No abstract

In the development of gyrotrons for controlled fusion experiments, two of the main goals are the increase of the radiated power and the wavelength shortening. Both tendencies lead to tubes in which the radial thickness of an electron beam may play a significant role not only in the efficiency degradation but also in the mode interaction. The second of those effects is analyzed in the present paper. The paper starts with a simplified analytical consideration of the effect of beam thickness on the mode coupling and stability in gyrotrons with soft and hard self-excitation. Then, this analytical consideration is illustrated by some examples which represent the results obtained with the use of the time-dependent, self-consistent code MAGY [M. Botton et al., IEEE Trans. Plasma Sci. 26, 882 (1998)].

No abstract

No abstract

A. N. Vkasov (SAIC, McLean, VA 22102) S. Cauffman and K. Felch (CPI, Palo Alto, CA 94304) For quasi-CW operation at the MW level and above, millimeter-wave gyrotrons should operate at very highorder modes. Selective excitation of these modes is an extremely difficult problem. Therefore a proper choice of the start-up scenario, which provides initial excitation of a desired mode and its drive into a high efficiency regime with simultaneous suppression of all undesirable parasitic modes, is very important. This problem was first formulated and studied a long time ago [I].Later, important contributions that clarified some issues in this problem were done in [2,3].There are several issues which are important for the correct understanding of this problem. First, gyrotron parameters, which correspond to the minimum start current, are different from the parameters optimal for the efficiency. So, it is desirable, first, to pass through the region of minimum start current where OUT operating mode will be excited, and then reach the region of maximum efficiency. Second, numerous calculations and experimental studies showed that the gyrotron maximum efficiency corresponds to the regime of hard selfexcitation where oscillations cannot be excited from the noise level as in the case of soft self-excitation. So, the device should, first, pass through the soft excitation region and then amve at the desired high efficiency operating point in the hard excitation region. Third, in pulsed gyrotrons it is impossible to vary magnetic fields during a single pulse even when water-cooled solenoids are used because they have a large inductance. (Clearly, it is also impossible in the case of superconducting solenoids.) Also, the beam current cannot be varied in one pulse, because electron guns used in gyrotrons operate in the regime of temperature limited emission, and therefore the current reaches saturation at rather low voltages. SO, the only way to provide a proper start-up scenario in pulsed gyrotrons is to manipulate with mod-anode, V,~,, and beam, Vb, voltages during the voltage rise time. Fortunately, in practically all pulsed gyrotrons the voltage rise time greatly exceeds the cavity fill time, Q / w , (the latter is on the order of nanoseconds). Therefore, the evolution of gyrotron oscillations during the voltage rise time can be considered as a slow, quasi-stationary process.A typical example of possible trajectories and selfexcitation conditions is shown in Fig. I [4] in the plane x = / 2 0 vi., and y = p2 / 2 0 v,. Here a shaded region shows the self-excitation zone for the operating mode. Self-excitation zones for the low-frequency and high-frequency parasitic modes are shown, respectively, on the right and left from it. Dash-dotted lines show the boundary of hard excitation regions in which the maximum efficiency point is located. Figure Ib corresponds to a denser spectrum of competing modes than Fig. la. As one can see, the point of final destination lies in the region of self-excitation of the low-frequency parasitic...