Radio Broadcast (May 1927-Apr 1928)

JANUARY, 1928 HOW THE ' 'SYNCHROPHASE1 1 SEVEN WAS DEVELOPED 233 grid-filament capacity within the tube be so nullified that the purchaser can use whatever tubes he may choose without fear of upsetting the tuned circuit balance obtained in the factory when the receiver was first tested. With many receivers the neutralizing condenser (or condensers) has to be readjusted whenever any changes in the tubes are made. The frequency response curve of the average tuned radio-frequency amplifier shows maximum response on some short wavelength (high frequency) with a falling characteristic as the wavelength is increased (frequency lowered). The system represented in Fig. 1 by the variable condenser Q and the two resistances, Ri and R2, was developed and incorporated to attain the two objectives mentioned at the beginning of this paragraph. The action of these units is twofold. First, they eliminate the effect of the gridfilament tube capacity upon the tuned circuit, particularly on the low settings of the tuning condensers, in such manner that tubes may be changed without affecting the original resonance setting. Second, they control the voltage being give sufficient sensitivity and selectivity. Furthermore, the sideband suppression characteristics of the radio-frequency system are known and the introduction of a variable constant, which would be encountered with a variable regenerative detector, would upset the balance between the radio and audio amplifiying systems. THE AUDIO CHANNEL THE audio amplifying system is closely associated with the radio-frequency system ; in fact, it must be, for the reason that the frequency characteristics of the audio-frequency amplifying transformers are governed by the sideband characteristics of the radio-frequency system. If the sideband suppression of the upper audio register is great in the radio-frequency amplifier, the audio system must possess a certain rising characteristic. The slope of the rise is governed by the degree of suppression in the radio-frequency amplifier. Hence the two systems are closely associated. Being familiar with the sideband suppression in the radio-frequency FIG. I Circuit diagram of the "Synchrophase" Seven fed into the grid-filament circuit of the amplifying tube, so that the radio-frequency voltage fed into the tube is practically uniform over the complete tuning scale and the frequency response curve of the tuned system is sufficiently flat. With this arrangement, and the inherent lack of regeneration in the system, a high degree of stability and amplification is afforded. The four stages provide ample selectivity and sensitivity and are'designed to possess sideband characteristics with minimized suppression above iooo cycles. This consideration is very important, and the presence of excessive regeneration would tend to nullify all the effects of scientific design. But the properties of the fieldless coil guard against this deterimental effect. With sufficient spacing between the inductances, the very small amount of external field, which cannot be eliminated completely, does no harm. Hence the regeneration present is uniform over the scale and is at no time sufficient to cause uncontrollable oscillation. The detector system is the grid leak-condenser arrangement, affording maximum signal sensitivity and intensity. The compensating system utilized in the radio-frequency stages, is also resorted to in the detector input circuit, thus permitting the use of any detector tube without unbalancing the tuning system. A non-regenerative detector was decided upon because the four stages of tuned radio-frequency amplification system, two audio stages are decided upon after a study of the amplifying powers of the system; these two stages possess sufficient magnifying power to afford a satisfactory output. The overall response curve of the audio system is shown in Fig. 2. The gain in transmission units is shown on the ordinate and the frequencies are shown on the abscissa. The curve is plotted on a logarithmic scale. Particular consideration was given to the feminine speaking voice, the frequencies of which are difficult to amplify, and to the overtones of the high-frequency producing musical instruments. The result is that the transformers were designed to function satisfactorily on audio frequencies above 8000 cycles. This is easier said than done. A great deal of work was entailed before suitable transformers were produced. In order to realize satisfactory amplification on the upper audio register it was important to reduce the distributed capacitance of the secondary winding. This was accomplished by the use of three layers of insulation between each layer of winding. The distributed capacity of the secondary winding is approximately 18 micromicrofarads. The importance of a low distributed capacity can be appreciated when one realizes that the higher it is, the more limned will be the frequency range of the amplifying unit. An example of vision, and a knowledge of the buying public's whims and fancies, is displayed in the inclusion of a device which permits tonal flexibility. The problem arose during the process of development when the sales staff mentioned the fact that the aural fancy of the listener-in was apt to vary over a wide range. Could not some device be incorporated which would permit variation of the tone of received speech or music, so as to satisfy the individual tastes of the multitude? Some fans prefer a preponderance of low tones, while others are not so anxious about these low frequencies. The engineers decided that the best location for such a unit would be in the audio amplifying system, but a continuously variable change in the physical structure of the audio-frequency transformers to produce different response was impractical. Hence the "tone color" unit, consisting of a number of fixed capacities which can be shunted across the secondary of the second-stage audio-frequency transformer to change its operating characteristic, was originated. The "tone color" is controlled by a knob on the front of the panel and, by its manipulation, the listener is able to adjust the tone to suit his own taste. The capacities in the "tone color" vary from 0.00008 mfd. downwards. THE PROCESS OF TESTING ENGINEERING and originality has made possible the manufacture of all the necessary equipment, exclusive of the cabinets, in the Grebe plant, at Richmond Hill, Long Island. The manufacture and testing of the tuned radio-frequency inductances is of especial interest. The winding is spaced yet the winding form is not grooved. This is made possible by means of a grooved slider which carries the wire as it is wound on the winding form. The grooves on the slider space the wire, and the turns are kept in place by means of a layer of clear lacquer which is sprayed upon the coil before assembly. Ingenuity in testing now manifests itself. The wire, as mentioned before, is "litz," and it is extremely important that all the turns remain intact. If a single strand is broken it will result in a steep rise in the radio-frequency resistance of the wire, with consequent increase in losses, and lower selectivity. The condition of the finished coils is tested on a d.c. bridge, accurate enough to show one broken strand. The total resistance of all the strands is balanced against a known resistance. One broken strand in the "litz" cable will deflect the meter in the bridge, in which case the coil is rejected. The satisfactory coils' radio-frequency resistances are then measured. The inductance value of a completed coil should be 310 microhenries. The condensers are matched on a capacity bridge, each one being individually tested against a standard. The condensers are then grouped according to their respective capacities. A control is arranged which shows a variation of 7 micro-microfarads for the complete scale, and extremely small variations are, therefore, detectable. By means of this control it is also possible to determine increased effective resistance of the condenser under test. The bridge is fed from Pi Sec imary Inductance 90 Henries ondary Inductance 500 Henries 400 1000 FREQUENCY FIG. 2 The overall response curve of the "Synchrophase" Seven audio channel