Bunch merging and splitting techniques in the injectors for high energy hadron colliders

One problem in the design of TeV hadron colliders arises from the choice of the radio frequency. To produce economically the short bunches needed during collision, a high frequency (400 MHz or more) is essential. On the other hand, in the injector chain at lower energies, lower frequencies are generally preferred to alleviate space charge and instability problems and, partially also, for historical reasons. The classical solution is to transform the bunch structure by passing via a debunched beam state, during which the whole machine circumference becomes filled with particles and the beam is subjected to induced parasitic fields and is often prone to microwave instabilities. Bunch merging and bunch splitting have therefore been developed as alternative methods that allow the number of bunches to be changed without passing via that state. Bunch merging has been used in the CERN-PS since 1989 for the anti-proton production beam. The reverse process of bunch splitting was first proposed in the frame of the CERN injector complex for LHC, as a means of quasi-adiabatically changing the time structure of the beam, increasing the number of bunches from four at low energy in the PS Booster to 16 at high energy in the PS. When the CERN accelerator complex re-started in March 1998, this technique became a routine operation. New plans are now being considered to replace the debunching-rebunching process that is still needed at 26 GeV/c in the PS to generate the train of 84 bunches required by the SPS and LHC. The experience gained in the application of bunch merging and splitting techniques is reviewed in this paper together with the extensions now envisaged and the analysis of the expected advantages with respect to the more classical process of debunching-rebunching.