4.3.2 CASCADED VS. PARALLEL CONFIGURATIONS The amplitron and klystron have quite different characteristics in a number of important respects. Of particular interest is the capability of feed through for the amplitron, i.e. , if prime power is removed, the RF drive power feeds through and provides a significant output power. This is not true of a klystron. With no de power the klystron is not a 0 dB gain amplifier but is an attenuator of significance (typically 90 dB). The second characteristic of interest is the phase sensitivity for changes in source power supply voltage. Phase sensitivity for the amplitron, using constant current regulation, is . Without this regulation it would be , resulting in . Also of interest in the case of the amplitron is the phase change from the feed through to the amplifying mode; i.e. , the phase change in output rf with and without dc power applied. This is typically 12°. Phase shift across the klystron for the MPTS is nominally 5000°. Change in phase with source voltage is approximately c. For the klystron it is proposed to operate unregulated, and this effect must be compensated by an offsetting change in phase command at the input. Major voltage changes at the power source can be prevented in the case of a solar photovoltaic source by varying sun angle and/or by switching parallel-series combinations of cell arrays (needed for amplitron startup). A third feature to be considered is the effect of loss of rf drive power on the amplifier. A totally different effect occurs in each case. An operating amplitron from which rf drive is removed inherently generates a noisy oscillation, which because of the uncontrolled rf voltages will result in rf arcing, and resulting dc arcing. In the MPTS the dc arc will initiate a crowbar (circuitbreaker) cycle, which removes the dc voltage for 100 microseconds. When dc is reapplied, there is no rf voltage available to initiate conduction, and conditions will remain stable. Thus the amplitron, in conjunction with the crowbar circuitry, is inherently self protecting for loss of rf drive. In an operating klystron, however, no instability occurs when rf is removed, there is simply no rf output and the total beam power is dissipated in the collector. The effect of the large increase in collector dissipation was not evaluated but must be considered in detailed design either by thermal overdesign or by protective circuitry incorporation. The latter would likely be the better approach.
RkJQdWJsaXNoZXIy MTU5NjU0Mg==