Orbit stability is of paramount concern for high brightness synchrotron radiation sources. Methods of overcoming both slow and fast variations were discussed. Several comments were made on the general topic of how best to implement orbit feedback systems. At present there is insufficient evidence whether photon BPMs can be made reliable enough, including the effects of ID gap changes, and so the use of electron BPMs has also to be considered. The question of whether local or global feedback is better, given that the electron BPMs are moving themselves, was discussed. Various opinions were expressed, such as that on the one hand global correction cannot perfectly correct the orbit and so leaves some residual errors, but that local correction is more sensitive to individual BPM movements and also that there could be a build up of errors if there are many loops active.

Slow Orbit Variations

Slow closed orbit movements due to changing synchrotron radiation load and cooling system fluctuations were reported at ELETTRA, MAX II and SRS. At ELETTRA, BPM and quadrupole movements during ramping were eliminated by means of additional cooling, while stress on the BPMs required development of a new gasket with an inner ring of copper and outer ring of stainless steel. Further orbit variations are also observed at ELETTRA after a ramp which can be correlated with water temperature variations, and which are overcome with an automatic slow local bump correction. Occasionally twenty-four hour orbit variations are also observed, the source of which is still unknown. At MAX II during the ramp the power dissipated in the quadrupole coils varies by a factor of ten and an associated periodic closed orbit variation is observed due to expansion of the magnet yokes. At the SRS, 100 µm horizontal movements are overcome with a global technique, while vertical movements are offset with local feedback; some difficulty is experienced however with the build up of the number of correction loops. The general conclusion is that full energy injection is clearly an advantage for orbit stability. To overcome effects due to magnet temperature variations, the interesting suggestion was made of maintaining a constant mean temperature, rather than stabilising the water inlet temperature.

Fast Orbit Fluctuations

At the ESRF the typical orbit fluctuations are 8 µm and 2.5 µm rms in the horizontal and vertical planes respectively, with a main component at the girder eigenfrequency of 7 Hz. The source is unknown, but not thought to be power supply ripple or induced by the cooling water. Provided data taking is slow, the orbit fluctuations give rise to a negligible change in effective source size, and so fast orbit correction may not be necessary. Some users however take data at a faster rate, in which case the sensitivity is greater and the usual 0.1 s criterion applies. Occasional beam vibrations can also be a problem, for example during closure of a beam shutter the closed orbit was seen to ring with the characteristic frequency of the magnet girders. Both local and global fast feedback systems are being developed, as well as APS-type girder damping.

At the SRS there is generally no need to worry about fast fluctuations because of the large beam sizes, however one user needs beam stabilisation at high frequencies (60-120 Hz) because of the way in which data is taken. Fast fluctuations are observed at ELETTRA in the 20-50 Hz range. A feedback development program is in progress using DSP's in conjunction with powerful mathematical software packages. The system uses photon BPMs with the option of switching to electron BPMs.

prepared by C.J. Bocchetta and R.P. Walker