Radio Broadcast - 1927, May - 61 Pages, 4.9 MB ... - VacuumTubeEra

42 RADIO BROADCAST MAY, 1927Shows effect of changingratio of L to C150200 300 400WAVELENGTH -METERSFIG. 2purchased on the open market, and may be takenas reasonably representative of the commercialcoils now available.One of the first things investigated, in thelight of the previous discussion, was the effectof changing the ratio of inductance to capacityin any circuit using coils of nearly identicalconstruction. A single-layer solenoid, consistingof 45 turns of No. 24s. c. c. wire was wound ona cardboard form of 3!! inches in diameter.The values for the gain of this coil over thebroadcast range, using a condenser with maximumcapacity of 35ommfd., were plotted as shown inFig. 2, labeled I_o= 0.250 mh. Wire was thentaken off this coil until the coil resonated to 550meters (545 kc.) with a maximum capacity of485 mmfd. The values for the gain were againdetermined over the broadcast band and plottedas shown in the curve Fig. 2, labeled = Lo 0.178mh. These curves, with others, seem to show thatthe gain in any given oscillating circuit does notchange greatly with moderate changes in theratio of inductance to capacity. There does,however, seem to be some slight advantage inusing the higher values of inductance.Fig. 3 shows the value for the gain for a certaincommercial "figure eight" coil. It was woundin the conventional manner, the winding beginningat one end and progressing in a seriesof "figure eights" to the other end. The coil wasmade up of 80 complete loops of No. 26 s. s. c.wire and had an inductance of 0.250 mh approximately.The coil was if inches long and ifby 4 inches in the other directions. The gainfor this type of coil was surprisingly low, as itvaried from 100 to 130, as shown by the curve.The curve for a typical commercial typecoil is shown in Fig. 3 (b). This coil was a singlelayeropen solenoid of 0.245 mh. It was madeup of 58 turns of No. 23 s. s. c. wire closely woundso it could be made self supporting with the aidof a binder. It was 3!" in diameter and if"long. The curve shows that the coil is not whatone would call a bad coil.Recent work of the Bureau of Standards hasshown that in general the best type of solenoidconstruction is the type known as the "loosebasketweave." The curve, Fig. 3 (c), shows thegain for such a coil. It was wound with 53 turnsof No. 24s. c. c. wire, and was 35 inches in diameter.This coil is a very good coil and would givethe best of results if it is not placed too close toother apparatus.Much better coils can, of course, be built by500 600the use of litzendraht wire, but the writer hasonly considered the solid wire-wound coil sofar. A graduate student getting his Master's degreeat Yale in 1926 succeeded in building asolenoid out of "litz" with which he obtainedan average value for the gain of over 450 for theentire broadcast band. Such coils are, of course,exceptional and are more or less laboratory curiositiesbecause of their size and large externalfield. They cannot be used in the vicinity ofother apparatus without increasing the lossesand hence decreasing the value for the "gain."This is because their field sets up either eddycurrents in the metal parts, or causes dielectriclosses in insulators. These losses naturally decreasethe value of the gain.THE IDEAL COILTHIS brings up consideration of the ideal coilor high-frequency inductance. Considerablethought has been given to this question andthose points consideredthe most important, ornecessary, are tabulated below. The requirementsin their order of importance follow:(a) The "gain" of the coil as previously defined,and hence also, its selectivity, should ingeneral, be made as large as possible but shouldnot exceed the limit of 250 for good quality.This item in itself says a great deal, implicitly,since in order for the gain factor to beeven as large as 200, say, all known methods tokeep the coil losses down must be employed.This would mean, in general, that the coil wouldbe self supporting, it would be wound with relativelysmall wire since small wire serves to keepthe eddy current losses down, and it would beso wound that individual turns were reasonablywell separated. It seems to have been prettydefinitely established that the best coil is obtainedwhen the wire is spaced by an amountabout equal to the wire diameter. It would meana great many other smaller things which canall be grouped under the statement given in thepreceding paragraph.(b) The exterior field of the coil should bezero or, certainly, very small. This is essentialin order that there shall be as little stray energyexchange within the various circuits of the radioreceiver as possible. Putting an ordinary coilhaving an exterior field in a metal can is to bediscouraged, since it will greatly reduce the valuefor the "gain." For the same reason, the gainof an ordinary coil may be greatly reduced bymounting the coil in the vicinity of metal or poordielectric.(c) The distributed capacity of the coilshould be very low, and the high- and lowpotentialends of the coil should be well separated.The terminals of the coil should be wellseparated so that connecting wires to the coilwill not introduce an excessive amount of distributedcapacity into the circuit.(d) Mechanically the coil should be strongand able to withstand a reasonable amount ofabuse.(e) Its physical structure should not be excessivelylarge taking into account item (a).(f) A commercial coil should have some markingwhich would give its "gain" over the broadcastband.200 300 400500 600WAVELENGTH - METERSFIG. 3