ULTRA-HIGH-FREQ UENC Y OSCILLA TIONS 



289 



of a wire extending across a diameter of the wave guide, parallel to the 

 Hnes of electric force in an Hi wave, there is placed a small crystal of 

 boron. Connection to the crystal is made by fine platinum wires, 

 melted into two small globules on opposite sides of the crystal.* By 

 virtue of its small size and the fine leads connected to it, small amounts 

 of power dissipated in the resistance of the crystal will raise its tem- 

 perature materially, with a consequent large change in its resistance. 

 With a stopping condenser, an ohmmeter connected as shown in Fig. 6 

 serves to indicate the resistance of the crystal when absorbing high- 

 frequency power, and calibration curves showing resistance as a func- 

 tion of power absorption can be obtained with direct current. 



JUL 



W?^^/^/////^^////^/////////2/j/^/^^^^j//////^^?^^j^r^ , 





^^//^JJ^^/JJ^^/^/77 



& 



^^^^ 



"^?////n 



Fig. 6 — Apparatus arrangement for "Loaded" test of diode no. 24. 



Power output and efficiency data obtained during these measure- 

 ments are shown on Fig. 7. Power outputs of a few tenths of a watt 

 at efficiencies ranging from one to two tenths of a per cent are 

 obtainable. 



In consideration of the wave-length of the oscillations generated by 

 these two diodes, it will be recalled that they were designed nominally 

 for a wave-length of about 10 centimeters. For diode No. 24 the 

 wave-length was close to 10.6 cm. (2830 mc.) and for diode no. 37 it 

 was somewhat higher, about 11.55 cm. (2600 mc). This difference is 

 of the order to be expected from the difference in the dimensions of 

 the two tubes. 



While the wave-length should be fixed largely by the dimensions of 

 the coaxial resonant circuit built into the diode, it is to be expected 

 that it will be affected to a small extent by the applied voltage and by 



* These were developed by Mr. G. L. Pearson of the Bell Telephone Laboratories. 



