E. Peschardt

Longitudinal Instabilities of Bunched Beams in the ISR

Microwave instabilities occur in bunched beams in the ISR leading to a dilution of the phase space density and limiting the longitudinal density of the stacked beams. According to D. Boussard this instability can be described as a coasting beam instability inside bunches. Experimental investigations of this microwave instability support this theory and give a high frequency impedance ¿ZL¿/n ¿ 14 ohms. Injecting large currents in bunches of large area increases the threshold of this instability. The larger currents can produce coupled bunch mode instabilities which can be cured by a higher harmonic cavity.

Ultimate performance of the LEP RF system

The LEP Superconducting RF system reached its maximum configuration of 288 four-cell cavities powered by 36 klystrons in 1999. In 2000, this system, together with 56 cavities of the original copper RF system, routinely provided more than 3630 MV, allowing the beam energy to be raised up to 104.5 GeV. This not only required operating the cavities more than 15% above their design gradient, but has also demanded a very high operational reliability from the entire system. This paper will describe the operation of the LEP RF system during 2000, including new features, operational procedures and limitations.

The LEP Radio-Frequency Low Power System

A brief description of the low power and controls electronics for running the 128 cavity assemblies in the initial phase of LEP is given. Virtually identical hardware will be used for the future extension of LEP with superconducting niobium cavities. One complete RF unit consisting of 16 cavities and all associated drive and controls electronics has been successfully tested.1

Global voltage control for the LEP RF system

LEP RF system is installed as independent 16 cavity units. In addition to the eight copper cavity units originally installed 12 units with super-conducting cavities are being added for the LEP200 energy upgrade. The total RF voltage determines the synchrotron tune (Q/sub s/) and must be controlled precisely during energy ramping. Local function generators in each of the RF units are pre-loaded such that when triggered simultaneously by ramp timing events transmitted over the general timing system the total voltage varies to give the Q/sub s/ function required. A disadvantage is that loss of RF in a unit at any time after the loading process cannot be corrected. As the number of RF units increases automatic control of the total RF voltage and its distribution around LEP becomes desirable. A global voltage control system, based on a central VME controller, has recently been installed. It has direct and rapid access to the RF units over the LEP time division multiplexing system. Initial tests on operation and performance at fixed energy and during energy ramping are described, as well as the implementation of a Q/sub s/ loop in which Q/sub s/ can be set directly using on-line synchrotron frequency measurements.<<ETX>>

The ISR Impedance between 40 KHz and 40 GHz

Different current dependent effects have been measured in the ISR over the years and used to obtain information on the impedance of the beam surroundings over a large frequency range. The transverse impedance at low frequencies f < 20 MHz was obtained from the response of the beam to an excitation (beam transfer function) and at medium frequencies 30 < f < 200 MHz from the growth rate of the head tail instability. Potential well bunch lengthening and synchrotron frequency shifts gave the longitudinal reactive impedance over low and medium frequencies while the observation of the microwave instability (turbulent bunch lengthening) gave its resistive part at high frequencies f ~ 1 GHz. The measured energy loss of an unbunched beam gave an integral over the resistive impedance up to frequencies of 7 to 70 GHz depending on the beam energy. To analyse all the measured data the approximate relation between transverse and longitudinal impedance was used and a model impedance was fitted which has the expected behavior at low, intermediate and very high frequencies where the impedance is dominated by the smooth resistive wall, the effect of higher frequency resonances and by diffraction respectively.

Longitudinal Bunch Dilution Due to RF Noise

The effect of phase noise on a tightly bunched proton beam is investigated taking into account the frequency spread in the beam, the wall impedance and the phase feedback loop. Under normal conditions the measured dilution rates in the ISR correspond typically to a doubling of the bunch length in one to a few hours, and are consistent with the measured noise spectra. Amplitude noise is not investigated.

Longitudinal feedback in LEP

Dipole coupled bunch oscillations were observed at an early stage of LEP commissioning for currents above about 150 /spl mu/A per bunch. An improvised feedback system, acting on the phase of some of the accelerating cavities was developed and has been in operation for about three years. However, due to the small bandwidth of the RF cavities this system can only be used with four bunches or less per beam. With plans for eight bunch operation (the Pretzel scheme) the construction of a dedicated longitudinal feedback system was approved in 1991. The system operates at 999.95 MHz with phase modulation of a 200 kW klystron feeding four seven-cell cavities. The necessary bandwidth of 260 kHz is obtained by heavy over-coupling. With a total cavity voltage of 1.9 MV a damping rate of about 450 s/sup -1/ is obtained with phase excursions of one radian. The system has been in routine operation since July 1992 with a feedback cavity voltage of 1.2 MV and a damping rate of about 100 s/sup -1/. Longitudinal feedback eases operation and usually increases the maximum currents which can be accumulated.<<ETX>>

Stochastic cooling in the CERN ISR during pp colliding beam physics

When the ISR is used as a pp collider a high-intensity proton beam in R1 collides with a low-intensity antiproton beam in R2 for periods of up to two weeks. The luminosity lifetime is increased with a vertical stochastic cooling system in R1 designed for currents up to 10 A with a total bandwidth of 3.3 GHz. Cooling rates up to 0.7%/h have been obtained. The R2 antiprotons are cooled vertically with a 100-600 MHz system which decreases the initial beam height by up to a factor 7 and increases the luminosity by a factor 1.3-1.4. A momentum cooling system in R2 (frequency range: 55-155 MHz) creates empty space within the p stacking aperture. This allows several stacks from the antiproton accumulator (AA) to be stored in the ISR. With this system which uses the Palmer method it has been proved experimentally that momentum cooling and horizontal betatron cooling are obtained simultaneously if the betatron phase between pick-up and exciter is an odd number of half-betatron wavelengths. The cooling rate of the momentum cooling system which is power-limited is typically 4%/h at 26 GeV/c. The layout of the cooling systems is shown schematically in Fig. 1. The position of the various elements is mainly determined by the available free space and the lattice parameters. Future applications include vertical cooling of α-particles and cooling of antiprotons in all three planes in an experiment where a circulating 3.5-7.5 GeV/c p beam collides with a hydrogen gas jet target.

On-Line Q Measurement during Phase Displacement Acceleration in the CERN ISR

A new technique for measuring Q-values in a coasting beam traversed by RF buckets has been developed based on the RF knock-out principle. At the same time as the bucket traverses the stack a sweeping oscillator excites the beam transversely inside the stack. The resonance created can be detected from a pick-up signal when its radial position coincides with the bucket position. The exciting oscillator is locked to this signal via a phase-locked loop and the Q-value can be measured and stored by a local data acquisition system. At the end of the RF sweep the data is transferred to a computer for further treatment. Automatic control of the working line during phase displacement acceleration is then possible.