Radio broadcast .. (1922-30)

RADIO BROADCAST. diameters of the various sizes of enameled wire required. The computation of coil size follows— core diameter allow JQ" clearance 1.750" 0.062" 1.812" Our diameter at the outside of the secon- dary is then 2.379" + 0.282" = 2.661" A wrapper 0.025" thick is satisfactory between the secondary and 281 filament windings and between the various ter- ting it run continuously for five or more hours, measuring the resistances of the various windings at the start and finish. The temperature rise of a winding may be computed from the following equa- tion— We must choose the primary tubing from experience, and the choice is entirely one of rigidity, hence dependent upon the size of wire. Likewise, the choice of the paper between layers depends upon the size of wire more than upon break-down. The amount of insulation between leads and the body of the coil and between pri- mary and secondary and the filament winding for the 281-type tubes and other windings depends largely upon break- down as well as mechanical requirements. Experience is the best guide in the choice of these values and the table in Fig. 6 give values that have been found satis- factory in transformer design. This table tells us that for our required pri- mary of number 18 wire we will need a pri- mary tube 0.080" thick, and an 0.007" paper between layers, preferably of .gummed kraft. The space taken by the primary wind- ing is figured as follows— No. 18 wire diameter'(sce tables) 0.0419" Paper between layers 0.0070" 25 per cent, of paper added for varnish 0.0018" At = 234 T C R - 4 - TH Fig. 6 tiaries, and about 0.030" will be sufficient for an outside wrapper. We continue our computations— Add 10 per cent, for bow 0.0507" 0.0051" 0.0558" Then each primary layer will take up •0.0558", or will add twice this, or 0.1116'', to the diameter of the coil. Inspection of Fig. 5 shows our •coil can be 3 \" long, and this allows us a 3" winding space. We can compute the number of primary turns per layer by dividing the wind- ing length by the wire diameter, and multiplying by the space factor, which varies from 85 per •cent, for small wires to 90 per cent, for larger sizes. This Outside of secondary Wrapper 0.025 X 2 281 fil. winding No. 21 Wrapper 250 fil. winding No. 18 Wrapper 227 til. winding No. 15 Doubled Outside wrapper 2.661" 0.050" 2.741" 0.050" 2.791" 0.042" 2.833" 0.050" 0.059" 2.942" 0.060" 3.002" Our coil will, therefore, be 3.002" or approximately 3" in diameter, and as we have a 3J" window we can easily get the R, - Calibrated B 2 - Volume control-500,000 ohms TOA.C Ripple = 64 turns per layer, or our 226- turn primary will require 4 layers, and the space taken up will be 4 X 0.1116 = 0.4464". Our •coil diameter at the outside of the pri- mary then is 1.812 + 0.446 = 2.258. The usual tubing between primary and secondary is about 0.040" —0.045" thick, so our diameter at the inside of the secon- dary is: 2.258" tube 0.090" A" for bow 0.031" 2.379" The secondary space is figured in like manner— Wire diameter 0.0097" Paper 0.0025" + 25 per cent, 0.0006" 0.0128" + 10 per cent. 0.0013" 0.0141" — Potentiometer M, - 1.5 Volt A.C.Meter Mj-1.5 Volt D.C.Meter 201-A Fig. 7 coil in it, allowing for commercial varia- tion. If the coil had turned out oversize it would have been necessary to use a larger stack of laminations, cutting down turns correspondingly. We can tabulate our coil thus— Core Outside of pri. Outside of sec. Outside of 281 fil. Outside of 250 fil. Outside of 227 fil. Outside diam. 1.750" 2.258" 2.661" 2.741" 2.833" 2.942" 3.002" Added diameter per layer » 0.0282 3.00 Turns per layer 0.0097 2700 -_ X 0.88 = 270 Number layers required - -5=7- = 10 layers Secondary space = 10 X 0.0282 = 0.282 If it is necessary to design a choke coil, the same process serves to compute the space required, although the process is simplified considerably by having only one winding. Similarly it may be used in determining size of field coils, audio trans- formers, etc. The Heat Run Having designed our transformer and chosen a choke coil we must check our design by placing the apparatus on a heat run. This consists of setting the pack up under actual working conditions and let- Where T c is the room temperature at the start of the run, TH the room tempera- ture at the end, R c the cold resistance of the winding, RH its resistance at the end of the run, and At the temperature rise in degrees centigrade (all tempera- tures centigrade). It will be found that if the rise is not over 50 degrees centigrade the heating will -not be excessive. We are now ready to assemble our pack, properly dispose the component parts, and choose values of filter capacity. It is here, probably, that the hardest part of pack design enters, as, while it is fairly easy to evolve a good filter by brute force means, it is not as easy to get down to an economical minimum of capacity and parts and still have a satisfactory absence of hum. In the measuring of a.c. ripple it is necessary to use a very sensitive indicat- ing device which is capable of measuring potentials of a few millivolts and up to about a volt. This is best obtained by the use of a vacuum-tube voltmeter with an associated amplifier, such as is illustrated in Fig. 7. This employs a three-stage im- pedance-coupled amplifier feeding the customary vacuum-tube voltmeter. Im- pedance coupling is advocated because resistance coupling requires more B bat- tery potential for its operation. It would be advisable to use wet B batteries and to have as much plate by-pass con- denser as possible, as is quite obvious when we consider that this amplifier must be practically flat from 25 to 120 cycles. The chokes LI should be as high in inductance as possible for the plate current they must carry, and chokes Lj can be of the order of transfor- mer secondaries. The comparison method is used, the switch being connected so that the a.c. ripple is impressed on the amplifier, and the volume control and R? adjusted to give a good reading of M 2 . The switch is now thrown the other way and the atten- uator adjusted to give the same reading of M 2 as before. Our a.c. ripple will then be the product of the attenuation ratio and the reading of MI. This gives us a convenient and sufficiently accurate means of measuring ripples from a few millivolts up to a volt. The Hum Problem We are now prepared to tackle our hum problem. Connect the pack to the set for which it is designed and connect the measurement set across the loud speaker. This gives us a visual indication of hum intensity and shows up small vari- ations that the ear would not detect. Of course, experience will be necessary be- fore the measurement of hum in milli- volts can be correlated to the resulting (Concluded on page 50) 48 • • NOVEMBER 1929