Radio broadcast .. (1922-30)

Further Notes on the Inverse Duplex Pertinent Data Respecting the Audio Channel — Why Combining Two Efficient Transformers Will Cause Distortion, and How It Is Eliminated in the Inverse Duplex Circuit —Choosing Correct Transformer Ratios — The Prevention of Overloading By DAVID GRIMES IN THE early days of broadcasting, the ques- tion of quantity rather than quality com- pletely dominated ourdesiresof achievement. The mere novelty of receiving anything at all without the aid of intervening wires was suffi- cient to offset completely the fact that those weird sounds emanating from crude loud speak- ers resembled only in a very minor way the original sounds at the studio. Standard trans- former coupling was universally used, but we would blush to glance at the jagged curves of some of these embryonic audio coupling units! Transformer coupling certainly delivered the volume, and the imperfections in most horn type loud speakers did not disclose the quality limita- tions. As broadcasting stations improved, and the cone speaker came into use, the trans- formers were placed aside, and almost every conceivable resistance and impedance coupling arrangement came into existence. Most of these combinations were quite satisfactory from an audio quality standpoint but were consider- ably lacking in volume output. A three-stage resistance- or impedance-coupled amplifier was usually required to equal even a low-ratio two- stage transformer-coupled amplifier. The resist- ance and ordinary impedance amplifier also appeared to choke up when required to deliver a great deal of volume so that their high quality was confined to rather modest output. This last fact caused the return of the trans- former, but the new type transformer coupling was a vast improvement over its predecessor. It brought in the bass notes as well as the high tones, and it certainly delivered the volume up to the tube limit without itself choking up. But a rather peculiar thing was noted in the extensive tests conducted by the author on the improved modern transformer. If one stage of transformer amplification only were to be used, the quality tf^^^^^^m was all that could be desired, and the actual results checked with the theory. An amplifica- tion curve of such a high-grade one-stage transformer coupling is shown at "A" in Fig. i. When two such stages are then connected in tandem, or cascade, theory would give curve " B," which would be very desirable. Actual results, however, were more like those presented in curve "C." The transformers themselves were obviously not to blame—it was a circuit trouble. The distortion and tendency toward "peanut whistling," in- dicated by the peaks in "C," were found to be due to audio regeneration. The circuit in Fig. 2, consisting of three stages of audio amplification, the first and third transformer-coupled, with an inter- mediate resistance-coupled stage, overcame all of the distortion shown in curve "C", Fig. i, and was very stable in operation. This system is employed in the latest Inverse Duplex circuit. In addition to its stabilizing influence, the resistance-coupled stage was made to deliver real amplification by employing a 25,ooo-ohm resistance in the plate circuit, and operating the tube on an impressed potential of 135 volts. The 25,ooo-ohm value is most efficient as this practically equals the internal impedance of the tube and permits the plate to receive a full 70 volts, which is quite satisfactory for the plate of a first audio tube. The choking of resistance stages before full volume is reached is not present here as the last transformer-coupled audio stage is called upon to deliver the real volume. This audio circuit now produces excellent qual- ity even under super-selective radio circuit conditions, if good quality is delivered at its ALTHOUGH this is the second of a series of articles on the new Grimes Inverse Duplex System, the first of which was printed in the January RADIO BROAD- CAST, the data contained herein are complete and the reader does not have to de- pend altogether upon the January article. It is recommended, however, that those readers who intend experimenting with this Inverse Duplex System, or who wish to build a receiver from the published details, should read the whole series of articles, the third of which will appear in an early issue. In essence, the following is the information contained in last month's paper: The receiver described ts a four- tube set in which the first tube acts in the dual capacity of first r.f. stage and second audio stage (resistance-coupled); the second tube acts as the second r.f. stage and the first audio,stage (transformer); the third tube is the detector stage, and the fourth tube is a power audio stage. Tuned radio frequency amplification is employed. Amplification and selectivity are excellent, and equal for all frequencies, a special filter arrangement providing for this. Increasing the amplification on the longer waves, has not, as one might suppose, the effect of broadening the tuning. This is one of the unique features of the new Inverse Duplex Circuit. Hand capacity effects have been obliterated. In some preliminary remarks concerning the audio amplifier, the author stated last month that the transformers should have a rising characteristic, i.e., they should exaggerate the higher tones slightly, to compensate the slight cutting of side bands in the radio frequency amplifier. — THE EDITOR. input. Eliminating the relatively few poor quality broadcasting stations from the discus- sion, still another problem presents itself before good quality reproduction can be obtained from strong local broadcasting stations. This prob- lem arises from the overloading or choking of the detector tube, and the omission of the low bass notes resulting therefrom. A perfect audio amplifier would utterly fail to register the bass notes if the detector grid was being choked by too strong energy from the local station. This has led many engineers to abandon the sensitive grid leak method of detection for the less efficient plate rectification system. But the real answer is found by employing three audio stages for the desired volume without being compelled to force the detector beyond its limit, and then, on local reception, to provide means for reducing the antenna input so that detector choking will not take place. A tapped antenna primary winding arranged according to Fig. 3 has been found to be satisfactory for this means. The antenna taps are taken off in a geometric manner so that each successive tap doubles the number of primary turns. In this way, not only does this switch act to reduce the input on local stations, but it also tunes the antenna circuit somewhat to the benefit of distant reception. Incidentally, by reducing the antenna coupling, its resistance loss is partially removed from the set and the selectivity is proportionately in- creased. This antenna switch performs three functions, then, and its proper adjustment is essential for best results. It must be remem- bered that a change in the position of the an- tenna tap switch necessitates a re-tuning of the first or left-hand tuning condenser. The tuning condenser will drop below the proper settings on the other two condensers upon an increase in the number of antenna primary ._ turns. AUDIO TRANSFORMER RATIOS THE next discussion in order pertains to proper audio transformer ratios for the several stages in the Inverse Duplex Circuit, as there appears to be a great deal of misinformation on this particular subject. The average radio experimenter uses transformer ratios in a rather haphazard manner, tending usually to place the highest ratio right after the detector. A few tests showed this to be abso* lutely wrong, as was pointed out by Julian Kay in the November, 1924, RADIO BROADCAST. The low ratio should be placed im- mediately after the detector and the high ratio in the last stage, especially when using the new