The linear theory of the Clustered-Cavity Klystron

In the Clustered-Cavity Klystron, the individual intermediate cavities of a multicavity klystron are replaced by pairs or triplets of artificially-loaded cavities having Q's reduced to one-half or one-third that of the single cavity they replace. The cavity spacing within the multiplet is as close as possible. The spacing from the center of a multiplet to the center of the next is about one-quarter of a space-charge wavelength. A simplistic way of explaining the improved bandwidth is to realize that the two or three cavities in the multiplet are tuned to essentially the same frequency and are excited by essentially the same RF current so they act on the beam in unison to form a new bunch, in the same way a single cavity might, but over two or three times the bandwidth. The actual physics is more complex, and in particular, when cavity spacings in a klystron differ from one-quarter of a space-charge (Langmuir) wavelength, velocity currents become important, These have not normally been included in the analytic theory of broadband klystrons. In this paper we point out the necessary corrections to the theory of broadband klystrons. Using the clustered-cavity principle, the bandwidth of two high-power klystrons has been doubled without changing the outside dimensions of the tubes. The theory in this paper predicts that bandwidths in the order of 30 percent can be obtained in megawatt klystrons using about fifteen intermediate cavities arranged in triplets. In ultrarelativistic klystrons, the added gain-bandwidth product provided by the cavity clusters can be used to provide high gain together with reasonable bandwidth in a reasonable length. By using cavity clusters, it is possible to build a fairly-short very-relativistic klystron that still retains considerable bandwidth.< >