Opinion
Thomas G. Haight, of Jersey City, N.J., Abel E. Blackmar, Jr. (of Sheffield & Betts), of New York City, and William G. Mahaffy, of Wilmington, Del., for plaintiffs.
George I. Haight and M. K. Hobbs, both of Chicago, Ill., John B. Brady, of Washington, D.C., and Ayres J. Stockly (of Hastings, Stockly & Duffy), of Wilmington, Del., for defendant.
NIELDS, District Judge.
The parties are before the court a second time upon a motion for a preliminary injunction seeking to enjoin defendant from further infringement of the two De Forest patents Nos. 1,507,016 and 1,507,017 for the 'feed-back' invention.
When this suit was brought, a motion was made for a preliminary injunction under the two De Forest patents. The radio transmitter then sold by defendant, known as the 4-A transmitter, utilized the feed-back invention to produce the alternating currents used in transmitting. On that motion defendant asserted that De Forest claims were anticipated by the disclosure of the application for the Goddard patent No. 1,159,209 and by the Lindredge article of September, 1912. The court held the patents valid. Radio Corporation of America v. Collins Radio Co. (D.C.) 13 F.Supp. 976. Thereafter defendant manufactured and sold the 45-A transmitter. Whereupon plaintiffs filed a supplemental bill charging infringement of the De Forest patents by the 45-A transmitter.
The two De Forest patents were issued in September, 1924, after extended interference proceedings in the Patent Office. Both patents have been repeatedly before the courts. They have been sustained twice by the Supreme Court. Westinghouse Electric & Manufacturing Co. v. DeForest Radio Telephone & Telegraph Co., 278 U.S. 562, 49 S.Ct. 34, 73 L.Ed. 507 (on certiorari to the Circuit Court of Appeals for the Third Circuit); Radio Corporation v. Radio Laboratories, 293 U.S. 1, 55 S.Ct. 928, 79 L.Ed. 163 (on certiorari to the Circuit Court of Appeals for the Second Circuit). Not only have the De Forest patents been repeatedly sustained but the inventions thereof have been repeatedly held by the courts to be of the utmost importance.
The De Forest patents relate to a vacuum tube and its associated circuits, which are so arranged as to progressively feed back a portion of the energy in the plate circuit to the grid to build up the electric currents in the circuits. Patent No. 1,507,017 covers broadly the feedback circuit, and patent No. 1,507,016 covers that circuit when so arranged and adjusted as to generate continuous electric oscillations.
Since the issuance of the preliminary injunction, defendant does not challenge the validity of the De Forest patents. On the present motion defendant replies that the 45-A transmitter is the same as that shown in certain figures in the drawing of the application for the Goddard patent. Defendant also replies that it does not utilize two of the elements of the claims of the De Forest patents. Therefore the questions in the case are: (1) Whether defendant's apparatus i.e., the 45-A transmitter, is the same as that disclosed in the drawing and described in the specification of the original application for the Goddard patent. (2) If the 45-A transmitter is not the same as Goddard, whether defendant utilizes two of the elements of the claims of the De Forest patents or equivalents thereof.
In De Forest a battery impresses upon the plate a positive potential. The filament emits particles of negative electricity which pass over to the positively charged plate. As a result, an electric current flows in the plate circuit. De Forest found that the current flowing in the plate circuit could be controlled by the grid or control electrode. Further, that a very small amount of electricity applied to the grid would control the flow of a much greater electrical power between the filament and the plate. This control is accomplished by establishing an electrostatic field between the grid and the filament. The more negative the grid with respect to the filament, the smaller the current in the plate circuit. The less negative the grid the greater the plate current. In his early patents De Forest showed the grid in various positions. (1) Directly in the electron stream between filament and plate. (2) Outside the stream but still in the tube. (3) Outside the tube.
De Forest No. 1,507,016 discloses the 'feed-back' circuit connected to a vacuum tube and so adjusted as to generate continuous electrical oscillations. The 'feed-back' is secured by so arranging the plate circuit and the grid circuit that variations of current in the plate circuit will react upon the grid circuit. In the arrangement shown in Fig. 1 of this patent, 'feed-back' is accomplished by inductively coupling the coil L-2 in the plate circuit to the coil L-1 in the grid circuit. In this manner amplified energy from the plate circuit is fed back to the grid circuit.
In this system, the generation of oscillations begins as soon as the operator completes the circuit. This produces a change in the potential impressed upon the grids and causes a corresponding variation in the magnitude of the electron stream flowing between filament and plate. These variations occur at a definite frequency determined by the amount of inductance and capacity which are present. Due to the amplifying action of the tube, the power in the plate circuit is greater than that in the grid circuit. The larger variations in the plate circuit react through the coupled coils upon the grid circuit, and therefore the original variations in the grid circuit are cumulatively affected by the current in the plate circuit. These increased variations applied to the grid further increase the variations in the plate circuit due to the control action of the grids on the current flowing in the tube from the filament to the plate. This is a cyclic process and the variations in the current, occurring at the frequency determined by the constants of the circuit, continue and thus become sustained oscillations.
De Forest No. 1,507,017 shows several arrangements utilizing a De Forest tube and a 'feed-back' circuit. The claims in suit broadly cover the 'feed-back' whether or not used as a source of continuous oscillations. The operation of the system shown in Fig. 1 of this patent is substantially the same as that of Fig. 1 of the other patent. The oscillations start with the completion of the circuits. There is an increase in the power and amplitude of the oscillations until tube and circuit limitations prevent a further increase in power. The oscillations are produced and sustained because the plate and grid circuits are coupled to each other in such a manner as to react upon one another and accomplish 'feed-back.'
Capacity is present whenever two conductors are so placed with respect to each other that an electrostatic field exists between them. Capacity is often provided by means of a condenser which has two metal plates separated by a small air-gap. However, capacity in an electrical circuit is often present because of the proximity of different conductors. Capacity provided in this manner is referred to as 'natural' or 'inherent' capacity. In some circuits natural or inherent capacities are depended upon for the operation of the circuit, their presence being just as essential as if a physical condenser were used.
Similarly natural or inherent resistances are often present in an electrical circuit. Conditions may be such that an electrical charge will leak off from one part of the circuit to another, across a path which is conductive, although such path may not be provided by a wire. Such a current is referred to as a leakage current. In the prior 4-A transmitter the 'feed-back' current was across the inherent capacity between the filament and grid in the vacuum tube. Also, in early vacuum tube circuits the current path from grid to filament was often composed of natural resistances either within or outside the tube.
Defendant's Apparatus.
The apparatus manufactured and sold by defendant is a radio transmitter known as type 45-A. The transmitter includes a vacuum tube and associated circuits so connected as to generate continuous radio frequency oscillations. This portion of the transmitter is known as the 'oscillator.'
The remainder of the transmitter comprises apparatus by which the power of the generated oscillations is built up by amplifiers. It also comprises means for impressing upon the oscillations an intelligible signal. In use, the transmitter is connected to a radio antenna and the amplified oscillations carrying the signals are sent out in the form of electromagnetic waves. There appears to be no necessity for considering this part of the transmitter.
The oscillator utilizes a special type of tube known as the 'Collins C-100-A' connected in a push-pull circuit. The push-pull system has been known a long while. It consists of two tube systems, arranged so that one is pushing the generated alternating currents (acting on the negative part) while the other is pulling (acting on the positive part). A push-pull oscillator employs two tubes and oscillator circuits, or their equivalents, operating as a single oscillator, so that the output of each half of the system is combined with the output of the other half to produce resulting oscillations of greater power.
The Circuit Court of Appeals for the Second Circuit held that a 1915 Colpitts patent was invalid because there was no patentable invention in employing the push-pull amplifier circuit with the De Forest amplifier tube. Western Electric Co. v. Wallerstein, 60 F. (2d) 723, 728. Moreover, a push-pull oscillator is shown in Langmuir No. 1,778,454 applied for in October, 1913. The disclosure of this Langmuir patent illustrates how in the push-pull circuit certain elements may be combined and, if the electron streams to the two plates are independently controllable, the two tube structures may be placed in a single bulb. Langmuir was in interference with De Forest for the feed-back inventions. It was decided that De Forest was the prior inventor. Langmuir v. De Forest (D.C.) 18 F. (2d) 345. Therefore, the use of the De Forest oscillator in the push-pull circuit was known long before this defendant put out type 45-A.
Defendant's C-100-A tube is two De Forest tubes with a single bulb. The tube is composed of a cylindrical glass bulb within which are placed six filaments located immediately inside the glass and spaced at equal intervals around the tube. Each wire extends from the bottom to the top of the tube, and all of them are connected together in parallel so as to constitute the source of the currents flowing within the tube. In the space between the filament wires, two mental cylinders are placed one above the other constituting the plates. The grids in the C-100-A tube are placed outside of the glass. One of them is a metal cylinder surrounding the upper halves of the filament wires and the upper plate. The other grid surrounds the lower half of the bulb and therefore surrounds the lower halves of the filament wires and the lower plate.
Defendant points out that the grids are outside the bulb and not inside. The apparatus operates by electrical effects. It is wholly unimportant where the grid is located, provided it accomplishes its electrical function. Grids on the outside of the bulb were not new. Plaintiff's expert deposes that the grid had been variously located; and in certain instances in the art had been placed outside of the bulb. The usual location has been within the bulb between the plate and filament because the maximum control of the electron stream is secured thereby. Control, however, can be effectively secured by a grid on the outside of the electron stream and outside of the bulb. This is recognized by Collins in his first affidavit on the prior motion in this suit:
'The presence of the glass wall of the vessel does not influence the effect of the electromagnetic field upon the electron stream and it exerts only a very minor influence, and that one of degree, in the case of electro-static control. For example, if an electro-static control element is positioned to have the glass wall between it and the electron stream the dielectric constant of the glass would change the intensity of the field slightly, depending upon the thickness of the glass, and would have to be taken into consideration in any precise calculations of a quantitative nature, but the effect would not be large enough to be measurable under ordinary experimental conditions.'
The C-100-A tube acts as though it were two separate three element tubes. Defendant's tube has two plates, a grid for each plate and, in effect, a separate filament for each plate. The six filament wires constitute the filament. They are connected together and extend parallel to the plates, so that each half of each filament wire is located adjacent to one of the plates and supplies electrons to it. Each grid is spaced by only the thickness of the glass from the filament and governs the number of electrons passing from the adjacent portions of the filament wires to the adjacent plate. In other words, each grid controls the strength of the current flowing through the corresponding half of the tube.
Due to the use of the push-pull circuit, the potentials of the two grids with respect to the filament vary in opposite directions. As one grid becomes more negative, the other grid becomes less negative and vice versa. This is characteristic of the push-pull operation whether the two tubes be combined as in Collins or are separate. As a result, the current in the tube is a maximum first in one half and then in the other half.
The circuits of defendant's oscillator enable the combination of tube and circuits to act as a generator of sustained electrical oscillations or alternating currents. The coil L-1 is connected in the plate circuits and the coils T and T' are connected in the grid circuits. The coil L-1 is inductively coupled to the coils T and T' in such manner that a portion of the energy in the plate circuits is fed back to the grid circuits. In this manner feed-back is secured and continuous oscillations are generated by tube and circuits by means of the feed-back connections.
Goddard Oscillator.
Defendant claims that its present oscillator operates by diverting the tube current instead of by controlling the strength of the current in an unchanging path. It also claims that its C-100-A tube is that of Figs. 3 and 4 of the original Goddard application.
On the prior motion the court found: 'Goddard has never gone into use and is a totally impractical device for the production of oscillations. ' It was necessary for defendant to reorganize the arrangement suggested by Goddard in order to obtain a practical oscillator. Defendant made substantial changes in converting the tube shown in Figs. 3 and 4 of the Goddard application into its C-100-A tube. Defendant took the totally impractical device of the Goddard application and made it into a commercially practical oscillator by applying to Goddard's disclosure knowledge acquired from the teachings of the De Forest patents, supplemented by many years of intensive research as to the construction and operation of the De Forest vacuum tube.
In the Goddard tube the entire filament is constantly emitting a stream of electrons in the direction of the two plates. Goddard provides means for diverting the entire current so that it flows first to one plate and then to the other. Defendant's oscillator does not operate by deflection. In the C-100-A tube the current flows first in one half of the two tube system and then in the other. This is not deflection. It is the action of a push-pull oscillator circuit. The C-100-A tube is in effect two tubes. Each grid controls the strength of the current flowing in the tube from the adjacent portion of the filament to the adjacent plate. When one grid is at its most negative potential, the adjacent portion of the filament is paralyzed. In other words, the defendant's tube contains one De Forest tube comprising a plate and the upper portion of the filament. The current flows between these electrodes controlled by the adjacent grid. Defendant's tube also contains a second De Forest tube composed of the other plate and the lower half of the filament. The current here flows between these electrodes being controlled by the adjacent grid. Defendant's tube does not operate by diverting the current through the tube from one path to another and back again. Defendant's tube comprises two separate tubes partially inclosed within a single glass envelope. Each of these tubes operates independently, and in each the strength of the current flowing through the tube in an unchanging path is controlled by a corresponding grid.
Defendant's grids are electrically interposed between filament and plates. It is immaterial where defendant places its physical structure. 'Interposed' in the De Forest claims means interposed so far as its electrical effect is concerned. That the grid might be electrically interposed without being physically interposed was known and recognized in the art. The grids of defendant's oscillator, although located outside of the tube, perform the same electrical function in the same way as the grid of the De Forest oscillator located within the tube. Neither the specification nor claims of the De Forest patent require that it be limited to a tube having the grid physically located inside the tube. Considering the pioneer character of the invention and the range of equivalents to which it is entitled, such a limitation is forbidden. Eibel Co. v. Paper Co., 261 U.S. 45, 63, 43 S.Ct. 322, 328, 67 L.Ed. 523; Smith v. Snow, 294 U.S. 1, 14, 55 S.Ct. 279, 284, 79 L.Ed. 721.
Upon this motion the court finds that defendant's oscillator is not a Goddard oscillator and is a De Forest oscillator. The court further finds that defendant's type 45-A transmitter infringes the claims in issue of each of the De Forest patents in suit.
Plaintiffs' motion to strike out paragraphs 22, 23, 24, and 25 of defendant's answer is granted. Defendant claims it is entitled to conduct an inquiry into the conduct of plaintiffs in asserting the Armstrong patent, while having rights in De Forest's applications and then dropping the Armstrong patent after De Forest issued. Defendant treats the plaintiffs as owners of both the Armstrong patent and the De Forest applications and patents at the time it alleges that plaintiffs asserted the validity of the Armstrong patent. The reports of the decisions of the Supreme Court show that the question as to whether the Armstrong patent or the De Forest patents covers the inventions of the patents in suit was bitterly contested until the latter part of 1928. Indeed, the question was not conclusively settled until 1934. It would be a strange doctrine which would deny to licensees of the De Forest patents the right to assert those patents after their validity had been sustained over the Armstrong patent, merely because, as licensees under the Armstrong patent, they had attempted to exercise their rights as such licensees by bringing suits thereon. Centaur Company v. Robinson (C.C.) 91 F. 889, 890.
Plaintiffs' motion to strike out defendant's counterclaim is granted. Defendant's counterclaim charges infringement of defendant's Riggs patent by R.C.A. Manufacturing Company, a wholly owned subsidiary of the plaintiff Radio Corporation of America. There would be no testimony common to plaintiffs' suit on the De Forest patents and defendant's counterclaim on the Riggs patent. In the exercise of its discretionary power, the court strikes out the counterclaim because it would delay and complicate the trial of the issues raised by the bill of complaint and the supplemental bill in this suit.
The preliminary injunction should issue under the supplemental bill of complaint.
This opinion contains a statement of the essential facts and of the law applicable thereto in conformity with Equity Rule 70 1/2 (28 U.S.C.A. following section 723).