Radio Broadcast (May 1928-Apr 1929)

206 RADIO BROADCAST ADVERTISER carry you safely to all "Front 99 page events With a new, wideawake Cunningham Radio Tube in every socket of your set you are "among those present" whenever and wherever things happen. With these faithful sentinels on duty, you are reliably radio-informed. Look for the monogram on the top of each tube and insist on them by name. E. T. Cunningham, Inc. NEW YORK CHICAGO SAN FRANCISCO No. 254 Radio Broadcast Laboratory Information Sheet January, 1929 A. C. Tubes EFFECT OF FILAMENT VOLTAGE TT IS becoming increasingly common to find manu1 facturers designing the filament windings on power transformers to supply voltages somewhat less than those rated for use with 226 and 227-type a.c. tubes. One parts manufacturer is marketing a filament transformer designed to supply 2.25 volts to the filament of a 227-type tube, although the rated voltage of this tube is 2.5 volts. A study of the circuit diagrams of manufactured receivers published in Radio Broadcast will bring to light other cases where a.c. tubes are supplied with somewhat lower than rated voltage. The life of a vacuum tube depends very much upon the filament voltage with which it is supplied, and frequently a very small increase in voltage above the rated value will cause a considerable shortening in the life of the tube. With a.c. tubes this problem has assumed especial importance, for these tubes are subjected to variations in filament voltage in accordance with any fluctuations of the line voltage. If the line voltage becomes somewhat higher than that value at which the set is designed to operate, the various tubes receive excessive filament voltage and their life is shortened to a marked extent. It is for this reason that manufacturers have designed the power transformer to deliver somewhat lower than rated voltage to the tubes so that even if the line voltage rises above normal the tube filaments will not be overloaded. A.C. tubes, types 226 and 227, will give entirely satisfactory operation at less than the rated voltage. The table on this sheet, obtained from figures in the Cunningham Tube Data Book, gives the characteristics of the 226-type tube with a filament voltage of 1.3 volts and 1.5 volts, the latter value being that at which the tube is rated. The slight increase in plate resistance and decrease in mutual conductance which results when the tube is operated at 1.3 volts is not sufficient to affect its operating characteristics. The 227-type tube saturates at about 1.9 volts on the filament and, therefore, it also may be operated at somewhat less than its rated voltage with satisfactory results. TUBE FILAMENT VOLTAGE PLATE IMPEDANCE MUTUAL AMPLIFICONDUC CATION TANCE FACTOR 226 226 1.3 10,000 1 . 5 9,000 750 830 8.3 8.3 No. 255 Radio Broadcast Laboratory Information Sheet Band-Pass Circuits January, 1929 WIDTH OF BAND TDAND-PASS filters, as used in radio receivers, consist of an arrangement of coils and condensers which produce a resonance curve of a form approximating that illustrated in the drawing on this sheet. It is possible to design a circuit to have a band-pass characteristic by the use of two separate tuned circuits, each tuned to exactly the same frequency and coupled. The coupling may be produced by condensers, by a separate coil, or by simply placing the coils of the tuned circuits in such relation that there is some coupling between them. One of the most important characteristics of a band-pass circuit is the distance between the two peaks in the curve, marked <*>i and J. H. Morecroft in Principles of Radio Communication gives some formulas for coupled circuits. If two circuits are coupled inductively, then the width in kilocycles of the bandj between <0i and is equal to the resonant frequency of either circuit alone multiplied by the percentage coefficient of coupling, k, between them. For example, we might take two coils and two condensers, arrange them in the form of two tuned circuits adjusted, say, to 1000 kilocycles. When there is 1 per cent, coupling between them then the width of the band will be equal to band with=(0i x k = 1000 x 0.01 = 10 kc. The width of the band is, therefore, 10 kilocycles. It should be noted that the band width is directly a function of k>i (or o>> since they are both tuned to the same frequency). Therefore, if the percentage coupling remains constant then the width of the band at 500 kc. is 5 kilocycles and at 1500 kc. is 15 kilocycles. The fact that the width (U2 of the band varies over the broad cast band in a ratio of 3 to 1 (5 kc. to 15 kc.) is a disadvantage, it being desirable, of course, that the width of the band should be constant over the entire broadcast range. If the circuits were capacitatively coupled the characteristic would be opposite to that when inductive coupling is used, i.e., at 1500 kc. the band width would be 5 kc, at 1000 kc. the width would be 10 kc, and at 500 kc. the band width would be 15 kc. No. 256 Radio Broadcast Laboratory Information Sheet January, 1929 Power Output HOW MUCH IS REQUIRED? T_TOW much available power in the output tube of a radio receiver does one need for ordinary home reception when using a standard loud speaker? This is a question about which one can find many diverse opinions. In Laboratory Sheet No. 245 we quoted George Crom to the effect that the usual loud speaker requires an input of 1 to 1 .5 watts for a volume of reception slightly above normal. In the Cunningham Tube Data Book (which costs $2.50 and which we recommend that you purchase, if possible) we read, "For home reception, with a speaker of average sensitivity, a tube capable of supplying at least 100 milliwatts (0.1 watt) maximum undistorted power output is recommended. The use of a tube giving lower output is almost certain to result in distortion appreciable to the listener. It is very desirable to have additional reserve power available, up to approximately 500 milliwatts, if the "B" power required can be conveniently supplied. Under such conditions the quality will not suffer if the volume is turned a little above normal, as may be required in a large room or for dancing, or if the loud speaker is somewhat low in sensitivity." The average of George Crom's figure is 1.25 watts and Cunningham recommends 0.500 watt. The mean of these two is 875 milliwatts, 0.875 watt. If the table of Laboratory Sheet No. 246 is referred to it will be found that the smallest power tube giving approximately this output is the 171a which is capable of supplying a maximum of 700 milliwatts to the loud speaker. It, therefore, seems fair to state that any installation using a power tube or combination of tubes in the output such that the available power is about 0.7 watt, that this amount of power will be sufficient to permit loud-speaker reproduction at fair volume without overloading.