+A   

3rd International Workshop on Beam Orbit Stabilization - IWBS2004

Announcement First / Second Accommodation Registration Participants Program Overview Details Orals by Session Orals with Abstracts Download PDF (gs) Download for PDAs Presentation   Upload / Download Summary Support Information No. #1 / #2 / #3 Pictures Gallery #1 / #2 / #3 Upload / Download SLS & Sponsors WWW psi.ch Updated: 25.01.2005		FACILITY REPORTS The session ``Facility Reports'' made up a large fraction of the workshop program. Reports on orbit stability at 10 operating and 4 future ring based light sources were given. The session concluded with a presentation on orbit stabilization plans for the Large Hadron Collider (LHC) presently under construction at CERN. The SPring-8 report focussed on the influence of the ``top-up'' operation mode on orbit stability which was introduced in May 2004. X-Ray beam stability at the experimental stations clearly improved and systematic current dependences of the orbit motion were reduced. Problems in the precise measurement of electron and photon beam positions remain. They are induced by systematic electron beam position monitors (BPMs) and X-Ray BPM (X-BPM) effects. BPMs in the vicinity of the Insertion Devices (IDs) exhibit ID gap, total current and filling pattern dependences and X-BPMs show a significant ID gap and phase dependence presently preventing SPring-8 from performing a ``hard correction'' on their readings. Local correction tests are planned in 2005 after improvement of the BPM hardware at the IDs. With ``top-up'' operation day/night related orbit variations can now be clearly resolved. The RMS fast orbit stability in the range 0.1-100 Hz at the IDs with $\sigma_{x/y}$ = 360/5 $\mu$m is presently x$_{rms}$/y$_{rms}$ = 1/4 $\mu$m. The RMS slow orbit stability $<$0.1 Hz has been found to be of the order of 1-3 $\mu$m/week. The central energy is kept constant within 2$\cdot$10$^{-5}$ full width. Although SPring-8 is ``only'' running a sub-Hz Slow Orbit Feedback (SOFB), it is able to near the sub-micron regime through careful noise suppression, indicating a Fast Orbit Feedback (FOFB) is unnecessary. The PF ring at KEK features a global FOFB with 12 ms cycle time which will be redesigned in the course of an upgrade from 7 to 13 straight sections in the PF ring in 2005. A local FOFB (50 Hz to be upgraded to 1 kHz) in combination with a feedforward (FF) has been successfully tested to suppress distortions induced by a new undulator for the PF-AR ring which provides circular-polarized light by mechanical switching at 0.8 Hz. In the PF-AR ring a new injection scheme facilitating a single pulsed quadrupole has been successfully tested reducing orbit transients to $\approx$100/10 $\mu$m in the horizontal/vertical plane. It is foreseen to employ such a device for ``top-up'' injection into the PF ring and to reduce the residual oscillations further. The Taiwan Light Source at NSRRC was upgraded in 2004 to suppress coupled-bunch instabilities by means of RF gap voltage modulation, super-conducting (SC) RF and a coupled-bunch feedback system. Orbit stability will be improved by ``top-up'' operation (already tested), mechanical/electrical source suppression, temperature stabilization and a global FOFB featuring a closed loop bandwidth (feedback BW) of $>$100 Hz. ``Top-up'' operation is scheduled for the end of 2005. The Brazilian Synchrotron Light Source suffered from a longitudinal dipole oscillation which appeared as a horizontal orbit distortion proportional to the second order dispersion which could be cured by applying a phase modulation to the RF voltage. BPMs which were affected by an increased electromagnetic noise floor in the hall received additional shielding boxes. BESSY is facing several new challenges. The femto-second slicing bump needs to be reproducible to within 0.5 % (10 $\mu$m) of the total amplitude which is difficult to achieve with the present correction hardware. The mechanical/magnetic hysteresis of a SC wave length shifter (WLS) leads to 0.1 mm orbit excursions in the frequency range 0.5-10 Hz. Operation with low alpha optics settings is delicate since the horizontal orbit response to energy changes is amplified by a factor 10-200 making the use of the RF frequency as a corrector difficult. A local FOFB has been implemented for an APPLE II type ID regulating on two adjacent X-BPMs utilizing dedicated broad-band correctors. At 1.6 Hz a $\approx$30 dB photon beam jitter suppression could be achieved. But there is a clear need for a global FOFB. A feedback on the horizontal Split Mirror Unit (SMU) of a beam-line operating at an APPLE II double undulator has been established in order to improve the two-beam overlap. With feedback (2 Hz feedback BW) an angular stability of $<$0.1 $\mu$rad has been measured. In 2004, ELETTRA underwent a complete machine re-alignment including the IDs to $<$100 $\mu$m RMS. Since re-alignment, tune and chromaticity changes cause very little orbit distortions, the RMS corrector strength is largely reduced, orbit corrections are converging faster and a quicker calibration of the ID corrector coils can be performed. A FF system for the Electromagnetic Elliptical Wiggler (EEW) allows to change the radiation polarization in AC mode up to 100 Hz without measurable orbit perturbation. Two local FOFBs with digital feedback electronics (sampling rate 8 kHz) are in routine user operation. They each involve two low-gap BPMs (14 mm) mechanically monitored to $<$50 nm with respect to a Carbon fiber column and four dedicated correctors. The PID controller is enhanced by a low pass filter with a cut-off frequency of 150 Hz to avoid power supply nonlinearities and harmonic suppressors at 50 Hz plus harmonics increasing the FOFB gain at these particular frequencies considerably. Sub-micron position ($<$0.2/0.9 $\mu$m) and angular ($<$0.02/0.2 $\mu$rad) stability has been achieved from 0-10/10-250 Hz. There are plans for a global FOFB involving all BPMs since crosstalk between many local FOFBs could become an issue. A full energy booster is under construction allowing for ``top-up'' operation. The APS is a mature facility with very sophisticated orbit control capability. A canted ID geometry (0.5/-1/0.5 mrad) in 3 sectors of the APS allows to suppress the dipole radiation background, which otherwise would significantly contribute to the systematic errors of the adjacent X-BPMs. By integrating pairs of vertically sensitive dipole X-BPMs into the global FOFB (in operation since 1997) excellent long-term pointing stability of $<$1 $\mu$m (20 m from the source point) and $<$0.2 $\mu$rad has been achieved. Dipole X-BPMs are perfectly suited for integration into a FOFB since the photon beam profile does not change. A new ``hard X-Ray'' BPM is being designed to achieve 100 nrad long-term pointing stability which would make local steering for the beam-lines obsolete. At the SLS ``top-up'' operation has proven to be an important prerequisite for high orbit and energy stability. The global FOFB running in user operation since $\approx$1 year ensures a complete decoupling of the ID operation up to 100 Hz. Slow ($<$1 Hz) X-BPM feedbacks running as an integral part of the FOFB following a cascaded feedback scheme guarantee sub-micron stability of the photon beam positions $\approx$10 m from the source point of presently 2 IDs. Several incidents related to the malfunctioning of the SLS cooling system have demonstrated how difficult it is to maintain the same high level of stability over long periods (weeks-months) if the operating conditions of the accelerator and the beam-lines cannot be kept constant. In Spring 2004 a global FOFB employing digitized Bergoz [4] BPM signals (sampling rate 1.1 kHz, 60 Hz feedback BW) was introduced to user operation at the ALS. A SOFB (sampling rate 1 Hz) running at the same time reports proposed corrections to the FOFB which inhibits possible crosstalk between both feedbacks. X-BPMs are only used by feedbacks on the beam-line optics (1 h$^{-1}$ to $\approx$ 10 kHz at infrared beam-lines). ALS is planning to upgrade to a full energy injector in order to be able to run in ``top-up'' mode at 500 mA with a granularity of 0.3 %, injecting single bunches every 30 s at a rate of 1 Hz. Gating signals will be provided to the beam-lines. First ``top-up'' tests revealed the necessity for improving the ring septum (leakage fields). Elliptically polarized undulators (EPUs) cause large beam size variations due to gap and phase dependent skew quadrupole fields of the IDs. Skew quadrupole FFs involving dedicated skew quadrupole correctors are being implemented. For ALS beam size stability is often more important than orbit stability. Until the end of 2004 ESRF operated a global SOFB ($\le$0.1 Hz feedback BW) with 224/96 BPMs/correctors, a global vertical FOFB (0.1-150 Hz feedback BW, correction deadband to decouple SOFB from FOFB) based on 16 BPMs/correctors (air coils) and 4 horizontal local FOFBs at the most sensitive IDs. Eventually the observed crosstalk between local FOFBs, due to non-closure of the 4-corrector bumps at high frequencies, motivated the upgrade to a global FOFB with 200 Hz feedback BW for both planes involving 32/24 BPMs/correctors. The new FOFB is operational and outperforms the old local FOFB arrangement. In spring 2006 SOLEIL will commence user operation with 10 initial beam-lines. Sub-micron orbit stability will be required at the IDs on the scale of a few ms to hours which makes a FOFB and ``top-up'' operation indispensable. The lowest eigenmode (measured) of the girders/dipoles has been shifted from 44/12.7 Hz to 46/27 Hz thanks to a more rigid fixation of the dipoles on the girders. The global FOFB will be based on digital BPM electronics from ``Instrumentation Technologies'' [5]. It is planned to have a global SOFB (0.01 Hz feedback BW) involving 120 BPMs and 56 slow correctors per plane ready for commissioning and a global FOFB (0.01-100 Hz, 8 kHz sampling rate) utilizing 120 BPMs and 46 fast correctors (air coils) in each plane a few months later. SOFB and FOFB will run simultaneously and will be integrated using a master/slave scheme as is the case at APS and ALS. The straight section BPMs are decoupled from the vacuum system by means of bellows and mounted on independent steel (later thermally stabilized or from Invar) supports. DIAMOND will be open to users in 2007. Similar stability requirements as in the SOLEIL case have led to comparable solutions. The very flexible girder design allows for movement in 5 degrees of freedom but as a result the structure exhibits the first eigenmode at ``only'' 29 Hz. The positions of the girders will be measured using a horizontal (HPS) and vertical (VPS) linear encoder (self-calibrating) based positioning system. They are intending to have one global FOFB using 168 BPMs and 168/96 slow/fast correctors in each plane where 4 of the fast ones will be located in each straight in order to allow for arc decoupled ID beam steering. The site of ALBA/CELLS has been selected. The proposed site, although not ideal, is not as bad as preliminary measurements first indicated. As it is often the case the criteria used to select the site were not of a technical nature. In mid-2007, DESY will start to transform the present HERA injector PETRA II (circumference 2304 m) into a synchrotron light source. One octant of the machine will be rebuilt to serve 13 undulator beam-lines (E = 6 GeV, $\epsilon_x$ = 1 nmrad, I = 100 mA). User operation is scheduled for 2009. The low emittance, $\epsilon_x$, is achieved by introducing 2 damping wiggler sections of 2$\times$40 m length. This imposes tight constraints on the allowed spurious dispersion in these sections ($\eta_{x/y}$ = $<$20/5 mm, similar values for ID locations) thus requiring dispersion correction to be an integral part of the planned SOFB (feedback BW $<\approx$0.1 Hz). Sub-micron stability requirements for the ID octant enforce the necessity for a ``local'' octant FOFB (feedback BW ~0.1-100 Hz). The FOFB extends into the adjacent ``old'' octants in order to be able to maintain the small vertical emittance (1 % coupling assumed). The Large Hadron Collider (LHC) (27 km circumference) at CERN is scheduled to be ready for start-up in 2007. Since it uses SC magnets with a somewhat ``extreme'' design in order to provide a dipole field of 8.3 T at 7 TeV, a fast beam loss of $<$10$^{-7}$ of each of the beams with 350 MJ stored energy at 7 TeV may quench a magnet (recovery time $\approx$6 h). Due to very limited apertures in the interaction regions the primary collimators must be closed to 5-7 $\sigma$ ($\sigma$ = 300 $\mu$m in the arcs) of the beam which leads to tight (for the proton world) constraints on the orbit stability at collimators and absorbers of $<\approx$50-70 $\mu$m. 528 BPMs/ring retrieve orbit data at a sampling rate of 50 Hz with a resolution of $<$1 $\mu$m. 280 slow (BW 1 Hz) SC correctors/plane/ring are employed for orbit correction. The planned SOFB which will have a centralized structure must meet high reliability requirements in order to minimize the risk of a magnet quench.