US1965522A - Tuned radio frequency amplifying system - Google Patents

Description

July 3, 1934. 1.. JONES TUNED RADIO FREQUENCY AMPLIFYING SYSTEM 2 Sheets-Sheet 1 Original Filed May 27, 1924 INVENTOR r L iones AT RNEYS i m N I 3 I Q wn mm mm nu W FT:

July 3, 1934. 1.. 1.. JONES 1,965,522

TUNED RADIO FREQUENCY AMPLIFYING SYSTEM Original Filed May 27, 1924 2 Sheets-Sheet 2 Patented July 3, 1934 UNITED STATES PATENT oFFICE TUNED RADIO FREQUENCY AMPLIFYING SYSTEM Lester L. Jones, Oradell, N. J., assignor of onethird to Maxwell James, New York, N. Y.

5, 1928, Serial No. 244,602.

In Canada June 13 Claims. (01. 179-171) This invention relates to tuned cascaded circuits embodying one or more electron discharge devices or tubes, and relates more particularly to an improved tuned radio frequency amplifying 5 system.

The invention has special reference to the provision of a tuned radio frequency amplifying system designed and constructed for producing a maximum amplification efficiency combined with a maximum selectivity over a large wave length range, and for permitting efficient controlling, neutralizing or compensating of the capacitive feed back reaction which is due to the capacitive coupling between the grid and plate of the tube.

This application is a continuation in part of the following of my copending applications: Capacitive coupling control system, Ser. No. 607,- 046, filed Dec. 15, 1922; and variometer, Ser. No. 614,404, filed Jan. 23, 1923, now Patent 1,664,513, granted April 3, 1928, and is a division of my application to Tuned radio frequency amplifying system, Serial No. 716,124, filed May 27, 1924, now Patent 1,712,214, granted May '7, 1929.

A prime desideratum of my present invention centers more specifically about the production of an amplifying system employing tunable circuits coupled in cascade in which the tuning of one of the circuits is accompanied by a predetermined change in the coupling coeflicient between said circuit and an adjacent circuit in order to obtain the maximum amplification efiiciency together with a constant and the greatest possible selectivity for each wave length.

A further prime object of my present invention 35 comprehends the provision of an amplifying system of the nature referred to embodying one or more electron discharge tubes in which the capacitive coupling coefficient between the input and output circuits of the tube and the voltage on 40 'the plate are maintained constant over the whole wave length range, producing a system in which the energy retransfer from the output tothe input circuit due to the grid-plate capacity of the tube is held invariant. Further and correlated objects of the invention include the provision of tuned radio frequency amplifying circuits in which the tuning is accomplished without produoing any material variation in the shunt capacity of the grid or input circuits of the system, and the provision of a system of this nature in which the first input circuit is rendered independent of the size of the antenna and is virtually decoupled therefrom.

A still further and correlated prime object of my present invention centers about the production of tuned cascaded circuits embodying one or more electron discharge devices in which any undesirable retransfer of energy from the output to the input circuit of the tube due to the gridplate capacity is effectively controlled or compensated for in a uniform mannerover the whole wave length range of the system.

To the accomplishment of the foregoing and such other objects as will hereinafter appear, my invention consists in the elements and their relation one to the other as hereinafter particularly described and sought to be defined in the claims; reference being had to the accompanying drawings which show the preferred embodiment of my invention, and in which:

Fig. 1 is a wiring diagrammatic view showing some of the principles of my invention applied to a radio receiving system of the tuned radio frequency amplifying type, and showing the use of my variometer and variotransformers for linking the circuits,

Fig. 2 is a View showing diagrammatically the construction and arrangement of the interlinking variometer and variotransformers,

Fig. 3 is a view showing the manner in which the parts of the variometer are preferably connected,

Figs. 4. and 5 are views respectively of the design of variotransformers employed in the first and second radio frequency stages of the system, and

Fig. 6 is a wiring diagrammatic view of the radio receiving system having two stages of radio frequency amplification and one stage of audio frequency amplification showing the principles of my invention applied for producing inductive tuning of the interlinked circuits and for controlling the capacitive feed back of the electron discharge devices.

Referring first to Fig. 1 of the drawings, I show the principles of my invention applied to a radio receiving system of the tuned radio frequency amplification type embodying preferably though not necessarily two stages of'tuned radio frequency amplification. This system comprises a plurality of electron discharge devices a, a and a each includinga cathode or filament 20, a grid 21, and an anode or plate 22,. Theelectron dischargedevice a is provided with an input or grid circuit generally designated as 1 connected to an energy receiving circuit such as the antenna circuit in, and is provided with an output or plate circuit generally designated as 0 which is linked to an input or grid circuit i associated with the electron discharge device a by means of a variotransformer V described more in detail hereinafter and embodying my invention as disclosed and claimed in the aforementioned copending application Ser. No. 614,404.

The electron discharge device a is also provided with an output or plate circuit o which is linked to an input circuit 2' associated with the electron discharge device a by means of a second variotransformer V The electron discharge device a and the linking structure V forms the first radio frequency amplification stage, and the electron discharge device a and the linking structure V forms the second radio frequency amplification stage, the electron discharge tube or device a forming the detecting stage, this tube being provided with the output circuit or telephone receiver circuit 0 In the preferred embodiment of my invention, each of the input circuits is inductively tuned, and preferably to the frequency of the radio waves desired to be selected. For inductively tuning the first input circuit 2' which connects the grid 21 to the negative terminal of the battery A, the input circuit is provided with the variable inductance or variometer V of my invention, the said variometer comprising a stator S and a rotor R-each composed preferably of a plurality of coils of the double D type as hereinafter described more in detail, the construction diagrammatically depicted in Fig. 1 showing the double D sections of the rotor R and the stator S arranged coaxially and connected across the input circuit, the opposite ends of the variometer being connected to the antenna n and to ground 9.

For inductively tuning the input circuit 12, the variotransformer V comprises a variable secondary L2 having a construction similar to that of the variometer V and including the coaxially arranged rotor R and stator S each of the double D section type, the said variable secondary being connected to the grid 21 of the tube a by means of the conductor 23 and to the negative terminal of the battery A by means of the conductors 24, 25 and 26. The primary of the variotransformer V comprises a coil L1 which is inductively related to the variable secondary L2 and preferably coupled to the stator coil S thereof and arranged coaxially with both the stator and rotor of the secondary, all as will be described more in detail hereinafter and as is diagrammatically depicted in Fig. 1. The primary L1 is arranged in the output circuit 0 of the first tube a and is connected to the plate 22 by means of the conductor 27 and to the positive terminal of the battery B by means of the conductor 28, the A and B batteries being connected together by the conductor 29as shown in the drawings.

For inductively tuning the input circuit 1 the variotransformer V is provided with the variable inductance or secondary L2 which includes the rotor R and the stator S of the double D section type having a construction similar to that heretofore described in connection with the variotransformer V, the said secondary L2 being connected to the grid 21 by means of the conductor 30, the integrating device 31 and conductor 32, and being connected to the positive terminal of the A battery by the conductors 33, 34, 35 and 36. The primary L1 of the variotransformer V is also arranged coaXially with the rotor and stator of the secondary as diagrammatically depicted in Fig. 1, and is also composed preferably of a coil of the double D section type connected to the plate 22 of the tube a, by means of the conductor 37 and to the positive terminal of the B battery by means of the conductor 38.

The output circuit 0 is provided with the current detecting means such as the .telephone receivers 39 connected at one end to the plate 22 by means of the conductor 40, and at the other end to the positive terminal of the B battery by means of the conductor 41, the usual. condenser 42 being connected in shunt to said phones. For heating the filaments of all the electron discharge devices to incandescence to produce the electron emission, the same are connected to the A battery, the filament 20 of the device a being connected to the terminal of the battery by means of the conductors 43 and 36, and to the other terminal of the battery by means of the conductor 44, rheostat 45 and conductors 46 and 26, the filament 20 of the device a being similarly connected to the battery A and to the rheostat 45 by means of the conductors 47 and 48, while the filament 20 of the device a is connected'to one terminal of the A battery by means of the conductors 49, 34, 35 and 36, and to the other terminal by means of the conductor 50, rheostat 51, and conductors 52, 25, 46 and 26.

As heretofore stated, a prime object of my present invention centers about the production of a system employing tunable circuits coupled in cascade, in which the tuning. of one of 'the circuits is accompanied by a predetermined change in the coupling coeflicient between said circuit and an adjacent circuit. I have" discovered by empirical determination in tuned radio frequency amplifying systems, that in order to obtain the maximum amplification efliciency, together with a constant and the greatest possible selectivity for each wave length, the coupling coefficient between adjacent circuits should be varied in the same direction and progressively with the wave length variations so that at the smaller waves the coupling should, be relatively loose, and at the longer waves relatively close or tight.

To accomplish the desired results, in the preferred embodiment of my inventionthe coupling between the cascaded circuits is made of the inductive type and the tuning of the circuits is effected inductively as already explained. Although in the preferred embodiment I show the inductive coupling combined with the inductive tuning, it will be understood however that the invention is not limited to the inductive coupling and tuning type, and that the principles of the invention may be carried out with other types of coupling and tuning. The inductive type is preferred, however for the numerous advantages offered thereby, which will become apparent as the description of the various features of my invention proceeds.

In this preferred inductive coupling and inductive tuning of the invention, the desired coupling coeflicient and tuning variations are pro,- duced by the design and construction of the variotransformers used in the first and second tuning stages. These .variotransformer units generally shown in Fig. 1 are diagrammatically, depicted in Figs. 4 and 5 of the drawings, Fig. 4 relating to the variotransformer used in the first stage, and Fig. 5 to that used in the second stage, the construction of the variable secondary of these variotransformers being more accurately illustrated'in Fig. 2 of the drawings, and the connection thereof in Fig. 3 of the drawings. 1

Referring first to Fig. 2 of the drawings, the variable secondary or variometer constructionally comprises a rotor including-a'set-of coils and a stator also including a set or co'ils,' the coils of the stator set and the coils of the rotor set being arranged in interleaving alternating relation after the manner of the stator and rotor plates of a variable air condenser, and as shown diagram- "matically in Fig. 2 of the drawings, the rotor R The rotor'a'nd stator coils, as will be understood,

are arranged in closely spaced superposed relation, and are shown in Fig. 2 displaced laterally only for purposes of clarification and for facilitating the tracing of the circuits in the'independent stator and rotor coils.

The stator coils areeach wound so as to produce a winding in a single plane, the winding comprising two coil sections for producing magnetic fields in opposite directions, each of the coils to this end comprising the two coil sectionslO and wound in opposite directions and presenting a double D 'formation, this being indicated by the arrows showing the momentary direction of flow of current in the windings to produce a field in the coil section 10 in one direction as indicated by the .circles 11 and a magnetic field in the coil section 10' in the opposite direction as indicated'by the cross 11. The stator coils are connected together so that their mutuals add, and to this end the oppositely positioned coil sections 10 and 10' are electrically connected by conductors 12 and the end coils are connected to the terminals 13 and 13 by means of the conductors 12a and 12b respectively. With this arrangement and interconnection, it will be seen that the magnetic flux threads up through the coil sections on one side and down through the coil sections on the other, producing a magnetic field which is closed substantially within the confines or dimensions of the coils, thus minimizing any external magnetic field.

The rotor coils T1 to T4 are each similarly constructed, comprising two sections 14 and 14' producing a double D coil, the sections being wound in opposite directions as clearly indicated by the arrows in the figure to produce oppositely directed magnetic fluxes as shown by the circles 15 and the crosses 15.

Theserotor coils arealso interconnected so that the mutuals therebetween' add,

and to this end some of the coil sections 14' are connected by means of the conductors 16 and some of the coil sections 14 are connected by means of 'the conductor 16, the end coils being connected to the said axis.

the terminals 1'7 and 17 by means of the conductors 16a and 16b respectively. To obtain the desired direction of the fluxes with this interconnection, it will be noted that the coils T2 and T4 are inversed with respect to the coils n and T3, the connection between the coil sections of the firstsetofcoils .being on one side of the rotor axis, and the connection between the coil sections of the other set being shown on the other side of With this arrangement it will be also noted that the magnetic flux threads through all the rotor coil sections in a manner to minimize an external magnetic field.

Still referring to Fig. 2, the rotor coils are shown arranged in alternation with the stator coils, and when the same are superposed in the position shown in Fig. 3, the mutuals oppose to produce a minimum inductance. With thisconstruction it will be seen that as the'rotor is rotated from the position shown to a position of 180 thereto, the mutuals will change from' full opposing to full aiding, and in the latter position the fields of shown in Fig.2 and described hereinbefore.

equal magriitude'will aid each other to produce a maximum inductance value. ToaccompIish this latter end, the rotorcoils are preferably made equal in number tothe stator coils, and are similarly wound to produce like physical dimensions.

The-stator and rotor coils may be connected either in parallel or in series, and when it is deby the condenser C being somewhat greater than that for the parallel connection.

Referring now to Figs. 4 and 5 of the'drawings, I show the preferred design and construction 'of the variotransformers, each consisting of the primary L1 of the double D type and the secondary L2 which is in all respects like the variometer The double D rotor and stator coils are shown in Figs. 4 and. 5 as 1'1, 1'2, r3, m; and s1, s2, s3 and $4, in a conventional manner for purposes of simplicity, it being understood, however, that the design is of the double D type more accurately depicted in Fig. 2. It will be noted that for the first stage variotransformer (Fig. 4) the primary L1 is arranged at the end of the unit and is coupled to a stator coil (.91), while for the second stage variotransformer (Fig. 5) the primary L1 is'arranged between the end stator and rotor coils s1-r1, and is coupled to the statorcoil for reasons that will become clearer hereinafter.

To produce the desired coupling and tuning variations, the variotransformer units V and V must be accurately-designed, and this includes a determination of 1. The maximum and minimum inductance of the secondary L2,

2. The positioning of the primary or plate'coil L1, and

3. The magnitude of the mutual M between the primary L1 and the secondaryLz and the resulting value or magnitude of the primary, each of which is determined by one or more factors as follows.

(1) Factors determining the -maatimum "and mimmum values of the inductance of the sec ondary The maximum-wave length desired and the grid shunt capacity to be used, as hereinafter explained, determine the maximum inductance -ofthe secondary of the variotransformer; and in the preferred construction thisis made 640 michrohenrys. termined by the shortest wave length required; and this minimum is obtained by reducing the The minimum inductance is de- '8;5-to' l isused,- this being found-ampleand efficient for covering the broadcast range of wave lengths.

(2) Factors determining the position of the primary or plate coil The position of the primary or plate coil relative to the variometer or-sec'ondary coils depends the variation of the mutual inductance is with a higher power of the wave length, the nearer the plate coil is to the center of the variometer coils.

For example, in the variotransformer of my invention I provide two primary coils L1 and R-P (see Figs. 4 and 5) only one of which is used as the plate coil. One of these coils is at the outer end of the variotransformer and is separated from the other by the rearmost stator coil 31. The second primary coil is therefore between the rearmost stator and a rotor coil of the variometer. The manner in which the mutual inductance varies with the Wave length depends upon which of these two primary coils is used as the plate coil.

In the first stage variotransformer (Fig. 4) I use, as already stated, the outside coil as the plate coil, with the result that the mutual inductance varies somewhat faster than proportionately with the wave length and therefore the coupling coefficient increases less than proportionately with the wave length, as will be seen from the following tabulation:

Actual mutual Theoreti- Variometer in cal micromutual henrys Maximum 97 97 90 58 72 Minimum 25. 5 33' 60 and in this second stage variotransformer, the inside primary coil is used for the plate coil, with the following results.

Actual mutual Theoreti- Variometer in cal micromutual henrys Maximum 128 128 Minimum 12.85 15 It will therefore be seen that in the one case I use a coupling coefficient which increases slightly with the wave length, and in the second case I use a coupling coefiicient which increases almost exactly proportionately with the wave length, and also that the variation is sharper the nearer the plate coil is to the center of the variometer unit.

(3) Factors determining the value of the mutual inductance and inductance of the plate or pri-- mary coil Having determined the position of the plate coil, from the above considerations (namely the type of variation of mutual inductance or coupling coeflicient desired) I next fixthe inductance of the plate coil. This is selected in general so as to get a high amplification efficiency at the long (or any) wave length, and so as to avoid so high an inductance as would create a primary circuit nearly resonant to the shortest wave length. Such a resonant circuit might be formed from the plate and compensator coils shunted by the tube capacity and a portion of the previous input circuit inductance and capacity.

- For determining the magnitude of the mutual M,

recourse must first be had to the factors of amplification eificiency and selectivity.

(a) the factor of amplification cificiency and conditions governing the same I have empirically discovered that the highest energy amplification from practically any type of tube is had by inserting an impedance in the output circuit of the tube equal to approximately one half the plate filament impedance of the tube. Stated in terms of conductance, the input conductance of the output transformer at resonance is arranged so that it is substantially higher in value than the plate conductance of the vacuum tube. Present day tubes operating with proper C batteries have plate filament impedances ranging from 20,000 to 40,000 ohms so that the tube output circuit impedance should be approximately 10,000 to 20,000 ohms and I select an average of 15,000 ohms.

In order to have high amplification efficiency in the tuned amplifier operating over a broad wave length band, it is highly desirable to hold or maintain this impedance of the output circuit of the tube up to approximately 15,000 ohms selected. The impedance of the primary coil is relatively low, having usually an average value of about 500 ohms for the broadcast wave length band. The greater part of the output impedance must therefore be the equivalent secondary impedance reflected into the primary. When the secondary is tuned to resonance the resistance reflected in the primary is:

where M is the mutual, w is 2 1r times the oscillation frequency and R2 the ohmic resistance of the secondary circuit. R2 being determined by considerations of sharpness of tuning, this equa tion determines M for any given wave length, and therefore the inductance of plate or primary coil. To maintain high amplification efiiciency this reflected resistance must be held constant over the wave length range, to the output impedance empirically determined, namely, 15,000 ohms or more (up to 30,000 ohms). With this value of the reflected resistance empiricallyarrived at, the values of the mutual M and the sulting decrement will be .0116.

ohmic'resistance R2 of the secondary may. now be determined.

(b) the factor 'ojselectivity and conditions youerning the same 'Before completely determining the value of M, it"is necessary to consider the secondary resistance because this is intimately related to the selectivity of the amplifier.

Assuming that it is desired to make each stage as selective'as possible,'it'is necessary to'find how far to go. General theory teaches that in order to amplify modulated waves without distortion, the amplifier must transmit all of the wave frequencies carrying the modulated energy. This band of frequencies consists of that group ranging from the carrier frequency less the highest modulation frequency, to the carrier plus the highest modulation frequency. In a practical way this band is'one lying 3,000 cycles each side of the car-, rier frequency. This condition may be translated into my tuned amplifier design by postulating that the resonant circuit shall be so broadly tuned that the highest and lowest frequenciesof this band shall'produce' voltage effects no less than of the effect produced by the mean or carrier frequency. To further relate this condition to the resonant circuit of my invention, I may say that the decrement of the resonant circuit for this broad casting band should be of the order of .02 to satisfy this. condition. For example, if a resistance of 2%; ohms is given to thesecondary circuit at 200 meters, where the inductance is microhenrys and the capacity 150 mfds, the re-. The sharpness of resonance (i. e. the detuning'factor for voltage effects) of such a circuit corresponds to 2 620 cycles change of frequency at this wave length and to only 850 cycles at 600 meters showing that such low decrement would produce appr'eciabledistortion on the long waves. Itwill be seenyther'efore, that the amplifier should be operated with appreciable ohmic resistance in the secondary circuit.

Having found by experiment, that approximately 2 /2 ohms of real ohmic resistance in the secondary circuit at'20fi meters is required, for stable operation, Iinse'rt this value in Equation (1), and find ,that the necessary mutual is approximately 20 m'icrohenrys. This value together with the secondary inductance and'the predetermined position of the plate coil as hereinbefore arrived at determines the inductance of the plate coil. In mypresenst embodiment this inductance is about microhenrys." e

The foregoing considerations aidin explaining the physicalbasis of the feature of my invention which, relates to the simultaneous increase of coupling coefficient with wave length. In the inductively tunedsecondary the resistance of the secondary, besides being appreciable as above indicated, should increase with the wave length. The increase of this resistance is determined by two factors, namely, the characteristic of constant selectivity and the characteristic of the greatest possible selectivity for each wave length. For obtaining constant selectivity, the log. dec. should be constant, and for obtaining the greatest possible selectivity on all Wave lengths, the log.

dec. should vary proportionately with the wave length, as given in the following equation:

o i For constant-selectivity since L2, the inductance of the secondary increases asthe square of the wave lengtlnR must increase with the wave length to hold this log. dec. constant. For obtaining-the greatest possible selectivity-on all wave lengths and the log. dec. varying proportionately to the wave length as stated, it follows that R must vary as the square of the wave length.

The resistance R of the secondary is composed of three parts:

1. The ohmic resistance of the secondary circuit which is designated R2.

2. The resistance of the primary circuit (substantially the filament plate resistance of the tube) reflected into the secondary circuit which is designated as R2.

3. The negative resistance introduced into the secondary through the capacitive feed back action of the succeeding plate circuit. This is introe ducedin the form of a voltage acting onthe secondary circuit in phase with the secondary current. This negative resistance is designatedas R2P. Y

I have found that good results are to beob tained when the resistance 'refiectedfrom the primary into the secondary (R2') is of the same order of magnitudeas theohmic resistance of the secondary. 1 By this I mean that R2 shall be from 1 to 2 times R2. This division leads to freedom from local oscillations through all practical variations of the tube impedance. The sum of these two is reduced by 50% to 75% because of the subtraction of R2P.

Now ifwe refer to Equation (1) M V T; we see that in order to obtain the constant selectivity, the mutual (M) must vary as the 3/2 power of the wave length, since R2 (which with the values selected substantially equals R) varies as the wave length. It is also apparent from this Equation (1) that to obtain the greatest selectivity atany wavelength, the mutual M must vary as the square of the wave length, since R2 varies as the square'of the wave length.

It therefore will be seen that for each stage, to obtain the optimum conditions, the I mutual should vary between the 3/2 power andthe square of the wave length. In the'practical embodiment of the invention this range of values is adopted, and as heretofore shown, I select for the second tuning stage of the system a mutual varying as the square of the wave length, while for the first tuning stage I select a mutual which varies somewhat less than the 3/2 power of the wave length; the lesser value being selected for the purpose of compensating for changes? in the losses of the first input circuit which take place with change in wavelength, as will be further explained hereinafter.

The above con'siderations'are theoretical and apply more closely to practical cases the greater the number of cascaded stages, assuming that all external feed back be eliminated, this for the reason that all 'intermediate'stages have input circuit resistances which are reduceable by an amount corresponding to the tube capacitive feed back reaction, whereas the laststage usually has a higher decrement than the intermediate stages. These theoretical considerations areimportant, however, because the selectivity of the system tends to become that power of the selectivity of the average of the selectivities of each stage corresponding to the number of stages. 1 r v The principles of my invention thus far described show the adaptability of my system to tuned radio frequency circuits in which the capacitive feed back reaction due to the grid-plate capacity is compensated for, neutralized or suitably controlled.

In cascaded circuits in which the secondary circuit is capacitively tuned, the capacitive coupling coefficient between the grid and plate circuits of a tube where Cgp is small compared with C -|.C'p) instead of remaining constant, will vary With the tuning of the secondary,increasing sharply with the'decreasing Wave length, with the result that the energy retransfer or feed back increases, this energy retransfer being proportional to K' V where V is the voltage on the plate. The ideal condition is where both K and V are constant, since it is desired that the feed back due to the capacitive coupling be a constant residual amount (about 3 micro-micro-farads) just sufficient to make up for losses in the input circuit of the tube, and not more. Capacitive feed back has the effect of decreasing the apparent resistance of the input circuit and thence increasing the current and voltage therein. This should be increased up to a certain point and not beyond,

since the more the current is increased, the louder the signal and the more selective the circuits, until a limit is reached, the limit being the stage at which incipient and sustained oscillations take place. The ideal system, therefore, is one in which the energy retransfer (K' V) is maintained constant over the wholewave length range of the system; and the capacitively tuned systern does not meet the requirements because of its variable capacitive couplingcoefiicient K.

In prior systems in which the capacitive gridplate feed back is controlled, not only is the value of K variable, but the plate voltage V is variant and in the same direction as K, producing a squared. increase of energy feed back with decreasing wave. length, instead of holding the energy retransfer constant as desired. This is due to the factthat the inductive coupling coefficient between the coupled or linked circuits is held constant, causing the output impedance in the plate circuit to decrease with the wave length and consequently the plate voltage increases with the decreasing wave length. The combination of these two factors, therefore, produces a negative resistance reaction in the grid circuit which diminishes at least as fast as the square of the wave length, leaving the input circuit with an extremely, high decrement on the longer waves.

With the principles of my system, however, I am enabled to design the circuits so as to maintain both the capacitive coupling coefficient K and the plate voltage V constant over the whole wave length range, producing a. system in which the energy retransfer due to the grid-plate capacity of the tube may be held substantially-invariant over the Whole wave length range, and maybe controlled for tubes of different gridplate characteristics.

The features of my system which permit of this capacitive feed back control are the inductive tuning and the inductive coupling coefiicientv variations produced. By means of my inductively: tuned circuits, I am enabled in a manner to be described presently to maintain the capacity of the tunable circuit constant during tuning thereof, permitting the capacitive coupling coeficient K to be held constant, and by means of the above described variation in the inductive coupling coefiicient K, the output impedance (the reflected resistance R1) is constant over the whole wave length range, and this latter produces a plate voltage which remains constant, the plate voltage being the product of the output impedance and the oscillating plate current. The plate voltage herein referred to is not the direct current plate voltage but the alternating voltage on the plate which is the product of the output impedance and the oscillating plate current.

To obtain the optimum conditions for holding the capacitive coupling coefficient K constant over the whole wave length range, correlated to the inductive tuning of the input circuits I provide means for producing a shunt capacity in each of the input circuits which remains substantially constant during the tuning of the same, this being provided for the antenna input circuit z and for the first and second stage input circuits i and i To achieve the desired results the total shunt capacity of each of the input circuits is made from ten to twenty times and preferably fifteen times the grid-plate capacity of the tube. I have determined that this may be accomplished by placing a condenser in shunt withthe secondary L2 of the variotransformers V and V of a value of about micro-micro-farads to bring the total shunt capacity of the input circuits 1" and i to about 150 micro-micro-farads and referring to Fig. 1 of the drawings, these shunt condensers are indicated respectively as m and m respectively.

The provision of these shunt condensers each having the above stated comparative magnitude possesses in general the following advantages:

(a) It makes the wave length calibration of the circuit independent of the tube capacity,

(12) It creates a capacitive feed back coupling of the right order of magnitude which is a coupling that will neutralize the input circuit resistance due to the feed back reaction,

(0) It offers a ready means of adjusting the total circuit capacity to, apredetermined value. This is commercially important in order to have all of the stages of the tube amplifier calibrate the same way.

Although I have selected a value of 150 micromicrofarads for the total shunt capacity, it will be understood that this may be varied within limits. Experimentally I have found that the best results are to be secured by using a fixed capacity across the grid circuit of approximately 100 micro-micro-farads. The useful range for present-day amplifying tubes and for the engineering materials available for the construction of the tuned transformers is from 50 to 150 micro-micro-farads. This applies more particularly to the present broadcasting range of 200 to 5'75 meters; and for higher wave lengths the shunt grid capacity should be increased approximately proportionately with the wave length. In the present form of my invention I employ, however, a total shunt capacity of 150 microrriicro-farads as stated. A higher efiiciency may be secured by the use of a total shunt capacity of 100 micro-micro-farads, but this optimum value is not desirable because of cost considerat-ions.

The selection of this value of shunt capacity is determined on the one hand by the fact that smaller shunt capacities are open to variation in calibration due to a change of vacuum tube, are subject to change of tuning due to change of filament current, and are subject to excessive capacitive feed back which is not easily controllable. The selection is determined on the other hand by the fact that larger shunt capacities make it extremely difficult to obtain the high amplification efliciency with the engineering materials commercially available. This is primarily due to the fact that the high output impedances (10,000 to 20,000 ohms) desired to match the output plate-filament impedance are diificult to build up with high capacity circuits.

The antenna input circuit 1' is similarly provided with a shunt capacity generally designated as m which is placed in shunt with the variometer V, the shunt capacity having a value of about 60 micro-micro-farads, this capacity functioning to render the tuning of the antenna circuit independent of any capacity changes.

In the preferred embodiment of the invention, the antenna or first input circuit 2 besides being constructed to maintain a constant capacity during tuning, is designed so as to be rendered independent of the size of the antenna, the construction including means for effecting a virtual decoupling of the first input circuit fromthe antenna n. This is accomplished by providing a condenser called a series condenser m in series with the antenna n and the variometer V, the size of this condenser being about the capacity of av small antenna, or the capacity of a large antenna. With this construction and with the total shunt capacity of the first input circuit 1, the capacity of the input circuit has a magnitude which is from four to' five times the capacity of the antenna and the series condenser combined. With this recited construction I have found that the variometer V tunes independent of the size of the antenna and with substantially constant coupling to the antenna over the whole wave length range, the capacity of the first input circuit being therefore localized to the variometer. Besides the many advantages of this tuning system hereinbefore described, the rendering of the capacity of all of the input circuits independent of the tuning permits the same dial settings for each of the variometer or variotransformer units, so that a mono-control system is afforded.

With the capacities of the input circuits all maintained invariant over the whole wave length range, it will therefore be seen that the antenna capacitive coupling coefficient is held constant and the capacitive coupling coefficients between the grid and plate circuits of both tuned radio frequency stages are also held constant, producing in combination with the constant plate voltage characteristic, a system in which the product K V is substantially constant for each stage; and with this product constant, and assuming other factors such as the resistance of the circuits constant, the energy retransfer or feed back due to the grid plate capacity may be held invariant. In the preferred embodiment of the invention, the energy retransfer in the second tuning stage of the system is held constant, while the energy retransfer in the first tuning stage is made to deviate somewhat from a constant value and more specifically to decrease with increase of wave length so as to compensate for the decrease of losses in the first input circuit which take place with increase of wave length, it being desired that the energy retransfer be a constant percentage of the losses in such input circuit. It is to compensate for such change or variation of losses that the mutual for the first stage is made to vary ata rate less than the 3/2 power of the wave length as heretofore set forth. The change in energy retransfer in this first tuning stage, however, While it is made to vary for compensating purposes, changes in the order of the power of the wave length and may be regarded as substantially constant. 7

This result of constant capacitive energy feed back produced with my system permits the employment of cascaded tube circuits without means for neutralizing the grid-plate capacity with tubes having a low grid-plate capacity such as 3 micromicro-farads. I have discovered that for such tubes and with shunt capacities of micromicro-farads the feed back is of the right order throughout the wave length range.

With tubes of larger grid-plate capacity and approximately 8 micro-micro-farads, the above described features of my invention may be com- 98 bined with capacitive feed back control means for compensating for the remaining grid-plate capacity (5 out of 8 micro-micro-farads), and for maintaining the retransferred energy in such tubes constant over the whole wave length range; 100 and this, as heretofore set forth, comprehends a further prime object of. my present invention.

The capacitive feed back control means which I combine with the inductive tuning and change of coupling coefiicient features of my invention 165 is that described and claimed in my copending application Ser. No. 607,046 heretofore referred to, the resulting system being shown in Fig. 6 of the drawings. In Fig. 6 of the drawings, I show a radio receiving circuit having two stages of radio frequency amplification similar to that heretofore described in connection with Fig. 1, the parts of which are indicated by similar reference characters, the system in Fig. 6 of the drawings further including an audio stage fol- 11 8 lowing the rectifying device a The amplifying circuits in Fig. 6 are substantially the same as that shown in Fig. l, and are also indicated by similar reference characters, the variometer V being of the type shown in Fig. 3, the variotransformer V that shown in Fig. 4, and the variotransformer V that shown in Fig. 5 of the drawings.

For controlling, compensating or neutralizing the excess grid-plate capacitive feed back, I provide means inductively related to the variotransformers V and V for producing a charge equal in magnitude but opposite in sign to the charge carried to the grid from the plate and for im-. pressingsuch charge on the grid of each ampli- 1 30 fying tube. This is accomplished by first creating a potential opposite to the potential on the plate and impressing the said potential on a condenser which is connected to the grid of the tube, the product of the potential and the capacityof the condenser being made equal to the excess charge transferred to the grid from the plate due to the inherent capacity therebetween.

Referring now to Fig. 6 of the drawings, I show such neutralizing means as comprising a coil R--P hereinafter called the reverse potential coil, inductively related to the end stator s1 of the variotransformer V, to which stator coil the primary L1 is inductively coupled, the construction being such that a potential is created by the R--P coil which is made equal in phase and amplitude and opposite in sign to the potential on the plate 22; and the reverse potential coil is connected by means of a conductor 55 to a plate (1 of a condenser 56, another plate eof 5 which is connected by means of the conductors 57 and 58 to the grid 21 of the tube a, the other end of the reverse potential coil being connected to the B battery by means of the conductor 28. The plate 6 of the condenser 56 preferably comprises a rotor so that the capacity of this condenser may be varied to make the same equal to that capacity between the grid and plate of the tube which is desired to be compensated for. The reverse potential coil may if desired be coupled to the primary L1, but in the preferred construction both the reverse potential and primary coils are similarly and equally coupled to the end stator coil 81 of the variotransformer, as clearly shown in Figs. 4 and 6 of the drawings.

For neutralizing the excess grid-plate capacity feed back of the tube a, the variotransformer V is similarly provided with a reverse potential coil RP, which however is coupled to the end stator coil s1 by being arranged at the end of the variotransformer unit as clearly shown in Figs. and 6 of the drawings, this reverse potential coil being connected at one end to the B battery by means of the conductor 38 and at the other end to the plate d of the condenser 56 by means of the conductor 59, the plate e of which condenser is connected to the grid 21 of the tube aby means of the conductors and 23. Here too it will be noted that both the primary coil L1 and the reverse potential coil RP are coupled equally to the end stator coil s1.

By the provision of this means, it will be seen that the grid of each tube is supplied with an electric charge which is equal and opposite to the charge thereon resulting from the grid-plate capacity, and it will be further noted that the neutralizing charge varies in correspondence with the varying charge on the plate in the operation of the circuits so that the energy feed back is compensated for in all the frequency variations of the grid and plate circuits.

'As above stated, the neutralizing condensers 56 and 56 are made adjustable for the purpose of varying the charge impressed on the grid. For

the purpose of preventing the variation of capacity of these condensers from varying the tuning tween the plates e and f is increased while the plate e "change in capacity between the plates e and characteristics of thegrid or input circuit, my invention further includes means for varying the capacity of the input circuit simultaneously and in correspondence with the varying of the capacity in the charge equalizing circuit. This result is accomplished in the preferred practice of my invention by making the condensers 56 and 56' in the form of three-element condensers each having an additional plate I which is connected to the filament end of the grid circuit by means of the conductors 61, 62 for the condenser 56 and conductors 63, 64, 24, 25 and 62 for the condenser 56. The construction of the plates 6, ,f and d of each condenser is such that when the is moved in one directiomthe capacity between the plates 6 and d is increased and the capacity between the plates 6 and f is correspondingly decreased, and as the plate 2 is moved in the opposite direction, the capacity becapacity between the plates 6 and d is correspondingly decreased. In view of the negligible impedance of the reverse potential coil as compared with the impedance of the grid coil, the

is normally effective for modifying the capacity of the grid circuit, and by this construction the grid circuit capacity may be maintained constant during adjustment of the plate 6 and hence during the variation of the charge impressed upon the grid. The three-plate condenser of my invention preferably has a construction described and claimed in my copending application Ser. No. 680,465 filed Dec. 13, 1923 patented Sept. 6, 1927, No. 1,641,438.

In addition to the two radio frequency stages, I show in Fig. 6 an audiofrequency stage comprising an amplifying tube a having an input circuit 2' coupled to the output circuit 0 of the tube a by means of a transformer t, the audiofrequency tube having an output circuit 0 provided with the usual telephone switch 65. mit a shift of the telephones from the audio to the detecting stage of the system, the output circuit 0 is provided with the usual three-point switching device 66.

From the above it will further be seen that with my inductively tuned system, a stability of feed back is obtained. The above considerations show that the total secondary'resistance R should increase with the wave length and that the percentage which remains after the tube feed back reaction should be approximately equally divided between real and primary reflected re sistances R2 and R'z respectively. With the engineering materials commercially available it is not always possible to obtain low enough values of R2 and R2. Advantage should be taken, therefore, of the reduction of resistance R2P due to the tube grid-plate feed back reaction. The great advantage of the inductively tuned system is that, as already remarked, with low capacity tubes the percentage reduction is approximately correct over the whole wave length range. With high capacity tubes using my feed back compensator system, a fraction of the feed back reaction may be left active and the same fraction is effective for the whole wave length range.

Two conditions should be satisfied in order to secure stable operation, by which is meant avoiding oscillation or incipient oscillation conditions of the amplifier. The first is that R2 plus R2 be always greater than RZP. The second is that R2 as above stated be comparable in magnitude with R2. The first condition means that the secondary circuit always has a net positive resistance; for, when the net resistance is zero or negative any oscillation set up in the secondary circuit persists indefinitely or increases its amplitude until the voltages exceed the linear limits of the tube characteristics. The second condition is one which permits the first to be practically realized. If a circuit. be constructed having low R2 and with a relatively high Rz and the sum of these be almost completely negatived by RzP, then the balance will depend primarily upon the constancy of R2. This reflected resistance, however, is again, dependent upon the filament plate resistance, which varies somewhat with the amplitude of the signal, with the B battery voltage, and quite sharply with the filament current. The usual and best way of controlling volume is by adjustment of the radio frequency amplification through the adjustment of the filament current.

Decreasing the filament current increases the electron emission, increases the plate filament resistance and therefore decreases R2. A slight variation of the filament current of the preceding tube will carry the net resistance of the secondary circuit down to zero or to a negative value with resulting oscillations.

As heretofore stated, it is a further principal object of my invention to provide a radio receiving set having a number of inductively tuned stages In order to per-' arranged in cascade, in which intermagnetic coupling between the stages is effectively obviated, permitting independence of adjustment for each stage and permitting the units of all the stages to be mounted closely together without producing magnetic coupling troubles. I have discovered that when the double D variometers of adjacent stages are turned at right angles one with respect to the other with their axes parallel or with their axes coaxial, substantially all residual magnetic coupling between the stages is effectively eliminated. It will be understood that the variometers of the double D type have little intermagnetic coupling to begin with, especially so when the variometers are set with their axes parallel, the residual coupling diminishing very rapidly with increasing distance between the variometer axes. When the double D coils of adjacent variometer or variotransformer units, however, are turned at right angles, this residual is practically nullified. The reason for this will be readily appreciated when it'is seen that if the coils of all the units were arranged-in the same way, the eifect of one of the D coil sections of one unit upon a D section of an adjacent unit would be different than the effect of the other D section of the first unit upon the said D section of the adjacent unit. Therefore by rotating one of the units 90 relatively to the other, the D section of one unit has the same coupling eifect on the 2 Us of the adjacent unit, resulting in magnetic neutralizing effects, the neutralization being the result of the fact that the magnetic fields in the two D sections of a given unit are in opposite directions. With this arrangement, therefore, all residual magnetic coupling between the units is practically eliminated or at least minimized to the lowest degree; and the units may be closely mounted without having magnetic coupling troubles.

Where a variometer and a plurality of variotransformers are arranged in series, as in Figs. 1 and 6, I have found that the optimum condition of elimination of residual coupling is reached when the D coils (that is to say the stator D coils and the rotor D coils for the same setting) of the intermediate unit are arranged at right angles to the D coils of the outside units. This phase of my invention is clearly depicted in Fig. 1 of the drawings, in which it will be seen that the coils of the variotransformer V are arranged at an angle of 90 to the coils of the variometer V and the variotransformer V Although the position of the rotor coils varies when the stage is tuned, it will be understood that the rotor coils of the different stages bear to one another the desired relation, since the circuits are designed, as hereinbefore fully set forth, to tune in the same way and with substantially the 1 same rotor settings.

There are four sources of regeneration or feedback in amplifiers, which cause local oscillations and distortion of tone. Or to put it less technically, there are four different channels through which some of the pressure of the amplified current from the plate circuit may push against the grid of that same tube. These are resistance coupling, direct magnetic coupling, inductive coupling, and capacitive coupling. Resistance coupling is mostly due to high internal resistance of the B batteries. is due to long wires connecting the B batteries. In these cases part of the voltage variation in the plate circuit of the last tube is communicated to the first tube because the B batteries and Direct magnetic coupling leads are common to both. Inductive coupling is due to magnetic linkages between radio frequency transformers and tuning coils or loop receptor. Capacitive coupling is mainly due to the electrostatic coupling between grid and plate of the vacuum tube.

These various feed-backs are eliminated by the various means disclosed herein. Inductive coupling is eliminated by arrangements whereby there are no magnetic linkages between the load or tuning coil and the radio frequency transformers or between the radio transformers themselves as above described. The capacitive coupling' is eliminated by the employment of the compensating means also described in detail above. Resistance and magnetic coupling are best eliminated by connecting a large condenser (about 1 mid.) across the A+ and 3+ terminals of the receiver or amplifier with the shortest possible connecting wires. This is shown in Figs. 1 and 6 of the drawings as the condenser connected across the A+ and 3+ terminals. This condenser 70 therefore shunts substantially all of the common portion of the output circuits and is of suiiicient capacity to keep the reactance voltage, tending to occur in the common portions of the output circuits, below the value at which it may cause self-oscillation.

The manner of making and using the tuned radio frequency system of my invention will in the main be fully apparent from the above detailed description thereof taken in conjunction with the operations as heretofore outlined. It will be further apparent that while I have disclosed in detail the construction and arrangement of the system and the parts thereof with the particular magnitudes or sizes of the capacities and inductances involved, that my invention is not to be limited thereto, and'that changes'may be made within wide limits as hereinbefore set forth without varying from the principles of my invention as defined in the following claims.

I claim:

1. The combination of an electron discharge device having a cathode, an anode and a grid, an input circuit connected to the grid, an output circuit connected to the anode, the said output circuit having an impedance which is equal substantially to one-half the plate-filament impedance.

2. The combination of an electron discharge device having a cathode, an anode and a grid, an output and primary circuit connected to the anode of the device and an input and secondary circuit coupled to the said primary circuit, the said circuits having resistances of a magnitude such that the resistance reflected from the primary into the secondary is of the same order of magnitude as the ohmic resistance of the secondary.

3. The combination of an electron discharge device having a cathode, an anode and a grid, an output circuit connected to the anode of the device and a next input circuit coupled to said output circuit, and means for tuning the said next input circuit, the said output circuit having an effective impedance which is substantially onehalf the plate filament impedance of the device.

4. The combination of a plurality of electron discharge devices each having a cathode, an anode and a grid and means for linking the devices in cascade including an output circuit connected to the anode of one device and an input circuit coupled to the output circuit and connected to the grid of the second device, means for tuning the said input circuit, the effective impedance of the said output circuit being substan tiallly one half the plate-filament impedance of the device to which said output circuit is connected.

5, The combination of an electron discharge device having a cathode, an anode and a grid, an output and primary circuit connected to the anode of the device and an input and secondary circuit coupled to the said primary circuit, means for tuning said secondary circuit, the said circuits having resistances of a magnitude such that the resistance reflected from the primary into the secondary is of the same order of magnitude as the ohmic resistance of the secondary.

6. The method of operating a tuned radio-frequency amplifier including a vacuum tube and an Out ut transformer associated therewith, which consists in arranging the input conductance of said output transformer at resonance so that it is substantially higher in value than'the plate conductance of said vacuum tube.

'7. In a tuned radio frequency amplifier, a vacuum tube and an output transformer pertaining thereto whose input conductance at resonance is substantially higher than the plate conductance of said vacuum tube.

8. In a tuned radio frequency amplifier including a plurality of amplifying stages, a vacuum tube in each stage, and a step-up transformer in the plate circuit of said vacuum tube, the input conductance at resonance of said transformer being substantially higher than the plate conductance of said vacuum tube.

9. In a radio frequency amplifier, a Vacuum tube and a step-up output transformer pertaining thereto whose effective ratio of turns is substantially greater than the square root of the ratio of the plate conductance of said vacuum tube to the conductance of the secondary circuit of said transformer at resonance.

frequency of the signals transmitted, and transformers having primary and secondary windings for linking adjacent stages, the primary winding of each transformer being in close physical coupling relation to a portion of the secondary wind- 3 ing thereof, the ratio of the number of secondary turns to primary turns of the windings of said transformers being such that the input conductance of each transformer at resonance is substantially higher than the plate conductance of the vacuum tube connected to the primary coil thereof.

11. The method of operating a radio frequency amplifier including a vacuum tube whose plate circuit is connected to the input terminals of an electric coupling system, and a second vacuum tube whose grid circuit is connected to the output terminals of said coupling system, which consists in tuning said coupling system, and in arranging said coupling system to step-up the voltage at resonance and to have an input conductance at resonance substantially higher in value than the plate conductance of the first mentioned vacuum tube.

12. In a, radio frequency amplifier, a tunable electric coupling system having input terminals and output terminals, a vacuum tube whose plate circuit is connected to said input terminals, and a second vacuum tube whose grid circuit is connected to said output terminals, said coupling system having at resonance :1. step-up voltage ratio and an input conductance substantially higher than the plate conductance of the firstmentioned vacuum tube.

.13. A tuned radio frequencyamplifier having aninput circuit and including a vacuum tube having grid and plate electrodes, means for tuning said input circuit, a radio-frequency transformer having a primary winding, with a low number of turns connected in the plate circuit of said vacuum tube, having a secondary winding with a number of turns great with respect to the number of primary turns connected in the input 1 circuit of a second vacuum tube and having means for tuning said second input circuit, whereby the input conductance of said transformer at resonance is substantially greater than the plate conductance of said first mentioned vacuum tube, the high-potential terminal of said secondary Winding being of opposite polarity to the highpotential terminal of said primary winding, the.- means for tuning said last-mentioned input circuit being operable over a frequency range for which said amplifier is unstable when said input circuits are tuned thereto, and means for maintaining the amplifier stable notwithstanding said input circuits are tuned throughout said frequency range.

LESTER L. JONES.

Images