Opinion
November 2, 1927.
Charles Neave, Stephen H. Philbin, and A.E. Blackmar, Jr., all of New York City, for plaintiffs.
Walter H. Pumphrey, of New York City, for defendants Grebe and Bunnell.
William R. Ballard, of New York City, for defendant American Telephone Telegraph Co.
In equity. Suit by the Radio Corporation of America and others against J.H. Bunnell Company, Inc., and others, for infringement of patents. Decree for plaintiffs.
This suit is brought for the alleged infringement of the Lowenstein patent, No. 1,231,764, and of the Mathes patent, No. 1,426,754, relating to radio apparatus. The owner of the patents is the defendant American Telephone Telegraph Company. The plaintiffs, Radio Corporation of America, General Electric Company, and Westinghouse Electric Manufacturing Company, are the licensees under the patents in suit. The defendant A.H. Grebe Co., Inc., is a manufacturer of the apparatus complained of, and the defendant J.H. Bunnell Co., Inc., is a dealer in the alleged infringing apparatus. The American Telephone Telegraph Company, owner of the patents, was made a defendant after it had declined to join in the suit as party plaintiff. The American Telephone Telegraph Company, as owner, is interested in sustaining the validity of the patents, but not otherwise specifically interested in the relief prayed for in the bill. It has filed its answer, substantially admitting the allegations of the amended bill of complaint. The real defendant in the action is the Grebe Company.
The action was commenced November 21, 1922, and has twice been before the Circuit Court of Appeals in this circuit. The defendants Bunnell and A.H. Grebe Co. originally filed a counterclaim in their answer, which, on plaintiffs' motion, was dismissed pursuant to an opinion by me, dated August 13, 1923. The Circuit Court of Appeals dismissed an appeal therefrom. 298 F. 62. Later a motion was made by defendants to dismiss the bill of complaint because of alleged lack of proper parties plaintiff. This motion was denied by Judge Bondy, by opinion dated February 10, 1925. An appeal was taken by the defendants, which appeal was dismissed by the Circuit Court of Appeals for this circuit, on April 27, 1925. The Supreme Court of the United States thereafter refused to grant defendants' petition for a writ of certiorari. 269 U.S. 565, 46 S. Ct. 24, 70 L. Ed. 414.
Judge Bondy's opinion consisted of the following memorandum:
"Defendant's motion to dismiss the amended bill of complaint is made on contentions all of which have been overruled in Radio Corporation of America v. Emerson (C.C.A.) 296 F. 51, and Radio Corporation of America v. Independent Wireless Telegraph Co. (C.C.A.) 297 F. 521. The motion, therefore, is denied."
No opinion filed.
It is contended by the defendants Grebe and Bunnell that the manufacture of the alleged infringing apparatus was discontinued some time since, but the answers interposed by them press the defenses of invalidity and anticipation by the prior art.
The Lowenstein Patent.
The Lowenstein patent, which will first be considered, was applied for April 24, 1912, and issued July 3, 1917. The claims relied on are Nos. 1 and 7:
"1. Telephone apparatus comprising the combination, with a talking circuit, of a suitably energized relay circuit, including an anode and a cathode separated by a conductive gap, a modulating device interposed in said gap and electrically connected with said talking circuit, means for impressing upon said modulating device a potential more negative than that of said cathode, and a translating device arranged to be energized from said relay circuit."
"7. The combination, with an audion having its anode and cathode included in a suitably energized circuit, of means for impressing upon the audion grid a potential more negative than that of the audion cathode."
The Lowenstein invention discloses a three-element vacuum tube or "audion" for amplifying speech currents in radio apparatus. The "audion," in the form that it is widely used to-day, is (a) for the purpose of detecting signal carrying radio frequency currents; and (b) also for the purpose of amplifying currents both of radio and of audio frequency. The contribution of De Forest to the art has been considered by the Circuit Court in this circuit. Marconi Co. v. De Forest Co., 243 F. 560. The vacuum tube of Lowenstein does not differ materially from the basic patents to De Forest, but the object that Lowenstein aimed at was to improve the reproducing qualities of the vacuum tube for speech currents. He sought to make speech, not only audible, but intelligible as well, and approximating the naturalness of the human voice. This much sought for object is what is commonly referred to in discussions concerning the art as "freedom from distortion" in an amplifier. This object was accomplished by Lowenstein's invention by making the grid of the vacuum tube normally negative. The alleged infringing device, the so-called C R-9 receiver, it is claimed, uses the inventions which are the subject of this suit in the audio frequency amplifier system.
Lowenstein discovered that, if the grid had a negative bias — more negative than any part of the filament — a more faithful reproduction of the sound could be accomplished. If this grid is maintained more negative than any part of the filament, no current can flow in the grid circuit. The electrons must flow from the filament to the grid, to complete the grid circuit and cause grid current. But electrons have a negative charge, and they will not flow to the grid, if it is more negative than any part of the filament. The plate, on its part, is always more positive than the filament. The fact that the grid is maintained negative while the signal variations are imposed upon it makes the tube of the Lowenstein class a potentially operated device, as distinguished from a current operated device, since no current can flow in the in-put circuit.
At the present time, it is recognized with practical unanimity, in both telephone and radio circuits, that the grid should be poled normally negative in amplifier circuits. Lowenstein accomplished his purpose of producing a negative bias of the grid by employing a grid battery, or C battery. The grid was connected to the negative pole, and the filament was connected to the positive pole of the C battery, thus giving a definite negative potential difference, which was impressed upon the grid with respect to the filament.
Plaintiffs' witness, Mr. Waterman, has testified clearly and convincingly as to the advantages of the negative bias in amplifiers. The quality is preserved, the tone is undistorted, and battery consumption is reduced.
A number of patents are presented to sustain the contention of anticipation.
The Stone and Cabot patent, No. 884,110, certainly does not anticipate the Lowenstein patent, if the court reads it correctly, for the reason that, if it teaches anything, it teaches exactly the reverse of what Lowenstein claims. Stone and Cabot advise making the grid bias positive by means of the battery Z. They were dealing with telegraph signals, not with the reproduction of speech currents; but we are not now concerned with speculating as to the results they accomplished. They did not anticipate Lowenstein.
The De Forest patent, No. 841,347, neither in the drawing nor in the specification, describes the specific poling of the battery (B) in the grid circuit. If De Forest was at all concerned with this question, his patent does not disclose that he had knowledge of the difference in effect between a positive and a negative bias. It was contended by counsel for defendants on the trial that certain symbols on the drawing (De Forest) should be accepted as indicating the negative or positive plates. Careful checking of the symbols does not enlighten us, nor tend to support the contention. This De Forest patent teaches nothing as to the desirability of the grid negative bias.
De Forest patent, No. 879,532, does not disclose a battery in the grid circuit to bias it one way or the other. It does show, however, a "condenser" in the grid circuit ( C'). Defendants relied, apparently, and urged with much plausibility, that this condenser was the equivalent of a battery, so far as giving a negative bias to the grid was concerned. With the court's approval, defendant's expert set up an apparatus in court, and demonstrated, or attempted to demonstrate, the correctness of this contention. At first blush, it seemed that there was justification for the contention; but the disconnection and removal of the condenser and the readings of the meter satisfied the court that the contention was without satisfactory basis. The characteristic relied on in the demonstration, the court is satisfied, was purely incidental, and in practical manufacture of the tube not at all relied on to bias the grid in actual tube operation.
I am convinced that it would be highly erroneous to say that a condenser can be used to constitute "means for impressing" a negative potential upon the grid, or "a source of potential" more negative than the cathode, as called for in the Lowenstein claims referred to.
De Forest patent, No. 995,126, as the court has studied it, does not describe any means for impressing a negative bias on the grid, in any of the tubes shown in the figures of the drawing. Connecting the grid to the negative side of the filament does not impress a negative potential upon the grid with respect to the filament, as claimed in the Lowenstein patent, nor does it serve to accomplish the object described by Lowenstein.
The Von Lieben and Reisz patents (United States No. 1,038,910, and French No. 1,376) are quite different from the Lowenstein disclosure, and do not anticipate.
The Mathes Patent.
The Mathes patent, No. 1,426,754, is in the nature of an improvement on the Lowenstein invention. It discloses a specific or particular way of getting a negative potential on the grid. Claim 8 is relied upon:
"8. In an electric translating circuit, an electron discharge device having a cathode, an anode and an auxiliary electrode, said auxiliary electrode and said cathode being in the in-put circuit of said tube, a source of current connected to said cathode, a resistance, a circuit containing said cathode, said source and said resistance in series, said resistance being also included in the in-put circuit of said tube in such a manner that the potential of said auxiliary electrode is normally maintained lower than any part of said cathode by an amount substantially equal to the drop in said resistance, said cathode being rendered thermionically active by current flowing in said series circuit."
Mathes accomplishes the result sought for by a resistance in the circuit of the battery that lights the filament. He connects this resistance on the negative side of the battery and the filament on the positive side, and then connects the grid circuit to the most negative end of the resistance element. By doing this, the grid will have a potential which is negative with respect to the filament. The A battery is used for biasing the grid, thereby eliminating the C battery. While there are two objects sought by the Mathes patent, this suit is only concerned with the first object which is set forth in the Mathes application, namely, to secure "a desired difference of the potential between the filament and the grid of a tube of the audion type."
The defendants have called the court's attention particularly to Arnold patent, No. 1,465,332, and Langmuir patent, No. 1,273,627; but I cannot find that either of these patents shows a resistance, or any equivalent element, interposed between the filament and the negative end of the filament battery, for obtaining an ultra negative point of connection for the grid.
The Round (British) patent, No. 28,413, does not suggest connection of the grid to the negative end of the battery, nor is there any suggestion in the Round patent that a negative grid bias is desirable. Indeed, defendants' expert, in substance, admitted while on the stand that there was nothing in the Round patent disclosing the Mathes invention.
Defendants' answer alleges that Grebe himself was a prior inventor with respect to the Mathes invention. Defendants' Exhibit Z, in evidence, is presented as being typical of audio frequency amplifiers sold by Grebe in August, 1916. It is in evidence that, by connecting the A batteries in one way in this exhibit, the Mathes circuit would result, but not if connected in another way. No instructions were issued by Grebe when these were sold as to how to connect the batteries, and, indeed, there is no evidence in this record that Grebe knew in 1916 the advantages of a negative grid bias. The court questioned Mr. Grebe at length as to whether, at the time he constructed these amplifiers, in 1916, he did so for the purpose of obtaining a negative grid bias; but Mr. Grebe was unwilling to state that he had any such purpose in view, and I am satisfied, beyond peradventure, that Mr. Grebe, in 1916, did not know of the advantages of a negative grid bias, or of the use of a Mathes arrangement to accomplish the desired result.
Anticipation is not made out by a prior patent, which incidentally shows a similar arrangement, where such arrangement was not claimed or designed to perform the function of the patent in suit. Gray Tel., etc., v. Baird Mfg. Co. (C.C.A.) 174 F. 417. A mere chance connection, even if it produces the result, does not constitute an anticipation. Defendants' Exhibit Z, not only does not show the Mathes invention, but it was not early enough to anticipate it, in view of the stipulated facts, which place the date of the Mathes invention as early as February, 1916.
I am satisfied that the Grebe C R-9 receiver infringes claims 1 and 7 of the Lowenstein patent and claim 8 of the Mathes patent. There are two stages of audio frequency amplification in the Grebe receiver, and they are identical, in that they use the inventions of Lowenstein and Mathes.
Summing the matter up, the Lowenstein patent covers an audion amplifier system, with means for impressing upon the grid a potential more negative than that of the filament. The Mathes patent (claim 8) has reference to an arrangement "for impressing upon" the grid a potential more negative than that of the filament. The discussion and the testimony regarding this claim by Mr. Waterman, a witness for the plaintiffs, is enlightening.
I am of the opinion that both the Lowenstein and Mathes patents, so far as the claims under discussion are concerned, are valid and infringed by the defendant Grebe Company, in the manufacture and sale of the C R-9 radio receivers, and a decree will issue in accordance with this opinion, and with the stipulations filed herein.