the design expectation, the solid line is the square-root fit used to extract the structure’s gradient-to-power ratio.
Reuse & Permissions
Figure 24
Amplitude transfer function of the non-IQ digital filter of the llrf system.
Reuse & Permissions
Figure 25
Phase slippage between output and input of the S-band traveling wave structures due to temperature variations of the structure. The ideal operating temperature, where output and input phases are equal, is found to be $41.7 ° C$ in this example.
Reuse & Permissions
Figure 26
Gun rf amplitude and phase with “rectangular” (left plots) and with “Hamming” (right plots) rf pulse shape excitation. The rectangular shaped forward pulse also excites the nearest undesired mode 8.1 MHz away from the $π$ -mode.
Reuse & Permissions
Figure 27
3D drawing of the bunch compression chicane showing the initial configuration featuring dipole magnets with return-yoke windings. The thick black arrow indicates the lateral movement of the central support for the two inner dipole magnets, which also houses a screen monitor and insertable collimation masks.
Reuse & Permissions
Figure 28
The two dipole magnet designs tested in the SITF bunch compressor. Top: original design with pole windings. Bottom: second design with larger pole width, later discarded because of the long-range stray fields resulting from the return-yoke windings.
Reuse & Permissions
Figure 29
Resonant stripline BPM pickup schematics (left) and mechanical drawing (right) [80].
Reuse & Permissions
Figure 30
Beam-induced resonant stripline BPM pickup signal voltage (vertical axis, $1 V / div$ ) versus time (horizontal axis, $2.5 ns / div$ ), sampled with a 20 GSample/s oscilloscope (screenshot). Only the 500 MHz fundamental mode is visible, higher harmonics are suppressed with a low-pass filter.
Reuse & Permissions
Figure 31
BPM system block schematics, including the linkage to the external trigger system and the EPICS control system (see Sec. 4g) [77].
Reuse & Permissions
Figure 32
Raw BPM electronics waveforms (ADC raw value versus sample number), sampled by the DRS4 chip at 5 GSample/s (a). The same waveforms after offset gain and time-nonlinearity correction, performed by the FPGA for each sample individually (b). Envelopes of the four BPM electronics signals as calculated by the FPGA (signal amplitude versus sample number) (c). The additional delay is due to digital filtering.
Reuse & Permissions
Figure 33
Technical drawing and photograph of the ultra-high-vacuum feedthrough insertion accommodating alignment targets, three scanning wires, two screens (crystal scintillator and OTR), and an rf shield.
Reuse & Permissions
Figure 34
Deflector calibration factor (top) and time resolution (bottom) as a function of the integrated deflecting voltage, measured in each case at the zero crossing of the rf phase, for a beam energy of 200 MeV and compared to model expectation [90].
Reuse & Permissions
Figure 35
Long-term drift measurement of a 500 m long CW optical link: raw data at 1 Hz (blue curve), and filtered with a time window of five minutes (red curve). See text for details.
Reuse & Permissions
Figure 36
Signal source analyzer screenshot illustrating the improvement of the jitter of the $Nd ∶ YLF$ laser: Using the in-house built synchronization unit with gain optimized within the controller bandwidth, the laser exhibits a jitter of 45 fs in the spectral range between 10 Hz and 10 MHz (bright yellow curve with arrows). For comparison, the original synchronization box results in a jitter of about 600 fs in the same spectral range (pale yellow curve).
Reuse & Permissions
Figure 37
Phase noise stability comparison between free running, rf-locked, and optically locked EDFL [109]. The values given in the upper part of the figure refer to rms jitters measured in the respective frequency ranges. The values in the lower part of the figure are jitters measured over the entire frequency range (10 Hz to 10 MHz).
Reuse & Permissions
Figure 38
Long-term in-loop optical lock stability of the EDFL OMO [109]. The integrated timing jitter over 60 minutes is 4.6 fs (rms) in the range between 1 Hz and 100 kHz. The clustering of the stability values at distinct values is a consequence of the limited resolution of the sampling clock.
Reuse & Permissions
Figure 39
Test of the stability of the optical phase detector by addition of a manually varied voltage offset to the error signal [109]. The initial curve shows stable operation following the induced changes, until the loop breaks after 2.5 minutes (indicated by the vertical line). The optical lock is recovered once the perturbation is removed and the optical gains are readjusted (curve to the right of the vertical line). The blue and green curves represent sliding averages over 20 consecutive measurements. The integrated timing jitter is 11 fs (rms) in the range between 1 Hz and 100 kHz. (As in Fig. 38, the resolution of the sampling clock leads to a clustering of the stability values at distinct levels.)
Reuse & Permissions
Figure 40
Top: Snap-shot of the first day of RADFETs monitoring radiation exposure of the undulator. The four curves correspond to the four RADFETs installed at the undulator entrance and exit. Significant doses (of the order of a Gray) can be accumulated in a few minutes, demonstrating the need for a fast interlock. When the machine settings are properly adjusted the remaining losses are moderate. Bottom: RADFETs monitor data for the entire period of the undulator experiment.
Reuse & Permissions
Figure 41
Noise (rms) measurements of the standard ICT (green stars), Turbo-ICT-2 (first pulse shown, magenta diamonds) and a stripline BPM located close to the ICTs (blue circles), as a function of bunch charge (as measured with a BPM).
Reuse & Permissions
Figure 42
Simplified schematic of low-Q cavity BPM rf front-end electronics, showing only one of its three input channels (one reference and two position signal channels) [79].
Reuse & Permissions
Figure 43
Cutoff view of the downstream end of the X-band structure showing the location and geometry of the wakefield monitor coupling cell relative to adjacent regular cells, high-power output coupler and cooling circuit [58].
Reuse & Permissions
Figure 44
Signal output measured by the upstream wakefield monitor at different frequencies for a varying structure offset. The structure is tilted on purpose by an angle of 0.5 mrad.
Reuse & Permissions
Figure 45
Front-view of the wire scanner vacuum chamber and the wire-fork. The wire-fork is equipped with 3 different pin slots where $5 μ m$ tungsten wires can be stretched (indicated as red and blue lines). The distance between the “wire vertex” and the vacuum-chamber center can be set to 8, 5.5 and 3 mm corresponding to the 3 different pin slots.
Reuse & Permissions
Figure 46
Top: Horizontal beam size versus off-crest rf phase as measured with the high-energy spectrometer, the YAG screen in the bunch compressor (BC-YAG), and the synchrotron radiation monitor in the bunch compressor (BC-SR) for a beam energy of 230 MeV and a bunch charge of 12 pC. Bottom: Relative energy spreads derived from the measurements, compared to the result of a numerical simulation based on Astra [127].
Reuse & Permissions
Figure 47
Example of a beam-size measurement for a bunch charge of 6 pC performed with the novel transverse profile imager. The horizontal and vertical rms beam sizes obtained from Gauss fits are about 15 and $20 μ m$ , respectively.
Reuse & Permissions
Figure 48
Schematic setup of the electron bunch length monitor based on coherent diffraction radiation. A window of high-resistivity float-zone (HRFZ) silicon is used to transmit the THz radiation.
Reuse & Permissions
Figure 49
Detector signals for two identical THz filters (a), relative signal fluctuation (b), detector signal noise as calculated from the correlation from the two detection channels (c), and noise contribution from read-out electronics (d), all shown as a function of the electron bunch charge.
Reuse & Permissions
Figure 50
Detector signals as a function of S-band and X-band rf phases controlling the bunch compression, shown for two different spectral bands: the signals of detector A, plots (a) and (b), are shown after a high-pass filter at 0.6 THz, those of detector B, plots (c) and (d), after a high-pass filter at 0.26 THz. For comparison, plots (e) and (f) show corresponding bunch-length measurements performed with the rf deflector. All signal values are displayed in relation to their corresponding noise levels as noted explicitly in the figures.
Reuse & Permissions
Figure 51
Schematic representation of the electro-optic bunch length monitor.
Reuse & Permissions
Figure 52
Comparison of bunch length measurements obtained with the electro-optic monitor and the transverse deflecting rf cavity.
Reuse & Permissions
Figure 53
Long-term bunch arrival-time measurement over a period of eight hours, using the BAM upstream of the bunch compressor at a bunch charge of 130 pC (top), arrival-time jitter over 100 shots (middle) with instantaneous resolution (red curve), and link length compensation applied by the motorized delay line (bottom).
Reuse & Permissions
Figure 54
Measured beam position variation along the injector in the horizontal plane in response to a change in rf gradient, before and after the beam-based alignment of the first booster structure. The beam position at the second BPM immediately after the first S-band structure is almost unchanged after the alignment, whereas further downstream changes persist due to the still misaligned structures. The locations of the four booster structures are indicated as gray boxes.
Reuse & Permissions
Figure 55
Measured spurious dispersion in the horizontal (upper panel) and vertical plane (lower panel) along the injector before and after the beam-based alignment (adjustment of the reference orbit).
Reuse & Permissions
Figure 56
Comparison of measurements of the $R_{56}$ parameter using the transverse deflecting cavity (TDC) and a downstream BPM (red points) and via bunch arrival-time using a BAM (blue points), for different bunch compressor angles.
Reuse & Permissions
Figure 57
Measured and modeled dispersion along the injector, with the bunch compressor angle set to 4.0°.
Reuse & Permissions
Figure 58
Example of a bunch length measurement with uncompressed beam, showing the calibration measurement (left plot) and the current profiles overlaid on the images of the streaked beam, as observed at the two zero crossings of the rf deflector (right plots).
Reuse & Permissions
Figure 59
Two examples of aperture scans showing core slice emittance measured for various laser spot sizes with the CTF gun (red asterisks) and the SwissFEL gun (blue circles). The dependence observed for the CTF gun features a distinct quadratic component (see text for details). In both cases the electrons were extracted from a copper cathode.
Reuse & Permissions
Figure 60
Normalized slice emittance measured with the SwissFEL gun for a bunch charge of 20 pC (red). The longitudinal charge profile is shown as gray bars.
Reuse & Permissions
Figure 61
Normalized horizontal (blue) and vertical (red) projected emittance obtained for a bunch charge of 200 pC from ten consecutive measurements with the CTF gun, illustrating the stability of machine and diagnostics chain. The geometric mean of the horizontal and vertical results is shown in black. The continuous horizontal lines represent the averages over the ten measurements.
Reuse & Permissions
Figure 62
Measured rms bunch length after applying the compression procedure, compared to the respective target bunch length. (Statistical errors, obtained from repeated measurements, are smaller than the marker circles.)
Reuse & Permissions
Figure 63
Examples of phase space fragmentation of compressed electron beams generated from a copper cathode with a flat-top like laser profile obtained with pulse stacking (see Fig. 4). The figures show particle densities of the vertically streaked beam as recorded on a screen in the dispersive section of the high-energy spectrometer for compression factors of 6 (top) and 12 (bottom). The magenta lines are projections of the distributions onto the horizontal (energy) axis, emphasizing the fragmentation effect.
Reuse & Permissions
Figure 64
Example of a compressed beam with smooth phase space distribution as obtained with a cesium telluride cathode using the same pulse-stacked laser profile as in Fig. 63. The applied compression factor is 6, corresponding to the upper panel of Fig. 63.
Reuse & Permissions
Figure 65
Current profiles (left plot) and three consecutive slice emittance measurements (right plot) for uncompressed and compressed beams (compression factor 7.5). The longitudinal extension of a slice is a fixed fraction of the rms bunch length.
Reuse & Permissions
Figure 66
Example of an undulator alignment scan showing the measured beam positions relative to the reference orbit, $Δ y$ , versus vertical undulator position.
Reuse & Permissions
Figure 67
Undulator entrance (black circles) and exit (red triangles) kicks measured as a function of the undulator $K$ value in the horizontal plane.
Reuse & Permissions
Figure 68
Position and angle changes of the electron orbit in the vertical plane at the undulator exit ( $Δ y$ and $Δ y^{'}$ ), represented as vertical and horizontal components of small white arrows, as a function of undulator pitch and height, individually plotted for undulator $K$ values ranging from 0.3 to 1.8. The underlying colors correspond to the lengths of the arrows, i.e., the deviations from zero orbit change, and illustrate the dependence of the natural focusing strength and center on the $K$ value. The plots for the different $K$ values are individually normalized to illustrate the focusing and bending effects.
Reuse & Permissions
Figure 69
Intensity distributions of SASE FEL pulses measured in the long-bunch (top) and the short-bunch configuration (bottom), shown as histograms. The curves overlaid represent fits to gamma distributions with the $M$ parameter left free (red) and fixed to simulation results obtained for equivalent beam conditions.
Reuse & Permissions
Figure 70
Undulator $K$ value as a function of undulator gap as derived from magnetic measurements using Hall sensors (black circles) and from spectral measurements of the generated SASE FEL radiation (red diamonds).
Reuse & Permissions

Physical Review Accelerators and Beams