Search

The transverse coherence of the source is an important property for FEL experiments. Theory and simulations indicated different features for seeded and unseeded FELs but so far no direct comparison has been pursued experimentally on the same facility. At FERMI one has the unique possibility to test both configurations (SASE and seeding) within the same operating conditions. In this contribution we present the experimental results of the characterization of transverse coherence with special attention to the evolution of such property.

Improvement of the longitudinal coherence in the proposed Soft Xray FEL, the SXL, for the MAX IV Laboratory is an important design aspect to enhance the user case. One of the main considered methods is HBSASE. However the final compression in the MAX IV acceleratos is done at full energy, and thus leaving an energy chirp in the electron pulse. This chirp in longitudinal phase space has to be removed for an efficient implementation of HBSASE. In this paper we show in simulations how the phase space is improved by first overcompressing the pulse, and then correct it by a two-plate wake field de-chirper. The resulting pulse is then shown to have qualities such that, by HBSASE, a significant narrowing of the FEL bandwidth is achieved at 1 nm.

Accurate and high resolution detection of the Longitudinal Phase Space (LPS) of the electron beam is a great advantage for operating and setting up a FEL. In the case of the soft X-ray FEL being proposed at the MAX IV synchrotron facility in Lund, this information is mainly supplied by a Transverse Deflecting Cavity (TDC) which is currently being installed and scheduled for commissioning in the autumn. Performing the LPS measurement with the future TDC is limited in two regards: it is destructive and may be low in resolution as compared to the maximum compression possible in the MAX IV linac. In this project we propose using machine learning tools to implement a virtual diagnostic to retrieve the LPS information non-destructively using fast, non-invasive measurements and critical set-points in the linac as inputs for a neural network. In this paper we summarize the current progress of this project which has thus far focused on simulation studies of the TDC and the training of a virtual diagnostic using the TDC's simulated output.

The existing MAX IV 3 GeV linac could drive, with minor improvements, a soft X-ray Free Electron Laser and the aim of the SXL project has been so far to deliver a conceptual design of such a facility in the 1—5 nm wavelength range. The project was initiated by a group of Swedish users of FEL radiation and the design work was supported by the Knut and Alice Wallenberg foundation and by several Swedish universities and organizations (Stockholm, Uppsala, KTH Royal Institute of Technology, Stockholm-Uppsala FEL center, MAX IV laboratory and Lund University). In this paper we will focus on the baseline FEL performance based on two different accelerator operation modes (medium and short pulses) and give some hints of future developments after the first phase of the project such as 2 color/2pulses and HB-SASE.