FIGURE 67. - SCHEMATIC OF ITT GILFILLAN-BUILT RADAR SYSTEM obtained in a microwave configuration having no special provision for filtering or diminishing the harmonics. These would appear to be typical data and can be used as a base for assessing the probable performance of the proposed SSPS tube. The third kind of energy — energy occurring during turn-on and shut-down — is produced at such times because the voltage applied to the tube traverses values which are coincident with other electronic interacting modes of the tube. Since it is not physically possible to swing the voltage from its running value to zero in zero time, there will be a finite amount of time spent at a possible oscillating condition. This time is determined entirely by the transient properties of the de power source. The frequency of the oscillation will be approximately 15% below the running frequency. In pulsed radar tubes, the voltage passes rapidly through the oscillating value on every pulse and a short burst of energy results. When the rate of rise and fall of the voltage pulse is fast, no energy is produced. However, in practice, the 15% frequency separation makes it feasible to filter out whatever energy does occur. Under de operating conditions, the voltage, of course, always remains at the running value and no signal of this type is produced in normal steady operation. The last type of energy can be categorized as discrete spurious energy related to microwave resonances in the tube and its environment. The microwave properties of a typical CFA are shown in Figure 69, which indicates a well-matched, operating passband in the mid-region of frequency and mismatches or resonances in the outer regions of the band. Sharp resonances tend to peak up background noise energy and produce concentrations of it at certain discrete frequencies. A spectrum analyzer presentation of this noise (Figure 70) was taken on a pulsed tube having a platinum cathode similar to the one under consideration for the proposed tube. Precise measurements of this noise have been made recently on a number of pulsed tubes with the data taken between the spectral lines of the signal and mathematically converted to an equivalent CW value. Table 15 presents CFA spectrum noise measurements. The best values obtained are about 55 dB/MHz below the main signal and were taken near the limits of the instrumentation. There is some indication that the intrinsic limit in the CFA may be better than presently observed, but this has not yet been confirmed. Very sketchy data on CW operation are available. These consist of one test on a low-power amplifier and another on a high-power oscillator. The oscillator data are consistent with the pulsed measurements, while the low-level amplifier data are about 30 dB better. Further data are needed.
RkJQdWJsaXNoZXIy MTU5NjU0Mg==