Radio Broadcast (May 1928-Apr 1929)

STRAYS from THE LABORATORY iVej© Trends in Radio Design IF THE a.c. screen-grid tube is made available before next year's receivers are designed, it is our bet that many of them will find their way into 1930's receivers, just as they have found their way into the best English sets. Over there 39 per cent, of all the types of sets manufactured during 1928 were three-tube sets — and 30 per cent, of the types used r.f. amplification. Screen-grid tubes were used in 80 per cent, of the sets with r.f. amplifiers. Data on English receivers: The above figures are but a part of an interesting analysis which was published in Wireless World, November 14, 1928, on the receiver situation in England. The other data show that the fivetube receiver, probably the most popular number of tubes in this country, constituted only 14 per cent, of the several types of receivers made in England during 1928. Seven per cent, of the receiver types employed no r.f. amplification at all, 89 per cent, used a grid leak and condenser type of detector, 70 per cent, used transformer coupling in the a.f. amplifier, and 79 per cent, connected the loud speaker directly into the plate circuit of the last tube. What will be used in next year's receiver? We have read recently several excellent articles on the use of a screen-griu tube as a detector. These articles described the research of J. R. Nelson, of the Cunningham laboratory, and were published in Radio Engineering, October, 1928; Proceedings, I.R.E., June, 1928; and Lefax 35, December, 1928. The great advantage of the C-bias, screengrid detector lies in its ability to handle a relatively large input r.f. voltage without overloading, and its relatively high output. According to Mr. Nelson, the first stage of audio can be done away with provided we use a screen-grid tube as a detector, and provided the r.f. amplifier supplies from 2.17 to 18.6 times as much amplification as when an ordinary general-purpose tube is used as detector with a two-stage transformercoupled a.f. amplifier. Now when the 322 detector tube is compared to a 327 tube connected as a grid leak and condenser detector, the amplification of the r.f. end of the receiver must be 18.6 times as much. Of course, a C bias is used with the 322 detector tube. Compared to the use of a 201a as a C-bias detector, the. 322 and one stage of audio amplification requires 3.17 times as much voltage amplification from the r.f. amplifier. If, then, we use screen-grid r.f. amplifiers, say two stages of them, the gain per stage must lie between 1.78 and 4.32 times as much as with present circuits, and if three of them are used, the increased gain per stage must lie between 1.47 and 2.65 for the two cases. Now it does not seem difficult to design an r.f. amplifier, using screen-grid tubes, that could produce 18.6 times as much amplification as present-day sets, which used with a screen-grid detector, will work directly into a power tube thereby eliminating one tube. This means an additional voltage gain of about 35 db to our present-day r.f. amplifiers. We do not consider this impossible — nor are we convinced that it is desirable to substitute r.f. gain for a.f. gain. One point of importance regarding the use of the screen-grid tube as a detector has not The following are among the subjects discussed in "Strays" this month: 1. New Trends in Radio Design 2. Two New A.C. Tubes on Way 3. Accuracy of Variable Condensers 4. New Radio Tubes in England 5. Importance of Reducing A.C. Hum 6. Receiving on 600 Meters 7. Selectivity of Browning Drake been discussed — what kind of a frequency characteristic can be obtained with it? Selectivity versus sensitivity: The problem is not to get mora amplification into our r.f. amplifiers. We have plenty now. The problem is to increase their selectivity without damaging their fidelity of response. It is our bet that the only reason why people tolerate present-day receivers — and their loud speakers which bring out the low notes — is because of the lack of competition from, and comparison with a really high-quality receiver. One only has to look at the selectivity curves, published in Dr. Hull's article, " Measurements on Broadcast Receivers." in February Radio Broadcast, or Mr. Jarvis' article in January Radio Broadcast, to see how few notes above 3000 cycles we are getting, and anyone has to listen but once on any good night in practically any home in the Middle West to long for a more selective receiver. The only answer is the solution that engineers have suggested time and again, and which the Members of Congress who meddle with radio affairs, do not seem to understand. This answer is to eliminate about 1.0 mfd half of the present broadcasting stations. The myth of having 40 cleared channels is amusing indeed to anyone who listens under average conditions. This does not mean listening-in in New York City, or near any great center of broadcasting, but say in Ohio. We had the dubious pleasure of listening-in there during the Christmas week. The receiver was undeniably less selective than many of the two-, three-, and four-stage r.f. amplifiers now on the market, sold as having perfect tone quality. An hour before sunset in Ohio, we heard stations as far away as Winnepeg, cky, on a four-tube set of the Lab. Circuit type which has been described many times in this magazine. We could tell at once that the practice of putting two large stations, kdka and wbz for example, on adjacaent channels is bad. If you live near kdka it works out fairly well because you can't hear wbz on the adjacent channel, but if you live equidistant from the two, you can't listen to either of them. Two New Tubes on The Way A.C. D.P.D.T. . Sw. 15mmfd. 55 mmfd. i+90 Fig. 1 — The diagram of the Lab. Circuit receiver with fixed condensers which increase wavelength range to 800 meters IF THERE is anything more interesting than radio gossip, it is speculation on how much of what you hear is true. We are glad to chronicle the gossip regarding two new tubes which — so they say ■ — are soon to appear on the American market . One is an a.c. screen-grid tube of somewhat better characteristics than our present d.c. tube with its fragile and microphonic filament. The tube has a typical heater filament, 2.5 volts and 1.5 amperes. At 180 volts on the plate, a screen-grid potential of 75 volts, and a control-grid bias of minus 1.5 volts, the plate resistance is about 400.000 ohms, its amplification factor about 400, and its mutual conductance about 1000 micromhos. This is considerably better than the d.c. tube with a mutual conductance — which is about all that matters in a tube of this kind — of not much over 300 micromhos. The grid-plate capacity is in the order of 0.01 mmfd., its input capacity about 5 mmfd., and its output capacity about 13 mmfd. The plate current is about 4 mA. and the screen-grid current under normal conditions about 0.3 mA. The other new tube is a cross between a 171a and a 250, i.e., a tube with about, double the power output of the 171a at a maximum plate potential of 250 volts. Many thousands of listeners who overload a single 171 on loud low-note passages, and yet who do not want to overload the house or the neighborhood with the racket from a 250-type tube with 450 volts on the plate, will be pleased with this new tube. Its filament consumes 1.5 amperes at a filament potential of 2.5 volts, and is not of the heater type. Its normal grid bias will be about 50 volts, plate current about 32 mA., amplification factor about 3.5, and power output of 1500 milliwatts. We have not been able to confirm the rumors that such tubes are going to be announced — but rumors mean that such tubes are in the process of development and that is the im • march, 1929 page 300 •