Comparisons of the CODD and of the Normaliser Systems For Closed Orbit Measurements in the PS

The PS ring is equipped with 40 position PUs distributed around the ring. These are connected to the CODD system, which performs trajectory measurements of any single bunch, on two consecutive turns, and to 40 Normalisers, which deliver an averaged orbit. CODD uses a beam-synchronous timing system, which tracks a given bunch all through the acceleration cycle, but needs resynchronization after harmonic changes. It measures at injection or at any given C-timing, albeit with no less than 5ms between acquisitions. It is blind during harmonic changes. A Normaliser, using a technique originally developed for radial loop control, produces a slow signal, proportional to the average position of the bunches. At beam injection a settling time of 1.5ms is needed; then it follows orbit changes with a 200us time constant. Thus, it will not show rapid position changes, such as betatron oscillations. It does not require accurate timing. In the PS installation, the Normaliser outputs are simply sampled at 1ms intervals. CODD affords measurements on single bunches right from injection, but needs a complicated timing system and delivers results with mediocre resolution. The Normaliser, on the other hand, has simple timing, but produces the average orbit over the whole beam over a few hundred microseconds, thus missing the injection event. The averaging lends it a better resolution than CODD, though not by as much as could have been expected, based on its time constants. The aim of this document is to compare both systems and show the superiority of the Normaliser for closed-orbit measurements.