C. C. Kuo kuo@bd09.srrc.gov.tw
SRRC, Taiwan3mm
The Synchrotron Radiation Research Center (SRRC) is located in Hinschu,
Taiwan, where a 1.3 GeV low emittance storage ring, the Taiwan Light Source
(TLS), has been operated routinely for synchrotron radiation users
and for machine physics studies since its successful commissioning in
1993.
The accelerator system of the TLS consists of a 50 MeV linac, a 1.3 GeV booster synchrotron and a 1.3 GeV storage ring. The storage ring is a six-fold symmetric, TBA, combined-function lattice with six long straight sections, each of 6 m. The circumference is 120 m and the nominal emittance is 20 nm-rad at 1.3 GeV. The storage ring performance is as follows:
At present, the routine operation for users' shifts is at 1.5 GeV with initial beam current of 200 mA and beam lifetime of more than 9 hours. Injection takes a few minutes and the injection interval is 6 hours.
Recent beam dynamics activities include: (1) machine parameter optimization for users' runs, (2) investigations of the transverse and longitudinal instabilities, (3) development of both transverse and longitudinal damping systems, (4) suppression and reduction of the beam orbit perturbation sources, (4) development of a digital global orbit feedback system, (5) design of a harmonic cavity, and (6) study of beam-ion inteaction.
From the beam spectrum, we observed that the girder vibration sources
such as the turbo pumps and fan for air conditioning (17 
30 Hz) in
the first two-year runs. We also indentified that the 60 Hz noise was mainly
from correction power supplies. In addition, a slow beam orbit oscillation
(0.01 Hz) due to the temperature variation of the cooling water for magnets
also appeared. In the first half of 1996, we removed these perturbation
sources and, as a consequence, the beam orbit fluctuation has been much
reduced to a few micrometres.
The need for longer beam current lifetime in the storage ring requires better vacuum conditions and smaller beam charge density. The vacuum reading was about 1.5 nTorr on average at 200 mA and improvement work is in progress. To reduce the charge density only a few empty buckets were left and the vertical emittance was enlarged by increasing the coupling strength with skew quadrupoles. A harmonic cavity is under development to lengthen the bunch so as to favor the beam lifetime. Moreover, ramping the beam energy from 1.3 GeV to 1.5 GeV also caused a gain of beam lifetime by more than 20%. In all, the beam lifetime achieved with 4% emittance ratio was about 9 hours at 200 mA and a beam energy of 1.5 GeV.
We have investigated the possible further reduction of the natural emittance in the last several months. With the existing magnet arrangement, a decrease of the emittance by a factor of two has been achieved by changing the quadrupole strengths. However, the drawbacks are the increases of the dispersion function in the long straight sections and chromaticities. Further study is still going on.
Transverse beam instabilities were observed for the multi-bunch filling mode with the threshold current around 50 mA for well tuned cavities. It was attributed to the beam-ion interaction. We also observed an increase of the vertical beam oscillation amplitude during the severe outgassing process of a beamline front end. In the single bunch mode, no transverse instabilities were found up to 30`mA. A wide-band bunch-by-bunch transverse damping system has been constructed and employed routinely.
The threshold current of the longitudinal coupled-bunch instabilities
for our DORIS I cavities was as low as 10 mA. At 200 mA, the increase of the
energy spread could be twice the natural energy spread 
. To control
the energy spread, the cavity modes were adjusted with an extra tuner in
each cavity and with more precise cavity temperture control. Besides, a
bunch-by-bunch longitudinal feedback system is currently under construction.
The microwave instabilities were observed using a streak camera. The
measured threshold current was around 4 mA. The measurements suggested the
broadband longitudinal ring impedance was about 0.5 ohm.
There are six long straight sections in the TLS storage ring. One is occupied by the injection elements and the other is reserved for the installation of RF cavities. Therefore only four long straight sections are allocated for the insertion devices. These devices are a wiggler (W20), plane undulators (U5, U9) and an elliptically polarizing undulator (EPU5.6). More aggressive plans are to install a superconducting wiggler as well as superconducting bending magnets. The investigations of the beam dynamics effects of these insertion devices are important activities. Some tracking tools are used for the linear and nonlinear beam dynamics behaviour and to search for optimum lattice working conditions. To correct the linear perturbation of the lattice function, quadrupole families will be replaced with more independent power supplies next year.
The beam orbit was shifted when the magnetic gap of the insertion devices was varied due to the residual field errors and perturbed linear lattice. To reduce the orbit change, we generated feedforward correction tables for all insertion devices. However, to keep the orbit within a few micrometres during the change of the magnetic gaps, a digital global feedback is needed. We have constructed such a system and the performance was very impressive. The orbit could be maintained within a range of a few micrometres, which was close to the resolution of the beam position monitors.