Radio broadcast .. (1922-30)

432 RADIO BROADCAST APRIL, 1928 8175Q an opera singer himself once, before he started conducting for them. Now and then he goes flat, but it does not bother him, for after all his singing is supererogatory. He. annoys the control operators, who- maintain that his singing inter- feres with their getting a balance during the rehearsal. But none of them has ever had the nerve to go out and tell him so. They are right; but if a man wants to sing it is generally dangerous to interfere with him. He will let you criticize his wife, his children, and the shape of his head be- fore he will let you stop him from singing when the spirit moves him. An Engineer's Embarrassment IT IS true that the more a man is educated, soundly, the less he is surprised at any- thing. And conversely, if he knows little, he is frequently astonished at perfectly rational be- havior which happens to be beyond his compre- hension. In the last few months I have had occasion, quite often, to venture into unfinished studios where painters were at work, there to warble various notes or to clap my hands to get some idea of the reverberation. The results have been almost as dreadful as when Mr. Hanson and I invaded a church for a similar purpose, as recounted in these columns a while ago. Some of the painters have nearly fallen off their scaffolds, causing me no small amount of anxiety lest 1 should give rise to a damage suit z=200Q against my company. Others have apprehended that I was mocking them, and the looks they directed toward me said 81.75 Q as much as that they were ready to paste me in the eye or to go on strike. In most cases I have slunk out after banging my hands together much fewer times than I had intended. Then the painters would glance at each other, and 1 knew as soon as the sound-proof door closed they would say, "The poor nut!" Well, they may be right, but not on the grounds they thought. What I was doing was just as rational as painting. Radio Inspectors—Fine Fellows WITHOUT doubt there is something magical about radio. Have you ever considered the remarkable infrequency of bureaucrats among the government employees in the business? The fact that where I have my office now no trans- mitting apparatus exists, so that the radio in- spectors do not visit the place, has made me think of them. The men in the U. S. Super- visors' offices are a fine lot of fellows. It is rarely that an operator is high-hatted by one of them. On the contrary, the inspectors and their as- sistants frequently go out of their way in order to help some boy in difficulty about his ticket. They form a marked contrast to the poisonous snobs in the naturalization offices, for instance. Maybe part of the difference is due to radio; 1 like to think so, anyway. Commercial Publications THE Daven Radio Corporation, of 158 Sum- mit Street, Newark, New Jersey announces "Super-Daven" resistance units, wire wound, available in values from 10,000 to 3,000,000 ohms, and accurate to within one per cent. The inductance is given as "practically negligible," with the distributed capacity likewise minimized. The temperature coefficient is o.oooi. At a slight additional cost, resistors are furnished with a closer tolerance than one per cent, and with zero temperature coefficient. These units mount in clip holders. The line is offered for use as lab- oratory standard resistors, voltmeter multipliers, plate and grid resistors, high-voltage regulators, and in telephone and telephoto work. No doubt broadcasters will find uses for them. I. E. JENKINS & S. E. Adair, of 1500 North J Dearborn Parkway, Chicago, describe in their Bulletin No. 5 a wire-wound gain control, Type GL 35. This instrument contains n sepa- rate resistance units, each wound non-inductively with enameled Nichrome wire. The total resist- ance is 350,000 ohms, arranged in logarithmic steps, resulting in a straight-line TU variation. The resistance values are good to less than one per cent. The housing is an aluminum shield, the end pieces being three inches square; a depth of 4! inches is required behind the panel. IN MARCH we started a review of the pam- phlet entitled Samson Broadcast Amplifier Units, issued for limited distribution by the Samson Electric Company of Canton, Massachu- 81.750 163.5Q 163 5 Q 40.4 Q Z=?OOQ Z=200Q 40.4 Q Z=200Q 81.75Q (A) (B) FIG. I setts. This detailed review will now be continued. On Page 18 we find a discussion of the design of a volume control arranged in 2-TU steps, with a maximum resistance of 100,000 ohms. The 2-TU drops may be secured by tapping off 79 per cent, of the voltage each time, or 79 per cent, of the resistance, as long as the voltage is pro- portional to the resistance. The first tap would therefore be at 79,000 ohms.- The description is not entirely clear in the pamphlet, and there is an error in that the first tap is given as 59,000 ohms. On Page 19 there is further discussion of gain controls, and the design of one covering a range of loTU in i-TU steps is given. This starts with a minimum, or first tap value, of 161,000 ohms, and continues in progressively increasing steps until the whole 10 TU have been covered. Another potentiometer affords a range of 50 TU in 10 steps of 5 TU each. The total resistance of each device is 500,000 ohms. One of these may be used across the grid and filament of one tube of an amplifier and the other across the input of a succeeding tube, thus providing fine and coarse regulation of gain as required. The material presented under "Pad Design" will prove extremely useful to many broadcasters although there is no attempt to derive the formu- las given nor to adhere to a conspicuously logical sequence in the discussion. The relationship be- tween the impedance on either side of a con- ventional resistive T- or H-network, the im- pedances of the legs, and the TU loss introduced, are given at length and worked out to a practical conclusion in the form of a table (Table III). From this tabulation, knowing the TU drop desired, one may read the value of Zi and Zz for Z equals 200 ohms and Z equals 600 ohms. Z is the impedance on either side of the pad (only bi- laterally symmetrical pads are discussed); Zi is the total value of the series or X-legs; Z 2 is the value of the shunt or Y-member. For example, if one requires a 20 TU pad presenting 200 ohms each way, one finds from the table that Zi must be 327 ohms, while Z 2 is 40.4 ohms. The result, for an H-network, would be the pad of Fig. i-A, while if a T-network with the same electrical characteristics is chosen the arrangement will be that of Fig. i-B. Numerically the difference is simply one of splitting Zi into four parts or into two parts, for the corresponding number of legs, depending on whether a balanced network is required or not. On Page 21 there is apparently an error, L] and Lj having been printed for Zi and Z 2 , re- spectively. The article as a whole seems to have been written by an engineer who knew what he was about but had no unusual skill in presenting the subject for the education of others. This de- fect is not extremely important, inasmuch as most of the people who consult the pamphlet will be more interested in the results than in the procedure required to reach them. In some ways the treatment is more extensive than that given in the article devoted to the same subject in this depart- ment for September, 1927, par- ticularly in the working out of the table, but in other respects it is less thorough than the latter, and readers who are much interested in pad design might go back to the Septem- ber, 1927, RADIO BROADCAST article after reading the Samson pamphlet, and also to the dis- cussion of the General Radio write-up on pads in the Janu- ary, 1928, RADIO BROADCAST. The limits within which all these pad designs hold good are well stated in the Samson booklet, which broadcasters may secure by writing for it on their letter-heads; this portion is well worth quoting: "It is assumed that the transformer is ideal, that is, that it has. a negligible resistance in its wind- ings, and an infinite input impedance with the output windings open-circuited, and vice-versa. It is also assumed that the line or other circuit element into which the transformer is working has pure resistance of the impedance value given as the impedance which the transformer is de- signed to match. It is also assumed that the transformer has no leakage reactance, that is, that all magnetic flux which links one winding links the other. In order that all these different things may hold, it is necessary that the con- sideration be based on a certain range over which the transformer practically meets all these con- ditions, and where those not depending on the transformer hold. Of course, as the frequency is reduced more and more, or as it is increased more and more, below and above this range respectively, these assumptions must fali down. Therefore, it must not be thought that a 'pad is a pad' over all ranges of frequencies. However, over the range that the circuit element is de- signed to work, these ideas hold very closely, and the simple addition of the attenuation of individual pads to obtain the total attenuation gives the desired working result, and that is the justification for its wide use in communication circles."