change coupled with the use of a different diode made from a different wafer resulted in less reflected power at the 50 milliwatt level (corresponding to the one milliwatt per square centimeter lower limit specified in the work statement! It was further found that very good efficiencies could be obtained at even lower power levels while operating into DC load resistance of 1000 ohms and that the reflected power remained low percentagewise. As curve 5 in Figure 3~1 indicates, seventy percent efficiency was achieved at the 10 milliwatt input level with a reflection of less than one percent. The efficiency at the 1 milliwatt level was still 45% with a reflected power of 5%. One of the practical problems in the use of the X/4 impedance transformer is that it makes the element too long to incorporate into the two- plane construction preferred because of its much more economical construction. This led to an approach in which the second filter section of the RXCV type element was converted into a matching transformer of lower impedance than the X/4 length of line. It is not possible to obtain as high a characteristic impedance in this section as in the X/4 length of line because of limitations imposed by the fragility of the inductive section of the filter. However, when attention is focused upon data that has been taken in the 500 milliwatt to 1.5 watt region where a great deal of the SSPS rectenna action is slated to occur, it is noted that a de load resistance of the order of 300-500 ohms seemed appropriate when used with a diode having a C^o capacitance of 1 picofarad. Matching into this requires only the increase of the characteristic impedance of the second filter section by a factor of V? rather than 2 and this is consistent with what can be accomplished in a physical structure. Consequently we constructed a rectenna element which incorporated a second stage filter with a characteristic impedance approximating V~2”x 120 ohms or 170 ohms as shown in Figure 3-3. A comparison of the performance of this rectenna element with that of the two filter section plus X/4 impedance transforming line and utilizing diode 40904 X WB3 for both sets of measurements is given in Table 3r2 for 1 watt of power input. The above data indicates that the two-section filter with a 170 ohm second stage is as efficient as the standard two section filter with an additional X/4 section of line. Moreover, this same diode was found earlier to have had an 88. 1% efficiency into a 450 ohm load, or 1.6% higher than reported in Figure 3-4. It is believed that the diode may have changed since another diode 40904 X WB4 was found to be more efficient than diode 40904 X WB3 in a new set of measurements whereas the reverse had been true when the diodes were checked several weeks previously. The conclusion is that the two section filter with the second section modified may be acceptable for operation of the 1 picofarad diodes at the 1 watt level into a resistive load of a nominal resistance of 400 ohms; thereby resulting in a better efficiency than with the 140 to 180 ohm load resistance normally used and in a design compatible with the two-plane construction.
RkJQdWJsaXNoZXIy MTU5NjU0Mg==