A perturbation technique is described for finding phase velocities and coupling impedances in a traveling wave tube for an arbitrary distribution of dielectric material. A model of the sheath helix is presented. Tape helix results will be presented in a separate paper. In all cases presented, without adjusting the dielectric constant, the calculated perturbed phase velocity provided a better answer than the homogeneous dielectric solution, or Naval Research Laboratories' Small Signal Gain Program. Deviation from theory versus experiment is reported by stating the average sum of the squares difference between theoretical calculations and a second order least squares fit of the measured data. Phase velocities can be calculated for uniform dielectric support rods where the average sum of the squares /spl les/1.19/spl times/10/sup /spl minus/5/. For cases with notched dielectric support rods phase velocities can be calculated where the average sum of the squares /spl les/1.94/spl times/10/sup /spl minus/5/. For NRL's SSG program the average sum of the squares was /spl les/1.01/spl times/10/sup /spl minus/4/ by comparison. For uniform dielectric support rods the perturbation does not significantly alter the basic shape of the predicted dispersion curve. For notched dielectric support rods applying the perturbation does alter and flatten the shape of the predicted dispersion curve in agreement with experiment.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">></ETX>
Read full abstract