Radio Broadcast (Nov 1926-Apr 1927)

Building the R. G* S» Inverse^Duplex A FRONT OF PANEL VIEW OF THE R. G. S. RECEIVER Constructional Data on a Fout'Tube Receiver so Made as to Qive Six'Tube Efficiency — Oper> ating Instructions, Emphasis Being Laid on the Correct Usage of the Antenna Tap Switch — Test Results Obtained with This Receiver 'HIS article is the third in a series describing for readers of Radio Broadcast the latest circuit developments in the new Inverse Duplex System. The present article gives the constructional details for the adaptation of these developments to a radio receiver— the new R.G.S. receiver. The parts selected for the R.G.S. were chosen only after many tests on many different makes of apparatus had been made, and overall performance, based on the particular requirements of these circuits, was always the deciding factor. Particular attention was devoted to the design of the audio channel. Too much emphasis cannot be placed on this phase of the circuit in view of the great demand for receivers delivering good audio quality. It is quite essential to have bass notes, but, at the same time, the high notes cannot be sacrificed in this sudden rush for the deep pitches. Many receivers have this limitation present so that, while they are resonant and mellow on music, they are indistinct for vocal reproduction. By pronouncing the letters in the alphabet, it will be noticed that many of them are merely the vowel "e" prefixed by some high-pitched sound produced by the lips or teeth. Such letters are b, c, d, g, p, t, v, z. W ithout excellent reproduction of the high audio frequencies, all of these letters would sound alike. Tone quality is only one of the several features on which the real performance of the new Inverse Duplex is built. By far the most important factors are equal r.f. amplification and uniform r.f. selectivity throughout the broadcast band. From descriptions in the preceding articles, the reader should be well aware that the r.f. filter feedback circuit is responsible for this innovation. Yet, the satisfactory performance of the radio circuit is not based only on the filter coil. The electrical constants of the r. f. filter coil were determined solely with a definite type of r.f. tuned transformer. The design of the filter coil and the design of the tuned transformer are correlated because the object of the feedback circuit is to compensate the inherent deficiencies in the luned transformers. To make the con struction easy, the simplest type of tuned transformer was decided upon before the final values were determined for the filter coil. So please do not adopt some other type of tuned transformer because it is the "ideal" type in some other circuit. The same coordinate design extends throughout the receiver, so all of the values given have been determined by considerable research. These values, while not critical, should be adhered to if the remarkable performance which this receiver is capable of, is to be expected. In a word, do not attempt substitution as. in this case, will be poor economy in the end. Having determined the circuits and the proper apparatus to use, the next step is to mount correctly the units so that no detrimental interaction occurs. Reference should here be made to Fig. i, which shows the top view of the base panel, on which is mounted the equipment. The relative positions of the various parts are indicated in the diagram. The R.G.S. receiver has been arranged for twocontrol tuning by combining the second and third tuning condensers on one grounded shaft. This allows the antenna tuning condenser, Ci, to be adjusted separately, which tends toward greater flexibility. Thus, any change in the antenna tap switch which necessitates a change in tuning of the associated tuning condenser, can be made without reference tothe double-tuning condenser. In order to allow for any variations in coils, condensers, wiring, or tubes, and still benefit by the excellent selectivity of the receiver, a small vernier variable condenser, C4, is placed across the detector tuning unit. This tuned detector A.F. Transformer 6-1 r.f. m Tuning gs coil m .1 FIG. I In this sketch the placement of all the parts entering into the construct clearly shown. The base, which is supplied with the kit, comes drilled may be readily mounted as shown circuit is not quite as sharp as the other two tuned coils, so the adjustment of the vernier, when located here, is not critical. Its best adjustment should be found, however, especially if maximum selectivity is desired. The adjustment of this vernier may be slightly different at the various wavelengths. One section of the physical arrangement will need explaining at once. It appears on the surface that all known engineering rules have been violated in the mounting of the second and third tuning coils. Under ordinary circumstances, coils so mounted would result in instability and oscillation, due to the third coil feeding back into the second. But this all depends on the polarity of the feedback. It is quite true that trouble would occur if these two coils were similarly mounted — so they are not. The third coil is mounted upside down with respect to the second coil. The grid end of the third coil is at the bottom near the baseboard, while the grid end of the second coil is at the top. This causes a slightly opposing feedback to occur, but due to the very loose coupling between the coils, this opposing feedback is only noticeable at the short waves. The stabilizing influence is needed when no shielding is employed as oscillation tends to take place by means of electrostatic feedbacks without the shields. Thus, mounting these two coils in the manner outlined, entirely dispenses with the need for r.f. shielding to obtain stability. The antenna coil is placed at right angles to the other two tuned coils so that no magnetic feedback is possible, and the remote location of it precludes any possibility of electrostatic feedback from the other coils. The antenna coil is slightly different in its construction. The primary winding is preferably located inside the filament end on a separate section of tubing, while on the other two coils, the primaries are wound on the same tubing as the secondaries, at the filament end. Details of the construction of these coils are repeated here in Fig. 2. The two radio-frequency tubes and the detector are located near their associated tuning units so that the grid leads may be kept short. The ganging of the two condensers thus up ion of the receiver is so that all the parts