US2293384A - Band pass coupling system - Google Patents

Description

Aug. 18, 1942. R. DE COLA BAND PASS COUPLING SYSTEM 2 Sheets-Sheet 1 Filed April 22, 1940 ATITENTUATION m DECIBELS PHASE ANGLE-- Aug. 18, 194-2. 5 2,293,384

BAND PASS COUPLING SYSTEM Filed April 22, 1940 2 Sheets-Shet? ceiver.

a frequency of 8.25 megacycles.

Patented Aug. 18, 1942 BANl) PASSCOUPLING SYSTEM Rinaldo De Cola, Chicago, IlL, assignor to Belmont RadioCorporation, Chicago, 111., a cor-' poration of Illinois f Application April 22, 1940, Serial No. 330,881

16 Claims.

The present invention relates in general to band pass coupling systems, and more in particular to coupling systems of this character which are suitable for use at the intermediate frequency stages in the video channel of a television re- Specifically, the invention is a new and improved intermediate frequency transformer adapted to pass the comparatively wide band of frequencies used for video signals and having great attenuation for an adjacent frequency In order to show the actual requirements, it may be stated that the general practice is to use an intermediate frequency band extending from about 8.75 megacycles to 12.75 megacycles for the video signals, which brings the audio signals for the television channel undergoing reception in at The spacing of the television channels is such that the audio signals for the next adjacent television channelappear at an intermediate frequency of, 14.25 megacycles. The intermediate frequency stages in the video channel of the receiver must therefore pro...

vide for uniform transmission over the range 8.75

" to 12.75 megacycles, and at the same time must provide adequate attenuation for the sound signals at frequencies of 8.25 and 14.25 megacycles.

The problem of producing the desired attenuationfor frequencies adjacent to the picture signal frequency band has received some attention frequency transformer which will give normal transmission of the picture frequency band, and at the same time produce relatively great attenue ation of the sound channel. Moreover, attenuation is provided over a fairly good range of frequencies including the sound channel frequency, so that there is no danger that the latter will shift outside the range of attenuation during reception, as might otherwise be the case due to 'band on which audio signals are transmitted.

slight changes in the receiver oscillator frequency.

The frequency of maximum attenuation may be lower or higher than the frequency of the picture'signal transmission band. In the case of a transformer designed to produce attenuation at a lower frequency, the frequency of maximum attenuation is made to coincide with the frequency of the sound carrier which is associated with the television channel undergoing reception, while if the transformer is designed to provide attenuation at a higher frequency, the attenuation frequency will be the same as the frequency of the 7 sound carrier of the next adjacent television channel. The transformers at different stages may therefore be of two types, designed to atten uate frequencies below and above the picture signal band, respectively, whereby both adjacent sound carriers may be eliminated.

heretofore, but the solutions proposed have not been entirely satisfactory. One'type of proposed solution utilizes so-called trap circuits-which act as short circuits for the frequency to be attenuated, but such systems fail to produce the requisite attenuation due to the radio frequency resistance of the trap elements. Other arrangements which have been proposed are objectionable because of the extremely narrow band of frequencies over which attenuation is obtained.

The present invention affords a solution which is substantially free of the objectionable features of prior systems. It provides a coupling system for use at an intermediate frequency stage, that is, an intermediate frequency transformer, having two different circuit paths between the input and output circuits, and having the constants of the system so adjusted thatat the frequency at Further features of the invention will be ex plained in the course of the ensuing detaileddescription, and with reference to'the accompanying drawings, in which:

Fig. 1 is a diagrammatic circuit drawing of an intermediate frequency transformer according to the invention, as used at one of the intermediate frequency stages of a television receiver;

Fig. 2 is an equivalent theoretical circuitdiagram of the transformer shown in Fig. 1, which is useful in discussing the operation thereof;

Fig. 3 is a diagrammatic circuit drawing showing a modification, and Fig. 4 is the equivalent .theoretical circuit diagram;

Fig. 5 is a diagrammatic circuit drawing showing a-further modification, Fig. 6 being the equivwhich attenuation is desired the currents in the two-circuit paths induce equal opposing voltages in -the output circuit and produce cancellation.

At the same time the necessary flat response characteristic over the frequency band which is to be transmitted is maintained.

The invention thus provides an intermediate alent theoretical circuit diagram; I

Fig. '7 shows the physical arrangement of the transformer windings in thatform of the invention which is shown in Fig. 1; v v Figs. 8 and 9 show winding arrangements that may be used with the modification of Fig. 3, the arrangement of Fig. 9, with a different coil spacing, being also usable" with the modification of Fig. 5;

Figs. 10 to 13, inclusive, are graphs which show selectivity curves such a may be obtained with variable resistor l5.

transformers constructed in accordance with the invention; and

Fig. 14 is a graph which shows typical phase angle curves over'a range including the transmission band and attenuation frequencies.

Referring to Fig. l, the reference character 2 indicates a vacuum tube, preferably of the pentode type, which may be an amplifier tube at an intermediate frequency stage of a television receiver. The vacuum tube 2 is provided with input terminals 3 and t, at which connections are made from the preceding stage. A resistor 5 may be included in the cathode circuit to provide the proper bias for the control grid. The connections for the screen and suppressor grids are not shown, as they are of the usual character.

The plate circuit of tube 2 includes the battery 9, representing -any suitable source of plus B potential, and also includes a tuned circuit/2i comprising the inductance H and the variable condenser l0. This tuned circuit may be regarded as the input circuit of the transformer.

Adjacent to the tuned circuit 28 there is shown a second tuned circuit 22, comprising the inductance 93, the variable condenser Hi, and the This circuit is preferably grounded as shown.

A third tuned circuit 23 is also shown, which comprises the inductance l6 and variable condenser l8. The circuit 23 may be regarded as the output circuit of the transformer and may be connected to the grid of the pentode 20, which may be similar to the tube 2. A suitable load resistor I9 is used, shunting the condenser iii. The complete grid circuit for tube 20 includes the automatic volume control lead, as indicated.

The transformer also includes a fourth tuned circuit 24 comprising the inductance l2, inductance I I, and variable condenser 25. This circuit may be grounded as shown in the drawings.

may be small trimmer condensers of the mica compression type, having a range of about 5 to 15 mmf. The load resistor l9 may have a value of about 2000 ohms. The value of the resistor I5 is indicated as variable, but a fixed resistor may be used, after the correct value has been ascertained.

The inductances ll, I2, l3, l6, and I7 are the windings of the transformer and are shown in their proper physical relation in Fig. '7. The socircuit 24. In these circumstances it will be clear that voltages will be induced in the winding l6 due to the currents in circuits 22 and 24, and that the effective voltage'will be equal to the vector sum of these voltages. The operation of the transformer depends on the fact that the induced voltages are essentially aiding over the picture frequency band, but at the frequency of the sound, carrier, where great attenuation is desired, these induced voltages are equal and opposing, thereby reducing the effective voltage and the current in the output circuit 23 substantially to zero.

The manner in which the desired result is accomplished can be explained more conveniently with reference to the theoretical circuit Fig. 2. In this drawing the several tuned circuits are indicated by the same reference numerals as in Fig. l, and can readily be identified. The mutual couplings between the circuits and other constants are indicated by conventional symbols.

In general, it may be stated that the selective action of the transformer depends on the proper arrangement of the polarities of the mutual coupling reactances, upon the relative phase shift between currents I2 and I4 which occurs over the range which includes the band of frequencies which are to be transmitted and the frequency to be attenuated, and upon changes in the relative amplitudes of the currents which take place over this frequency range.

.40 The condensers shown in the tuned circuits called universal winding'is preferably used but is t not absolutely essential. The coils are mounted on a small Bakelite tube 26 and are slidablerelative to each other so that the coupling can be adjusted. From the positions of the coils as shown in Fig. 7, it will be seen that coill l is inductively coupled to coils l2 and 93, while coil I6 is inductively coupled to coils I3 and IT. This relation is also indicated in Fig. 1, where the input circuit 2| of the transformer is shown as inductively coupled to circuits 22 and 2d, while the two latter circuits are shown as inductively coupled to the output circuit 23.

From the foregoing it will be appreciated that there are two circuit paths over which energy may be transmitted from the plate circuit of tube 2 to the grid circuit of tube 20. One of these paths includes the tuned circuits 2|, 22, and 23, while the other path includes the tuned circuits 2| 24, and 23. Stating it another way, the input circuit 2| of the transformer is coupled to the output circuit 23 by way of two intermediate or link .circuits, one of which is the tuned circuit 22'while the other is the tuned The relation between these factors and the manner in which they affect the operation of the system can perhaps be best explained and understood with the aid of a brief mathematical analysis of the theoretical circuit, Fig. 2, which will now be proceeded with. Taking into account the four coupling reactances involved, and assuming that coupling M4 is negative, the equation for the current I: in the input circuit 23 can be written in the form where D and D' are the determinants of the system. Maximum attenuation will be obtained when 11 becomes equal to zero, or when I3=O=D=M3M4Z2M1M2Z4 (2) where M1, M2, M3 and M4 are the mutual coupling reactances, and Z2 and Z4 are the impedances of circuits 22 and 24. written with good accuracy as Z2=R2(1!,7'262Q2) and similarly Z4 can be written as Z4=R4(1+7'264Q4) where 62 is equal to Now Z2 can be and 64 is equal to the frequencies is and f4 being ,the resonant fre- 0: (K3K4L4R2K 1K 2L2R4) y'zLzLdKsKqfiz-KrKzfiQ (5) The right hand portion of Equation 5 compositive, with the exception of M4.

other combinations, which will reduce current I:

I prises .two expressions which must each equal zero if the right hand portion ofthe equation is equal to zero. The first expression reduces to zero when i mm c. (6)

the symbol Q2 being written for the expression 21rfL2/R2, and Q4 being written for .the expression 21rfL4/R4. hand portion of Equation 5 reduces to zero when (resistance balance) m (reactance balance) (7) A simultaneous balance is obtained, that is, Equation 5 is fully satisfied, when are From Equation 8 the equation for the frequency of infinite attenuation can readily be derived and may be written in the form of In the foregoing discussion of Fig. 2, all the mutual coupling reactances were assumed to be Various to zero at a particular frequency are possible, as may be perceived from inspection of Equation 2. All these combinations satisfy the requirement that for any given combination of couplings in the two circuit paths all the couplings, except one, have the same polarity. Unless this require- The' second expressionin the right fl and is indicated at f. Curves 2 and a4 show the manner in which the corresponding currents I: and I4 lead or lag the impressed voltages at the different. frequencies. At the resonance frequency the current I: is in phase with the voltage,

as indicated by the fact that curve t: crosses f:

on the zero line. At frequencies higher than In current I: has a lagging characteristic, while at lower frequencies it leads the voltage. Current I4 behaves in a generally similar manner, as indicated by curve 4, although the phase angle changes more rapidly, due to its lower dissipation or resistance. At frequency f currents I: and I4 are both leading at the same angle and are in phase with each other due to the fact that the voltages are ,in phase (couplings M1 and Mzbeing both positive), which gives the correct phase relation for cancellation with the negative coupling M4.

As regards the current magnitudes, it will be seen also that at frequency f, currents I: and I4 M4 are equal.

ment is met, cancellation in the output circuit 23 cannot result; that is, current I3 cannot be reduced to zero.

It will be of advantage now to consider one to zero at some particular frequency where currents I: and I4 are in phase, whereby opposing voltages are induced in L3, provided that the current magnitudesand the concerned mutual couplings are so related that the induced voltages are equal. This last condition is satisfied if I27'WM2 is equal to I4iWM4. The foregoing states generally the conditions as regards phase and magnitude relation of currents I2 and I4 which obtain ,at the frequency of maximum attenuation, when using the particular coupling combination under consideration. i

The phase relation between currents I: and I4 may be discussed more. in detail with reference trated in Fig. 14.

are equal in magnitude, which satisfies another condition for cancellation, assuming that M: and

It will be understood that currents I2 and.I do not necessarily have to be equal at frequency 1, so long as the mutual couplings M2 and M4 are such as to produce equal induced voltages in circuit 23. It will be noticed that-at frequency 1 both the phase angles and magnitudes of I2 and I4 are changing rather slowly, which accounts for the excellent attenuation over a fairly wide range including frequency f.

Since currents I2, and I4 are in phase at frequency f, the frequency of maximum attenuation,

it follows that they must be out of phase, or differ in magnitude, or both, at frequencies within the picture transmission band, which includes frequencies f: and f4. This condition is also illus- As previously intimated, the attenuation frequency f of the transformer does not necessarily correspond to the sound channel frequency of 8.25 megacycles and can be shifted to a'point above the picture signal frequency band, if desired, to' the sound carrier frequency of 14.25 That such shift is megacycles, for instance. possible will be evident from Equation 9, which shows that ,if frequency fl is lower than, frequency h, then frequency i will be lower than frequency f4, while if frequency I4 is higher than frequency f2, then frequency f is higher than frequency f4. This means that if frequency {is 1 to be shifted to a point above the picture transto Fig. 14, which shows graphically the phase angles at which the currents Ia and I4 lead or lag the corresponding voltages over. the frequency range with which we are concerned.- In this figure, I2 is the resonance frequency of circuit 22 (Zn) and i4 is the resonance frequency of circuit 24 (Z4), while I2 and I4 are the corresponding resonance curves. The frequency of maximum attenuation is assumed to be lower than I: and

mission band, the .relative positions of frequencies f4 and f2 as shown in Fig. 14 mustbe. reversed; that is, frequency f4 must be higher than frequency f2. Frequency f4 can be adjusted to the properpoint' by varying condenser C4. It may also be necessary to relocate frequency f2, which can be done by adjusting condenser C2. The other constants, including resistance R2. will of course require readjustment. These adiijus ments will shortly be discussed more in deail.

It should be noted that in any case. and this is true of the modifications which will be described later on as well as of the ,one under discussion, the frequency of maximum attenuation, does not correspond to the resonance frequency of any of the tuned circuits. This is an important and fundamental characteristic fea-- ture of the invention, which diffentiates it from known coupling systems, in which the resonance frequency of one at least of the tuned circuits always coincides with the attenuation frequency.

The calculation of the response curves of current I: from Equation 1 over the concerned frequency range and with the'conditions required by Equations 6, 7, and 8 is possible but is exceedingly tedious. The proper adjustment of the constants of the transformer which is required in order to satisfy the equations is p'referably determined experimentally, therefore.

.A primary consideration, of course, is uniform transmission over the required. frequency range, which is obtained by properly adjusting all four coupling factors and by adjusting the condensers C1, C2, and C3. These adjustments fix the limits of the frequency range at the desired points, and the load resistor 89 associated with tuned circuit 23 produces the necessary flat top characteristic over this range. The phase characteristic from the input side of the transformer to the output side has been checked and has been found to be quite uniform over the transmission range.

The next step is to locate the attenuation frequency ,1, which is done by adjusting the condenser C4. The adjustment of. this condenser fixes the location of frequency ii, which in turn determines frequency ,1, from Equation 9. In other words, the frequency f at which attenuation is to be obtained is fixed at the desired point by properly locating frequency f; with reference to frequency ii. The adjustment of condenser C4 therefore produces the required frequency relation to satisfy Equation 7, and its effect can be visualized by reference to Fig. 14, from which it can be seen that a shift inthe location of frequency f4 will correspondingly shift the phase angle curve 4:4. Thus the phase angle curve m is made to intersect the phase angle curve 2 at the desired attenuation frequency f.

Having made the necessary adjustments to insure that currents I2 and I4 will be in phase at the desired attenuation frequency, the next step is to bring the current magnitudes into the properrelation to satisfy Equation 6, which is done by adjusting the resistance R2. Here again the operation can be visualized from Fig. 14, from which it can be seen that if the curve I: were to be raised, the intersection of the current curves would occur at a frequency higher than frequency f and cancellation would not take place because the phase curve intersection and the current curve intersection would no longer be at the same point on the frequency axis. If the curve of current 12 were to be lowered the same condition would exist, except that the current curves would intersect at too low a frequency. The value of resistance R2 determines the height of current curve I2, and by adjusting this resistance the curve can be shifted up .or down until it intersects curve 14 at the desired frequency f. The phase angle curves and the current curves now intersect at the same frequency, Equation 8 is satisfied, and maximum attenuation is obtained.

The results actually obtained in practice with the transformer of Fig. l are illustrated by the selectivity curve Fig. 13, where attenuation is decibels is plotted against frequency, In plotting this curve the point of zero attenuation is arbitrarilyfixed' at the lowest point of the curve 'which is within the frequency range to be transtion is obtained at the frequency of the accompanying sound'carrier, or at a frequency of 8.25 megacycles, the attenuation at this frequency being on the order of nearly 60 decibels. While the frequency range over which the great attenuation is obtained is narrow compared to the picture signal range, consideration of the frequency scale will show that at points not far from the bottom of the curve the attenuation band is many times the width of the sound channel. At 50 decibels, for instance, the attenuation band has a width ofat least 25Q,000 cycles, while the sound signals cover a range of only about 10,000 cycles. I

The selectivity curve Fig. 18 also shows a good order of attenuation for a frequency of 14.25 megacycles, which is the frequency of the sound carrier of the adjacent television channel. The excellent attenuation at 14.25 megacycles is accounted for by the direct coupling between L1.

and and can be predicted mathematically if an additional coupling factor designating the coupling between Li and L3 is introduced into Equation 1.

In the foregoing explanation it was assumed that all mutual couplings are positive except M4, and it has been explained in detail how cancellation is efiected at a particular frequency. The transformer, operates in a similar manner if the mutual coupling M2 is negative, all other couplings being positive. Coupling M2 can be made negative by reversing coils it and ll, Fig. 1, corresponding to inductances In and 12/4, Fig. 2.

A somewhat difierent case is presented if coupling M is made negative by reversing coil 82, Fig. 1, corresponding to inductance L'-'4, Fig. 2, all other couplings including M4 being positive. The phase relation between currents I2 and I; may be essentially the same as depicted in Fig. 14, except that current I4 is shifted in phase relative to current Is by 180 degrees, due to the fact that L"4 is reversed. The phase angle curve 4 in Fig, 14 must therefore be regarded as being shifted down to a new zero axis which is 180 degrees out of phase with the zero axis shown. With this understanding, it may be stated that at frequency f currents I2 and I4 lead their respective voltages by the same phase angles, but are 180 degrees out of phase. As couplings M2 and M4 are both assumed to be positive, the voltages induced in La oppose each other, and cancellation results if the proper relation exists between the current magnitudes and couplings as previously explained.

A case similar to the last described case is presented if coupling M1 is made negative, which can be done by reversing coils II and I2, Fig. 1,

corresponding toinductances L1 and L4, Fig. 2.

It will be clear that in this case currents I2 and I4 will again be out of phase by degrees at frequency .f and that the operation will be similar to the operation just described.

All these variations come under the general rule which states that in order to effect cancellation at some desired frequency all the mutual couplings, except one, must have the same polarity, and the mathematical treatment is the same in each case.

'Attention may now be directed to the modification which is shown in Fig. 3. Like Fig. 1, this figure shows two vacuum tubes 32 and '43, which are assumed to be located at adjacent'intermediate frequency stages in a television receiver. The plate circuit of tube 32 is connected to the grid circuit of tube 43 by means of a modieludes inductance 40 and condenser 4|. The inductances are the windings of the transformer. Two arrangements of the transformer windings or coils are possible, as illustrated in Figs. 8 and 9. 7

According to Fig. 8, windings 38 and 39 are combined and form a single coil which is located between the windings or coils 31 and d0. With this arrangement it will be seen that circuit 5| is inductively coupled to circuit 53 and also to circuit 52, while the latter circuit is also coupled to circuit 53. The couplings M1, M2, and Ma are 0 indicated in Fig. 8. Two circuit paths are provided between the plate circuitof tube 32 and the grid circuit of tube 43, one including the tuned circuits 5| and 53, and the other including the tuned circuits 5|, 52, and 53.

In the arrangement according to Fig. 9, the windings 38 and 39 a're separate coils which are located outside of coils 31 and 40, respectively. As in Fig. 8, there are three mutual couplings, M1,,Mz, and Ma. Also as in Fig. 8,'two circuit paths are provided between tubes 32 and 43, one including the tuned circuits 5| and 53, and the other including all three tuned circuits. The arrangement shown in Fig. 9 has certain advantages over that shown in Fig. 8, as will be explained presently. a

The operation of the transformer shown in Fig. 3 is similar in principle to the operation of the transformer shown in Fig. 1; that is, it depends upon the generation of substantially equal 40 i and opposing voltages in winding 40 at some particular frequency where attenuation is desired, which reduces the current in tuned circuit 53 to a minimum at that frequency. The operation of Fig. 3 differs in detail from that of Fig. 1, be- 5 cause of the fact that in Fig. 3 the two circuit paths through the transformer include an unequal number of inductive couplings, whereas in r Fig. 1 the two circuit paths include an equal, number of inductive couplings. The details may be brought out so far'as necessary by continuing the discussion with reference to Fig. 4, which corresponds to Fig. 3 like Fig. 2 corresponds to Fig. 1

It may first be assumed that the winding arrangement is as shown in Fig. 8. With this ar- 55 rangement it will readily be seen that all the mutual couplings are positive, or at least the effect is the same as when they are, for reversing any one coil will simultaneously reverse two couplings. Reversing any two coils produces the same condition as reversing the remaining coil, and of course is equally ineffective. Assuming all couplings to be positive, therefore, the equation for the current I3 in the output circuit 5 can be written in the form E v (10) Defining Q2 and 62 as specified in the previous discussion of Fig. 2, and assuming that Q1 and Q3 are each equal to Q, Equation 10 can be rewritten in the form (1 An inspection of Equation 11 indicates that current I: can never be reduced to zero. However, the value of 13 will be a minimum when (12) Equation 12 can be rewritten in the form K K 2 v 2 l(l and from the latter equation may be derived the equation for the frequency of maximum attenuation in the form A of the operation of the circuit, Fig. 4, from a functional standpoint. Bearing in mind that the three coil structure of Fig. 8 is under discussion, and that the mutual couplingsare all positive, it will be appreciated that at the desired frequency of attenuation currents I1 and I2, must be 180 degrees out of phase with each other to produce the required opposing voltages in In. Since the voltage in L2 (coil 3839) lags the current I1 by degrees, the required phase relation between I2 and I1 canbe obtained if current I2 lags behind the voltage in L2 by 90 degrees at the attenuation frequency. This theoretical condition cannot. be attained in practice, due to unavoidable resistance in circuit 52, as can beseen from Equation 11. It can be approximated close enough, however, so that practical results can be secured. But the required lagging characteristic of current I2 can only be obtained at a frequency higherthan the resonance frequency is, which lies in the picture transmission band, and hence the attenuation frequency I must be located above the transmission band, as stated at the outset.

Fig. 10 shows a typical selectivity curve such as may be obtained with the three coil arrangement of Fig. 8. It will be noted that the maximum attenuation is not obtained at a frequency of 14.25 megacycles, but at a somewhat higher frequency. The attenuation frequency could be brought closer tothe transmission band, as indicated by Equation 14, but it has been found in practice that it cannot be done without adversely affecting ,the required flat top characteristic of the transmission band. The curve shown indicates about the best results that can be expected 1- though the frequency of maximum attenuation with the arrangement under discussion.

is .not at the optimum point, the curve does have a much steeper slope on the high frequency side. i

From the foregoing it will be seen that the practical results which can be secured with the three coil arrangement of Fig. 8 fall somewhat short of the results which might be expected from the theoretical discussion, although the selectivity of thecoupling system is distinctly improved. The partial failure is due to the inflexibility of the coil arrangement as regards the adjustment of the mutual couplings. With only three coils, any change in one of the couplings changes another coupling at the same time, as can readily be seen from Fig. 8. if we change coupling M2 by moving coil 31, we necessarily change coupling M1 at the same time. Adjusting coupling M3 by moving coil 40 also changes coupling M1, and of course we cannot For instance, I

adjust coupling M1 by moving coil 31 or coil 40 without also changing coupling M: or cou- P M3.

The difliculties referred to above are obviated by the four coil arrangement of Fig. 9, which is entirely flexible as regards independent adjustment of the'couplings, as will be apparent from inspection of Fig. 9 and the circuit drawing, Fig. 4. To give just one example, coupling M1 can be adjusted by moving coil 39, and this adjustment will be without effect on couplings M1 and M2, for the relative positions of the coils involved in the latter couplings is not disturbed. The freedom in coupling adjustment afforded by the four coil arrangement gives greater flexibility in the choice of the attenuation frequency, which can be located close to the transmissionband without affecting the flat top characteristic which is essential for the correct reception of the picture signals.

The four coil arrangement of Fig. 9 also has the advantage that the attenuation frequency may be located either above or below the picture transmission band, being somewhat comparable in this respect to the modification which is shown in Figs. 1 and 2. This freedom as to location of.

particular coupling combination used determines Q the location of the attenuation frequency above or below the transmission band, -as will be explained.

If all the mutual couplings are positive, the operation is similar to the operation when using .the three coilarrangement insofar as the purely mathematical treatment is concerned and the case is governed by Equations 10 to 14, inclusive. The attenuation frequency is higher than the resonance frequency f: of circuit 52 and lies, above the transmission band, as indicated by Equation 14. v

For the reasons stated, however, considerably better results are secured with the four coil arrangement as compared to those obtained with the three coil arrangement. Fig. 11 is a typical selectivity curve, which shows excellent attenuation on the high frequency side of the transmission band, with a definite maximum at the.

and Equation 14 for the frequency of maximum attenuation changes to the'form Equations 15 and 16 hold independent of the location of the negative coupling. The operation of the transformer is also fundamentally the same, although the phase relation of curdifferent in the two cases, depending on whichv coil in circuit 52 is reversed in order to produce the negative coupling. This will be explained Lz (coil 39), currents I1 and I; must be in phase the voltage in L: with respect to 11, gives ,a H

rents I1 and I: at the attenuation frequency is at the frequency where attenuation is desired, in order to produce opposing voltages in La. Since the voltage in L"2 lags the current 11 by degrees, the desired phase relation between currents I1 and I: can be obtained if current I: leads the corresponding voltage by 90 degrees. This condition cannot be reached in practice, due to the resistance in circuit 52 but can be approached near enough to produce effective attenuation at the desired frequency. Since a leading current is required, the attenuation frequency must lie below the picture transmission band, as confirmed by Equation 16.

If coupling M2 is made negative by reversing L": (coil 38), couplings M1 and M: being positive, currents I1 and I: must be out of phase by degrees at the frequency of attenuation. The reversal of L": produces a phase shift of 180 degrees, which when added to the phase shift of phase shift of 270 degrees for current I: with respect to current 11. This phase shift will be reduced to 180 degrees at the frequency of attenuation if current I2 leads the voltage by 90 degrees. As in the previous case, the required theoretical condition can only be approximated, but near enough for practical results.

Fig. 12 shows a typical selectivity curve for the transformer Fig. 3, using the four coil arrangement of Fig. 9.and a negative coupling. The curve shows a considerable attenuation on the low frequency side, with a frequency of maximum attenuation at 8.25 megacycles, where the transmission is down about 35 decibels. The curve of Fig. 12 is generally similar to the curve of Fig. 13. The results secured are not as good, but under certain conditions they will be entirely satisfactory.

Referring now to Fig. 5, the further modification shown therein may be described briefly. As in the previous modifications, the two vacuum tubes 62 and 83 may be located at adjacent intermediate frequency stages in a television receiver. The plate circuit of tube 62 is connected to the grid circuit of tube 83 by means of a transformer which comprises the four tuned circuits "ii, l2, l3, and I4. The tuned circuit H in-' cludes the variable condenser 63, a fixed condenser 64, and the inductance B'l. Tuned circuit 12 includes the variable condenser 65, the variable resistance 80, the inductance 66, and the fixed condenser 64. Tunedcircuit' "ll includes the variable condenser .15, the inductance 69, the fixed condenser 16, and the'variableresistance 8|. Tuned circuit-13 includes the variable condenser H, the fixed condenser I6, and the inductance 68. The condenser 64 is shunted by a resistance 19 of about 1000 ohms to afford a direct current path for plate current." 0n the output side the condenser 11 is shunted by the load resistor 18. The inductances 66, 61,158, and 69 are the windings of the transformer and comprise four coils the physical arrangement of which may be explained with reference to Fig. 9. Coils 68 and 68 constitute a pair of coils and may occupy positions corresponding to the positions'of coils 3| and 31, Fig. 9. Coils 66 and 68 accordingly are inductively related. Coils s1 and "constitute a second pair of inductively related coils and may is suflicient to substantially "eliminate inductive coupling between pairs; that is, there should be substantially no coupling between the two inside coils. If necessary to attain this object, the core on which the coils are mounted can be made somewhat longer, or the two pairs of coils canbe mounted on separate cores. a

As regards the coupling between the several tuned circuits, it will be seen that the first tuned circuit 1|, on which high frequency signal voltages are impressed byv the plate circuit of tube 62, is inductively coupled to tuned circuit ll by means of coils 61 and 69, while the latter tuned circuit is capacitatively coupled to tuned circuit 13 by means of condenser I6. Also, it will be seen that tuned circuit II m capacitativelycoupled to tuned circuit I2 by means of condenser 64, while the latter tuned circuit is inductively coupled by means of coils 66 and 68 to tuned circuit 13. Thus, as in the previous modifications, two circuit paths are provided between the input and output sides of the transformer, one including the-tuned circuits 1|, I2, and 13, and the other including the tuned circuits ll, I4, and I3. The arrangement is similar to the arrangement of Collecting the reactive, terms of Equation 19, we

derive the equation C4M1R4 52Q2 czMiRfifi (21) From Equations 20 and 21 we have which is the equation for simultaneous balance. From Equation 22 may be derived the equation for the attenuation frequency j in the following 23 1J3} I f2Q4, i which is identical with EquationB. It will be seen therefore that the attenuation frequency can be located either below or above the transmission band as desired.

In adjusting the transformer to secure maximum attenuation at the desired frequency, the variable condensers C: and 0'4 are adjusted in order to satisfy Equation 21', that is, to produce the requisite phase relation between currents I: and I4 at the desired attenuation frequency, after which the resistances R2 and R4 are adjusted to satisfy Equation 20, which insures the requisite Fig. 1 in that the number of couplings is the I same in the two circuit paths, being difierentin this respect from the arrangement of Fig. 3. The arrangement of Fig. 5 differs from both of the previous arrangements in the fact that a capacitative coupling is used in each circuit path.

The operation of the transformer may be discussed briefly with reference to Fig. 6, which is the equivalent theoretical circuit diagram, and will be understood without difficulty in view of the previous explanation. Suffice it to say that at the frequency of maximum attenuation, the currents I2 and I4 produce substantially equal and opposing voltages in circuit 13, which reduces the corresponding current I3 substantially to zero. In

, order to effect such cancellation, it will be clear that oneof the two mutual couplings M1 and M2 must be negative. Y

The mathematical analysis of the transformer is similar to that of thetransformer shown in Fig.

*1. "Thus we may write as the equation for the current I; in the circuit 13 IFTYI For cancellation of current 13, the determinant D must equal zero, and we may write M124 M222 C C Collecting the resistive terms of Equation 19, we derive the equation relationship as regards the magnitudes of currents I2 and I4.

No selectivity curve is shown for the transformer of Fig. 5, but it may be stated that the results obtained are comparable with those obtained with the transformer of Fig. l, insofar as attenuation at some selected frequency is concerned. Assuming that the transformer is adjusted for attenuation at 8.25 megacycles on the lower side of the transmission band, reference may be made to Fig. 13 to ascertain the amount of attenuation at this frequency. That is, the low side of the selectivity curve for Fig. 5 is substantially the same as shown in Fig. 13. There is, however, no additional attenuation on the high frequency side as shown in Fig. 13, but the curve on this side is about the same as is shown in Fig. 12. Of course, the frequency of attenuation f can be moved to the upper side of the transmission band, if desired, in which case there would be no additional attenuation on the low frequency side, the lower side of the curve being similar to the lower side of the curve in Fig. 10.

The invention having been described, that which is believed to be new and for which the protection of Letters Patent is desired will be pointed out inthe appended claims.

I claim:

1. In a band pass coupling system for transmitting frequencies within a given frequency band, an input circuit including acoil, an output circuit including a second coil inductively coupled tosaid first coil, a third circuit including a third coil inductively coupled to said first and second coils, and means including said couplings for tuning said circuits so that the frequencies within said band are transmitted with substantially uniform response while at a frequency close above said band the current in the third circuit has a lagging phase angle approaching degrees, whereby the currents in the input circuit and said third circuit differ in phase by an angle approaching degrees and the voltages induced in the output circuit approximately cancel each other.

2. In a band pass coupling system for transmitting frequencies within a given frequency band, an input circuit including a coil, an output circuit. including a second coil inductively coupled to said first coil, a third circuit including two coils which are inductively'coupled to said first and second coils, respectively, all said couplings having the same polarity, and means including said couplings for tuning said circuits so that frequencies within said band are transmitted with substantially uniformresponse while at a frequency close above said band the current in the third circuit has a lagging phase angle approaching 90 degrees, whereby the currents in the input circuit and said third circuit differ in phase by an angle approaching 180 de grees and the voltages induced in the output circuit approximately cancel each other.

3. In a band pass coupling system for transmitting. frequencies within a given frequency band, an input circuit including a coil, an'output circuit including a second coil inductively coupled to said first coil, a third circuit including a third coil inductively coupled to said first coil and a fourth coil inductively coupled to said second coil, said last mentioned coupling being negative and all other couplings being positive, and means including said couplings adjusted to cause the current in said third circuit to lead its voltage by a phase angle approaching 90 degrees at a frequency close adjacent the lower side of said frequency band, whereby the currents in said input and third circuits are approximately in phase at such frequency and the voltages induced in said output circuit approximately cancel each other.

4. In a band pass coupling system for transmitting frequencies within a given frequency band, an input circuit including a coil, an output circuit including a second coil inductively coupled with said first coil, a third circuit including a third coil inductively coupled with said second coil and a, fourth coil inductively coupledwith said first coil, said last mentioned coupling being negative and all other couplings being positive, said last mentioned coupling and the negative polarity thereof tending to produce a 2'70 degree phase shift between the currents in said third circuit and said input circuit, and

means for reducing said phase shift to approximately 180 degrees at a frequency close adjacent the lower side of said frequency band, whereby the voltages induced in said outputcircuit ap-- proximately cancel each other.

5. In a band pass coupling system for transtively coupled to said input circuit and capacitatively coupled to said output circuit, and tuning means for said circuits including said couplings so adjusted that substantially uniform transmission is obtained over said frequency band whereas at a particular frequency outside said band voltages are produced in said output circuit which substantially cancel each other.

6. In a transformer for a band pass coupling system, four coils arranged in pairs and so positioned relative to each other that the coils of each pair are inductively related whereas the coupling between coils of different pairs is substantially zero, an input circuit including a coil of one pair, an output circuit including a coil of the other pair, two link circuits including, respectively, the other coils of said pairs, and condensers coupling said link circuits, respectively, to the output and input circuits.

7. In a transformer for a band pass coupling system, four coils located in spaced parallel planes on a common axis, said coils being movable along said axis to vary the coupling between adjacent coils, input and output circuits including the second and third coils, respectively, a circuit including the first and fourth coils, for transferring energy from the input to the output circuit in addition to that transferred by the coupling between said second and third coils and means for tuning each of said circuits.

8. In a transformer for a band pass coupling system, five coils located in spaced parallel planes on a common axis, said coils being movable along said axis to vary the coupling between adjacentcoils, input and output circuits including the second and fourth coils, respectively, a circuit including the third c011 inductively coupled thereby to said input and output circuits, a circuit including the first and iii h coils inductively coupled to said input and output circuits by said first ing each of said circuits.

9. In a band pass coupling system for transmitting frequencies within a given frequency band, a tuned ouiput circuit, second and third and fifth coils, respectively, and means for tuncircuits inductively coupled to said output circuit, said couplings having unlike polarities,

means for dividing an incoming signal between said second and third circuits in such manner that the voltage components in said second and third circuits are in phase", and tuning means including said couplings for tuning said second and third circuits to difierent frequencies within said frequency band, said tuning means and other constants being so adjusted that in the vicinity of a predetermined frequency adjacent said frequency band the phase angles by which the can rents in the second and third circuits are shifted from their respective voltages are changing slowly and are equal at said predetermined frequency, while at frequencies within said band said currents differ in phase, the phase difference being so related to the magnitude of the currents that substantially uniform transmission is obtained over said band.

10. In a band pass coupling system for transmitting frequencies within a given frequency band, a tuned output circuit, second and third circuits inductively coupled to said output circuit, said couplings being of like polarities, means for dividing an incoming signal between said second and third circuits in such manner that the volt age components in the two circuits are degrees out of phase, and tuning means including said couplings so adjusted that in the vicinity of a predetermined frequency adjacent said frequency band the phase angles by which the currents in the second and third circuits are shifted from their respective voltages are changing slowly and are equal at said predetermined frequency.

11. In a band pass filter, an output circuit, second and third circuits coupled to said output circuit, means for dividing incoming signal currents between said second and third circuits, and means for tuning said second and third cir- -cuits.to different frequencies within a desired fredue to the difference in phase relation of the currents in the second and third circuits while near the extremities of the band the transmission is mainly due to the difference in the magnitude of said currents, the adjustment of said tuning means and other constants being such also that at a-fre'quency close outside said band the'said currents have the proper phase relation and mag.-

' niiude to induce equal and opposing voltages in said band at which they are equal in magnitude.

said last means being also effective to cause said currents to coincide in phase at a selected frequency outside said band, and means for adjusting the relative magnitudes of said currents so that at said selected frequency the opposing voltages induced in said output circuit are equal.

13 In a band pass coupling system for-transmitting frequencies .within a given, frequency band, an input circuit, second and third circuits connected to said input circuit byinductive couplings, said couplings being of like polarity, an output circuit connected to said second and'third circuits by inductive couplings, said last mentioned couplings being of unlike polarity; means for tuning said input and output circuits and the a said second circuit toflx the limits of said frequency band, means for tuning said third circuit to a frequency within said band but different than the frequencyto which the second circuit is tuned, means for adjusting the impedance resistance ratios of saidcircuits to different values correlated with the tuning and coupling factors,

whereby the currents in the second and third circuits are caused to differ in phase throughout 14. A band pass couplingsystem as claimed in claim 13, wherein the couplings of unlike polarity are associated with the input circuit rather than the output circuit, whereby the phase difference which exists between the currents n the in which I represents the f is to be located below or above said band, and

' second and third circuits at any frequency is v augmented by degrees.

15. In a band pass filter system for transmitting frequencies within a given frequency band,

first and third circuits constituting input and output circuits, respectively, second and fourth circuits disposed in parallel relation between said input and output circuits, inductive couplings connecting the input circuit with said second and fourth circuits, inductive couplings connecting the second and fourth circuits with said output circuit, one of said couplings being negative, and tuning means in said circuits so adjusted that substantially uniform. transmission is secured over said frequency band while at a frequency of maximum attenuation adjacent said band but outside thereof the currents in the second and fourth circuits induce substantially equal and opposing voltages in said output circuit. the frequency of maximum attenuation being determined ,by the equation 1 frequency of maximum attenuation, I: and if the resonant frequencies of the second and fourth circuits, respectively, and Q: and Q4 the impedanceresistance ratios of the second and fourth circuits, respectively.

16. A transformer for transmitting signals over I a given frequency band and having a non-symmetrical transmission curve showing a frequency of maximum attenuation only on one side of said band but either above or below the same, said transformer comprising an output circuit, second and third circuits coupled to said output circuit, means for dividing incoming signals between said second and third circuits, tuning means including the mutual couplings, capacity, inductance, and resistance of said circuits, the second and third circuits being tuned to different frequencies in said band and the second or the third circuit having the higher frequency depending on whether the frequency of maximum attenuation maximum attenuation but not at frequencies immediately above or below the same the currents in said second and third circuits are so related in phase and magnitude that substantially equal and opposing voltages are induced in said output circuit;

RIN'ALDO DE COLA.

Images