Radio broadcast .. (1922-30)

Reducing the Guesswork in Tuning 243 ing radio frequency ratios has been described in Circular 74, page 103. The method has been especially developed by Prof. H. Abraham whose apparatus is known as the multivibrator. The multivibrator has not been set up in the Bureau of Standards laboratory, but has been in use in the Naval Radio Research Laboratory located at the Bureau of Standards. Comparisons of this multivibrator and the other methods have been carefully made throughout the range from 18 to 3600 kilocycles. In method C, the frequency of a tuning fork is carefully measured with a chronograph, and ratios of this to frequencies in the radio range are measured by the use of the cathode-ray oscillograph. This device is a special kind of vacuum tube in which a beam of cathode rays produces a spot of light on a fluorescent screen placed in the end of the tube. When the beam is deflected the spot moves across the fluorescent screen. Provision is made for deflecting the beam in two directions at right angles to each other by two pairs of small condenser plates to which controlled voltages may be applied. When two alternating voltages are applied which have frequencies related by a simple ratio, the spot of light traces out a Lissajous figure on the fluorescent screen. The shape of the Lissajous figure tells the ratio; the examples in Fig. 2 show this. The two pairs of deflecting plates are connected in parallel with the condensers of two independent generators of currents of audio or radio frequency. In the first step of the standardization process, the "low-frequency" generator, Fig. 3, is an electron tube generator of approximately i kilocycle, the frequency of which is determined by the tuning fork. The "highfrequency" generator is an electron-tube generator having a frequency anywhere from 1 1 to 22 times this, the frequency being varied until successive Lissajous figures appear on the screen. The frequency ratio thus measured is extremely accurate, being in all cases closer than the accuracy of setting of the standard wavemeter, which is tuned to the frequency of the "high-frequency "genera tor. The method was largely developed by Mr. R. T. Cox. Its use will be described in a separate publication entitled, " Primary Radio Frequency Standardization by Use of the Cathode-Ray Oscillograph," by Misses F. Kenyon and G. Hazen. In method D, a direct measurement of the wavelength of short waves on wires is made by coupling a short-wave generating set (Fig. 4) to o GtNtKATOK I . TOH JJ.OOOkt RC.CE.IVING 3CT FOfl 33,000 kc GCNCRATOR I.FOR 340 TO I6.5OOK.C FIG. 3 Showing connections of generators whose frequencies are compared by the cathode-ray oscillograph a loop of wire connected to one end of two long parallel wires approximately 4 centimeters apart. A short wire at right angles to the parallel wires, and containing a thermo-element is moved along the parallel wires and the positions of successive maxima of current are noted. The distances between these positions are each a half wavelength. They are measured with a steel tape, and agree within a few hundredths of one per cent, for the wavelengths used, 9 and 16 meters. These wavelengths correspond to frequencies of approximately 33,000 and 19,000 kilocycles respectively. These frequencies are not so high as to be beyond practical application in radio telegraphy and telephony. Experiments at the Bureau have shown how to produce them and to use them for communication as well as for measurements. The method of harmonics has been used to step down from the very high frequencies thus produced and measured to radio frequencies in the whole range down to 340 kilocycles. This is done by placing between generators I and II (Fig. 5) a receiving set tuned to the frequency of generator I which produces the current in the parallel wires. Generator II is an auxiliary, the frequency of which is varied until one of its harmonics approaches the frequency of generator I. The frequency of Generator II is adjusted until the beat note heard in the receiving set becomes zero. The wavemeter is meanwhile tuned to the frequency of the generator II. The frequency for that setting of the wavemeter is then the frequency obtained from the wavelength measured on the parallel wires, divided by the number of the