Figure 4-26 shows both the body heat and the output-cavity heat as calculated (using Equations (E-ll through E-16), Appendix E) for tubes of output power up to 43 kW. The output cavity alone produces a kilowatt or more both in the 5.6 kW PPM tube and in a confined-flow tube of output exceeding 16 kW. While this power may be dissipated at a density of more than 100 watts/sq. cm. , it must be radiated into space at about 0.4 watt/sq. cm. (depending on the radiator temperature). At the start of this study it was hoped to use conduction cooling as in the amplitron. Unfortunately, solid conduction is impractical at these power levels (see Appendix G) and, in this space application, forced air or water cooling are unsuitable because of their weight and reliability. However, heat pipes can transfer kilowatts of power from source to radiator surface with temperature drops of only a few degrees (reference 2, pp. 121-125). (17-19) The output gap and particularly the tunnel tips present the major problems, as experience with actual tubes has shown. (20) In the PPM tube, the heat must be transferred past the focusing magnets and pole pieces, which are less effective heat conductors than the copper cavity walls and can operate only at 300°C or less. Even with heat pipes between the cavity walls and the magnets, the available power is seriously restricted in such a tube with passive cooling. The results in Figure 4-26 show that a 16 kW confined-flow tube is at least as good as the 5.6 kW PPM tube; the latter produces relatively more heat from the beam before collector depression, while the lower perveance requires a higher voltage at the output gap and hence there are higher skin losses there. Large-signal computer studies yield useful information on both the quantity and the distribution of the heating of the klystron body. For example, beam interception in the PPM tube dissipates 120 watts halfway along the second drift tunnel (see Figure 4-18), and a further 180 watts are distributed over the inside of the tunnel wall at the end of the output gap (see Figure 4-19). Skin losses are calculated from the gap voltages (Equation (E-12), Appendix E) where each bunching cavity of the PPM tube loses about 5 watts, while the 43 kW tube would lose about 120 watts per cavity. While the field pattern in the cavity determines the actual distribution of this heat over the walls, a simple model (21) treats two-thirds as uniformly spread over the two re-entrant tunnel walls (inside the drift tunnel) and the rest over the inside of the cavity.
RkJQdWJsaXNoZXIy MTU5NjU0Mg==