The motion picture projectionist (Nov 1929-Oct 1930)

January, 1930 The Motion Picture Projectionist 33 Fig. 3. Transformer Coupling less amplification is that the total amplification must be constant. This is a brief statement which may need considerable explanation. There are two factors on which the constancy of amplification ordinarily depends: (1) the magnitude of the voltage changes which are to be amplified, and (2) the rapidity of these changes. Consider first the condition that the amplification shall be independent of the magnitude of the voltage change. Let N designate the total amplification from the grid of the first tube to the plate of the last tube of the amplifier. Changing the input voltage from — 5 to — 4 will result in a change of N volts at the output, and changing the input voltage from — 9 to zero must, if the amplification is constant with respect to magnitude, produce a change of 9N volts at the output. Likewise any one volt change of grid potential at the first tube must produce a change of N volts at the output terminals whether that change is from — 9 to — 8, — 5 to — 4, or — -1 to zero. It is, of course, impossible to make such a condition hold true except between certain limits, and I am taking the limits, for purpose of illustration, as — 9 and zero volts at the first grid. The amplifier fulfilling the conditions just described will then be satisfactory provided at no time the input voltage causes the first grid to swing more than 4.5 volts above or below the mean potential of — 4.5. Figure 4 shows the characteristics of a tube plotted in terms of plate current against grid voltage. Curve I is for constant plate voltage. Curves II and III show the plate currents with two values of resistance in the supply circuit. The plate voltage corresponding to any point on one of these curves may be found by subtracting the voltage drop (current time resistance) from the supply voltage. It is evident that curve I represents a condition under which the tube cannot be a voltage amplifier, since the plate voltage is constant. Referring to Curve III and finding the corresponding voltages at the plate as shown by Ilia, we find that the tube is giving a voltage amplification of about 5.5. The condition for constant amplification (i.e., amplification independent of the input voltage) is that the tube shall be used only over such a range of grid voltage that the characteristic, Curve III, or Ilia, is practically straight. Fortunately a slight distortion is permissible, because the ear accepts some distortion of this kind without judging the quality of reproduction Jt ?, 1 / 100 * \ ?» 1 j % ^o x~7 1 >0 t i \ 1? f.1 \ / I-" ZC0 f \ ( 60 1 gu / \ p. (. k \ ■ JIG Ui 14. V / \i / V id / \ ' So ,' / 4 7 f f> / // / *? 1 /■ ? /-i s o ^ *£' GRtD VOLTAGE Fig. 4. Plate current and plate voltage as affected by grid voltage; 100 volt supply to be impaired. It becomes a matter of judgment to determine the exact maximum range to be used. Referring to Curve III, Fig. 4, it will be noticed that the characteristic is nearly straight for values of plate current above 0.1 milliampere or — 9 volts grid. Increasing the range of grid voltage to greater negative values would only slightly increase the output of the tube and would very rapidly increase the distortion. A conservative and a liberal estimate of permissible range would probably not differ by more than 20%. In the actual design of an amplifier it is usually found easy to provide ample margin in all the tubes, except the last stage or perhaps two stages in which the voltage swings become large. The smaller the swing the less is the distortion in a tube due to nonlinearity of its characteristic; hence distortion of this kind is usually confined to final stages of the amplifier. In fixing the range of grid voltage, under which a tube is to be worked, we practically always specify that the grid shall never become positive with respect to the negative end of the filament. So long as the grid is always negative it receives no electrons and absorbs no power from the preceding tube, but the moment its potential becomes positive it constitutes a resistance load across the circuit, and since this load is on during only a part of the cycle, distortion results. Putting the matter differently, the preceding tube tries to push the grid positive, but the effect is in part neutralized by the electrons or negative charges picked up by the grid. On the other hand, no such opposing effect is encountered when the grid is pushed in the negative direction. The effect on the output wave shape is just as if Curve III in Fig. 4 bent sharply toward the horizontal to the right of the zero grid volts axis. Since the range of grid voltage is always from zero to a certain maximum negative value, means are always provided in properly designed amplifiers to give the grid a mean potential of half of this maximum negative voltage. The voltage to be amplified is then superimposed upon this average or bias voltage, now adding to and now subtracting from the bias voltage. In resistance-coupled amplifiers this proper bias is established by connecting the grid through a high resistance or grid leak to a point or suitable potential, usually a biasing battery. When a transformer is used, its secondary winding constitutes the conducting path through which the grid is held at the desired bias voltage. If it is known that a tube has a permissible range many times the actual voltage which it will have to handle, as is often the case with the first stage of an amplifier, it is only necessary to make sure that the bias is somewhat greater than the extreme input voltage swings, so that the grid will never become positive. This bias may be much less than half the extreme negative value which the tube characteristics might permit. The design of amplifiers is conducted with the help of curves as shown in Figs. 4 and 5, but testing is usually carried out by impressing on the input circuit a sine wave alternating voltage and studying the wave shape of the output voltage. (To be Continued) so — — , -s°/ / / 1 i J / « i 1 / y 1 V 1 f 1 V > / / 9 ' / ' / / V JJ £ / / / / h / / / / / / id / / / / / A 3 X1 / f / h * ^ »n / '\ / / / / / 1"° ^\ / / / / * / c / f Fig. 5. Characteristics of small vacuum tube