Radio Broadcast (May 1928-Apr 1929)

CIRCUITS FOR PHOTO-ELECTRIC CELLS 33 Gauge 16 B & S "9 22 Resistance per Loop Mile 42.2 ohms 83.2 171. Capacity per Mile 0.062 mfd .062 .073 In cables of this type the inductance may be neglected, but General Radio also designs special artificial circuits to simulate loaded lines, in which case an appropriate amount of inductance is included in each loop mile, or open wire lines. The same issue of the Experimenter describes a 600,000-ohm potentiometer suitable for use in the grid circuit of an amplifier tube as a gain control. The circuit, which is familiar to most broadcast engineers, is shown in Fig. 7. Such a potentiometer requires a high resistance in order that it may not draw appreciable current from the preceding element. This particular design, known as the Type 452, covers 30 TU in 1 5 steps of 2 TU each. The August issue of the General Radio Company's bulletin contains a discussion of electrical filters, which are divided into four groups: (1) Low-pass filters which cut off all frequencies above a certain value; (2) High-pass filters which cut off all frequencies below a certain value: (3) Band-elimination filters which block all frequencies between certain limits; and (4) Bandpass filters which block all frequencies on either side of a section which is permitted to get through. Classes 1, 2, and 4 have been discussed previously in this department. Type 3 is new in this respect, and as it is not described in the General Radio article I might point out that an audio rejector of the sort shown in Fig. 8 is an elementary form of band elimination filter. In parallel with the speech coil of a loud speaker, for example, it may be used to smooth out a peak in the response. The effect of the resistance in series with the capacitive and inductive elements is to broaden the resonance curve and to limit the effect of the device. If the resistance is very low practically no current of the resonant frequency will get past the shunt circuit, but by adjusting the resistance one may balance the by-passing effect of the shunt circuit against the undesired peak in the instrument under treatment and get a flat over-all characteristic. Compare this use of the resistance element with its similar function in a line equalizer (Fig. 9) which is also a form of filter. Coming back from this digression to the General Radio discussion, we note there sketches of sections of high-pass and low-pass filters, here reproduced as Figs. 10 and 1 1, respectively. The Fig. 6 Input Fig. 7 Input Fig. 8 ~ A C Cell No. 1 Transformer Fig. 4 Transformer Fig. 5 PHOTO-ELECTRIC CELL CIRCUITS USED IN SOUND MOTION PICTURES types of filters regularly marketed by General Radio are of these classes, designed for impedances of 600 and 6000 ohms, and with cut-off frequencies of 500, 1000, and 2000 cycles. That is, one may order from stock a 6000-ohm filter of the high-pass type, for example, to cut off everything below 1000 cycles. It may be shown that, given the iterative impedance Z and the cut-off frequency F for which a filter is to be designed, the value of the elements is given for a high-pass filter section by 0.07958 Q = — — — farads 0.07958Z . L2 = henries while for a low-pass filter we have similarly 0.3 183 C>= — p£~~ farads 0.3183Z . Li= — — henries In each case F is in cycles per second, while Z is in ohms. The September issue of the Experimenter carries as the leading article, "Notes on Group Address Systems," by C. T. Burke. Since papers on public address technique have appeared in this department of Radio Broadcast Mr. Burke's discussion will not be abstracted here, but it 1 J' To Amplifier Fig. 9 Input Output Fig. 10 Input L, HC2 Output Fig. 11 L. P. Filler A GROUP OF INTERESTING FILTER CIRCUITS should be interesting as a general treatment for broadcast operators who have not devoted attention to' the subject. A schematic diagram is included. This issue also contains a description of the Type 426-A thermionic voltmeter, including a diagram of the bridge circuit employed with the vacuum tube. The range is 0-3 volts, and under the' usual conditions of use the calibration is maintained to within 0.5 per cent, for about 1000 hours, and is good over the whole audiofrequency range. Even at 20 kilocycles there is only a 2 per cent, error, and less than 3 percent, at 300 kilocycles. At broadcast and higher frequencies the calibration is no longer valid. The entire instrument, including the battery, is selfcontained. Another item in the September issue concerns the Type 532 Station Frequency Meter, the entire scale of which covers only 0.3 per cent, of the frequency of the designated station, the variable condenser being connected across a larger fixed capacity. This gives ten scale divisions per kilocycle. The resonant frequency is read by an ingenious method, which consists in connecting across the main condenser a small auxiliary capacity to shift the peak of the meter from one side of the station peak to the other. The frequency meter is adjusted until the indicating galvanometer does not change its reading when the button connecting this condenser is pressed. This gives a more accurate setting than an attempt to find the exact peak by the maximum galvanometer reading. The accuracy of this meter is certified as within 500 cycles with a temperature variation of not over 5 degrees Fahrenheit from the temperature specified in the calibration. The guarantee is for six months, after which the instrument must be recalibrated. The General Radio Experimenter is a very commendable publication, to my mind, and broadcast technicians should get their names upon its mailing list. It is sent free on application to the company's offices at Cambridge, Mass. It is a commercial publication, but the discussions are severely free from advertising blather and generally contain as much general theory on the subject as specific description of the General Radio Company product. A commercial publication which is scientifically and informatively written is better than a sensational medium whose ultimate commercial aims are less frankly revealed, and if you fail to extract a few dollars' worth of data from the Experimenter it is your fault.