Radio Broadcast (May 1928-Apr 1929)

Operating Data on a NEW AUTOMATIC VOLUME CONTROL SYSTEM By CHARLES WILLIAMSON Department of Physics, Carnegie Institute of Technology 4LL of us have been annoyed at some [\ time or other by a terrific burst of _/T^_ sound from the loud speaker when, in the course of tuning, we have happened to come across a powerful local station. Sudden loud speaker overloading may also occur with certain types of broadcasting, such as organ music, which exhibits an unusually wide range of volume. In many cases, the voice of the announcer is received at a much higher volume level than that of the music. But the worst case of unexpected sound peaks arises when fistenin^ to a distant station that is fading badly. When it comes back, it usually does so with a roar; and static crashes are at their loudest because the sensitivity control has been advanced fully in order to make the program audible when the station fades. An Automatic Control LAST August the writer designed a simple and inexpensive device to cure these troubles. It has been tested on several receivers and in each case the difficulty was corrected almost perfectly. The addition of the writer's device to any d.c.-operated receiver does not require that the circuit of the set be altered in any way. (When using this device in connection with an a.c.-tube receiver the results may not be entirely satisfactory as an increase in hum may result, particularly when the set uses 226-type tubes — Editor.) The device is connected across the loud speaker terminals and is operated by the a.c. voltages appearing across the loud speaker. It functions by reducing the plate voltage on the r.f. tubes automatically as the volume from the loud speaker rises. Since the device described in this article functions due to the changes in voltage across the loud speaker, it will have the effect of contracting the volume range — in other words it does automatically at the receiving end what the monitoring operator does (or should do) at the broadcasting end. Serious distortion of the signal currents will occur only if the device operates with a time lag so short as to be comparable with one fourth of a cycle at the lowest frequency likely to be handled by the audio system. An examination of the circuit (Fig. 1 diagram a) will show that the time constants of the choke-condenser and resistancecondenser combinations are of the order of 0.1 second or greater. Dr. Charles Heinroth, the eminent organist, assures me that he would not regard a systematic contraction of the dynamic range as undesirable distortion of his programs. In fact, he thinks it would be better than the hit-or-miss monitoring sometimes met with! Hence I think we can regard this type of automatic volume control as distorting only in a formal sense. Not even a musician can detect it in the receiver's output unless he has an unmonitored output with which to compare it. Manual adjustment of the plate voltage of an r.f. tube will show that large changes may be made over a certain region without noticeably affecting loud speaker volume. If, as is often the case, the amplification of the r.f. stage under control is partly regenerative, it will fall off rapidly at first, as oscillation is left behind. After this there will be less change, since the a.c. plate resistance and the mutual conductance of the tube will change slowly with diminishing plate voltage over a certain region. Later, however, they both begin to change faster, and the amplification of the r.f. circuit is reduced rapidly. Thus, in a receiver of this type, this automatic volume control has the advantage that a loud signal throws the receiver out of oscillation instead of into oscillation. List of Apparatus THE parts required for the construction of the volume control described by the writer follow: Ri One high-resistance unit (See Table I for proper value) ; B-2 One rheostat or filament-ballast unit; Ci One fixed condenser, 200-volt, 4-mfd.; Li One choke coil, 30-henry; One tube socket, ux-type; One power tube, 17iA-type. The circuit diagram of the automatic volume control device is shown in diagram a of Fig. 1. This circuit also shows the method of connecting the volume control to a receiver having an output transformer. Diagram b shows the input of the device connected to a receiver having a choke-condenser output circuit. It will be noted that in each of these cases the loud speaker return must be made to the negative 45-volt C-bias terminal rather than to the negative A wire as usual. Diagram c shows the device connected to a receiver using a push-pull output circuit. Two condensers, C2 and C3, are required, as shown. In all of the above cases, the control tube may be operated from the same A, B, and C supply as the rest of the receiver, and its filament may be heated by a.c. if desired. Also, with all types of receivers the r.f. tubes ob tain their plate potential from the "B+B.F." terminal of the volume control unit. The high-resistance unit, Ri, should have a value of 50,000 ohms if a plate potential of 180 volts is available, and if there is a negative bias of 4| volts on the grid of the first r.f. tube. For lower values, either of B or grid bias potentials, the following table shows the proper size resistor to use: R.F. Grid Bias (Volts) -4§ -3 0 Table I Plate Voltage 180 160 140 50,000 40,000 30,0001 n 40,000 30,000 20,000 }■ fjl 30,000 20,000 10,000 ) unms RECEIVER AUT0MATIC-V0LUME-C0NTR0L Loud , . Speaker Fig. 1 — Diagram A shows the automatic volume control connected with the output transformer of a standard receiver. Diagrams B and C show the input of the device connected with other types of receivers The values of the resistor Ri are not especially critical; they are determined by the restriction that the first r.f. tube should not receive more than 90 volts of plate potential. If the r.f. B-plus lead supplies two tubes, the resistances in the above table should be divided by two. If a variable-resistance unit is used in place of the fixed resistor, it may be set to the best value by placing a high-resistance voltmeter (1000 ohms per volt) across the plate and filament terminals of the first r.f. tube, and adjusting the knob of the variable resistor until the meter reads 90 volts (or whatever other voltage is normally placed on the r.f. tubes). This must be done with the receiver fully turned on, but not tuned into any station. The choke coil, Li, used in this device, if not bought as a unit, may be made up of almost anything at hand, since, if only one r.f. tube is controlled, it need not handle more than 5 milliamperes at 90 volts. If a speaker filter is available, it can be used in place of Ci and Li. With the receiver in operation, there is nothing to suggest the presence of the automatic volume control, except a gratifying absence of speaker overloading. The original volume control on the set, whatever its type, is not interfered with in any way. It might be supposed that the volumerange of the music would be brought to a dead level; such is not the case, however; the range is merely reduced to an extent that the audio system can handle without noticeable distortion. If desired, the amount of such compensation can be regulated by turning down the filament of the control tube; and, of course, the device can be cut out entirely by turning it off. As to the sensitivity of the set, this is in no way impaired; for the full 90-volt potential is available for the r.f. tubes until a signal begins to come in; and it is not materially reduced until the signal becomes loud. Both calculations and trial indicate that a 112aor 120-type tube may be used for the control circuit instead of a 17lA-type tube with some loss of efficiency. In these cases the negative grid bias for the control tube should be 9 and 22| volts, respectively. Of course, the r.f. tube or tubes under control may be of any type whatever except the type 226, which hums badly if its plate voltage is changed. • march, 1929 page 299