The long-term performance of the S-band klystron modulator system in the CERN LEP pre-injector

The Large Electron-Positron collider (LEP) is the final machine in a chain of four accelerators that are used to create particle collisions for high-energy physics experiments. LEP collides bunches of electrons (e/sup -/) with bunches of positrons (e/sup +/) that have originated in the LEP Injector Linac (LIL). These particles travel around the 27 km circumference of the LEP ring in opposite directions at velocities close to the speed of light. When bunches of particles collide, bursts of very high energy are created during a tiny fraction of a second, emulating the state of the early Universe. Four huge detector assemblies record the tracks of particles created in this way, and provide the physicists with a means of looking at the behaviour of matter at these high energies. LIL is at the front end of this chain and is used to produce the sequence of e/sup -/ and e/sup +/ beam pulses that are accumulated in 4 or 8 bunches, at a 100 Hz rate, in the Electron Positron Accumulator (EPA). The klystron-modulators, utilising commercial S-band technology, were designed for producing the peak (35 MW) pulsed RF accelerating fields at 2.9985 GHz frequency. This paper reviews the performance and equipment evolution of the klystron-modulators and associated systems, from when the first e/sup -//e/sup +/ beams were available in 1987 until 2000.