DISTRICT COURT, E.D. NEW YORK
October 6, 1932
THCHNIDYNE CORPORATION et al.
The opinion of the court was delivered by: GALSTON
GALSTON, District Judge.
This is a patent suit in which infringement is alleged of ten letters patent, all relating to radio receiving apparatus. The inventions described are susceptible of conjoint use.
The following are the patents involved; the first nine listed below were issued on the applications of Lester L. Jones, and the tenth on that of Charles E. Bonine:
No. 1,673,287 granted June 12, 1928, for an electron discharge tube amplifier system, application filed June 11, 1927.
No. 1,713,130 granted May 14, 1929, for a method of, and means for, controlling energy feed-back and electron discharge devices, application filed October 8, 1924.
No. 713,132 granted May 14, 1929, for a radio frequency amplifying system, application filed August 21, 1925.
Reissue No. 17,915 granted December 30, 1930, for a radio frequency amplifying system, application filed August 21, 1925.
No. 1,732,937 granted October 22, 1929, for a transformer and coil system, application filed June 1, 1927.
No. 1,770,525 granted July 15, 1930, for a radio receiving apparatus, application filed July 15, 1927.
No. 1,779,881 granted October 28, 1930, for an amplifier, application filed October 5, 1929.
No. 1,788,197 granted January 6, 1931, for radio frequency circuits, application filed October 5, 1929.
No. 1,791,030 granted February 3, 1931, for a radio receiving system, application filed March 6, 1928.
No. 1,696,263 granted December 25, 1928, for a radio apparatus, application filed August 12, 1922.
The alleged infringement consists in the sale by the defendant of radio receiving sets known as the Sparton Equasonne receivers, manufactured by the Sparks-Withington Company of Jackson, Mich.
The Jones patents all involve highly technical specifications, exceedingly lengthy, oftentimes prolix with needless repetitions, and with numerous claims. To present the issues and keep the opinion within reasonable limits invites a certain grouping of the patents involved.
The first group, consisting of five patents, may be said to relate to means for coupling a selector to an amplifier; the second group of four patents relates to a fixed tube radio frequency amplifier; and the remaining patent, to a detector and single stage audio frequency amplifier combination. It will add to a logical consideration of these patents in their respective groups to consider them in the order of their application dates.
The First Group.
Jones patent No. 1,713,130. This patent relates to electron discharge tube circuits and for means controlling the feed-back from the output circuit to the input circuit due to the capacitive coupling between the grid and the plate of the tube.
Its particular object is to suppress or eliminate oscillations arising out of such feed-back energy. It was known that such feed-back gave rise to distortion of the input voltage and to oscillations either incipient or sustained in the grid circuit.
The inventor describes various methods that have been known to the prior art for the control of this feedback. The alleged invention consists in the discovery that the feed-back energy from the plate to the grid circuit may be controlled by suitable resistance means inserted in the plate circuit, such resistance means being associated in a predetermined manner with the characteristics and constants of the electron discharge tube. Such resistance, the inventor found, should preferably be inductance and capacity free, though such condition need not necessarily exist. The resistance should preferably be inserted in the plate circuit lumped at the plate. It is said that such a resistance has the effect of absorbing energy from the preceding input circuit at substantially the same rate, as energy is fed back from the plate to the grid circuit; and hence that the resistance is effective for neutralizing the feed-back reaction.
Claims 3, 4, and 6 are in issue. It will suffice to consider claim 3. It reads: "3. In combination, an electron discharge tube relay having a filament, a grid and a plate, a variably tuned grid circuit, a plate circuit coupled to the grid circuit through the grid plate capacity of the tube and having a load for producing an energy feed-back through the grid-plate tube capacity, and resistance in said plate circuit for producing a predetermined energy transfer or feed forward from the input circuit to the output circuit through the medium of said coupling and the relay action of the tube for substantially equalizing or neutralizing the energy retransfer or feed-back which takes place from the output circuit to the input circuit due to the capacity coupling of the circuits."
Means for neutralizing energy feed-back were known prior to the Jones' patent. Generally, they are shown in the Hartley patent No. 1,183,875, Rice patent No. 1,334,118, and the Hazeltine patent No. 1,489,228.
In Dr. Miller's article in the Bureau of Standards, issue of November 21, 1919, occur the following passages:
"It will be shown that when the load in the plate circuit is a resistance or capacity the input impedance can be represented as a positive resistance and capacity in series. Thus the tube is not a pure voltage device, but absorbs power."
"When the load is inductive the input impedance can, in many cases, be represented as a negative resistance and capacity in series. This represents regeneration through the tube itself, and is of importance in the regenerative effects and oscillations in amplifiers."
With the foregoing as a base, and what the art knew of resistances positioned in the plate circuit, it is hard to see how Jones effected an inventive contribution to the art.
There was no novelty in introducing a resistance in the plate circuit, for stabilizing an amplifier. It is found, for example, in British patent to Round, No. 149,433.
In Fig. 1, he shows a radio frequency amplifier having two tubes coupled through a transformer P' S' composed of resistance wire with an air core, which he called a resistance transformer, and arranged so that the ratio of inductance tube capacity is as great as possible.
The curves in this patent show that the purpose of the resistance is to stop oscillations in the transformers. Round says: "By increasing the number of valves
and of transformers the magnification of the signals may be increased or until the series begins to self-heterodyne."
It is true that no teaching is suggested with respect to a difference of the resistance in the primary and the resistance in the secondary of the transformer, but the fundamental disclosure presented is a resistance which operates in accomplishing the same result and in substantially the same way as that of Jones.
The object of United States patent to Carlson, No. 1,613,741, was to provide an amplifier in which the tendency to oscillate is reduced to a minimum. The inventor says that in radio frequency amplifiers there is a tendency for the individual stages to oscillate locally and for the several stages to oscillate collectively. These oscillations are caused by capacity coupling between the high potential terminals of the circuit, resistance coupling in the common battery circuit, and irregular transformer output voltage characteristics.
Carlson's invention was to use iron core transformers with primary and secondary windings upon the iron core.The disclosure certainly is of a resistance in the plate circuit for preventing local oscillations of feed-back. Carlson syas: "The primary and secondary windings 8 and 10 are constructed of high resistance wire in order of 10 to 20 times the resistance of an equivalent copper wire construction."
It may be noted that the value of the resistance resulting from these windings would be, as Mr. Hogan points out, from 1,500 to 3,000 ohms, which is approximately the value suggested by Jones.
An article in the Radio Digest by H. J. Marx, of January 27, 1923, is referred to by the defendant's expert to show that the writer was aware of the advantages of a resistance in the plate circuit of an amplifier for the purpose of preventing feed-back. The article contains this passage:
"The variable resistance sometimes added in the plate circuit is used for stabilizing the various stages through which two different frequency currents are passing. * * *
"High variable resistance such as potentiometers (400 to 1,000 ohms or even more) are inserted in series in the plate circuits of the second and third tubes."
Jones points out that the use of resistances in the Marx article is for reflex circuits. Though it is probably true that the circuit constants in a reflex circuit and the operational characteristics of the related audio and radio frequency circuit parts are different from that of a circuit in which it is desired to neutralize the feed-back reaction which takes place through the grid-plate capacity of the tube, yet Mr. Hogan seems convincing in contending that the behavior of a radio frequency amplifier in respect to regeneration is not effected by the double use of the tubes.
In the issue of Q.S.T. of March, 1923, appears the Phelps' article on the Invert Duplex System of Amplification. Fig. 4 shows the employment of an iron core transformer coupling. The receiving set illustrated therein was constructed as a model and offered as defendant's Exhibit F. The measurements on this model clearly indicate its operability in respect to neutralization.
Generally, on the subject of invention, Jones was not entitled to a patent because he discovered a function in the resistance which was not pointed out in the prior art. He is not entitled to a patent for discovering a particular function of the resistance. In arranging his resistance and determining its magnitude with respect to the characteristics of the particular tube in the circuit of which it is positioned, so as to produce, as claim 3 says, "a predetermined energy transfer * * * from the input circuit to the output circuit through the medium of said coupling and the relay action of the tube," Jones as a skilled engineer took what the art taught and made certain calculations to bring about certain results. It is true that in his specifications he refers to his empirical investigations, which would suggest not the work of a mathematician, but of an experimenter. Despite that reference in the specification, I cannot avoid the conclusion that his work was mathematical rather than inventive.
Accordingly, I hold this patent to be invalid.
Jones reissue patent No. 17,915 is for a radio frequency amplifying system. Three groups of claims are involved: 1 and 10; 3 and 5; 7, 13, 14, 15, and 25. The main defense is invalidity.
The inventor sought here to devise a system in which the number of cascaded radio frequency stages might be increased beyond that found practicable in the prior art.
A further object was to produce an increased power audion amplifying system with high stability and possessing the characteristics of ease of tuning and high degree of selectivity with the use of a minimum number of control elements.
The alleged discovery defined in the patent is that there were certain inherent limitations in tuned radio frequency systems, which can be avoided by the cascading of untuned radio frequency stages to tuned radio frequency stages, and that the reactions between the two stages may be so controlled as to produce a suitable radio frequency system having an amplifying power from ten to twenty times greater than the systems of the prior art.
Jones proposed determining the natural frequency of the output stage with relation to the range of frequencies for which the system is to be designed, so as to prevent production of any damping reaction, and so as to produce only an oscillation-creating feed-back reaction throughout the whole wave length or frequency range for which the system is built. He selects the characteristics of the transformer for the untuned stage, so that the natural frequency of the untuned output circuit is greater than the highest frequency of the predetermined frequency range through which the tuned stage is tunable.In such a selection of the natural frequency or wave length of the untuned stage, the system will oscillate during all changes in tuning of the tuned input circuit in the absence of any feed-back neutralizing means. A definite control may then be introduced to neutralize the feed-back-creating oscillations through the whole range of frequencies which the system is capable of transmitting.
Jones found empirically that under these conditions the reaction between the output circuit of the transition tube, such as the second radio frequency tube of Fig. 1, to the tuned input circuit, is substantially constant over the whole wave length range.
As the second step in his invention he provides means associated with the untuned stage for neutralizing or balancing the feed-back reaction.
Stating the matter as briefly as possible, in view of the involved specification, it would appear that the alleged invention, as shown in Fig. 1 of the drawings, consists in a plate circuit of the tube 2 R.F. having a natural wave length lower than any operating wave length so as to create an inductive reaction to all waves to which the input of 2 R.F. may be turned, and thus having a tendency to oscillate; and secondly, the amplifier is to be provided with some neutralizing means to prevent the oscillation from occurring.
Claim 3 will serve as well as any other for the purpose of considering the prior art. It reads as follows: "3. An amplifying system comprising an electron discharge device having an input circuit and an output circuit, means for tuning the input circuit through a predetermined range of frequencies, the output circuit being untuned and embodying means for coupling the same to a subsequent circuit, the load of said untuned output circuit combined with the load reflected thereinto by the subsequently coupled circuit comprising an impedance having a natural frequency greater than the highest frequency of the whole frequency range through which the input circuit is tunable whereby feed-back reactions from the output circuit to the input circuit are produced for the whole frequency range, and means for controlling the feed-back reactions."
The text of Dr. Miller's article in the Bureau of Standards publication heretofore referred to is relied upon to show the knowledge of the prior art, that when the plate load is inductive the input impedance can in many cases be represented as a negative resistance and capacity in series. "This represents regeneration through the tube itself, and is of importance in the regenerative effects and oscillations in amplifiers."
It is also stated that, when the load in the plate circuit is a resistance, the input impedance can be represented as a positive resistance and capacity in series, so that the tube is not a pure voltage device but absorbs power. Mr. Hogan interprets these passages as showing that Miller had expressed exactly what the patentee in this patent asserts, that in order to get away from the transition from a capacitive output to an inductive output, it is necessary to stay on one side of the resonant point. It is the passing of that resonant point which produces the transition. Hogan said: "One may define the transition as that from capacitive to inductive output or as from damping to regeneration or as from shorter to longer wave lengths, it makes no difference."
It was this knowledge that was used in the practical design of many types of amplifiers prior to the application for the original patent herein on August 21, 1925.
The fact that in passing through the natural wave length or natural frequency of the circuit, the circuit changes from a capacitive to an inductive phase, was known by others.
In Principles of Radio Communication, a publication by J. H. Morecroft, of 1921, in Fig. 70 are shown curves drawn between frequency along the resistance and reactance. The curve marked "inductance" at or about the resonant point falls to zero, and the curve marked "capacity" then begins and carries on to the right hand side of the sheet. Mr. Hogan is of opinion that this figure is illustrative at audio frequencies, power frequencies, but that the phenomena are the same as in radio frequency circuits.
Reference is made to an article by J. Scott-Taggart, published in the Annapolis Radio Station, "The Design of Multi-Stage Vacuum Tube Receiving Circuits," December, 1919. Fig. 1 shows a three stage amplifier with a periodic auto-transformer coupling arrangement. Coils L -- 3 and L -- 4 are the plate coils of this system, and their natural wave length controls the natural wave length of the circuit. The writer says: "The inductances L -- 3 and L -- 4 should have a natural wave length less than the wave length to be received." And adds: "If the range of wave lengths to be received is wide, tappings may be taken from L -- 3 and L -- 4."
The article in Q.S.T. of March, 1923, by Phelps has been referred to. In Fig. 5 the coil L has a single layer of windings of 100 turns adjusted to have a natural period of about 210 meters. The drawing shows a condenser of .001 mfd. Mr. Hogan says that "when a coil of 210 meters natural period at L (inductance) is connected with another similar coil and a condenser of this size, the effect is to reduce the natural wave length of the entire circuit so that in a specific case the natural wave length of the output circuit of the first tube set up in this way was reduced from 210 meters to 158 meters."
Mr. Hogan's assumption, however, in this calculation is that the inductance CL is equal to L; and Mr. Jones points out that the inductance CL is a radio frequency choke coil described in the caption and is designed to keep the radio voltages from leaking through the audio-transformer. The author indicates that CL is an inductance similar to L. That hardly justifies the inference that it is equal to L. Mr. Jones was of opinion that this similarity was one of form only; that is, that it should have a single layer winding on a cardboard tube. There seems force in the interpretation that this was merely a suggestion to an amateur to wind his coil single layer. Hogan admitted that the reference to the author's 200 meter experimental work would include a band of from 190 to 220 meters, so far at least as Fig. 5 of the article is concerned.
I think it clear that the output natural wave length for the circuit shown in Fig. 5 was not below the 200 meters amateur working wave length for which the circuit is suggested, but on the contrary, within the intended wave length band of from 190 to 220 meters.
The article published April, 1924, in the Bell System Technical Journal by Friis and Jensen on High Frequency Amplifiers, is also referred to, but its pertinency to this patent was not developed by Mr. Hogan and there is nothing but a passing reference to it by name in the defendant's brief.
Finally, reference is made to Rawsthorne, British patent No. 233,053, for the purpose of showing that the use of small coils tuned to natural frequencies higher than the working range is not limited to circuits in which the inter-tube coupling is invariable. Hogan relies on the following paragraph: "The primary coil P together with the circuit to which it is connected, is in all cases not tuned to resonance with the desired frequency either directly or indirectly, but is arranged to have a natural frequency which is considerably greater than the frequency which it is desired to receive.This is accomplished by making the inductance of the primary coil small and by excluding from its circuit any capacities that would be sufficient to cause its natural frequency to equal or approximate to the frequency which it is desired to receive."
Jones endeavors to distinguish this reference on the ground that the difference results obtained by him and Rawsthorne indicates difference in invention. Rawsthorne endeavors to prevent self-oscillations; Jones sought to produce these oscillations by means of a coupling between the circuits, and providing means for neutralizing them.
In Jones' arrangement, the means for coupling together two tubes in untuned; that is, following the tuned stage o, i, the stage which couples together the second and third radio frequency tubes, 2 R.F. and 3 R.F., is an untuned stage. Rawsthorne on the other hand, discloses a tuned stage that couples P and P. In Rawsthorne the primary of the transformer coupling device is untuned, but the secondary is tuned by the condenser C; whereas in Jones, the whole transformer T is the untuned transformer. In Rawsthorne, as appears from the quoted passage, the primary when untuned has a frequency greather than the frequency which it is desired to receive; whereas in Jones, the entire coupling means (i.e., the primary and the secondary of the transformer in connected circuit) and all which follows, has the predetermined natural frequency greater than the range of frequency.
In his specification, page 3, lines 94 et seq., Jones says:
"The output circuit o of this tube and the coupled input circuit i of the third radio frequency tube 3 R.F. being non-tunable, these circuits being coupled by the transformer T having a natural frequency in circuit greater than the highest frequency through which the input circuit i is tunable. Viewed from another aspect, the natural wave length of the output transformer T in connected circuit or the natural wave length of the coupled circuits o -- i is smaller than the lowest wave length that can be transmitted through the system by adjustment of the input tuning condenser C."
For the foregoing reasons I find the combinations of the claims in issue are not disclosed by the cited prior art and are, therefore, valid, showing as they do patentable subject-matter.
Plaintiffs' Exhibit 5, diagrammatically illustrated as Fig. AB in plaintiffs' Exhibit 8, embodies the combination of these claims. For example, "an amplifying system comprising an electron discharge device having an input circuit and an output circuit" is the tube T with its input and output circuits; "means for tuning the input circuit through a predetermined range of frequencies" is the tuning condenser C; "the output circuit being untuned and embodying means for coupling the same to a subsequent circuit, the load of said untuned output circuit combined with the load reflected thereinto by the subsequently coupled circuit comprising an impedance having a natural frequency greater than the highest frequency of the whole frequency range through which the input circuit is tunable whereby feed-back reactions from the output circuit to the input circuit are produced for the whole frequency range" is the output circuit o' in connected circuit; and "means for controlling the feed-back reactions" is the resistance R'.
I find, therefore, that this patent is infringed
Claim 3 is of the same group as claims 5, 7, and 13, and is likewise infringed.
The other claims involved are sufficiently similar in terms and scope to call for the same finding.
Patent No. 1,620,661. This patent relates more particularly to the art of cascade amplicification.The object sought is to eliminate distant stage feed-back.Jones says: "A disturbing reaction may take place on the tunable input circuit due to the tuning of a circuit arranged posterior to the untuned or non-tunable stage of the system."
This disturbing reaction may not be of great magnitude and is to be considered as secondary and distinguished from the primary feed-back reaction. His aim, therefore, is to overcome these secondary disturbing reactions.
Broadly, the means employed consist in tapping a coupling condenser of the untuned impedance network to different parts of the resistance R'.
Claims 5, 6, and 7 are in issue. Claim 5 reads as follows: "5. An amplifying system comprising a pair of electron discharge tubes, each having input and output circuits capacitively coupled through the electorstatic coupling of the tube elements, means for tuning the input circuit of the first tube through a predetermined frequency range, the output circuit of the first tube being coupled by non-tunable means to the input circuit of the second tube, said non-tunable coupling means comprising an inductance common to the coupled circuits, means for producing an energy feed-back reaction from the non-tunable coupled circuits to the input circuit of the first tube for the whole predetermined frequency range and means for neutralizing said feed-back reaction."
From the specification, as well as from Jones' testimony, it would appear that the tapped resistance R', connected at some intermediate point to the coupling condenser K', is the essential means disclosed for eliminating distant stage feed-back.
Infringement is denied on two grounds: That the defendant's set does not contain a condenser K' tapped into an intermediate point of the resistance R; and because it is without an inductance common to the output circuit of the first tube and the input circuit of the second tube.
That the condenser is not tapped into an intermediate point of the resistance is admitted by Mr. Jones. The condenser is tapped to the plate end of the resistance. But none of the claims -- 5, 6, and 7, limits the position of such tapping; so I cannot see that infringement would be avoided by positioning the tap at any point other than the intermediate point referred to.
But I think Mr. Hogan is right in his contention that the inductance L' is not common within the meaning of the patent to the output circuit of T and the input circuit of T. Mr. Hogan testified: "The output circuit of tube T goes directly through the shunt circuit comprising the resistor and coil L' in one branch, and the primary of the transformer M in the other branch. The grid circuit is coupled to the primary of the transformer M, but inductively so, though the coupling of that transformer and the direct grid circuit is simply through the secondary of the transformer M.Community of circuit elements as generally understood involves such a community as is shown in Fig. 1 and Fig. 2 of the patent, and not an association of the type existing in the Exhibit 5 set."
Plaintiffs, on the other hand, contend that L' is "an inductance common to the coupled circuits" within the meaning of the patent. The specification says: "The election discharge tubes 1 R.F. and 2 R.F. are coupled together by non-tunable means which in the present embodiment of the invention comprises a coupling inductance L' common to the coupled circuits which are the output circuit o' of the first tube and the input circuit i of the second tube, the input circuit i also including a fixed condenser K' in series with the coupling inductance L'."
Mr. Jones testified:
"The term 'common' applied in this connection means that the currents that flow in the output of the tube are similar to the currents that flow in the input of the next tube. The selection of a bifilar transformer M in Plaintiffs' Exhibit 5 connected across the small inductance L and the resistance R does not detract in any way from the action of L being common to the output of T and the input of T. The transformer M has a large value of inductance compared to the coil L, and the currents that flow in it are negligible compared to the currents that flow in the coil L'.Therefore it is entirely proper to consider the inductance L' as being common to both circuits."
"Another way of looking at that problem is from the viewpoint of coupling. Elements not common to both circuits occur when the coupling is loose, but with the bifilar transformer that is used here the coupling is unity and there is no way in which other oscillatory currents of any appreciable magnitude may arise in this system."
And again: "The current through L' is made up of two parts, one part which goes through the tube T and the other part which goes through the tube T. And of course there is a very small additional amount of current which goes through L' which also goes down through the transformer M; that is small in comparison with the total."
The inductance L' of the defendant's set is certainly a nontunable coupling means, and within the meaning of the patent seems to be common to the tube coupled circuits.
Since at least some novelty seems to be admitted by the defendants, this patent will be held valid and infringed.
Patent No. 1,770,525, granted July 15, 1930, to L. L. Jones, on an application filed July 15, 1927, is for a radio receiving apparatus.
The invention relates to that type of apparatus embodying separate selective and amplifying systems. The selection of the electrical oscillations of a desired frequency is obtained prior to amplification. The patentee has sought to provide, therefore, an improved amplifying system adapted to be coupled to a separate selective system of such radio receiving apparatus.
The amplifying system is of the so-called untuned type having a selective or tuned input circuit. The object of the invention is the production of a high and reasonably constant degree of amplifying system having fixed circuit adjustments, with a view to avoiding those reacting effects on the sharply tuned selective input circuit which tend to change the selectivity to values other than those that correspond to the natural damping of the input circuit.
The specification, not unlike most of the other patents in suit, is very lengthy, and the claims numerous. Of the 38 claims, the following, in three groups, are in issue: In the first group, Nos. 1, 2, and 3; in the second group, 12, 13, 14, 20, and 37; in the third group, 25, 27, 28, and 29.
Fig. 1 shows a radio receiving apparatus consisting of selector and amplifying units. The selector system comprises a plurality of circuits S, S, and S, coupled together inductively by variable coupling means. The adjustable coil units of the coupling means are simultaneously adjustable for the purpose of producing similar degrees of coupling between the circuits S -- S and S -- S.
The first selector circuit S is coupled to the energy receptor ant by means of a coupling condenser, whereas the last selector circuit S is coupled to the filament grid circuit of T, the first tube of the amplifier system. Each of the selective circuits is tunable.
To accomplish the objects of his invention the inventor proposes an amplifier embodying a series of electron discharge tubes, of which the first pair in the series is so organized as to eliminate all damping reactions and oscillating creating tendencies at the input terminals of the first tube of the series. This pair of tubes is to be coupled together and to subsequent circuits of the series so as to eliminate all of the reactions at the input terminals of the tube T throughout the wave length range of the receiver.
The tubes are of the usual three-electrode type, grid, plate, and filament. The filaments are connected by means of filament circuits to the A battery. The B battery is connected to the A battery and to the output circuits of each of the tubes T and T, a bypass condenser being employed connecting the plus side of the B battery to the minus side of the A battery.
The circuit network of the amplifying tubes T and T is organized to act as a pure reactance, either capacitive or inductive, and should be designed, says the inventor, to produce a constant reactance at the input terminal of the input circuit i'. When this circuit network is made to react as a pure reactance over the whole wave length range, and, as the inventor says, "all positive and negative resistances effective on the input circuit i' being eliminated for the whole wave length range," it is said that the selectivity of the selector system remains unimpaired and is capable of optimum operating results; and with a constant capacitance, uni-control of the selector system will be accomplished.
To attain this balance the inventor adopted the expedient of fixing the constants of the output circuit o' of the first tube, in accordance with the principles described and claimed in his patent No. 1,713,132, or in his patent No. 1,620,661, both of which patents are involved in this suit. The natural frequency of this output circuit in connected circuit is to be greater than the highest frequency of the whole frequency range through which the input circuit L' is tunable. By the natural frequency of the output circuit is meant the natural frequency of the impedance network of the plate of the tube or of the impedance L' of the output circuit o in connected circuit. Through this fixation there are produced feed-back reactions from the output circuit o' of the first tube to the input circuit thereof for the whole frequency range of the receiver.
The neutralize the feed-back reaction, resistance means are provided in the plate circuit of the first tube. This resistance must be of such a magnitude as to cause a feed-forward action from the input to the output circuit equal to the feed-back action. This resistance, designated as R' in Fig. 1 of the drawings and arranged "preferably" between the plate and the inductive component of the output circuit, may be of the order of 1,600 ohms, and is selected in accordance with the teachings of patents in suit, No. 1,713,132, No. 1,620,661, and No. 1,713,130.
With this determined characteristic the output circuit o' is next coupled to the input circuit i of the second tube by means of a coupling condenser K and an inductance M.The condenser K is tapped at the mid-point of the resistance R' and designed so that the condenser and inductance in series have preferably a natural frequency lower than the lowest frequency of the band to be transmitted through the amplifier.
To overcome residual reactions in the effort to secure exact balance of the feed-forward and feed-back effects, due largely to the distant stage feed-back, Jones sought elimination by selecting the constants of the output circuit o' so as to neutralize on the long waves, thus producing a residual over-damping on the short waves, and then neutralizing such residual damping reactions by producing varying input capacity for the second tube T. Jones observes that the input capacity should increase at the longer wave lengths so as to neutralize the damping reaction at the short wave lengths. The former increase he obtains by selecting the output circuit to have a natural frequency lower than the lowest frequency of the wave length band.
Following the coupling tubes T and T, the amplifier is made to provide a plurality of so-called untuned audion amplifiers, three in number, followed by a detector tube which is connected to the audio amplifying system.
The tubes T, T, and T are constructed in accordance with another patent of Jones described in patent No. 1,673,287 in suit. The constants of these circuits of tubes T, T, and T are selected so as to produce an automatic tuning of the system as a whole. This is accomplished by coupling the tube structure T and T with a transformer p s having a natural wave length of predetermined meter, then coupling the tube structures T and T with the transformer p s also with a natural predetermined wave length, but in excess of the transformer; and finally coupling the tube T with the detector tube by means of a transformer p s and having a natural predetermined wave length in excess of the transformer p s.
The inventor says: "The principle underlying this arrangement of amplifying tubes (T, T and T) is that the higher selected natural wave length of the last transformer of the series (p s) has the effect of producing a rising input capacity change with increasing wave length for the last amplifying tube T which is suited or made to coincide with that desired characteristic of the input circuit of said tube T which would cause the said input to vary its resonance frequency and thus to automatically tune itself over a considerable wave length band. This same relation is carried out in the coupling of the tube T to the tube T, the natural wave length of the transformer p s being greater than the natural wave length of the transformer p s."
In designing the circuits consideration is given to the interrelations among them.
The claims in this patent are very voluminous. They are classified in three groups by the plaintiffs. Those in issue are 1, 2, and 3 of the first group; 12, 13, 14, 20, and 37 of the second group; 25, 27, 28, and 29 of the third group.
The defenses are invalidity and noninfringement.
Mr. Jones testified that the first step towards solving the problem of eliminating undesired reactions of the first tube of the selector and of making the apparent input capacity constant was to make the first tube output coil small, so that the resonance point of the output circuit was slightly lower than the lowest working wave length. This first step was not new, having been disclosed in both patents No. 1,620,661 and reissue No. 17,915 in suit.
This second step consisted in overcoming the regeneration due to coil L by adding the resistance R'. This second step was also described in patent No. 1,620,661, as well as in patent No. 1,713,130 in suit.
The contention of the plaintiffs is that these two steps in effect resulted in eliminating about 90 per cent. of the disturbing reactions, and that the remaining 10 per cent. was eliminated by the joint action of the tube T with the tube T and by the addition of the condenser D.
The selection of an output circuit for the tube T that will produce a rising input capacity on that tube with increasing wave length was certainly not new with Jones. The Miller article in the Bureau of Standards publication, in Fig. 3 and Fig. 7, is significant. These figures show that, as the plate resistance is increased, the input capacity of the tube will appear to rise, and secondly, how the input capacity in mfds varied over a range of from 20 to 80 mfds, as the inductance of the plate circuit was changed from a low value to a higher value.
From the Morecroft article, page 76 and the figure shown thereon, it may be inferred, as Mr. Hogan says, "that the impedance or resistance of the output circuit varied as the applied frequency changed; and it was known that the input capacity varied as the resistance changed; and consequently as the applied wave length changed. Those are the natural and inherent effects in normal circuits."
German patent No. 343,702 to Barkhausen shows a multi-stage amplifier in which the successive stages have no longer natural wave lengths. The inventor sought to avoid the whistling tendency caused by the tube capacity acting between grid and plate. The translation of this patent has the passage: "on the other hand, if the resistance of the output side is capacitive, that is, if the voltage lags behind the current in phase, the natural oscillations of the input side are damped, and therefore, a self-excitation and a tendency to whistling cannot occur. However, as is well known, a system of conductors consisting of a coil with a condenser connected in parallel forms an inductive resistance for oscillations which lie below its natural frequency, and a capacitive resistance for oscillations which lie above its natural frequency, as the phase of the current in passing through the resonance point changes through about 180 degrees."
By his invention he claims self-excitation is avoided by tuning the output side to a lower natural oscillation than the input side; and this is true in the multi-stage amplifier.
With that understanding of the patent and of the relevant prior art, we may consider the defendant's alleged infringing set. It consists, broadly, of a pre-selector unit and an untuned radio frequency amplifier, together with the power unit. The question is whether the residual 10 per cent. reactions, not controlled by the improvements set forth in Jones' patent No. 1,713,130 and reissue No. 17,915, is obtained by the defendant, and, if so, whether such control is effected by the means described in this patent.
Mr. Hogan's measurements of the defendant's set show that there is considerable reaction on the input circuit of the first tube. At a wave length of 250 meters, the reaction through the grid-plate capacity of the tube doubled the resistance of the circuit, while at the wave length of 450 meters, the resistance was half.
Defendant's receiver does not embody a coil N or condenser D shunted around a portion of resistance R'.
Claim 1 of the patent in suit reads as follows: "1. An amplifier embodying an electron tube having input and output circuits, and means in the output circuit for causing a predetermined uni-directional variation in the input capacity of the tube with a variation in wave length, said uni-directional variation being effective over the whole wave length range over which the amplifier is operative."
This claim, in the light of the Miller, Morecroft, and Barkausen references and the earlier Jones' patents, shows no invention, and I think, therefore, is invalid; and for the same reason, claims 2 and 3 of the same group are likewise invalid.
Of the second group claim 12 may be taken as typical: "12. An amplifier comprising an electron tube having an input circuit and an output circuit capacitively coupled by means of the tube structure to the input circuit, means for neutralizing the feed-back reactions due to the capacitive coupling between the circuits, and means for producing a rising capacity with increasing wave length effective across said output circuit for neutralizing residual feed-back reactions effective on said input circuit."
For the same reason, I think this claim and the others of its group, 13, 14, 20, and 37, are likewise invalid.
Claim 25 reads: "25. An apparatus for selecting and amplifying high frequency oscillations comprising an untuned amplifier including electron discharge tube coupled stages having fixed constants predetermined to maintain the stages automatically resonant over a given frequency band, a tunable selector circuit connected to and feeding into the input terminals of said untuned amplifier, and means for causing the untuned amplifier to act as a pure reactance at the said input terminals over a given range of frequencies."
The prior art cited does not meet this claim nor the others of the group, 27, 28, and 29; but since the resistor used by the defendant does not cause the amplifier to act as a pure reactance, the claims are not infringed.
The last of the first group of patents is No. 1,788,197. The primary object of this invention was to improve the reactionless coupling means described in Jones patent No. 1,770,525. In the circuit described in this patent, an inductance is supplied in the output circuit of the tube, of such magnitude and construction as to cause that circuit to be resonant naturally to a higher frequency than the highest frequency of the working frequency range in which the circuit is to be employed. The neutralizing means indicated may take the form of a resistance preferably free of capacitance and industance, and of a magnitude sufficiently great to cause a feed-forward current through tube 2. The succeeding tube is coupled to the output circuit of this first tube through a capacitance 12 and an inductance 14, the latter being connected across the cathode and control electrode of the succeeding tube. The capacitance is connected between the output circuit of the tube 2 and the control electrode of the said tube 10.
The claims in issue have been reduced since the trial to the following: the first group, claims 21 and 26; the second group, claim 8; and the third group, claims 1, 5, and 18.
The defenses of the patent are invalidity and noninfringement.
In his specification Jones says: "But of greater importance, in order to raise the gain obtainable in the transition or reactionaless like tube 2 I find it feasible to increase by approximately two-fold the resistance and inductance values suggested above, the resistance then being of the order of 2800 ohms, and the industance being of the order of 230 microh. This approximately doubles the impedance of the output circuit and approximately doubles the gain in the tube."
Engineers, it is contended by the defendant, long prior to the filing of this patent, understood the relation between the gain of the tube and the load impedance. It is, therefore, urged that there could be no novelty in merely doubling the resistance and inductance in the output circuit of the first tube in order to increase the gain.
Burndept, British patent No. 235,307, contains this passage: "A capacitive, non-inductive resistance becomes self-adjusting as will be hereinafter explained."
British patent to Round, No. 194,433, shows a capacity shunt across the resistance widing of the primary of the coupling transformer.
Australian patent No. 521 of 1926 to Priess discloses a damping resistance of the order of magnitude specified by Jones in the patent under discussion.
The advantages of transformer coupling over impedance coupling was recognized by Ballantine in his book on Radio Telephony for Amateurs.
Claim 21 of the first group of claims in issue reads as follows: "A reactionless amplifier comprising a vacuum tube, an inductance in the output circuit thereof arranged to make the reaction of the tube regenerative, a resistance in series with said inductance for opposing the regenerative reaction of the tube, said inductance and resistance being greatly increased in magnitude over the correct value for making the tube reactionless in order to increase the gain therein, and means arranged to prevent residual reactions through the tube caused by the said increase in the output impedance thereof."
The nearest approach to this patent are the earlier Jones' patents, but not one of those patents can be read upon this claim. There is some suggestion in the argument that no invention was involved but only engineering skill in doubling the impedance of the output circuit to double the gain in the tube. The question of invention thus viewed is close, but in my state of doubt, I must favor the patent. The prior art does not disclose the combination of the claim.
Jones sought to get the utmost amplification from the first tube and did it by increasing the load in the output of the tube. Jones thinks the rule to be that when the load impedance is substantially smaller than the plate resistance of the tube, the amplification varies almost proportionately.
Comparing the invention of the patent under discussion with the circuit and means described in his patent No. 1,770,525, Jones said of the invention described in the latter patent, that with the tube and arrangement shown in that patent, it would not be possible for him to increase the load impedance to obtain greater amplification, because by making the coil larger, the natural wave length of the coil is longer, and by bringing the wave length 200 meters the function of the coil would be destroyed in producing a regenerative feed-back over the whole wave length range.
On the other hand, in the patent under discussion the type of tube shows smaller inter-electrode capacities, so that even though the coil be made larger to secure increased amplification, the natural wave length of the circuit would be held smaller than the smallest wave length of the band.
Jones recognized that with these increases in the inductance and resistance components, as specified in this patent, residual reactions are present, most of which are taken care of in the same way as set forth in the patent No. 1,770,525, by selecting the output of the second linking tube to be naturally resonant with the capacitance, hanging thereon to a wave length higher than the working wave length range. He noted, however, a tendency toward regenerative feed-back in the middle of the working frequency band, and, therefore, had to correct it. In his specification he says: "One of such minor variations is a tendency toward regenerative feed back in the middle of the working frequency band. To correct this, in accordance with my present invention a capacitance 16 is arranged in shunt or parallel with the feed forward resistance 8, and this capacitance I have empirically found serves the function of causing a decreased regenerative action in the middle of the band."
Fig. 1 of the drawing illustrates the invention as applied to impedance coupled tubes. Fig. 2 illustrates the invention as applied to transformer coupled tubes. In the former the plate circuit of the first tube is coupled to the input circuit of the second tube capacitively through the condenser 12 and inductance 14; whereas in the second form, coupling is accomplished by bifilar transformer 20. These are equivalent forms.
In respect to the advantage of the particular form of transformer indicated in Fig. 2 and described in the specification, Jones says: "A fifth advantage is peculiar to the use of a bifilar transformer or its equivalent, that is, a unity coupling or very closely coupled transformer, and resides in the fact that such a transformer transfers the changes in input capacitance of the following tube directly across to the output circuit of the preceding or link tube, and therefore is ideally suited for correcting the residual reaction which it has already been explained tends to arise when the output resistance and inductance of the link stage is raised with a view to increasing the amplification gain therein."
Therefore, I think claim 21 is valid.
Claim 26, like claim 21, is directed to the reactionless amplifier and its combination in a radio receiving circuit, wherein the inductance in the output circuit and the regenerative opposing resistance in series are both increased in magnitude. This claim is narrower than claim 21, and adds, as to such increase in magnitude, "to about twice the correct value for making the tube reactionless over the whole frequency range in order to increase the gain in the tube." If such determination were an a priori conception or indeed just the result of a mathematical calculation, and, if the combination of means indicated in the claim had been set forth in any one patent or publication or device of the prior art, one might well hesitate to ascribe invention to such act. But such is not the condition revealed in the proofs, and so I think this claim likewise is valid.
Claim 8 is a broader claim. It reads: "8. A radio receiving circuit comprising a selector, a cascade vacuum tube amplifier, and means coupling said amplifier to said selector comprising the first tube of said amplifier, an inductive impedance in the output circuit of said first tube, a resistance in series with said inductance, a capacitance arranged in shunt with said resistance, and means coupling the output circuit of the first tube to the input circuit of the next tube of the amplifier."
It will be noted that the matter of increased magnitude referred to in claim 21 is omitted; but for the reasons stated heretofore, I think this claim is valid.
Claims 1, 5, and 18 are directed to the form of invention shown in Fig. 2 of the drawing, as embodying the use of the bifilar transformer; and under the view heretofore taken of this patent, are valid.
There remains the question of infringement. The form of coupling embodied by the defendant is that shown in Fig. 2 of the patent in suit. From the discussion heretofore had of the defendant's set, I think it embodies the combination of means in all of the claims in suit, and not merely those, such as 1, 5, and 18, limited to the employment of the transformer coupling, since this transformer coupling is merely specific to the form of Fig. 1 and is the equivalent thereof.
Patent No. 1,696,263 to Bonine. This patent is the first of the second group of patents referred to in this opinion, and relates to a radio apparatus for amplifying and detecting high frequency oscillations. The inventor deals with two or more stages of amplifying apparatus, "whose anode and grid circuits are coupled either by inductive reactance or auto-transformer, or, and preferably, by two-winding transformers, it being a feature of my invention that, particularly when the thermionic amplifiers are substantially similar in size or electrical characteristics, the impedance of the inductive reactance, auto-transformer, or windings of the two-winding transformers progressively change, and preferably increase in magnitude in the successive stages."
The radio frequency or transformer windings referred to are used in such relationship that the anode circuits of the amplifiers "are substantially aperiodic or non-oscillatory." The inventor further provides that with given windings the apparatus is operable over a relatively wide range of wave lengths, and adds, that the inductive reactance comprises windings "in single layer of extremely small conductor having the convolutions spaced from each other for minimizing distributed capacity, and in the case of a two-winding transformer the primary and secondary are positioned close with respect to each other for effecting close coupling."
Claims 3, 5, 12, and 15 are in issue. The defenses are invalidity and noninfringement. Claim 3 reads as follows: "3. In a multistage amplifying system, the combination with successive thermionic amplifiers, of inductive means for coupling said amplifiers, the inductance of the coupling means of a later stage being greater than the inductance of the coupling means of an earlier stage."
The "inductive means" referred to in this claim are the transformers. Referring to the drawings, Fig. 1 is a diagramatic view of one of the inventor's circuit arrangements. It discloses a loop receiver, an untuned radio frequency amplifier, two stages of audio frequency amplification working into a telephone. Bonine says: "The transformers associated with the amplifiers V -- V are radio frequency transformers whose primaries and secondaries are preferably close to each other to effect close coupling and the inductance and impedance of whose primaries P progressively increase with the succession of stages of amplification; that is to say, the inductance and impedance of the primary P in the anode circuit of the amplifier V is greater than the inductance and impedance of the primary P in the anode circuit of the amplifier V, and there is preferably a similar increase in inductance and impedance of each succeeding stage. The same is true of the secondaries S, the ratio of transformation being anything suitable or desirable, but preferably unity for relatively short wave lengths, as for example, 200 to 500 meters; but in the case of longer wave lengths the ratio of transformation is preferably greater than unity."
And again: "In general, however, the inductance of the primaries and secondaries progressively increase with the succession of stages, this relation having been found by me to the essential for best effects in multistage radio frequency amplification. And this progressive increase of inductance and impedance of the transformer windings is of particular advantage where the several primaries P are connected to one and the same stabilizer R, b, or equivalent, serves to stabilize the several stages with which the transformer windings are associated."
Invalidity is predicated upon the teaching of the Miller and Morecroft articles heretofore referred to, German patent to Barkhausen No. 343,702, and United States patent to Pohlmann No. 1,523,836.
The Barkhausen patent may be received only for the purpose of indicating the state of the art and not as an anticipation, for it was not set up in the answer.
The Barkhausen patent seems to have a very close bearing upon Bonine's teaching, even though it is contended that Barkhausen is dealing with radio rather than with audio frequencies.At audio frequencies the coils have a relatively large distributed capacity, and the change in input capacity of the tube would not operate to change the resonant frequency of the audio system. Nevertheless, Jones is forced to contend that, if one were to assume that Barkhausen's system was broad enough to include radio, as well as audio frequencies, then "the disclosure is incomplete and unworkable on account of the fact that any question of spacing of the relative frequencies of the transformers is not given."
But claim 3 makes no reference to the question of "spacing of the relative frequencies of the transformers" in the Bonine patent. Barkhausen says:
"According to the invention the self excitation of natural oscillations of the input side, and therefore the tendency toward whistling, may be completely avoided by tuning the conductor-system of the output side of a tube to a lower natural oscillation than the conductor-system of the input side. In multi-stage amplifiers each conductor-system is in like manner tuned lower than the preceding."
Barkhausen also says: "The transformers are, in known manner, 'coil resonance', i.e. wound so that the inductive resistance of the coil is equal to its capacitive resistance. * * * The transformers, together if necessary with the condensers belonging to them, are tuned with respect to one another so that the natural frequency of each one is lower than that of the preceding, i.e. the transformer i is thus tuned lower than the transformer h, the transformer h lower than transformer h, and the latter again lower than the input transformer a."
The publication date, November 7, 1921, of the Barkhausen patent is of a period in which certainly the art knew of the distinction that Mr. Jones sought to establish between audio and radio frequencies; and I think it unjustified to limit the teaching of Barkhausen merely to audio frequencies. There is no express limitation to that effect, nor does it follow from the suggested use of the shunt condensers that Barkhausen was dealing only with audio-frequencies. Miller taught that identical principles apply in both audio and radio uses.
On the showing of the Barkhausen patent, I find no invention in the amplifying system defined in claim 3 of the Bonine patent; and accordingly I hold it invalid.
Claim 5 is narrower than claim 3 in the limitation imposed, with "the number of primary and secondary turns in each transformer being substantially equal."
It cannot be said that in any of the references to the prior art there is specific mention of this limitation. But the Bonine specification distinctly says that this is a matter of preference rather than of invention, and in view thereof and of the teachings of Barkhausen and Miller, I think this claim likewise is invalid; and for the same reason I likewise hold invalid claim 15, the narrower claim, and claim 12, the broader claim, for want of invention.
Patent No. 1,732,937 to Jones is for a transformer and coil system. Jones in this patent set forth the difficulties encountered in the prior art with types of transformers subject to short wave oscillations in a frequency band closely adjacent to the frequency band for which the transformer is designed. He claimed to suppress the objectionable short wave mode of oscillation by eliminating the leakage inductance between the windings of the transformer: "This I have found can be accomplished by winding the primary and secondary of the transformer bifilarly, with the filaments of said windings arranged in such close juxtaposition that the leakage inductance between the windings is substantially completely eliminated."
Jones stated that another object of this invention centered about the provision of a transformer of the bifilar type, which is astatic; that is, he sought to provide a transformer "or an inductance coil system designed and constructed so that the external magnetic field is reduced to such a minimum that the coil is substantially de-coupled magnetically from surrounding apparatus and may be placed in close proximity to other similar coils of a radio receiving system without intermagnetically reacting with the same."
The validity of the patent is vigorously attacked; and reliance is had on United States patent to Vreeland No. 1,666,518; an article in the Bell System Technical Journal on High Frequency Amplifiers by Friis and Jensen; Whittingham patent No. 1,581,900, Reissue No. 17,194; United States patent to Horle No. 1,693,898; and an article in Q.S.T. of February, 1924, by Kruse, describing a Ballantine vario transformer.
It will be sufficient to discuss claim 1. It reads: "1. An astatic bifilar transformer comprising primary and secondary windings would bifilarly, similar parts of the filaments of said windings being arranged in close juxtaposition throughout their lengths, said windings comprising two axially spaced sections substantially equal in size and oppositely wound."
The specification itself admits the use of the bifilar type of coils in radio work prior to the filing of the application for this patent. Most important of the references is the Vreeland patent. The specification states: "The preferred arrangement shown in Figure 1 possesses an astatic feature whereby the external field of the transformer is practically eliminated. In this arrangement each of the coils of the transformer is made in two sections, would or connected in opposite directions so that their external fields oppose and neutralize each other, as fully explained in the application above mentioned."
Fig. 2 is a more detailed diagram showing physical construction of a type of coupling transformer. Here is disclosed specific axial spacing for astaticism. The primary is composed of two sections wound in opposite directions and axially spaced, as is also the secondary.
The functions of close coupling in a coil or transformer was also well known prior to 1927. This appears from the Bell System Technical Journal.
Mr. Hogan refers in that article to the following formula for inductive coupling:
f' = fo / root 1 - k,
f" = fo / root 1 k,
and points out: "Now if k is very small, one minus k and one plus k have substantially the same value, and therefore the two resonant frequencies have substantially the same value, that is, are close together."
As is perfectly obvious, if the value of k is increased to unity, then f' becomes indeterminate and f" has a definite value. The teaching clearly is that with samll values of k, the resonant frequencies are close together, and by increasing the value of k, they are moved apart.
The advantages of close coupling were also appreciated by Horle, as shown in his patent No. 1,693,898. The invention described in this patent relates to the amplification of feeble electric currents, and more particularly "to means for coupling the input and output circuits of amplifiers operating over broad bands of alternating current frequencies."
It describes the provision of a duplex winding for the circuits of the coupling device and the amplifying means. The transformer shown consists of two windings, a primary winding and a secondary winding, the conductors of which are laid on the winding in parallel and separated from each other by the thickness of the insulating material on the conductors. Figs. 1, 2, and 3 of the patent thus indicate a bifilar winding of a radio frequency tube coupling transformer.
In the publication Q.S.T. of February, 1924, a type of radio frequency transformer is disclosed. It is composed of two parts, one the stator and the other an internal rotor. The stator carries one part of the winding. This apparently, from what appears in Fig. 4 of the article, is made up of a double strand of wire. The rotor also carries a winding. By turning the rotor within the stator, the inductance of the system is changed because of the mutual inductance between the two halves of the tansformer, but the coupling between the primary and secondary is not changed. When the transformer is at its minimum inductance its two parts are in opposition, just as in the case of the astatic coil.
In the light of these pertinent references of the prior art, it is hard to see how the claims in suit disclose any invention. The art certainly showed an astatic bifilar transformer comprising primary and secondary windings. It showed closely spaced or closely coupled windings and also oppositely would sections axially spaced.
I believe, therefore, that the claims in suit are invalid.
Patent No. 1,673,287 is for an electron discharge tube amplifier system. The invention relates to the provision of a nontunable amplifier system for use in the amplification of high radio frequencies having a high amplification over a wide wave length band.
Untuned or nontunable amplifiers are those in which the circuit constants, capacity, inductance, and resistance of the circuit are fixed and nonadjustable. Jones says that such an amplifier system was known to the prior art, but was characterized by a lack of uniformity and inefficiency in some part of the band, due to the fact that the amplification varied with the frequency impressed upon the system, the maximum amplification being produced at the resonant frequency of the system.
Jones' invention, as set forth in the specification, consisted in substituting, for what he called the untuned system:
"A circuit arrangement or system which is self-tuning or is automatically tuned over a relatively wide wave length band resulting in the obtaining of a uniformly high amplification over said wide wave length band. * * *"
"From another aspect, the application of the principles of my present invention results in the transformation of a so-called untuned amplifier into an automatically tunable amplifier, or an amplifier which is self-tunable over the wave length band to produce uniformly high and efficient amplification over the band."
The invention, therefore, as described, consists in predetermining the constants (that is, capacity, inductance, and resistance) of the input and output circuits and the relation between such circuits.
Claims 1, 2, 3, 8, 10, and 14 are in issue.
The defenses are invalidity and noninfringement.
It is important to determine within the disclosure of the patent what automatic tuning is present in the defendant's receiver. Automatic tuning requires a rising capacity with increasing wave length. The measurements on the stages of the defendant's set, including the tubes T and T, show peaked amplification curves, and not the flat gain curve such as is shown in Fig. 3 of the patent. Mr. Hogan's curves show tube T and the output circuit, a voltage gain at approximately 5 at 200 meters, increasing to 9 at 260 meters, and then decreasing to about 4 at 550 meters. This certainly does not show a constant gain; and as to the tube T stage in which the input capacity of the tube T is effective, a varying gain curve is disclosed. At 200 meters the gain is a little over 7; at 265 the gain is 4 1/2, and it rises to 11 at 550 meters.
Mr. Hogan presented an interesting study of the gain of these two stages together. His is illustrated in defendant's Exhibit K. It shows a joined effect of the stages of tubes T and T. This study shows a high gain at about 250 meters, and the effect of the high gain in the longer wave is shown by the rising terminal value at about 550 meters.
I gather from Mr. Hogan's further exposition that the added capacity of the defendant's receiver in the stages T and T do not fall within the automatic tuning defined in the patent. He says, for example, referring to the changes of capacity on the input of tube T: "The variations from the capacity actually required run from a tremendously large percentage down through 51 at 250 meters, and then in the deficiency direction to 55 per cent at 300 meters, 76 per cent at 350 meters, and so on, up to 90 and 91 per cent at the longer wave lengths."
And again: "From this tabulation of the capacity produced in the input of tube T it may be noted that the added capacity is practically constant over the wave length range from 214 to 250 meters, and that this amount does tune the circuit at approximately 260 meters. That agrees fairly closely with the peak of amplification at 260 meters which is shown in the gain curve previously introduced, but because of the constancy of the added capacity there is no automatic tuning and can be no automatic tuning."
The plaintiff argues that certain deviations from perfection are possible in automatic tuning without departing from the teaching of the patent, citing such cases as Hazeltine Corp. v. A.H. Grebe & Co. (D.C.) 21 F.2d 643; Hazeltine Corp. v. Radio Corp. of America (D.C.) 52 F.2d 504. But the patent in suit clearly limits that allowable deviation: "Since the transformer structures are broadly tuned by this, that is, the change in response with frequency change is small, it is not necessary to have the terminal capacity exactly correct. A deviation of 10 per cent. or 20 per cent. in the actual capacity from that which is theoretically required would produce only 5 per cent. deviation in the resonant frequency from the impressed frequency. This change in response due to a 5 per cent. frequency change would be relatively small so that the sacrifice in efficiency due to lack of coincidence of these curves is very small. It is only necessary that the curves of Figs. 5 and 5 have approximately the same starting and ending points and that their form in between these points be generally similar."
But the operation of the defendant's set shows no such "automatic tuning" within the allowable deviation.
The defendant's expert analyzed Jones' exposition of the operation of the defendant's set, as illustrated in defendant's Fig. C of plaintiffs' Exhibit 13, and showed that the measurements of the capacity that was added as the frequency was changed, was at the half volume setting for the receiver. If the volume control is at full setting, the added capacity through the tube is greatly increased.
Relating the additional capacity required for tuning at various wave lengths with the actual capacity that is added at the full volume, it is possible to state the express capacity actually added in the set in terms of the percentage of the added capacity required. Mr. Hogan's figures thus show:
Wave Additional Actual Percentage
Length Capacity Capacity Excess over
Required Added Required
per Ex. C at Full
300 meeters 1 mmf 17 mmf 1700 per cent.
350 9 27 200
400 18 37 105
450 27 48 78
500 39 60 54
540 48 55 14.5
These figures, of course, do not comport with the tolerance allowed by the patentee. The measurements made on the changes of capacity of the input of tube T confirm the results noted above in respect to the matter of deviation. Defendant's receiver in the stages T and T does not show automatic tuning within the meaning of the patent.
Letters patent No. 1,779,881 is for an amplifier. It is among the latest of Jones' patents, and the specification includes a review of theory and summarizes the teachings of all the Jones patents previously considered. The specification covers fifteen pages of descriptive matter and claims.
Broadly, the invention relates first to the matter of dividing the so-called automatic tuning range over two tubes, and, secondly, the inclusion of a volume control which does not disturb the stability of the amplifier.
Fig. 1 of the drawings shows a radio receiving apparatus comprising a selective system and an amplifying system. The selector system comprises a plurality of circuits in cascade having variable constants to effect selectivity. The amplifier also comprises cascaded stages having fixed circuit constants coupled to the last selective circuit.
The principle underlying the arrangement of the two stages of automatic tuning shown in Fig. 1 is described with much detail on pages 6 and 7 of the specification. The inventor states that the longer natural wave length of the last transformer of the series p s has the effect of producing a rising input capacitance change with increasing wave length for the last amplifying tube T, which is suited or made to coincide with such desired characteristic of the input circuit of the tube as would cause the input circuit to vary its resonance frequency and so automatically to tune itself over a considerable portion of the wave length band. The same relation is carried out in the coupling of the tube T to the tube T, the natural wave length of the transformer p s being greater than the natural wave length of the transformer p s, and causing tube T automatically to be tuned to the lower wave length portion of the total wave length band.
With the values which Jones assigned in his consideration of the inter-relations, the last stage of this amplifying system tunes over a range of 350 to 550 meters, whereas the preceding stage tunes over a range of 320 to 450 meters.
The circuit arrangement of Figs. 2 and 8 are then described in the specification, showing the combination of the selector and the untuned amplifier employed with A.C. tubes and embodying the application of volume control. Ti is pointed out that the arrangement of the transformer in a progressively spaced relationship produces an inherently stable amplifier.
The problems presented in applying volume control to stages automatically tuned and the amplifier as a whole are said by the inventor to be:
"In the automatically tuned stages the output circuits are naturally resonant to frequencies lower than the input circuits, and the stages normally are therefore stable. This relation may be lessened and even reversed at some point over a wide range of volume control, and accordingly one object of the present invention is to so adjust the input and output circuits of these stages, within the range of dimension giving the desired automatic tuning, that the natural resonance frequency of the input circuit remains higher than the natural resonance frequency of the output circuit over the full range of volume control."
"In one of the amplifier stages, coming between automatically tuned stages, the input transformer is necessarily of greater inductance than the output transformer, and the latter is therefore naturally resonant to higher frequencies than the former, and therefore is inductive and regenerative in the working frequency range. However, their wide difference in frequency lends stability which tends to prevent oscillation in spite of the regenerative reaction in this stage, but variations in the constants of the circuits in the range of volume control may lessen this frequency difference and again lead to instability. Accordingly, a further object of my invention is to overcome this possible source of instability, which I do by making the input and output circuits more widely different in frequency than would otherwise be necessary, and sufficiently so to prevent instability over the range of volume control."
Jones believed that the volume control means should be capable of varying the amount of amplification over wide limits, and says: "For example, in a five-stage untuned amplifier the control of volume from a whisper to full loud speaker volume for all stations, both near and far, requires a variation of ampliciation ranging from full power or an amplification of 10,000 or more down to an amplification of less than one, i.e., an attenuative condition. Some methods of securing this variation of total gain are by varying the anode potential or by varying the bias on the control electrodes of the tubes. In either of these methods there occurs a condition in which, the tube impedance being greatly increased, the resistance load due to the anode to cathode resistance of the tube is reduced and effectively removed from the fixed transformers. At the same time the apparent input capacitance of the tubes at any given wave length is reduced by a large amount from the value it has at full amplification, which raises the natural frequency of the adjacent transformer, so that it may become inductive and regenerative in the working frequency range."
The difficulties encountered are cured by slightly reducing the inductance of transformer 28, which insures that the input circuit of tube T will remain naturally resonant to a frequency higher than that of the output circuit; and, secondly, slightly to increase the inductance of the transformer 24, which coacting with the decrease in the inductance of the transformer 28, thereby causes an augmented difference in the natural frequencies of the input and output circuits of tube T, which leads to greter stability.
The claims alleged to be infringed are in three groups: 1, 8, and 10; 3, 17, and 19; and, 4 and 22.
They are resisted on the ground of invalidity and noninfringement, but invalidity, as will presently appear, was but faintly urged.
The prior art relied on consists in the main of the Miller paper published by the Bureau of Standards, the Ballantine Physical Review, the Bonine patent in suit, the Friis and Jensen publication, Barkhausen patent No. 343,732, Australian patent to Priess, No. 521/26, Austrian patent No. 91,188, and United States patent to MacFadden, No. 1,639,042.
So far as the references to Miller and Ballantine are concerned, I confess I cannot follow Mr. Hogan at all, nor does the defendant's brief aid me in this respect. I fail to find in these articles the problem of capacity variation with wave length change in any such circuit as Jones proposes.
So also the discussion of the Bonine patent by Mr. Hogan and the reference to it in the brief are both cursory. I think it appears by any fair comparison of Bonine with the patent in suit, that Jones was correct in claiming as he did, that by utilizing the principle of automatic tuning in a cascaded amplifier, he succeeded not only in starting from the high point of each individual stage to the peak of amplification, which one gets with copper wire transformers, but that he was able to maintain that high amplification over a larger wave length range in each stage.
The Barkhausen patent is referred to only to show the state of the art and not as anticipation. Mr. Jones points out that there is nothing stated in Barkhausen as to a separation of the natural frequencies in the transformers shown, and he concludes that the three transformers were closely spaced one to the other. There is nothing indicated in the patent which shows "automatic tuning" in a radio amplifier unit.
Austrian patent No. 91,188 was also admitted, not as an anticipation, but to show the state of the art. Apparently its relevancy is not relied upon because it is not discussed in defendant's brief. The patent is obscure. Certainly that is so, if it is relied upon as showing anything equivalent to Jones' method of connecting the transformers in an automatically tuned system.
The Priess Australian patent, also not discussed in the defendant's brief, relates to staggering radio transformers. The most that can be said of this by Mr. Hogan is that the last stages may exhibit "an apparent input capacity increase of substantial amount." I cannot see the bearing of this patent on the question of invention.
MacFadden patent No. 1,639,042 relates to a radio system in which the input circuit is tuned and "whose anode circuit is the primary of a radio frequency transformer whose secondary is tuned and loosely coupled to and reversed with respect to the primary." This patent is in no way an anticipation of the claims in suit.
On the question of infringement the defendant rests its entire defense on the ground that it has not an automatically tuned system, but in this respect its position is not so convincing as in respect to patent No. 1,673,287 in suit. In the discussion of that patent it will be recalled that noninfringement was found owing to the deviation in the defendant's set from the limits specifically set forth by the inventor in the patent under consideration. There is no such limitation in patent No. 1,779,881. The defendant's set certainly does involve automatic tuning. This is, perhaps, best illustrated in the last stage. The output of tube T in defendant's set is made to tune to the input of tube T. Tube T is utilized for the upper wave length part of the range, and the preceding stage involving the tube T is utilized for automatic tuning over the lower wave length range.
The principle of the patent in respect to the automatic tuning range clearly was adopted by the defendant.
The defendant cannot blow hot and cold in this matter. In its own literature it states: "The amplifier unit is as revolutionary in its operation as the selector units. With 6 tubes, one of them operating as a detector tube, it amplifies the impulse received 15,000 times instead of the usual 5,000 times. Even the most sensitive of present day sets rarely go above 8,000. This amplifier is unique in that it automatically tunes itself to the incoming frequency no matter what the wave length."
So that claims 1 and 10 relating to the division of the working frequency range are infringed.
Claim 8 in this group seems to have been withdrawn from discussion; and, therefore, I have not considered it.
The volume control claims are resisted by the defendant because the defendant's receiver employs a volume control on tubes T to T inclusive, whereas it is claimed that Jones shows a volume control which is not applied to tubes T and T and states that oscillations or disturbances are produced if an attempt is made to apply such volume control to tube T and T. It is true Jones says in his specification: "I therefore prefer to apply volume control only to tubes T, T, and T, and not to the transition or link stages T and T."
But he adds: "However, it is possible to carry the foregoing precautions for stabilizing the amplifier further, and then to apply volume control to all of the tubes, with a slight sacrifice in selectivity."
Attention also may be called to Fig. 8 of the patent and to the description on page 12 beginning with line 17, which describe an amplifier in which the volume control is applied to all of the tubes and not merely to the last three tubes.
And Jones says: "I have illustrated volume control applied to all five tubes in Fig. 8 because it is possible to produce a stable amplifier even when this is done."
I, therefore, find the volume control claims 17, 19, and 22 also infringed.
Claims 3 and 4 are apparently withdrawn from discussion, and I have not considered them.
Jones patent No. 1,791,030. This patent relates to radio receiving systems with special reference to the provision of a high power radio receiver in a system embodying only a single stage of audio amplification.
The alleged discovery is that in this system a detector tube may be given such predetermined constants that the radio frequency amplifier may be related to the detector so as to generate sufficient audio output in the detector fully to load in a single stage of amplification. He states that he empirically determined that the constants of the audio frequency stage may be so related to the detector and radio amplifier stages that the audio tube may be fully loaded by the audio output developed by the detector without causing any overloading of the audio tube over any part of the range of the detection stage.
Jones states that, when the detector tube is operated by grid bias rather than by a leak grid condenser and is provided with a high biasing grid voltage and a correspondingly high plate voltage, and when this tube is supplied with a sufficiently high radio frequency input comparable to the high grid biasing voltage, then the detector tube is capable of developing an audio output sufficiently great to work directly into and adequately load a power tube audio stage. In a specific instance he says that when the bias on the grid of the detector stage is raised 10 to 20 volts, with a correspondingly high plate voltage of from 100 to 200 volts, and when the radio frequency amplifier is built to produce a high gain or amplification sufficient to develop a high radio frequency voltage at the input of the detector, over a range of 1 to 10 volts, then the desired high audio output is produced.
In applying these disclosures, he first determines the constants to be imposed upon the detector tube. These imposed constants include the biasing of the grid g of the detector tube negatively with a voltage of at least 10 to 20 volts, supplied from a battery C', and the use of high plate potentials derived from a source B'. The input voltage at the input circuit i of the detector is made high enough to produce a grid swing of the order of magnitude of the biasing grid voltage. This is accomplished by the use of a high gain radio frequency amplifier A.
By means of these constants he was able to produce a linear detection characteristic, as illustrated in Fig. 3 of the drawings, and then to provide a radio frequency input voltage for the detector sufficient to swing the grid over a voltage to utilize substantially the full linear portion of the detection characteristic. The detector stage is, therefore, enabled to yield an audio output to work directly into the single stage power tube C.
Then the constants of that single power stage are determined so that the high audio output of the detector is amplified to produce full volume operation of the loud speaker D, preferably so that at the saturation point of the detector stage B the power tube is fully loaded. The grid of the power tube is energized at a fairly high voltage, usually of the order of 50 volts, to deliver the full output.To generate a commensurate audio frequency voltage on the grid of the power tube, the ratio of the transformer T, the inventor points out, should be step-up ratio of approximately 3 to 1. With power tubes having an amplification factor of 3, the 50 volts developed on the grid g' correspond to 150 volts on the plate p'.
To interrelate the single stage audio amplifier with the preceding parts of the system, Jones first determines the maximum audio output voltage of the detector and then the maximum load of grid voltage for the audio or power tube, and provides a step-up transformer between the detector and the audio tube, so that its ratio is the ratio of the two voltages.
The application for this patent was filed March 6, 1928. At that time the radio art had available such power tubes as were known in the trade as UX 250, UX 210, and UX 171. In cross-examination Mr. Jones admitted that one stage audio amplifiers at the time he filed his application were old; also that grid bias detection was old. The use of a high input voltage on a grid bias, in order to obtain sufficient voltage on the output to operate a single stage audio amplifier, was also known to the prior art.
In an article published in The Wireless World, July 14, 1926, it is said: "From the foregoing it is clear that for receiving weak signals or for circuits where no high-frequency amplification is employed before the detector valve (unless the receiver happens to be very close to the transmitting station) the grid method of rectification is the more suitable, whereas for strong signals or where considerable H.F. amplification is used, the anode method of rectification will give better results. In both methods the truest reproduction will be obtained at those amplitudes where the efficiency is greatest, because the steepest part of each A.C. characteristic occurs where it is straightest."
An article published by John Scott-Taggart on Thermionic Tubes in Radio Telegraphyl and Telephony, in 1924, discloses the manner of grid bias detection without overloading. The author states: "It will be clearly seen that by taking any of the curves of Fig. 55, and by suitably adjusting the voltage on the grid, we can use the valve at any point on its characteristic curve. When we speak of using the valve at a certain point on its curve we mean that the vertical line through the given grid voltage cuts the curve at that point."
From the Fig. 55 and the 82.5 -- volt curve shown thereon, with no volts on the grid, the valve functions almost at its saturation point. At -5 the valve is working at a point midway along the straight steep portion of its curve; and at -12 volts, the valve is being used at the bottom bend of its curve.
The author writes: "It will be seen later that it is of the utmost importance that we should be able to adjust the valve to that point on its characteristic curve at which it best carries out the function desired of it."
Then follows the assertion that the same object may be achieved by varying the high tension voltage, that is, the plate voltage, or varying the filament current or by varying both together.
It seems clear from this that the author disclosed the relation of the grid potential and its choice in a negative direction up to values as high as 12 volts or more, and the relation of the grid potential for a given use in connection with the plate voltage applied to the tube. In this respect Jones contributed nothing new.
Scott-Taggart also discusses the tube as an amplifier.
Referring again to the 82.5 volt curve, he says: "This curve will do justice to our 8-volt alternations, and no distortion effects will be produced. It will, however, be necessary to give the grid a normal voltage of -8 volts in order to use the curve at its halfway point."
The paper of Moullin and Turner, "The Thermionic Triode as Rectifier" published in the Journal of the Institution of Electrical Engineers, 1922, may be referred to to show the understood advantages of grid bias detection over grid leak detection for strong signals.
In the article of Keith Henney in The Radio Broadcast, of December, 1925, "The Tubes, Their Uses and Abuses," there is recognition of the fact that one of the requisites of the detector is to have sufficient output to operate a single stage of audio amplification so as to attain sufficient volume for the loud speaker.
In The Radio World, April 19, 1924, is an article on Power Amplification on Only One Tube, by Charles H. M. White. The writer says: "One stage of power amplification will give you all the volume you desire and if you want more just add a few more volts of 'B' battery and watch the volume swell."
He appreciated the use of but a single stage of power amplification:
"The fewer tubes used for audio-frequency amplification the purer the tone production. Many radio fans who now have two stage audio-frequency amplifiers will notice that noise and distortion is squared every time another stage of ordinary amplification is added."
"Many radio fans have noticed that one stage gives about the best results, since it happens to be the best compromise between volume and quality. With some second stages it seems volume exceeds the quality. Now it is only logical that if this reproduction of the first stage is amplified through a power tube, much better results will be obtained."
He adds: "Several large set manufacturers have incorporated only one stage of audio-frequency in their sets because they realize that with sufficient volume for the detector this is sufficient when power amplification is used. As a matter of caution I advise that sufficient 'C' or grid bias battery be used."
In a second article entitled "One Stage of A.F., But Oh, What Power!" he adds:
"Unusually great distance has been covered using this type of audio-frequency amplifier and a receiver consisting of one stage of tuned radio-frequency amplification and a tube detector. * * *"
"It is then quite obvious that an amplifier consisting of one stage would have far less tube noise relatively than a three-stage amplifier. With power amplification the necessary volume for average use can be secured with one stage."
High gain amplifiers were used by Armstrong, as appears in his article entitled, "Some Recent Developments of Regenerative Circuits," published August, 1922, by The Institute of Radio Engineers. Such use was in connection with a grid bias detector and single stage audio amplifier.
The teachings of the prior art in the publications referred to disclose a complete understanding of the interrelations of a detector with a single stage audio amplifier; so that whatever Jones set forth in that relationship, considered in the advanced state of the art, involves nothing more than the application of the skill of a radio engineer.
And so far as utilizing the detector as a power limiting device is concerned, so as to prevent an overload in the audio stage, that also was well understood in the prior art. It is disclosed in the patent to Curtis, No. 1,415,999, even though Curtis makes no specific mention of a detector. Such a tube detector, however, is shown in the article by John H. Morecroft, "Principles of Radio Communication," published in 1927.
In conclusion, therefore, it may be said that whatever advance Jones made in the art, however meritorious, was the result of his very pronounced engineering skill and profound knowledge of the art, and not the result of an inventive act. I hold this patent invalid.
Plaintiffs may have a decree in accordance with this opinion.
If this opinion is not in sufficient compliance with the rule requiring findings of fact and conclusions of law, submit findings of fact and conclusions of law in accordance therewith.