Fig. 10. An example of a planar gyrocon. cavity-segment power-loss ratios are roughly the same for the superfish mode, and for the one in the gyrocon. In actuality, the superfish mode has no azimuthal variation (m = 0), whereas the gyrocon has a single traveling-wave period in the aximuthal direction (m = 1). Although a power density of 530 W/cm2 is above what heat pipes can easily cool, it is still within quoted heat-pipe possibilities (7). It is also comparable to power-loss densities in much lower powered klystrons. However, the location of the maximum power loss presents additional problems. We calculate maximum power-loss density to occur on the cavity inner wall between the beam slit and the axis. This surface must be accessed through an aperture less than 2 cm in diameter. Alternatively, the beam deflection can be increased. This enlarges the aperture, but also increases total power loss as well as drive power requirements. However, if a directly driven rf cathode could be developed, output cavities could be positioned at a deflection of 90°. Beam perveance might also be increased, and this combination could reduce both power losses and cooling difficulties. Such a system, called a trirotron (Figure 10) has been proposed (8), and some experiments are anticipated. C. Output-Cavity Field Fringing At S band, the beam slit in the output cavity can occupy as much as a third of the cavity width. This may lead to difficulties with fringe fields that introduce asymmetric beam slowing and additional power losses. One solution seems to be the operation of the output cavity in frequencymultiplying modes. These are higher order azimuthal variations in the output-cavity fields. They can be introduced so that the traveling wave remains synchronous with a subharmonically swept beam by simply widening the output cavity. This reduces the ratio of slot width to cavity width. It also raises the Q, raises the necessary beam deflection, and raises power losses.
RkJQdWJsaXNoZXIy MTU5NjU0Mg==