Radio broadcast .. (1922-30)

PEBRUARY, 1927 A. C. FOR FILAMENT LIGHTING 395 FIG. 12 nounced peak in the hum voltage at a point of .about half the normal filament voltage, and there is a very definite minimum at a voltage of about 15 per cent, below the normal voltage of 5 for this tube, and again a definite rise as the normal voltage is approached and exceeded. If now we change the operating conditions only by doubling the grid voltage, we obtain the curve shown in Fig. 9. Here we note that the hum peak has remained about the same, that the minimum point has risen to a considerable value, and that the upper maximum has increased about four times. If, instead of doubling the grid voltage, we leave the grid voltage at 4.5 volts and decrease the plate voltage to 90, we obtain a curve, shown in Fig. 10, similar to that for the nine-volt grid bias and 135-volt plate voltage. If we now reduce the plate voltage still further, 1067.5 volts (see Fig. 11), we notice a slight decrease in the lower peak, a further rise in the minimum portion, and a decided rise in the maximum portion. Going down to 45 volts we note in Fig. 12 that the lower peak has almost disappeared while the upper maximum has risen to a comparatively high value. If we now compare the hum curves of Figs. 8, 11, and 12, by drawing them together on one curve sheet, as shown in Fig. 13, we can at once see the general nature of the variation in the curves under the changing plate voltage condi- tions. With high plate voltage, the hum peak is predominant, and as the plate voltage is lowered, this hum peak decreases in amplitude while the upper maximum steadily increases. It may be observed in these curves that the peak of the hum curve always coincides with the point of maximum steepness in the filament voltage-plate cu re curves drawn with them. This appears to identify definitely this hum peak with a temperature variation cause. Since a given amount of filament voltage variation at the steepest point of this static plate-current curve produces the maximum change of plate current, it is quite reasonable to expect that, under dynamic conditions, the complete change from maximum to zero of the filament voltage would produce a periodic change in plate current whose frequency is double that of the filament exciting current frequency. And, since the great- est variation in plate current is produced at this filament voltage, it is obvious that the greatest at this point. We have, therefore, rather definitely identified the lower voltage peak with the temperature variation of the filament. This identification is still further strengthened by the fact that this hum peak cor- responds fairly well in amplitude with the slope of the filament voltage plate-current curve. The filament voltage of the 2OI-A type tube is the same as that of the 112, whose characteris- tics have just been shown. However, its fila- ment is made of thoriated tungsten designed for a o.25-ampere operating current, while the 112 tube has an oxide coated platinum filament de- signed for an operating current of 0.5 ampere. HUM CURVES IN FIG. 14 are plotted three hum curves of the * uv-2oi-A representing the two extremes and middle conditions as far as plate voltage is con- cerned. By an inspection of the curves it is easy to visualize very clearly the changes in hum characteristics with variations in plate voltage, and again indentify this hum with the filament temperature variation cause. It will also be noticed that the hum peaks are much higher than those obtained for the 112 type of tube. Similar curves for the 120 tube are given in Fig. 15. Proceeding now to the uv-iox) tube with a three-volt filament, taking only 60 milliamperes, we obtain curves such as shown in Fig. 16 and Fig. 17. With 90 volts on the plate and a nega- tive grid voltage of 4.5 (Fig. 16), the plate cur- rent characteristic is comparatively flat, while curiously enough the hum peak is very pro- nounced at the normal operating voltage of the filament. If we decrease the plate voltage to 45 and maintain the same grid voltage as before, we get an extremely flat plate-current character- istic, a very low hum peak, and a definite indica- tion of a minimum point at about 2.5 volts on the filament. See Fig. 17. Finally, we have two curves for the wo-12 type of tube. This tube has the lowest operating voltage of all, while the filament current is the same as for the 2OI-A, or 0.25 ampere. It has, however, an oxide-coated platinum filament as has the 112 tube, and operates at comparatively low temperatures. The first curve taken on this tube is shown in Fig. 18, with a plate voltage of 135, and 9 volts negative bias on the grid. The plate current curve is much steeper than any we have yet considered, while the hum peak is FIG. 15 about the same as for the 2OI-A. At a point near or beyond the normal filament voltage, the hum curve is still quite high and there is little indication of a minimum point. In Fig. 19 a curve is given with a plate voltage of 90 and a negative grid voltage of 4.5. A reduced hum peak is obtained, but in general the same hum and plate current characteristics are present. Curiously, there is a steep bulge in about the middle of the plate current characteristic that has not been noted on any of the other tubes under discussion. If we consider the grid and voltage ratings of the filament as indications of the thickness of the filaments, and consider that a thick filament will not fluctuate in temperature so much as a thin filament, we may compare filaments so far as their temperature varia- tion hum characteristics are concerned on a basis which we may call the "thermal inertia" of the filament itself. This is a time tempera- ture factor which is determined by the thermal characteristics of the filament. The cubical contents of the filament, its specific heat, its radiation constant, the conduction effects through lead-in wires, and some other factors, determine the value of this thermal inertia factor for any given filament. With a given material it is, of course, highest for a cylindrical type of filament, as compared for example, with a flat strip type of filament. It is greater for a mate- rial with high specific heat than it is for one with a lower value. It is greater for a filament hav- ing a surface with low radiation constant than it is for one having high radiation properties. It is greater for a filament of low temperature than it is for one of high temperature, because the radiation factor increases rapidly with high temperatures. If, therefore, we classify the various tubes we have thus far studied with reference to their filaments and select two ex- tremes and a medium, we will have the 112 tube at the one extreme with highest thermal inertia, the 2OI-A with medium thermal inertia, and the 199 with the smallest thermal inertia. THERMAL INERTIA IF WE plot the hum cutves of these three tubes 1 for the same plate and grid voltages, that is, 90 volts plate and negative 4.5 grid volts, as shown in Fig. 20, we may compare their hum characteristics directly as a function of the ther- ,'Et WX- II. Ep-l3SV Ei.4V If. US Amp.- WX-12. E t -»IV. E«—4.5T. FIG. l6 Ef FIG. F:G. 18 234 Ei FIG. IQ