MULTICAVITY POWER KLYSTRON AMPLIFIERS

tection for the body current and the collector current. In most systems, the collector and the body operate at nearly the same potential. Any potential difference is usually only the difference in voltage drop across the various metering circuits. MULTICAVITY POWER KLYSTRON AM- PLIFIERS.— The simple, two-cavity power klystron amplifier is not capable of high-gain or high-power output, or suitable efficiency. However, with the addition of intermediate cavities and other physical modifications, the basic two-cavity klystron may be converted into a multicavity power klystron, capable of both high-gain and high-power output. In addition to the intermediate cavities, there are several physical differences between the basic two- cavity klystron and the multicavity klystron. The cathode of the multicavity power klystron must be larger to be capable of emitting large numbers of elec- trons. The shape of the cathode is usually concave, which aids in focusing the electron beam. The col- lector must also be larger to allow for greater heat dissipation. In a high-power klystron, the electron beam may strike the collector with sufficient energy to cause the emission of X-rays from the collector. Many klystrons have a lead shield around the col- lector as protection against X-rays. Most high-power klystrons are liquid-cooled and must be constructed to facilitate cooling the collector. Klystron amplifies have as many as seven cavities, including five intermediate cavities. The intermediate cavities improve the bunching process, resulting in increased efficiency. Adding more intermediate cavi- ties is roughly analogous to adding more stages to an IF amplifier; that is, the overall amplifier gain is in- creased and the overall bandwidth is reduced—if all the stages are tuned to the same frequency. The same effect occurs with klystron amplifier tuning. A given klystron amplifier tube will deliver high gain and narrow bandwidth if all the cavities are tuned to the same frequency-this is called synchron- ous tuning. If the cavities are tuned to slightly dif- ferent frequencies, the gain of the klystron amplifier will be reduced and the bandwidth may be appre- ciably increased—this is called asynchronous tuning. Most klystron amplifiers that feature relatively wide bandwidths are stagger-tuned. The klystron is not a perfectly linear amplifier; that is, the RF power output is not linearly related to the RF power input at all operating levels. In other words, the klystron amplifier will saturate, just as a triode amplifier will limit if the input signal becomes too large. In fact, if the RF input is increased to levels above saturation, the RF power output will actually decrease. To better understand the reason for this decrease, remember that electron bunches are formed by the action of the RF voltage across the input cavity gap. This RF voltage accelerates some electrons and slows down other electrons, resulting in the formation of bunches in the drift tube region. Obviously, this speeding up and slowing down effect is increased as the RF drive power is increased. The saturation point is reached when the bunches are perfectly formed at the instant they reach the out- put cavity gap. This results in the maximum power output condition. When the RF input is increased beyond this point, the bunches are perfectly formed before they reach the output gap; that is, they form too soon. By the time the bunches have reached the output gap, they tend to debunch because of the mutual repulsion of electrons and because the faster electrons have overtaken and passed the slower electrons. This causes the output power to decrease. FOCUSING KLYSTRON AMPLIFIERS.— One very important item that is required for high- power klystron amplifier operation is an axial magnetic field (a magnetic field parallel to the axis of the klystron). In klystron amplifiers, which are physi- cally long, it is difficult to keep the electron beam properly formed during its travel through the RF section. The mutual repulsion between electrons causes the beam to spread in a direction perpendicular to the axis of the tube. If this is allowed to occur, elec- trons will strike the drift tube and be collected there, rather than passing through the drift tube to the collector. 2-14