Internal feedback and its effect on phase linearity in a forward wave crossed-field amplifier

Summary form only given, as follows. Two sources of internal feedback couple the input and output of crossed-field amplifiers (CFAs). Direct RF feedback occurs because the ends of the slow wave circuit radiate energy into the drift space connecting the output and input; the magnitude of this type of feedback may be measured at cold test. Electronic feedback, on the other hand, occurs only when the tube is operating, and is much harder to measure. It is due to the residual coherency retained by the beam after its passage through the drift space. In the forward wave case, this feedback can still induce significant currents at the beginning of the slow wave circuit that must be added vectorially, along with the direct RF feedback signal, to the input signal to produce an effective drive signal. This effective drive signal is therefore shifted in both amplitude and phase from the input signal, though the phase shift is generally more important than the amplitude shift in saturated amplifiers. The re-entrant current also contains certain random components that add to the input signal and contribute to CFA noise. As the input signal frequency of the amplifier is varied, the difference in electrical path length around the tube leads to a periodic variation of the total feedback signal relative to that of the input signal, resulting in a variation in phase of the effective drive signal, which in turn produces a periodic variation in phase of the output signal. This variation can have significant consequences for the system in which the CFA is used. The magnitude of this variation is very difficult to estimate other than by the use of a simulation code. We have applied our CFA simulation code, MASK, to this problem and have produced very good agreement with measurements of output phase versus frequency for a high power, forward wave S-band tube.