Radio Broadcast (May 1929-Apr 1930)

„ 4 5 3 < Eg = -i E§ = tg--z Co ( SCREEN-GRID PENTODE PI Es — ii ii ii >0 Es E> 10 2-5 150 200 will deliver about 2 00 0 milliwatts with a plate current of about 40 milliamperes at 250 volts, and requires a grid voltage (a.c.) of only 12 volts peak. This may be compared to a 245type tube which delivers 1600 milliwatts on approximately the same amount of d.c. power and with a grid voltage of 50 volts. TABLE II Fig. 5 250 300 Name Ef If Bp Mu Gin Ep Es Ec It Po Is Cossar 230 2 0.3 20 X 10' 40 2000 180 120 9 14 1.6 415 4 0.15 20 40 2000 180 120 9 14 1.6 Marconi pt 240 2 0.40 55 90 1650 150 150 9 16 500 6.0 pt 625 6 0.25 1850 250 200 15 26.5 2000 7.0 Milliard pm 24 4 0.15 28.6 65 2300 150 150 12 12 500 3.0 pm 24a 4 0.275 1550 300 200 21 18 2000 5.0 pm 22 2 0.3 62.5 80 1300 150 150 10 13 350 3.5 Six Sixty ss 230 pp 2 0.3 64 80 1250 150 150 10 13 350 3.5 415 4 0.15 27 60 2200 150 150 12 12 500 3.0 4pen 4 0.275 1550 300 200 21 18 2000 5.0 Mazda 425 4 0.25 2000 150 150 12 18 750 5.0 Philips C443 4 0.25 40 60 1500 300 200 15 28 PE 7 2.5 1.75 40 80 2000 250 250 12 40 2200 10.0 Ef — filament voltage Gm — mutual c onductance Ii — filament current Rp — plate resistance Mu — amplification factor Is — current to high-voltage grid Ep — plate voltage Es — high-grid voltage Ec — control-grid bias Ip — plate current Po — power output (milliwatts) Characteristic Curves Plate current-plate voltage characteristics of the Arcturus pe-7 tube are shown in Fig. 4. There is still some secondary emission at low plate voltages. The effect of varying the potential of the cathode grid is shown in Fig. 2. At high negative voltages on this grid, all the electrons are speeded back to the plate and there is no secondary emission. While it is probably not correct to use the usual method of laying out load lines on the Ep-Ip curves to determine the proper load resistance and the second harmonic distortion, some idea of the respective values can be discovered by so doing. These data are presented here with the knowledge that they may mean very little indeed. Thus it may be calculated that the Arcturus experimental tube with an internal resistance of about 40,000 ohms will work best into a load resistance of from 40Q0 to 8000 ohms. Within these two values the second harmonics (according to usual methods of calculation) will be less than 5 per cent., the usual criterion for distortionless amplification. At the same time the power output does not increase appreciably for values of load resistance in excess of 8000 ohms. (Fig. 3). It is probable that the third harmonics are the worst offenders in the pentode and at the present time there seems to be no generally accepted and easily worked method by which the percentage of third harmonics can be calculated from the characteristic curves. Use of the Pentode In Europe the pentode has been worked with magnetic loud speakers whose wellknown impedance characteristics are anything but straight flat lines. As a rule the impedance of these loud speakers increases rapidly with frequency. Since greater and greater distortion is the result -40 (Note change in grid current scale below abscissa.). Fig. 6 TABLE III Filament volts Filament amperes Control grid bias Space-charge grid volts Screen-grid volts Plate volts Mutual conductance Plate resistance Amplification factor Plate current Screen current Space charge current Maximum amplfication (into 100,000 ohms) New CeCo Screen-Grid Tube 2.5 1.75 (— ) 1-5 (+) 10 (+) 180 250 2300 250,000 575 3.9 ma 1.5 ma 5.0 ma 164 Present Screen-Grid Tube 2.5 1.75 (— ) 1.5 (+) 75 180 1050 400,000 400 4.0 ma 84 of using a highresistance load with the pentode, the fidelity of response is not very good. The high frequencies develop very high voltages across the load and within the tube, and are reproduced all out of proportion to the low tones,. In this country the magnetic loud speaker is practically on the shelf in favor of the electrodynamic type which has a flatter impedance characteristic. Since the pentode should be worked into a load lower in resistance than itself, it can be coupled to an electrodynamic loud speaker through the same transformer which usually couples it to a 4000-ohm tube. In the Laboratory the fidelity from such a tube and a Peerless loud speaker compared very favorably with that obtainable from a 245-type tube, and with a given grid voltage input the output was some 15 db higher. In other words, feeding 12 volts into a 245-type tube produced about 100 milliwatts but the same voltage fed into an Arcturus pentode produced about 2000 milliwatts. This is a very appreciable difference in volume. Practical Applications One of the important applications of this tube may be in the detector socket as a true power detector. Thus it may be used with the loud speaker in its plate circuit. However, whether it will make a good power detector tube has not been determined by experiment. A glance at its grid voltage-grid current curve (Fig. 6) indicates that it will make a good grid-circuit detector. The problem then becomes one of obtaining sufficient power from it. Development of more efficient loud speakers will make it possible to eliminate the audio-frequency amplifier entirely and to use only the pentode as a power detector working directly into the loud speaker. It has been suggested that the tube in this capacity might serve in automobile radio receivers and in other places where the space limitations are severe. The fidelity obtainable from a small loud speaker under the best conditions cannot be extremely good, and so the use of a small magnetic or electrodynamic loud speaker in connection with a pentode detector supplying perhaps 500 milliwatts may be an important application. In this particular tube, which may not (Continued on page 293) • RADIO BROADCAST FOR MARCH • • 255