required at the center of the transmitting antenna. The module power required at the edge is 10 dB down from this value. If we now look for simplicity in the transmitting antenna construction, it is desirable to dissipate the heat by radiation, thus eliminating any additional cooling structure. The line marked T, gives the substrate temperature in degrees centigrade on the assumption that the emissivity is 1 and the entire module area of X2 is at the same temperature and radiates from both sides. At this point some rough estimates have to be made for the temperature drop from the solid-state device to the substrate. Since the GaAs MESFET has the highest reported efficiency for a microwave amplifier, we use this device in the further calculation. For the GaAs MESFET there is reason to estimate a 60-80°C drop from the FET channel to the substrate. For long life, we assume that the channel temperature should be limited to 120°C. These temperature considerations lead to a design around T„ = 50°C with a module power of = 100 W/X2. As may be seen in Figure 1, by constructing a vertical line intersection Pm at 100, the system power is 2600 MW and the antenna diameters are 1.3 km for the transmitter and 6.6 km for the receiver. The solar array area required for this design is 27.5 km2. There are several major disadvantages of this design using solid-state devices. At the current state-of-the-art, the devices are inherently low voltage (15-25 V) and low power (3-5 W). For a rough calculation we consider that the FET gives 4 W at 20 V. A 100-W module would thus require 25 FETs. To reduce the power supply current and, therefore, the conductor loss, we would power the devices in series, requiring a bus voltage of 500 V. Compared to the Reference System Design using klystrons at a bus voltage of 40,000 V the solid-state design is at a serious disadvantage. The 80-fold reduction in voltage requires an 80-fold increase in current, and a 6400-fold increase in I2R conductor loss or a 6400-fold increase in conductor cross-section area and mass to keep the losses constant. This would indicate an unacceptable increase in mass since the specific mass (kg/watt) for klystrons and solid-state devices are comparable. Attempts to relieve this mass disadvantage, by providing a high voltage bus for the main distribution and a shorter low voltage bus must take into account the added mass due to the extra power-conditioning (dc-dc conversion and regulation) required. A second major drawback to this design derives from the need to combine the output power from 25 devices. Even low-loss microwave power combiners using current technology show a loss of about 3/8 dB per combination. To combine 25 (=25) devices requires 5 combiners for a loss of 15/8 = 1.875 dB and the 1000-W modules transmit only 0.65 x 100 = 65 W. This loss applies only to the center high power density region since the 10 dB Gaussian taper reduces the combiner requirements away from the center. To avoid these disadvantages of solid-state devices as replacements for klystrons in the Reference System, it is proposed to use the RCA SMART (2) modules. Each microwave module has its own solar cells and radiates its microwave power through its own antenna. (See Figure 2.) For preliminary calculations the power density may be taken as uniform across the transmitting antenna diameter. The system parameters for the SMART design are shown in Figure 3. Since the Goubau Gaussian-taper design is not used, the transmission efficiency has been reduced from the Reference System value of 89% to 80% since preliminary calculations show that with the proper phasing of the transmitting antenna modules this efficiency may be obtained with the receiving antenna diameter as shown. It should be observed that for a transmitting antenna diameter of about 3 km, the distance between antennas is the Rayleigh distance, D2/2X. Therefore, for smaller
RkJQdWJsaXNoZXIy MTU5NjU0Mg==