Radio broadcast .. (1922-30)

164 RADIO BROADCAST DECEMBER, 1925 on their way toward the plate. If the grid is negative it repels electrons and less plate current flows; if it is positive, it draws more electrons from the filament out into the space of the tube and the plate current increases. In this way the grid is essentially a controlling element. DETECTOR THEORY AND PRACTICE *~pHE theory of detection is complicated and ^ will not be described here. It is only necessary to say that 41 volts on the plate of the detector is about the correct value with modern highly pumped tubes; that the grid return should be connected to the positive side of the filament; that for grid condenser-leak detection, the proper values seem to be about .00025 mfd. capacity and two megohms, although other values may be used; that there is little use in using a C battery detector unless very powerful signals are to be I r Speaker FIG. 2 Signals from an antenna go through several electrical devices before they finally emerge from a loud speaker. This illustration shows the path of these signals. At the input and output of each amplifier the voltages and power levels differ, increasing as the signal approaches the loud speaker received, say in the second super-heterodyne detector. Often a detector that will not work on 45 or even 225 volts B battery will work very well indeed on 12 or thereabouts. If regeneration is not smooth, that i«, if advancing the tickler, or the condenser in capacity feedback systems, is accompanied with growls and low frequency clicking noises, the trouble lies in too much tickler, wrong grid leaks, or too much B battery. The tube should slide into operation without fuss, and if it does not, something is wrong. With low loss receivers, not much tickler is needed. The higher the resistance of the coil into which regeneration is being introduced, the more tickler will have to be used and the more erratic will be the operation. There is one point that may be mentioned here. It is a common statement that there is no necessity for low loss circuits in regenerative receivers since the addition of regeneration reduces the resistance of the circuit. Regeneration does reduce the effective resistance, making tuning sharper, and receiving more selective. If the receiver suddenly begins to oscillate after the regeneration has been set say when a crash of static comes along, or some loud signal, the Hill operator can look for a high resistance circuit in which the tuning is broad until much regeneration is added. Then it is time to read up on low loss circuits. The use of low resistance grid leaks, say one half megohm, will improve the quality of music received but on the other hand, low valued grid leaks will cause some loss in volume — which maybe made up in the audio amplifier. Various methods of obtaining regeneration in a detector circuit have been described (see RADIO BROADCAST for October) and all produce the same results. Increased signal strength, increased selectivity, and, if it is pushed too far, decreased quality. AMPLIFIERS: RADIO AND AUDIO ""THERE is little that one can do to a detector ^ tube or detector circuit beyond what has been mentioned above. When it comes to amplifiers, however, there is much to be said, and many false notions to be discussed. There are two kinds of amplifiers in the usual radio receiver, those which are working at very high frequencies, and those which work at low audible frequencies, and there is a league and a half of distance between them. In the first place there are two things to consider, voltage, and power amplification. These are two different things, and until quite recently little attention has been paid to the difference between them. Now that we have semi-power tubes appearing on the market from several tube manufacturers, we shall be able to plan our amplifiers with a little more engineering and a little less guess work. Fig. 2 is a diagrammatic method of showing a receiver with its component parts. We shall begin at the loud speaker and work up toward the antenna circuit. The speaker requires power — and there is a certain amount of power that is required by every good one to give a good, well-behaved sound. For example, the Western Electric 2i6-A tube, which until recently was the only semi-power tube available, has an output of .06 watts under the proper operating conditions, and if this is placed upon a good speaker, plenty of volume will result. Such volume will not be sufficient for a large auditorium, it will not be heard a mile or so up the street, nor will it drive any one out of the house — but who nowadays wants such volume? Let us say, then, that a good signal requires .06 watts and since this figure represents power, the last tube in the receiver should supply power. Now there is an expression, due to Van Der Bijl, which amplifier designers seem to have overlooked, that says that the power output of any tube will be as follows. (mu X input voltage)2 FIG. 5 Transformers are used to "match" impedances. In the case shown here, for maximum transfer of power from tube to load, the turns ratio of the transformer must be /Zp VzL FIG. 3 A two-stage audio amplifier and the voltages that must appear at various points along the circuit if the full output of the last tube is to be delivered to the loud speaker. If lower voltages are delivered the volume will be "down." If more than nine volts peak are applied to the grid of the last tube, overloading will occur and a cone type loud speaker will, in popular parlance sound, "awful" power = 8 X plate impedance Now, using this formula, let us figure out the maximum power obtainable from several tubes under the usual operating conditions, namely, 90 volts B battery, minus 4.5 volts C battery, and assuming that the input voltage peak is just equal to the C battery voltage. In other words we are working the tube up to the limit of distortionless amplification. Under these conditions the following table gives the power obtainable, 3-VOLT TUBE .0066 5-VOLT TUBE UX 1 12 .0135 .0184 Thus it is seen that none of the tubes ordinarily used will give sufficient output to operate a loud speaker at the desired level of .06 watts. The following table gives the powers obtainable from tubes under conditions of greater input and plate voltages. J-VOLT UX-I 12 2I6-A .0135 B-VOLTS 90 90 '35 '57-5 4-5 6.0 9 10.5 .058 .0328 .118 .185 .059 From this table it may be seen that sufficient power is not obtainable for satisfactory reception with a 5-volt tube until 135 volts are used on the plate and until 9 volts are placed upon the input to the last tube. Under the same conditions, the newer j-volt, one-half ampere filament tubes, such as the ux-H2, and similar tubes foi the same purpose, will deliver nearly twice as much power as is actually needed, and with 157.5 volts on the plate and 10.5 volts C bias will have an output that is still more favorable. These figures mean that it will not be necessary to crank up a receiver to the top notch to hear the average level of an orchestra; and to endure distortion, or to turn down the set when a player bangs down on his kettle drums unexpectedly, or when the orchestra rises to a maximum output level. In other words, a receiver properly operated with one of these semi-power tubes in the last — HI FIG. 4 The last tube in a receiver must deliver power to the load which is usually a loud speaker. If the impedances of the tube, Zp. and the load, ZL are alike, maximum power will be delivered stage may always be somewhere short of the overloading point, and the range in volume, without the distortion due to overloading, will be much greater. For practically the first time in radio broadcasting reception it is possible to avoid overloading distortion without going to the bother of installing power tubes and high voltage