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

34 RADIO BROADCAST MAY, 1927the average set constructor is that a real advantage Examination of the other filament circuits bias, the number of available biases was insufficient.A receiver wired in this manner veryisgained by this position of the detector tube. would indicate similar overloading.A second difficulty frequently encountered On account of these various difficulties involved,frequently blocked completely in one or morea study of the stages on account of improper grid bias. Whenin this method of series connection is that theproper values of negative grid bias are not always errors which existed in the straight series connectionand to makestages, amplification was again at normal values,C batteries were inserted in these troublesomeavailable. In certain commercial receivers, notablythe Western Electric 4-6 super-heterodyne, such alterations asbut as this receiver was intended to be a batterylessset, other means of grid biasing were sought.it is common practice to connect the grid return were required to bringlead of those tubes requiring negative grid bias the radio receiver toThe second difficulty was encountered in theto some preceding filament in the series. This its normal operatingr.f. stages, where regeneration often occurred.connection takes advantage of the voltage dropIn this situation the operator is unable to controlhis receiver on account of the incessant squealso B+ Max.and whistles of heterodyned carrier waves. Asimilar situation may also exist in a.f. stages.In this case it may cause an audio howl that willB+ Int. completely prevent reception.The means^^finally adopted in obtaining correctgrid biases was to place a 60- to 75-ohm adjustableresistance in series with each-4-0 filament, for3+ Det.199 type tubes, as shown in Fig. 5.This methodentirely satisfactory and has been used to someextent in Western Electric receivers. If adjustableresistances are used, the plate current of eachtube can be set at the right value when final testsare made on the receiver. Fixed resistances maybe used, however, if the constructor does notwant to make this refined adjustment. The platecurrent of each 199 type amplifier tube can bereducedA.C. Filament to about 2.2 mA., with 90 volts on theSupplyplates.For Power Amplifier TubeWhen the 6o-ohm resistances were placed inposition in the filament circuit, it was found thatthere was noFIG. tendency to regenerate, as out-3in the filament circuit, and employsit as a gridbias. It can be readily demonstrated that whilesome of the tubes in the series will be benefitedby such a connection, there will always be atleast one of the tubes for which the requirednegative grid bias is not available. The thirdobjection to the straight series connection is thatall of the 199 tube filaments, with the exceptionof the first, must have some current overload inthem, if the applied plate voltage exceeds 40volts. For an analysis of this situation, see Fig.4-In this diagram, with no potential applied toany of the plates, the filament current ammeter,A, could be adjusted to read 60 mA., and thecorrect current would then be passing througheach tube filament. However, when B voltageis applied to the plates, current flows in all theplate circuits, as indicated by the plate milliammeter,B. This current must return to the negativeB. If we check the current in the filament fortube No. 3 at the point (b) we would find that itwas carrying the regular filament current plusthe plate current of tubes Nos. i and 2, because,as stated above, these plate currents must returnto the negative B, and the most direct path isthrough the filaments. Consequently, the filamentof tube No. 3 would be overloaded by anamount dependingupon the plate currentof tubes Nos. i and 2.The data printed onthe diagram, Fig. 4,give the amount of currentin the filaments ofeach of the first fivetubes. The amount ofoverload depends onthe plate current of thepreceding tubes. Thus,in the case of tubs No.3 (the 2nd audio stage).the overload isequalto the sum of the platecurrents of tubes Nos.i and 2.1st. R.F. 2ndR.F. 2nd.A.F. 1st. A.F. DET. 3rd. A.F.Tube No. 1 -IfTube No. 4- lj-t-IP,+IP2+ IP 3-Tube No. 2- Ip- IP,TubeNo.5- If-t+ IP -t IP 1P, 2 3Tube No. 3- lf tlP,+ IFIG. 4Ir* Normal FilamentCurrentIP. Plate CurrentThe first step in this direction was to place the lined above. From a study of Fig. 5 it may benormal grid bias on each tube, with 90 volts on seen that r.f. currents in tube No. i must traversethe plates of the amplifier tube and 45 volts on the filament of tube No. 2; likewise, those currentsin tube No. 2 must pass through the fila-detector plates. It was frequently found that, ifthe various filament voltage drops were relied ment of tube No. 5.The question might ariseupon as the only source of grid bias, three difficultiesarose. First, because of the need of a tube ation. It was found, however, that the inclusionwhether or not this would be a source of regener-in the series ahead of each one requiring grid of the 6o-ohm resistances in the grid circuit ofthe respective tubeswas of the propervalue to eliminate thetendency tooscillation.This phenomenon is awell known fact, andin many modern radioreceivers a resistance isdeliberately includedin the grid circuit forthe purpose of stabilizingthe tendency tooscillate. The principleinvolved here is that,if the resistance of anoscillatory circuit is ofsufficiently large value,sustained oscillations