Radio doings (Dec 1930-Jun1932)

Page Twenty-six RADIO DOINGS March, 193 DEflGN €E /LPEC-HETCCCDyNE INTERMEDIATE TRANJTCRMEKT By RICHARD G. LEITNER A paper presented before the Electronic Clan, an organization of Los Angeles radio engineers engaged in the design of radio receivers and their integral units. Mr. Leitner is Chief Engineer for the ]ac\son Bell Manufacturing Company, Los Angeles THE imminence of the Super-Heterodyne Receiver as an important Figure 1 shows how these peaks, widely separated for large values factor in radio merchandising this year, and for some time to of K, gradually merge into one as K is reduced, the value of w' come makes it important at this time for those engaged in the — w" varying between zero and infinity. The value of 12 between design of commercial receivers to apply themselves closely to the de peaks falls less sharply than U. With a fixed current in the primary, Mgn problems and fundamental theory associated with this type of the value of ls rises slightly as K is reduced since energy in the cir receiver. cuit becomes more nearly confined to one frequency. A critical value In presenting this paper to this group, the writer has in mind of K will be found where the two peaks will be so spaced that a rather than the exposition of new theories and facts, a review of fairly uniform flat top response will be observed. The steepness of the principles involved, with which you gentlemen are familiar, but the sides of the response may be controlled more easily than any such a resume is considered timely. Specific design problems will he other way by the shielding material introduced into the fields of the considered in the order of their importance. coils as illustrated in some of the accompanying curves. Let us as' GENERAL PROBLEMS OF DESIGN sume now that a band width of 5 kilocycles is desired for maximum The transformer alone must be considered as a part of a complete selectivity (This may be broadened in favor of fidelity for local radio frequency amplifier from which certain predetermined results are reception by the introduction of resistance.) We first select two to be obtained. The performance of the amplifier as a whole may be peaks about 3 kilocycles apart. Then, for 175 kilocycles as the classified under three general headlines; overall voltage amplification center of the band we get from (4) and ( 5 ) . to be obtained, selectivity required, and method of control. These will \y " _ \\/ ' be treated in the order mentioned. ^ _ In order to obtain sensitivity of the order of one microvolt per ^ meter, which is necessary to make the Super-Heterodyne Receiver worthy of its name and reputation, voltage amplification in the in and a value of K in the neighborhood of .02. termediate frequency amplifier should be from 4000 to 7000. This f«r the determination of K, we have can be accomplished easily with two cascade stages, each having a M gain of from 65 to 85. Some receivers are being built with but one K = — — ■ (7) I. F. stage, and the lack of gain in the low frequency amplifier is V L, Ls partly compensated by additional audio or signal frequency amplifi Qr cation or both. However, the two stage I. F. amplifier is almost uni M = K V fj (8) versal at the present time, and will be considered as the accepted anJ where ^ h u whjcn Jg thg us'ual case,"then" standard. The same general theory of design, or course will apply __ j^j^ or (9) to any number of stages. _ . . r tlt^l ^ t j t SELECTIVITY measure M first measure Li, then L2, then connect Li and Ls Selectivity in this type of amplifier, as well as all other types, op >n serles assisting and call this L„ Then crating at high frequency, in order to meet the ideal standard, would Ls — Li — La be such that the interference ratio would be infinite at ten kilocycles M = (10) off resonance, and zero over a band exactly ten kilocycles wide. This, 2 of course, has never actually been accomplished, but in a low fre Or where Li equals L2 quency amplifier for use as the I. F. unit of a Super-Heterodyne Re _ L3 2Lt ceiver it can be more nearly approached than in any other type of M — ( ) receiver circuit. For reasons that become more apparent as the dis 2 cussion proceeds, the best form of transformers for the purpose is one The band width may then be predicted by in which both primary and secondary are tuned. The frequency of W"-W' = WK — (12) resonance for either one of the tuned circuits will be Fjgure 2 shows couplmg tOQ loose Thls transformer would atten , / , . uate high frequencies and be extremely critical to tune. Figure 3 shows the same transformer with coupling slightly increased. Note JtV the flat top, not quite wide enough in this case for good fidelity. Fig . r Hi . ure 4 shows the same transformer with same adjustments, but with When circuit one. whose frequency we will designate as wu is , . ., . , ~, , , .... ■ u~U*U. , , , , M ' n j „ a copper shield in place. The band width has been slightly increased. coupled to circuit two. whose frequency we will designate as u>2, mag 1 \ .. . »f „■{„„„ c,„.,.-„ t ii i . i • j . . l .. -if A * .-..7„ and the sides of the resonance curve are more uniform, figure 5 nctica y, due to mutual inductance, the system will respond to two ,. , . , , „„ ■ „ ■ c . . , ■„ i u l j • «. j ' ( a i„ shows the transformer with slightly increased coupling again, and frequencies which will hereafter be designated as w for the lower . .. , . b . ' , c. ,r , ° , ratt i "/ *l U'L c tl w ( ti,„„„ with one of the circuits slightly mistimed, figure 6 shows the result frequency and w for the higher frequency. The solutions tor these , b. 7 .. . ° D; , p,„,,,„ n ( u , . of correct alignment under the same conditions as figure 5. figure / frequencies will be shows ^ cffect of add)ng an lrQn shie]d wkh the arcult adjusted oj ' j^.-^uj : — ~~ as in figure 6. Figure 8 shows the effect of further increasing the ,/ 2 \J H'-^'^k'WwI- (2) coupling with the same shielding as Fig. 7. In this Fig. K is too large, V 2 (i-k2) and we find if' and w" beginning to spread, destroying the band an" pass effect obtained in Figure 7. i — r~ . In the case of two stages operating in cascade, great care must be \ Ul-2*"w>z2W ^ig--»)z4 4K2u)i2w>!.s (3) exercised in properly shielding all portions of the amplifier carrying ^°'-M f(.UY l ^jj,)-, freqUCncy current. The presence of regeneration may easily Equations (2) and (3) apply to cases where u>, and w are tuned change the shape of the overall response curve of the I F. amplifier alike. Since we are interested only in cases where they are tuned from the solid curve of Figure 7 to the dotted curve shown in the alike when w equals wi equals Wi we may use the greatly simplified same Figure. forms SELECTIVITY OF SINGLE CIRCUIT , ^ In a series resonance circuit maximum selectivity, that is the sharp , est resonance curve, is obtained when 1+K WL 1 and °r 18 sma11 w R WCR W" r= C>) By maintaining R and LC constant and increasing L with respect yj to C, then w WL/R is increased in value and the response becomes 1 K broad.