Abstract :
With the single exception, it is thought, of the E.M.I. equipment based on the gold-leaf electroscope, every pattern of feed-through wattmeter for high frequencies must undergo empirical calibration at the working frequency. This is best done by passing power through the wattmeter to a calorimeter load terminating the wave guide. Balanced calorimeter outfits suited to wave guides of pipe form are here described which were designed primarily for this purpose. The good match provided by the tapered water-termination to the pipe, and the ease and sureness of the thermal measurement, have extended the application of the apparatus to laboratory investigations where no wattmeter calibration is involved. Thus, while the calorimeter outfit described in Section 3 has been used in the development and calibration of a number of feed-through monitors? designed for incorporation in working radar transmitters, it has also formed the termination in various transmitter investigations and in the routine testing of magnetrons for checking their power output. In any continuous-flow calorimetric measurement, great dvantages in ease and rapidity accrue from the adoption of a oalancing system. In the present system the power is found as he square of an ammeter reading, without measuring the rate of low of the water or its temperatures. The tapered load and the balanced-calorimeter device are here escribed in combination; but the tapered load may be used without any calorimeter, and the balancing scheme may be pplied to any form of continuous-flow calorimeter. In calorimetric measurements of power, if the leakage of lectric power is wholly avoided (as it is here), the only hidden rrors are associated with imperfect mixing in the water and with leakages of heat. The heat leakages cannot be avoided. In view of the use of the present apparatus as a primary standard for power measurement, adequacy of water mixing, and the effect of heat leakages, have been closely investigated. The correction- required for the leakages can readily be determined experimentally; but they are in fact arranged to be nearly self-compensating. Powers of the order of 500 W at 3 000 Mc/s, or 50 W at 10 000 Mc/s, can be measured with ease to within about 1%; it is thought with a certainty that cannot be questioned. The earlier apparatus (Section 3) was designed for frequencies near 3 000 Mc/s. The principles of the method, and some experimental checks, are described with specific reference to this apparatus. Subsequently an outfit of the same general character was produced for lower powers at frequencies near 10 000 Mc/s. This is described in Section 5, with only such further comments as the changes call for.
