After 31 years of serving the user community with excellent results, on July 2nd 2025 the removal of the Italian third generation synchrotron light source Elettra (www.elettra.eu) will start to be replaced by Elettra 2.0 a fourth-generation one. The project is in full development and, being a diffraction limited light source, Elettra 2.0 will provide ultra-high brilliance and coherence to the experiments while at the same time the machine is designed to provide very short pulses for time resolved experiments. The project status and its possibilities will be presented and discussed

Elettra Sincrotrone Trieste Research Center (Elettra) is one of the Italian Institutions, together with Istituto Nazionale di Fisica Nucleare (INFN) and Consiglio Nazionale delle Ricerche (CNR), committed to the realization of the Italian in-kind contributions for the European Spallation Source. Elettra contributions are concentrated on the proton accelerator and more specifically they concern the construction of the conventional iron-dominated electromagnets and related power converters to be installed in the superconducting part of the linac and in the High energy Beam Transport (HEBT), the RF power stations for the superconducting spoke cavity linac section and the wire scanner acquisition system for the beam diagnostics. This paper provides a description of the contributions and an overview of the status of the construction activities.

FERMI is the seeded Free Electron Laser (FEL) user facility at the Elettra laboratory in Trieste, operating in the VUV to EUV and soft X-rays spectral range; the radiation produced by the seeded FEL is characterized by wavelength stability, low temporal jitter and longitudinal coherence in the range 100-4 nm. During 2018 a dedicated experiment has shown the potential of the Echo Enabled Harmonic Generation (EEHG) scheme [1] to cover most of this spectral range with a single stage cascade [2]. Such a scheme, combined to an increment of the beam energy and of the accelerator performances, could extend the FERMI operating range toward the oxygen k-edge. With this perspective, we present the development plans under consideration for the next 3 to 5 years. These include an upgrade of the linac and of the existing FEL lines, consisting in the conversion of FEL-1 first, and FEL-2 successively, into EEHG seeded FELs.

In the framework of the Italian In-Kind Contribution (IKC) to the construction of the European Spallation Source (ESS), Elettra Sincrotrone Trieste is in charge of providing some key equipment for the accelerator. Among them, there are the magnets and the associated power converters for the Linac Warm Units (LWU), the High Energy Beam Transport (HEBT) dogleg, the Dump Line (DmpL) and the Accelerator to Target (A2T) sections of the neutron source. Magnets and their power converters are complementary parts of common systems. Their design cannot be totally separated, requiring iterations for an optimal solution that should include also power cable standardization. This work will describe the power converters' solutions adopted for the magnets of the abovementioned sections. They comprise 4-Quadrant power converters for the correctors, compact DC power converters for the small quadrupole magnets and more powerful DC power converters for the dipole and large quadrupole magnets. INTRODUCTION: ESS “The European Spallation Source (ESS) is a European Research Infrastructure Consortium (ERIC), a multidisciplinary research facility based on the world’s most powerful neutron source. The unique capabilities of this new facility will both greatly exceed and complement those of today's leading neutron sources, enabling new opportunities for researchers across the spectrum of scientific discovery, including life sciences, energy, environmental technology, cultural heritage and fundamental physics.” [1]. The neutron source is currently under construction in Lund, Sweden, in the framework of a 14-countries InKind Contribution (IKC) agreement [2, 3] (Fig. 1). Figure 1: Aerial photo of ESS. Italy is one of the founding Countries of ESS ERIC. The Italian participation to ESS consists in the In-Kind Contribution for the construction of the Facility. Three Italian Entities are involved: INFN, Elettra and CNR. Elettra, in particular, will provide components for the linear accelerator and proton beam transport. MAGNETS AND POWER CONVERTERS Besides superconducting accelerating sections, the accelerator requires a significant number of warm magnets – quadrupole (Qx), dipoles (D1) and combined H+V corrector magnets (Cy) – along the proton Linac (the socalled Linac Warm Units – LWUs) and beam transport [4]. Part of the Elettra contribution consists in these magnets and the associated power converters (PC). Figure 2 reports a schematic view of the accelerator and transport line structure. In the boxes, there are the magnet types and their number. Figure 2: ESS structure with magnet types. Table 1 summarises the type and number of warm magnets, their nominal current, and the number of power converters (PC). The acronym TBD indicates that the nominal currents are still to be defined while TBC means that the nominal current is an input parameter in the design of the corrector C8 and has to be confirmed. Table 1: LWUs and Beam Transport Magnets and PC Magnet Type Magnet # Imag [A] PC # Q5 26 150 26 Q6 95 180 95 Q7 12 180 12 Q8 6 TBD 6 D1 2 TBD 1 C5, C6 68 ±16 A 136 C8 4 ±16 A/TBC 8 In our vision, a magnet and its associated power converter are two parts of a unique system designed and optimized as a whole, considering also their installation in ___________________________________________ †roberto.visintini@elettra.eu WEPVA051 Proceedings of IPAC2017, Copenhagen, Denmark ISBN 978-3-95450-182-3 3372 Co py rig ht © 20 17 CC -B Y3. 0 an d by th er es pe ct iv ea ut ho rs 07 Accelerator Technology

FERMI has reached its nominal performance on both FEL lines, FEL-1 (12 eV to 62 eV) and FEL-2 (62 eV to 310 eV). After a brief overview of the activities with users, we will describe plans for LINAC, FEL and beamline upgrades for 2016-2018 and beyond. This includes EEHG schemes for FEL-2. INTRODUCTION FERMI is the seeded Free Electron Laser (FEL), operating in the VUV to soft X-rays range, located at the Elettra laboratory in Trieste, Italy [1]. FERMI is the only FEL user facility designed to operate in the seeded HGHG mode [2] and has successfully demonstrated operation from 100 nm down to 4 nm [3, 4]. The VUV to EUV FEL line, FEL-1, started operation with external users, i.e. selected by the FERMI Review Panel, in December 2012. Since then a period of actual commissioning of the source with users has been conducted, in a strong interaction between machine and laser team and the users. This resulted in proposing and implementing innovative solutions and schemes for pumpprobe experiments. A portion of the seed laser is delivered as optical laser pulse to the experimental stations for pump-probe experiments with extremely low jitter to the FEL pulse, less than 7 fs rms [5]. Several schemes to produce two colour, FEL-pump and FEL-probe, pulses have been implemented [6-9]; the most recent FEL scheme makes use of two seed laser beams of different wavelengths and of a split radiator section to generate two extreme ultraviolet pulses from distinct portions of the same electron bunch [10]. A parameter under observation during user’s beamtimes is of course the actual availability of the FEL compared to the scheduled time. The uptime has been always around 85% of the scheduled time and, thanks to some improvements on the machine systems, has lately attained 90%. FEL-2 is the second FEL line of FERMI. It uses a double stage cascade scheme and the fresh bunch technique. First lasing was observed in 2012 [4] and the performance was gradually improved [11], in parallel to FEL-1 operation, until it reached nominal performances in 2014 [1]. The FEL emission of FEL-2 is characterized by excellent spectral line shapes and the transverse profile of the FEL pulses is close to the TEM00 Gaussian mode. At wavelengths below 5 nm the FEL tuning turned out to be critical, with large shot-to-shot energy fluctuations and with the necessity of a seed energy in excess of 40 J. Based on the experience of the operation for users on FEL-1 and of the commissioning of FEL-2, and with the aim of increasing the uptime of the facility beyond the 90% mark, a number of upgrade actions have been activated in 2015 and are being planned for 2016-2018. These upgrades shall further improve the performance of FEL-1 in the various multicolour schemes and should allow on FEL-2 similar tuning possibilities as on FEL-1, along with improved stability and flexibility of operation. OPERATION FOR USERS FERMI has been in operation for 6528 hours during 2015; 3520 hours thereof, that is 54% of the total operation time, were dedicated to users with a clear increase compared to the 43% reached in 2014. The remaining time has been divided between machine commissioning (1312 hours equal to 20% of the total operation time; was 40% in 2014) and machine and beamlines tuning (1696 hours, equal to 26% of the total operation time; was 17% in 2014). The average FEL uptime for users in 2015 was equal to 88.0% of the scheduled FEL time, confirming the effectiveness of the constant effort in increasing the reliability of the machine. In fact, in 2014 the uptime was equal to 86.1% and in 2013 to 84.7%. In the last run of 2015, that lasted three months under continuous operation, a value of 91% was registered. A total of 27 Users’ beamtimes, 20 peer-reviewed and 7 in-house, were allocated on the three beamlines operated in 2015, namely Diffraction and Projection Imaging (DiProI, 12 experiments), Low Density Matter (LDM, 7 experiments) and Elastic Inelastic Scattering TIMEX ____________________________________________ † email address: alessandro.fabris@elettra.eu MOPOW015 Proceedings of IPAC2016, Busan, Korea ISBN 978-3-95450-147-2 744 C op yr ig ht © 20 16 C C -B Y3. 0 an d by th e re sp ec tiv e au th or s 02 Photon Sources and Electron Accelerators A06 Free Electron Lasers (EIS-TIMEX, 6 experiments). In-house beamtimes in 2015 also included commissioning of the EIS-TIMER experimental station, which started in July [12] and continued in November, when the first evidence of a transient-grating signal was observed from a BaF2 sample. Also the commissioning of the TeraFERMI beamline, that collects the THz radiation (ultrashort pulses in the 0.1-15 THz range) naturally emitted by the electron beam already spent by the FEL undulators, has started in December 2015 and the first THz photons have been observed. The sixth beamline at FERMI is MagneDyn, dedicated to the exploration of the ultimate limits of magneto-dynamic processes in conventional and advanced magnetic materials. This beamline is still under construction, with the commissioning expected to begin at the end of 2016. After five calls the number of the submitted proposals by users requiring beamtime for experiments at FERMI is still steadily increasing. In the last call for experiments, at the beginning of February, 72 proposals were submitted, 21 of them are experiments using FEL-2. The estimated oversubscription rate is about 3.4. UPGRADE RESULTS 2016 The first upgrade program started between 2014 and 2015, focused mainly on the LINAC and on FEL-2. The results collected in a dedicated commissioning period scheduled at the beginning of 2016 are discussed in the following sections. LINAC FERMI is driven by a 200 meter long S-band LINAC. In the injector region, two 3-meter long forward traveling wave accelerating structures, coming from the old Elettra injector, were installed. In order to improve the electron beam quality, it was decided to replace the first two existing structures with two dual-feed accelerating structures [13]. The structures were manufactured by Research Instruments GmbH and delivered to Elettra in July 2015. Figure 1: The new accelerating structures in the injector. After the high power testing in the FERMI cavity test facility, in January 2016 the structures were installed, on schedule, in the injector region, as shown in Fig. 1. The old structures were consequently moved into the high energy region of the LINAC to further increase the final energy by approximately 90 MeV. In February 2016 normalized emittances of 0.7 and 0.9 mm mrad in the horizontal and vertical plane, respectively, were measured in the 100 MeV diagnostic section. These values at 700 pC are 10-15% smaller than the previous ones. Peak LINAC energies as high as 1629 MeV were measured. The maximum operating energy, with compressed and linearized electron beam phase space, at 700 A of nominal current, is about 1550 MeV (see Fig. 2). Figure 2: Maximum LINAC operating energy. FEL-2 In 2014 the design parameters of FEL-2 were demonstrated down to the lower end of the wavelength range, 4 nm. However, it appeared evident that upgrades were needed to improve the stability and flexibility of FEL-2, and to achieve a frequency tunability similar to the one available on FEL-1 (OPA operation). The first of these upgrades addressed the LINAC energy, as described in the previous section. The second upgrade, specific to FEL-2, consists in the installation of a second regenerative amplifier. It produces a shorter seed pulse (infrared pulse duration less than 40 fs, 60-70 fs in the UV) that allows for better filling of the longitudinal phase space in fresh bunch mode. The laser also permits a wavelength tunability in the UV (+/-2%) allowing a limited FEL wavelength tuning with the third harmonic generation setup. As well a second OPA laser system was installed, so that double pulse seeding is possible on FEL-1 with two fully tunable sources. Another important upgrade of FEL-2 consists in the installation of a third Elliptically Polarized Undulator (EPU) in the radiator of the first HGHG stage of FEL-2, completing the design layout of this FEL line, shown in Fig. 3. Figure 3: FEL-2 with the new installed EPU. Proceedings of IPAC2016, Busan, Korea MOPOW015 02 Photon Sources and Electron Accelerators A06 Free Electron Lasers ISBN 978-3-95450-147-2 745 C op yr ig ht © 20 16 C C -B Y3. 0 an d by th e re sp ec tiv e au th or s The additional EPU reduces the seed energy required for operation in the shorter wavelength range, providing the possibility to operate with OPA laser system in a longer wavelength range. The EPU was built by Kyma srl with the same parameters of the two already existing ones: 55.2 mm period length, 42 periods, 10 mm gap. In March 2016, after the upgrades, FEL-2 reached unprecedented shot to shot stability, below 6% rms at 4.2 nm, as shown in Fig. 4, and average intensity greater than 15 J/pulse. The energy requirement from the seed laser could be reduced to less than 20 J. The new undulator configuration also offers the flexibility to operate efficiently with different harmonic conversion factors, as 16x4 which is almost equivalent to 13x5 used in the past. Figure 4: FEL-2 pulse by pulse intensity at 4.2 nm. Reasonable energy per pulse was measured down to 3.8 nm. Radiation spectra could be measured in single shot even at harmonic 3.1 nm, as shown in Fig. 5. Figure 5: Single shot spectrum at harmonic 17x5=85. The plots in the inset of Fig. 5 represent spectra from multiple shots during the same sequence. The green line is the projection of the single shot CCD acquisition displayed in the figure. The vertical axis represents the unfocused vertical distribution of the FEL pulse. PLANS FOR THE NEAR FUTURE The upgrade plans for 2017-2018 at FERMI foresee exploring the possibility of replacing the aging S-band Backward Traveling Wave structures (BTWs) of the LINAC and to implement the Echo Enabled Harmonic Generation scheme (EEHG) on FE

FERMI is the seeded Free Electron Laser (FEL) user facility at the Elettra laboratory in Trieste, operating in the VUV to EUV and soft X-rays spectral range; the radiation produced by the seeded FEL is characterized by wavelength stability, low temporal jitter and longitudinal coherence in the range 100-4 nm. During 2018 a dedicated experiment has shown the potential of the Echo Enabled Harmonic Generation (EEHG) scheme [1] to cover most of this spectral range with a single stage cascade [2]. Such a scheme, combined to an increment of the beam energy and of the accelerator performances, could extend the FERMI operating range toward the oxygen k-edge. With this perspective, we present the development plans under consideration for the next 3 to 5 years. These include an upgrade of the linac and of the existing FEL lines, consisting in the conversion of FEL-1 first, and FEL-2 successively, into EEHG seeded FELs.

ESS, the European Spallation Source, will be a major user facility at which researchers from academia and industry will investigate scientific questions using neutron beams. ESS will deliver its protons to a solid, rotating tungsten target, which will in turn generate neutrons. A linear accelerator (Linac) creates protons at the ion source, accelerates them to an appropriate energy and steers them onto the target to create neutrons via the spallation process. Among the cryogenic accelerating sections, there are "warm" magnets for correcting the proton beam "optics". This paper reports the chosen strategies for the power converters in these LWUs (Linac Warm Units). Amongst the power converters, we will highlight the 4-quadrant corrector magnets ones, based on a high precision and high bandwidth MOSFET H-bridge chopper, and a proposed novel topology for pulsed quadrupole magnets, instead of traditional DC ones. We will also demonstrate the interest of using pulsed magnets in place of DC ones in low duty-cycle Linacs, reducing the magnet losses and therefore the power consumption up-to 90%.

The complexity of a particle accelerator implies the remote control of very large numbers of devices, with many different typologies, either distributed along the accelerator or concentrated in locations, often far away from each other. Local and global control systems handle the devices through dedicated communication channels and interfaces. Each controlled device is practically a "smart node" performing a specific task. In addition, very often, those tasks are managed in real-time mode. The performances required to the control interface has an influence on the cost of the distributed nodes as well as on their hardware and software implementation. In large facilities (e.g. CERN) the "smart nodes" derive from specific in-house developments. Alternatively, it is possible to find on the market commercial devices, whose performances (and prices) are spread over a broad range, and spanning from proprietary design (customizable to the user's needs) to open source/design. In this paper, we will describe some applications of smart nodes in the particle accelerators field, with special focus on the power supplies for magnets. In modern accelerators, in fact, magnets and their associated power supplies constitute systems distributed along the accelerator itself, and strongly interfaced with the remote control system as well as with more specific (and often more demanding) orbit/trajectory feedback systems. We will give examples of actual systems, installed and operational on two light sources, Elettra and FERMI, located in the Elettra Research Center in Trieste, Italy.

Modern particle accelerators, in particular “Ultra-Low-Emittance Light Sources”, often adopt small and compact electromagnets, with low current levels for minimizing losses. Compact power supply systems are required, as well, and characterized by high efficiency and high output current stability. These are the key features on which we have developed the A2720 Power Supply System described in the paper. The A2720 System can host one to four independent current-controlled bipolar DC-DC modules, each capable of delivering up to ±20 A and ±50 V in DC-operation, in a single 19-inch, 3U standard crate. Each module implements a fully digital control loop (PID regulation) with high-resolution PWM pattern generation that makes the system versatile and easy to “tune” to any load condition. The A2720 System is composed by the crate of the power supplies and up to 4 AC/DC “bulk” power supplies, connected electrically by a proprietary backplane. Remote communication is guaranteed by means of a 10/100 Mbps Ethernet autosensing socket present on each A2720 module front panel.

FERMI is the seeded FEL user facility in Trieste, Italy, producing photons from the VUV to the soft X-rays with a high degree of coherence and spectral stability. Both FEL lines, FEL-1 and FEL-2, are now available for users, down to the shortest wavelength of 4 nm. We will report on the completion of the commissioning of the high energy FEL line, FEL-2, and on the operational experience for users, in particular those requiring specific FEL configurations, like two-colour experiments. We will also give a perspective on the improvements and upgrades which have been triggered by our experience and are aiming to maintain as well as to constantly improve the performance of the facility for our user community.

FERMI is the FEL-based light source in operation for external users since 2011 at the Elettra Research Center in Trieste, Italy. FERMI@Elettra is the name of the project for the construction and commissioning of this source. The design strategies adopted in the project had to consider the extremely close presence and the routine operation for users of the synchrotron–based source Elettra. There are more than 350 magnets and coils distributed along the linear accelerator, the two chains of undulators and the electron beam dump. Almost each magnetic element requires a dedicated DC power supply. Magnets, power supplies, and the connecting cables constitute a system, strongly interconnected with the remote control system, the machine and personnel safety system, and the infrastructure. All this has to fulfill the requirements from the particle physics specialists. We adopted a “systemic” approach in the design of new magnets and the re-use of the old ones, as well as the choice and the design of the associated power supplies. The commissioning of the systems started early 2010 and almost all magnet power supplies are in operation since then. During these 5 years, we introduced few minor upgrades and patches while the adopted solutions proved their soundness in terms of performance and reliability, causing very little downtime to the FERMI operations.

Full Energy Booster injectors for Storage rings are complex systems and somewhat reduced versions of the ring itself. They provide accelerated electrons that usually have the same energy of the storage ring accumulated beam. Power converters used to deflect, focus and defocus the electron beam in the booster must carefully follow a waveform both in time and in current values, as it is mandatory to get the electrons accelerated at the right energy to be further injected in the storage ring. A highly reliable, precise and repeatable power converter is needed to operate correctly on the beam and get a well-defined and stable orbit during the acceleration phase

Ageing of devices and components phasing-out, as well as the increasing maintenance costs affect particle accelerators similarly to any industrial plant. A careful maintenance plan can cope with these problems in the medium-term and then a complete update of the oldest parts is required. The most recent technologies available on the market, together with a modular and open design approach are the basis of an upgrade program aimed at replacing the existing controllers of the Elettra storage ring magnet power supplies. The design considerations, the constraints and the first results are here reported.

Among the wide range of applications that power supplies are able to cover, the Particle Accelerator based Light Sources are one of the most demanding. In the specific case of the Elettra Storage Ring, a Synchrotron Radiation Facility operating since 1993, magnet power supplies with other critical devices are determining the quality of the photon beams and overall accelerator uptime. Because of that, they have to provide stable, reliable, precise and remotely controlled operation. From a strictly functional perspective, each magnet power supply can be seen as a power part, regulating the output current, and a control part to connect the unit to the remote control system. The original Elettra magnet power supplies are still reliably in operation but their control parts, based on VME boards and MIL 1553 interfaces, need an update to cope with ageing, components obsolescence and servicing. This paper presents the laboratory and on-field performance of the NewPSC (New Power Supply Controller), the board designed at Elettra to replace the VME-based power supplies control systems in the Storage Ring.

The effective management of several distributed devices, even in real time, is one of the mandatory conditions for the design, operation and upgrade of modern particle accelerators control systems. In order to cope with the performance demanded by such plants, the existing equipment has to be continuously improved or new strategies have to be considered to overcome its limitations. The impressive capability growth of mobile and handheld devices allows moving from a traditional distributed architecture, based on complex and expensive operating nodes, to a pervasive approach based on simple and cheap embedded systems, tailored for each specific application. Such pervasive trend is further favored by the open design approach, which is becoming more and more popular as it enables designers to easily adapt an already existing board to their specific needs. In this communication a brief description of Elettra and FERMI light sources in the context of controls, diagnostics and power supplies is given. The main applications developed and successfully deployed on the accelerators and based on a commercial open design embedded board are also presented.

Stringent demands must be fulfilled when designing power supplies (PSs) for particle accelerators, such as stability, efficiency, accuracy, and electromagnetic compatibility. In this context, the paper analyzes the basic requirements to be considered when designing a PS for such applications. After reviewing the main circuitry topologies used in PSs, the electrical specifications and functional modes necessary to set up high-performance PSs are addressed. Problems concerning the fulfillment of the electrical specifications are discussed, followed by an overview of the common PS control methods. Finally, the impact of some of the discussed design considerations is evaluated for the PS of the magnets of particle accelerators.

The Elettra Storage Ring is operating for external Users since 1994. High stability, low ripple, DC power supplies are needed for energizing the magnets used to keep the electron beam on the defined path with the required characteristics. The remote control of the power supplies is still based on VME standard and serial MIL 1553 network. A study for upgrading the local power supplies control system, adopting modern technologies and software tools is presented in the article, along with the inherent difficulties due to the large number of units and the impact on the operation of the accelerator.

FERMI@Elettra is the first seeded VUV/soft X-ray FEL source. It is composed of two undulatory chains: the low energy branch (FELl) covering the wavelength range from 20 nm up to 100 nm, and the high energy branch (FEL2, employing a double stage cascade), covering the wavelength range from 4 nm up to 20 nm. At the end of 2012 FELl has been opened to external users while FEL2 has been turned on for the first time having demonstrated that a double cascade scheme is suitable for generating high intensity coherent FEL radiation. In this paper we will share our experience and will show our most recent results for both FERMI FELl and FEL2 sources. We will also present a brand new machine scheme that allows to perform two-colour pump and probe experiments as well as the first experimental results. (C) (2013) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.