Summary
In Technidyne Corporation and Lester L. Jones v. McPhilben-Keator, Inc., 72 F.2d 242, it was said: "Many patents have been issued in this well-developed field, and, as indicated by the date of their issuance, all are of recent grants.
Summary of this case from Hazeltine Corporation v. AbramsOpinion
No. 294.
July 23, 1934.
Appeal from the District Court of the United States for the Eastern District of New York.
Suit by the Technidyne Corporation and another against McPhilben-Keator, Incorporated, for infringement of ten patents. From a decree sustaining four patents, holding them valid and infringed, and dismissing the bill as to six patents ( 1 F. Supp. 423), both parties appeal.
Modified, and bill dismissed.
James Franklin, of New York City, for plaintiffs.
Stephen H. Philbin, of New York City (W.E. Mueller, of Syracuse, N.Y., of counsel), for defendant.
Before MANTON, L. HAND, and AUGUSTUS N. HAND, Circuit Judges.
This appeal involves the validity and infringement of seven patents, including a reissue, to Jones and assigned to the plaintiff Technidyne Corporation, known as reissue patent No. 17,915 originally filed August 21, 1925, and issued December 30, 1930; patent No. 1,620,661, application filed March 6, 1926, issued March 15, 1927; patent No. 1,770,525, filed July 15, 1927, issued July 15, 1930; patent No. 1,788,197, filed October 5, 1929, and issued January 6, 1931; patent No. 1,673,287, filed June 11, 1927, and issued June 12, 1928; patent No. 1,779,881, filed October 5, 1929, and issued October 28, 1930; patent No. 1,791,030, filed October 5, 1929, and issued October 28, 1930; and one issued to Bonine, known as patent No. 1,696,263, filed August 12, 1922, and issued December 25, 1928. The eight patents relate to radio tubes and circuits.
Many patents have been issued in this well-developed field, and, as indicated by the date of their issuance, all are of recent grants. Since the radio art was well developed, the standard of inventive thought to sustain a patented radio tube or circuit necessarily is higher than in the earlier history of the art. Kintner v. Atlantic Communication Co., 230 F. 829 (D.C.), affirmed 240 F. 716 (C.C.A. 2); Dubilier Condenser Co. v. Aerovox Wireless Corp., 37 F.2d 657, 660 (C.C.A. 2); Radio Corp. v. Dubilier Corp., 59 F.2d 305 (C.C.A. 3); Western Electric Co. v. Wallerstein, 60 F.2d 723 (C.C.A. 2).
For an understanding of the patents in suit, a brief description of radio operation may be helpful. Radio signals are transmitted and received in the following manner: Speech or music produces sound waves, which have frequencies of alternation or pitch, ranging from about 50 cycles per second to 20,000 cycles. These are called audio frequencies and higher frequencies are termed radio frequencies. At audio frequency sound waves are applied to the diaphragm of a microphone, it vibrates in accordance with the sound waves, causing an electric current to alternate at the same frequency. Audio frequency currents are then applied to the radio frequency current, such as 1,000,000 cycles per second, modulating it. A voice (audio) frequency of 5,000 cycles would vary the radio frequency from 1,005,000 cycles to 995,000 cycles. The audio modulated radio frequency currents are then applied to a transmitting antenna which produces ether waves, corresponding in frequency to modulated currents. A receiving antenna intercepts the waves, thereby setting up radio frequency currents with the same frequency as that of the waves. The radio frequency currents are amplified in the receiver and are then subjected to the process of detection, whereby the audio frequency modulations are separated and appear as audio frequency currents; the radio frequency currents being eliminated. The audio frequency currents are then amplified, and actuate the diaphragm of a loud speaker or telephone receiver, thus producing the sound waves, which have the same frequency or pitch as the original sound waves.
The broadcasting band is from 550 KC to 1,500 KC. A radio wave of electrical current to-day is usually referred to in the terms of frequency; formerly they were referred to in terms of length. Frequency is readily convertible into wave length; either frequency or length can be obtained by a known method of calculation. In a receiver, the radio frequency currents in the antenna are amplified by one or more tube stages. The detector produces audio frequency currents, which are amplified by one or more tube stages and then actuate a loud speaker. A battery supplies energy to heat the filament, which then emits electrons that go to the plate. Another battery supplies energy to the plate circuit and another battery maintains a negative charge on the grid. Lowenstein patent; Western Electric Co. v. Wallerstein, 60 F.2d 723 (C.C.A. 2).
Tuning, or selectivity, becomes necessary because the antenna has as many currents of different frequency as it has intercepted radio waves of different frequencies. The user wishes to receive only currents of one frequency at a given time, and therefore tunes to that frequency so that it discriminates against all other frequencies. In modern broadcast receivers, tuning to stations of different frequencies is accomplished by varying the capacity (condenser).
In the second tube stage of the receiver, there may be no variable tuning and therefore at times it is referred to as an untuned stage. However, it is tuned broadly to a band of frequencies, although not narrowly to a single frequency. The same natural capacities are in the tuned circuit of the first tube, but their effect is very small compared to the large and controlling effect of the variable condenser. But such nonvariable tuning circuits are sometimes referred to as untuned.
The court below treated the patents in groups. The appeal has been argued with the same regard for these groups. The first group (Jones, reissue No. 17,915; No. 1,620,661; No. 1,770,525; No. 1,788,197) deals with tuned circuits, radio tubes, feedback or regeneration, and means for neutralizing or controlling feedback. A tuned circuit is a circuit tuned to a current of certain frequency of alternation. The circuit is tuned to the desired frequency or band of frequencies by adjusting its values of inductance, capacity, or both. Every electrical circuit has each of the three fundamental characteristics of inductance, capacity, and resistance, although the values of each differ greatly in different circuits.
A radio tube comprises a hot cathode which emits electrons that go to a cold anode (plate) both inclosed in an evacuated glass envelope. A grid, also inclosed, is located between the filament and the plate so that, when the varying potential of a signal is impressed upon the grid, the electron stream is correspondingly modified. The electron stream is a part of a powerful electrical current which is modified to produce great amplification of the feeble grid potentials. Regeneration or feedback can be produced in the radio tube when used as an amplifier, because it then has currents in its plate (output) circuit which are greater than in its grid (input) circuit. If the output is coupled to the input, a part of the amplifying energy can be fed back to be added to the in-coming signal currents, and, as modified, will be amplified further by the tube circuits, feedback again occurring, and this will continue on until the losses of the circuits put an end thereto. We have considered controlling a feed back heretofore. Armstrong v. DeForest Radio Telephone Telegraph Co., 280 F. 584 (C.C.A. 2); Radio Corp. v. Radio Engineering Laboratories, 66 F.2d 768 (C.C.A. 2).
In 1917, Rice improved neutralization by an auxiliary circuit through which energy was fed back to oppose or neutralize the positive feedback. Radio Corp. v. Twentieth Century Corp., 19 F.2d 290 (C.C.A. 2). In 1919, Hazeltine further improved neutralization. Hazeltine Corp. v. Wildermuth, 34 F.2d 635 (C.C.A. 2). Thus the art relating to tuned and untuned radio tubes, feedbacks, and the control of feedbacks was well known prior to the grant of the patents in suit. Men skilled in the art after knowledge of the results of these inventions were fully capable of controlling feedback in amplifiers, and the basic question on this appeal is whether what the inventors did in the patents in suit rose to the dignity of inventive thought or constituted engineering efforts of a man skilled in the art.
Reissue No. 17,915. The first patent of the group (Reissue No. 17,915) is to provide control of feedback in the same untuned circuits. The patent was the earliest of the group as well as the broadest, and has admittedly for its object an improvement of the old radio frequency system (Alexanderson and Rice), so that it will have "an amplification power substantially greater than that attainable in radio frequency systems known to the prior art." And Jones says that by this patent he wanted to use an untuned radio frequency receiver which did not have as much feedback as the usual receiver of that type, so he employed two expedients to reduce the feedback. The first expedient was to not tune the plate circuit of one of the tube stages, and the other expedient was the resistance. Below, this patent was held valid and infringed as to claims 1, 3, 5, 7, 10, 13, 14, 15, and 25.
By untuning the plate circuit (output) that is, by not tuning it to a frequency within the range, the feedback is reduced. This was old in the art before Armstrong tuned the plate circuit to obtain feedback, and the fact that feedback depended upon tuning the plate circuit was fully understood back in 1913. Armstrong v. DeForest Radio Telephone Telegraph Co. (C.C.A.) 280 F. 584.
Jones accomplished his result by fixing the natural frequency of the untuned output circuit and succeeding untuned input circuit at a higher frequency than that which would be encountered within the range of selectivity. He thus made the effect of untuned circuits constant always, a feedback, and then, having done this, he could neutralize this by a resistance in ways well known to the art. The natural frequency or resonance of an untuned output circuit was in several instances set above frequencies to be encountered therefor, as was done by Rawsthorne (British patent No. 223,053 [1925]) although his succeeding input was tuned. While it does not appear that he conceived the idea of making the frequency of both output and input untuned circuits above the wave length band, the tuning of the plate circuit to produce feedback was well known and proved. It did not require inventive thought in the then state of the art to think of counteracting this by a resistance. The use of a resistance to overcome the effects of regeneration was old, and the designing of transformers and having a frequency outside the range was old, and to use both in combination was not invention as the court below found. It was an engineering arrangement which one skilled in the art would be expected to make if he wanted the desired result. It was error to sustain any of the claims of this patent. Pearce v. Mulford, 102 U.S. 118, 26 L. Ed. 93; DeForest v. Genl. Electric Co., 283 U.S. 664, 51 S. Ct. 563, 75 L. Ed. 1339; Carbice Corp. v. Amer. Patents, 283 U.S. 420, 51 S. Ct. 496, 75 L. Ed. 1153; Powers-Kennedy Contracting Corp. v. Concrete Co., 282 U.S. 175, 51 S. Ct. 95, 75 L. Ed. 278; Western Electric Co. v. Wallerstein, supra.
Patent No. 1,620,661. Claims 5, 6, and 7 of patent No. 1,620,661 were sustained and held to be infringed. After the reissue patent had been developed, this patent was applied for "for improvements in the radio frequency cascade amplifying system." The inventor says that he proposed to cure a defect in this improvement by a particular way of connecting a condenser at a point between the ends of the resistance in the plate circuit which he used in his reissue patent to reduce feedback. The connection must be either at one of the ends or intermediate, and Jones says that he found that, when the coupling condenser is connected to the plate end of the resistance, there will be a "loss of selectivity," that, when connected to the other end, there may be "self sustaining oscillation therein," and that the connections "between the points (its ends) * * * results in the substantial disappearance of this disturbing distant stage feed back reaction." The patent teaches that, if the condenser is connected to the plate end of the resistance, the tuning of the selective system is affected, and, if it is connected to the other end of the resistance, there will be oscillation. By selecting some intermediate point of resistance, both these effects can be avoided. The defendant's apparatus, however, does not use a tapped resistance, but the condenser is connected directly to the plate side of the resistance in question. In other words, the apparatus of the defendant was doing exactly what the patent says not to do. The defendant does not use what the patent says is essential, instead it uses a nonintermediate connection. To constitute infringement, it should be shown that there is substantial identity, and here there is not. Silver Co. v. S. Sternau Co., 258 F. 448, 451 (C.C.A. 2); Linde Co. v. Morse Co., 246 F. 834 (C.C.A. 2).
Claim 5 specifies "means for neutralizing said feed back reaction," claim 6, "resistance in the output circuit of the first tube for neutralizing said feed back reaction," and claim 7 calls for a "resistance."
These phrases mean the same thing, because the tapped resistance is the resistance or neutralizing means disclosed. The patent warns against a nontapped resistance such as is used in defendant's set. It was error to find infringement, and the decree must be reversed as to this patent.
Patent No. 1,770,525. As to this patent, the claims were held valid but not infringed. Jones says he found the system of the patent just considered was defective, because the feedback has not yet been sufficiently decreased, so he (a) changed the first tube of his system by putting a coil and a condenser around a part of the resistance (this coil and condenser are not used by the defendant, as the court found); and (b) he made the plate circuit of the second tube capacitative. The features of this claim are the so-called automatic tuning or causing the input capacity to increase with increasing wave lengths by means for neutralizing feedback. The new thing claimed in this patent is the capacitative plate circuit of the second tube. The inventor claims that he finally obtained a reactionless, free from feedback, amplifier.
The Barkhausen German patent, No. 343,702 (1921), describes amplifiers which oscillate because of feedback, and it proposes the same expedient as Jones claims to be his invention, to control the feedback. Claim is made in the Barkhausen patent for "amplifier circuit with vacuum tubes characterized by the fact that the natural frequency of the conductor system of the output side (anode circuit) of the tube is lower than the natural frequency of the input side (grid circuit) of the same tube."
In view of the Barkhausen patent, there was no invention in this accomplishment, and the court below should have held the patent to be invalid.
Patent No. 1,788,197. The court below found claims 1, 5, 8, 18, 21, and 26 valid and infringed. Jones says the purpose of this patent was to improve the system of his No. 1,770,525 patent by adding "specific improvement," and the patent says that the object of the present invention was to improve "such reactionless coupling means used between any successive vacuum tubes, and more especially to improve the coupling means used in a radio receiving circuit such as is disclosed in the said co-pending application."
The alleged invention was for the purpose of doubling the resistance and inductance values, to obtain greater amplification. In patent No. 1,713,030, the inventor disclosed a system of amplification by inductance with corresponding resistance. He said: "I intentionally selected a large inductance and to make this possible intentionally reduced the capacitance in parallel."
It was known at the time that an increased inductance would amplify more than a lesser inductance and whether you reduce the capacitance for one reason or another is irrelevant at least after the reissue patent No. 17,917. Having an increased inductance, the inventor had to have a neutralizing increased resistance, and he added to his resistance. This was a detailed improvement consisting of a special selection of the constants of the inductance and resistance in the method of coupling used in all the patents. It is found between the first and second tubes; it is no more than good engineering and did not involve invention.
The decree holding this patent valid and infringed is reversed.
The second group of patents included Bonine and Jones, No. 1,696,263 and No. 1,779,881. They relate to automatic tuning, and the plaintiffs claim infringement by the third, fourth, and fifth amplifier tube circuits in the defendant's set. The court below held that Bonine was anticipated by the Barkhausen No. 343,702 patent. The first of the four radio frequency amplifier tubes has a variable tuned input (grid circuit), by means of a variable condenser. The next three tubes are not variably tuned, but the transformers connecting them are so constructed that they are automatically tuned over the desired tuning range. The tuning is by varying the amount of inductance in each transformer, as by having in the first transformer 90 turns and the second 112 turns and in the third 140 turns. The gist of the invention was "to find that building an untuned amplifier of a multi-stage type which is capable of amplifying over a wide wave length band and which uses fairly good amplifier tubes of the type commercially used in this country, that he could get a large amount of inherent stability in the amplifier by arranging the natural wave lengths of the transformer in definite relation one to the other, that relation being that the first amplifier transformer had a short wave length, the next one had a slightly longer wave length, and so on."
Each successive interstage transformer is so adjusted that its resonant frequency is lower than the preceding transformer; therefore the reaction through each tube is the reverse of that necessary to produce regeneration. The transformer may become a capacity by increasing the windings, and, if it does, there is no feedback. This was not a discovery of this inventor, for it had been known that feedback was dependent upon an inductance state of the output circuit. He did establish a stability over a larger wave length band by tuning his stages successively to longer and longer wave lengths; the output resonance always being lower than the input. This is what Barkhausen did in an audio multistage amplifier. It is claimed that the shunt condensers of Barkhausen would not do at all at radio frequencies, but the phenomena are the same at either frequencies; that is, the output circuit will never feed back if it is capacitative. There is no invention in adapting an idea so fully expressed by Barkhausen to a similar situation for precisely the same purpose. It did not amount to invention.
The decree of the court below as to this patent is affirmed.
Patent No. 1,673,287. This was filed five years after Bonine, and is for an automatic tuning for uniform amplification. The object of the invention is said to relate specifically to the production of a high and reasonably constant degree of amplification over a band or range of 500 to 1,500 KC. The claims sued on, Nos. 1, 2, 3, 8, 10, and 14, were held not to be infringed.
The patent describes the same type of fixed tube amplifier as Bonine. The inventor proposed to accomplish his purpose by constructing a plate circuit of capacitative (long wave or low frequency) type, so that the effect of the capacity reflected back into the grid circuit will aid the latter's tuning. He proposes to provide a transformer in the plate circuit whose capacity effect would change with frequency, so that, as the frequency decreased or increased, more or less capacity would be added to the grid circuit and thus aid in tuning it to whatever frequency was received. This increase in capacity of the input circuit was a tuning of that circuit as though a condenser had been put across that circuit. The inventor describes the process and left to the skill of an engineer to arrange the various capacities, resistance and inductance. Bonine had done this in his multiradio stage tubes, for his series was in a set of steps; the output circuit being always of lower natural frequency than the input. The patent says that the invention is "the obtaining of a uniformly high amplification over said wide wave length band" or "an amplifier which is self tunable over the wave length band to produce uniformly high and efficient amplification over the band."
By this patent the inventor did not discover any new principle of operation. He merely stated what was known to the art, namely, that the plate circuit had some tuning effect over the grid circuit, and he proposed that such effect be great enough to automatically tune, which is impossible. This patent was also invalid.
Patent No. 1,779,881 is based on the idea, as was the last patent considered, that the grid circuit can be automatically tuned by the plate circuit. The patent admits that the prior patent was mistaken in attempting to tune automatically over a range of frequencies for the present patent uses two tubes for this purpose; one being tuned over one part of the range and the other tube being tuned over another part. The court below held the patent valid and infringed as to claims 1, 10, 17, and 19.
One of the two disclosures of the patent is that, instead of trying to make one transformer do all the tuning, the range of frequency should be divided. The inventor found that he could not range a single output circuit to accommodate itself to all variations of frequencies, and he distributed it into two. This is the only new idea with respect to automatic tuning. It is a restatement of the earlier patents including the prior automatic tuning plan.
He also applied the old volume control of his automatically tuned amplifier in such a way as to prevent feedback. He found that his automatically tuned amplifier would oscillate when be used volume control which would be varied by the user, with a consequent change of currents in the set, and, as a radio engineer would be expected to do, he varied his feedback cause inductance until the circuits were tuned to the proper values. There was no invention in as late as 1929 varying inductance either to produce or prevent feedback.
It means to tune the inductance in the plate circuit so that its frequency would never exceed that of the grid circuit. An engineer familiar with the art could build up any type of set with any type of volume control, namely, to adjust the circuits so that they would not have too much feedback in the year 1929. The art knew from Armstrong in 1913 that tuning in the various inductances was one of the ways of obtaining or preventing feedback. This accomplishment was the result of engineering skill only and not inventive genius.
The decree sustaining this patent is reversed.
The third group consists of the patent to Jones, No. 1,791,030. It was held invalid by the court below. It relates particularly to a detector, the function of which is the separation of audio frequency modulations from other radio frequency carrier currents. It is, first, the use of a detector that requires only one audio frequency amplifier tube for operation of the loud speaker; and, second, a detector that will overload when loud signals are applied before the following audio frequency amplifier tube is overloaded. There was used in the first practice after grid bias detection was invented and was in use two stages of audio frequency amplification after the detector. As radio frequency amplifiers became more efficient, it was found unnecessary to use more than one stage of audio frequency amplification provided enough voltage was used on the detector. The use of the grid bias voltage in the detector was well known at the time of the application for this patent, and its special desirability for strong signals. Jones was not the first to use a detector of the grid bias type with a high radio frequency input as against a detector of the grid condenser type. Grid bias detection is effective for higher input voltages. Wireless World Article 1926.
Apparently Jones thought that the grid bias detection was new with him. The fact that it was desirable to operate a detector with a grid bias when strong signals are applied to it was known in 1928 with respect to tube characteristics and operations. Its functions were explained particularly with respect to voltages on filament and plate in 1906. Marconi Wireless Tel. Co. v. De Forest Radio Telephone Telegraph Co., 243 F. 560 (C.C.A. 2). Lowenstein, in 1912, explained the importance of grid voltage, Western Electric Co. v. Wallerstein (C.C.A.) 60 F.2d 723, and Armstrong, in 1913, described the interaction of grid circuit and plate circuit to produce feedback, Armstrong v. De Forest Radio Telephone Telegraph Co. (C.C.A.) 280 F. 584.
Jones is not entitled to his claimed monopoly on the use of a single audio amplifier stage, and the claims in suit were correctly held to be invalid.
The decree will be modified in accordance with this opinion, and the bill dismissed, with costs.