306 F.2d 319
STANDARD COIL PRODUCTS CO., Inc., Plaintiff-Appellant,
GENERAL ELECTRIC COMPANY, Defendant-Appellee.
United States Court of Appeals Second Circuit.
Argued June 6, 1962.
Decided August 6, 1962.
Samuel Ostrolenk, of Ostrolenk, Faber, Gerb & Soffen, New York City (Marc S. Gross, of Ostrolenk, Faber, Gerb & Soffen, New York City, on the brief), for plaintiff-appellant.
John Hoxie, of Davis, Hoxie, Faithfull & Hapgood, New York City (Harvey M. Brownrout, of Davis, Hoxie, Faithfull & Hapgood, New York City, and Donald N. Timbie, Syracuse, N. Y., on the brief), for defendant-appellee.
Before CLARK, WATERMAN, and MOORE, Circuit Judges.
CLARK, Circuit Judge.
Plaintiff-appellant, Standard Coil Products Co., Inc., appeals from a judgment of the district court adjudging its patent Nelson No. 2,775,659 invalid for want of invention and alternatively not infringed by defendant's accused television amplifiers. Since we conclude that plaintiff's patent is invalid, we affirm. We reach this conclusion by analyzing the technical problems of television reception here involved to find a lack of adequate disclosure of any new invention and a lack of novelty in what was stated in the patent. This is substantially the approach of the trial judge — with whose views we are in accord — except that he went on to find no infringement, an issue we do not reach.
The relevant portions of the patent in question deal with a coupling circuit in a television amplifier. The amplifier described is a two-stage type, utilizing two vacuum tubes known as triodes and linked in a manner known as a "cascode."1 Plaintiff claims that the Nelson patent prescribes a specific form of coupling between these tubes by the use of which optimum performance in the amplification of television signals can be achieved. To resolve the issues raised by this appeal we must turn first to the technical background of the Nelson patent.
The patent arose out of the attempts by the television industry to solve the problem of "snow" on television screens in fringe areas. "Snow" (or "noise") is what television engineers call the signal distortion caused by certain inherent irregularities in the components of television amplifiers.2 These intrinsic distortions have little effect when the signals are strong, but can prove serious in the case of weak signals. The presence of "snow" on the screens of sets in fringe areas where signals were weak created a problem in the early days of television. The solution to this problem, as reflected by amplifiers such as those of both the plaintiff and defendant, was to increase the signal amplification and thus decrease the relative importance of the distortion.
The cascode amplifier had been developed during World War II by a group at the Massachusetts Institute of Technology who were trying to solve the noise problem in the amplifiers used in radar receivers. It was found that the cascode arrangement gave the best performance from a noise-reduction standpoint. Since television, like radar, operates at high frequencies, it was clear that this arrangement might similarly solve the "snow" problem in television.
There is, however, a fundamental difference between the operation of radar and television amplifiers, due to the wide frequency band of the latter, which complicated the adoption of the cascode principle. The tubes in an amplifier have what is known as a capacitance — the ability to retain electrostatic charges. This capacitance has a tendency to impede the flow of electric current, and thus decrease the system's ability to amplify the current passing through it. The impedance caused by a tube's capacitance is not a fixed quality; it increases with the frequency of the signal passing through the tube. Thus the higher the frequency the less amplification a tube will give. This effect can be offset by the introduction of a coil into the circuit. A coil has a quality known as inductance. An inductance presents an impedance to the flow of current, which also varies with the frequency of the current passing through the circuit. But this impedance operates in the opposite direction as the capacitance of the tube and also decreases as the signal frequency increases. Since its effect is directly opposite to that of a capacitator the coil can offset the amplification-dampening effect of the tube. There is, moreover, one specific frequency where the opposing effects (reactances) of a given coil and capacitator are equal — this is said to be the resonating frequency, and the circuit is said to be resonant. If the elements are connected in series the circuit will produce the least opposition to current at that point; if they are linked in parallel the circuit presents the maximum opposition and no current will flow through it.
The principle of resonance can be applied to offset the effect on amplification of high frequency signals. If a coil is introduced which will resonate with the tube at the desired frequency, amplification will be increased in the vicinity of the resonating frequency. This principle was recognized in radar cascodes, which incorporated coils tuned to the operating frequency of the radar receiver. It is in the adoption of this resonance principle from radar to television, however, that the differences between the two became relevant. Radar operates on a very narrow range of frequencies, while television signals are broadcast on a wide frequency band; channels 2-6 range from 54 to 86 megacycles, and channels 7-13 range from 174 to 216 megacycles. Since a given capacitance and inductance resonate at only one frequency, it seemed as though the radar approach could be adopted to television only by using many separate coils which would be tuned to resonate with the fixed tube capacitances at various frequencies. Such a system, however, would necessitate an elaborate switching mechanism so that the amplifier could be shifted from one coil to another as the frequency increased.
In his patent, Nelson, an employee of the plaintiff, claimed that such an expensive switching mechanism was unnecessary. He claimed to have discovered a very simple circuit which would permit the use of the low-noise cascode amplifiers developed for radar and still give high levels of amplification at all the television frequencies. The heart of the circuit described by the patent was an inductor coupling the two triodes and "* * * being proportioned to provide a negligible effect upon the passage of lower frequency channels of [the lower band of TV frequencies] * * * and to provide an inductance value that resonates with said driven triode cathode capacitance at a frequency near the higher frequency portion of said second band to increase the system amplification of higher frequency channels of said second band." This is the statement in Claim 12 of the patent, the single claim in issue here.
The parties disagree on the interpretation to be given this claim. Resolution of this threshold question of interpretation is crucial to the outcome of this appeal, for the issues of the patent's validity and possible infringement hinge on a determination of the subject matter of this patent.
This is the issue: Defendant asserts that Nelson attempted to patent what amounts to a scientific principle — the principle that a coil which resonates with the capacitance of the second tube of a cascode television amplifier will, by virtue of such resonance, give broad band and increased high frequency amplification. Thus the patent is read by defendant as merely describing a rig in which such resonance occurs. On the other hand, plaintiff asserts that Nelson was not attempting to describe a principle or a device operating on a certain principle. It construes Claim 12 and the related specifications as establishing no more than a formula for selecting a coil which will give the desired results when utilized in a working system. The discussion of resonance is explained not as an attempt to describe what happens in the operating amplifier, but as a shorthand method of telling an engineer the calculations he must make in order to select the correct coil. In fact a well-known formula describes the relationship between a given frequency and the inductance and capacitance which resonate at that frequency. To take an example stressed by plaintiff's expert in his testimony, given a capacitance (the rating of tube 2) and a frequency (216 megacycles or thereabouts), a simple computation will yield a figure for the resonating coil. It is that computation, and nothing more, that plaintiff claims Nelson "invented," and thus constitutes the subject of its patent. Plaintiff disavows any assertion that the amplifier achieves the desired functional results because of an actual resonance of the kind mentioned in the patent. Indeed, plaintiff's expert witness admitted that the resonance principle alluded to in the Nelson patent is not the principle upon which its own amplifier or defendant's accused amplifiers work.3 As plaintiff conceives Nelson's contribution, it was not in discovering that a coil would do the job, or why a coil would do the job, but in teaching the method to use to pick the right coil.
Thus plaintiff reads the patent as saying: A single coil interstage coupling will give the high amplification desired — to pick the right coil, treat it as if it were to resonate with tube 2 at 216 megacycles. For scientific reasons (not wholly made clear) such a coil gives the right results. Defendant says no; the patent states: Build an amplifier with an inductance which actually resonates with tube 2 at 216 megacycles or thereabouts and one will get the desired amplification.
After a lengthy trial and a full examination of the patent the trial court concluded that defendant's interpretation was the correct one. We agree with that decision. The claim quoted above and many of the specifications indicate that Nelson was describing the scientific principle or actual operation of the amplifier. The patent repeatedly uses words of action. Referring to the interstage coil, labeled Coil 27 in the drawings, the patent specifications state: "Coil 27 is made to resonate at a frequency in the neighborhood of channel 13 or 216 megacycles" and "* * * the series coil 27 by using a suitable value of inductance can be made to exactly resonate this capacity at one frequency (preferably around 11 or 12 channel) within the high band." (Emphasis supplied.) Moreover, in plaintiff's own answer to defendant's interrogatories, plaintiff stated that, according to its patent, "[f]or a given driven [second] triode selected, the inductor 27 is `proportioned' to produce therewith the desired resonance action, as emphasized in the specification and claims," and also in analyzing defendant's circuits that the interstage coil and the second tube coact "to produce * * * resonance." Thus at one stage in this litigation plaintiff itself adopted the "actual resonance" interpretation. For these reasons we conclude that the trial court's finding that Nelson's patent refers to an "actual resonance" was correct.
Plaintiff's main attack on this interpretation is to assert that it is "meaningless" and abstract. Putting aside plaintiff's own apparent reliance on the interpretation at one point in the litigation, we must say that there is merit in that claim. As was pointed out by both Nelson himself and Dr. Macnee, plaintiff's expert witness, the amplifier circuits here contain capacitances other than that of the second triode, all of which must be taken into account in determining the actual resonating frequency or effect of the circuit.4 Nevertheless, the patent clearly speaks as if a resonating effect occurred between coil 27 and tube 2 in isolation. Whether this is a result of imperfect understanding at the time the patent was applied for, or faulty drafting, or can be explained in some other fashion, we cannot of course say. But the patent clearly reflects a belief that this resonance is an operating quality of the amplifier.
Courts, with whom the responsibility for construing patents lies, will and should resolve ambiguities in a manner which will preserve the patent. Apex Electrical Mfg. Co. v. Maytag Co., 7 Cir., 122 F.2d 182, certiorari denied Maytag Co. v. Apex Electrical Mfg. Co., 314 U.S. 687, 62 S.Ct. 297, 86 L.Ed. 549. They cannot, however, add words that are not there or construe the patent in the light of subsequent discoveries which may have changed the patentee's understanding of what his own invention was. Cf. Smith, Kline & French Laboratories v. Clark & Clark, 3 Cir., 157 F.2d 725, certiorari denied Clark & Clark v. Smith, Kline & French Laboratories, 329 U.S. 796, 67 S.Ct. 482, 91 L.Ed. 681; Myers v. Beall Pipe & Tank Corp., D.C. Or., 90 F.Supp. 265, affirmed Fruehauf Trailer Co. of Cal. v. Myers, 9 Cir., 181 F.2d 1008, certiorari denied 340 U.S. 827, 71 S.Ct. 63, 95 L.Ed. 607. Since the patent clearly envisages an operational resonance between coil 27 and tube 2, we can adopt no other interpretation than that of the district court, even though, as will appear, on this interpretation the patent must be held invalid.
We turn, therefore, to a resolution of the substantive issues of this appeal. The first of these is whether or not the patent sufficiently disclosed the nature of Nelson's invention. There was evidence that plaintiff was able to produce a functioning low-noise cascode amplifier at the time Nelson's patent was applied for. If this was due to the patent, then even if Nelson attributed the functioning of the amplifier to the wrong principle, he may nevertheless have made an invention, albeit one which at the time he imperfectly comprehended. It is not necessary that an inventor understand all scientific principles involved in the device or process patented. General Motors Corp. v. Swan Carburetor Co., 6 Cir., 88 F.2d 876, certiorari denied 302 U.S. 691, 58 S.Ct. 49, 82 L.Ed. 534; Union Oil Co. of California v. American Bitumuls Co., 9 Cir., 109 F.2d 140; 2 Walker on Patents § 253 (Deller Ed. 1937). Failure to disclose or even grasp the full theoretical basis of an invention will not invalidate an otherwise valid patent if knowledge of such data is not necessary to the successful operation of the process or construction of the device.
For a patent to be valid, however, it must adequately disclose the invention which has been made. The test of whether disclosure is adequate is an operational one; the description must be of sufficient clarity so that one skilled in the art or science involved could make and use the invention. 35 U.S.C. § 112; A. B. Dick Co. v. Barnett, 2 Cir., 288 F. 799. Thus a failure to set forth the correct electronic principles governing the operation of the interstage coupling would not invalidate Nelson's patent if an ordinarily skilled television engineer could on the basis of the patent construct an operating cascode amplifier embodying Nelson's discovery.
The first difficulty with Nelson's patent, however, is that the confusion over "actual resonance" results in a disclosure insufficient to permit construction of such an operating system. As both sides admit, an inductance value calculated theoretically to produce resonance with a known tube capacitance at a stated frequency in actual practice will not be the inductance value which results in such resonance. All systems have stray capacitances which are difficult to measure, but which affect the operation of the circuit. Thus an engineer who has made theoretical calculations to get a preliminary inductance value will have to use experimental, or "cut and try," methods to find the operating inductance value which will produce resonance at a given frequency. If Nelson had provided enough information so that an engineer could reach the point where as an ordinarily skilled practitioner of the art he should realize he must use "cut and try" methods, Nelson might well have sufficiently disclosed an invention. But by telling the engineer to achieve actual resonance between coil 27 and the capacitance of tube 2, Nelson set him to establish a relationship which even plaintiff now admits is nonexistent. The steps that would be taken to achieve the working resonance are different from those that would be followed if only an initial rule of thumb were prescribed. To attempt to establish an operational resonance, the engineer, as defendant's expert testified, would quite naturally include in his computations factors, such as the cathode loading resistance of tube 2, which he would not apply if he believed he should use "resonance" only as a rule of thumb. Since the only fair reading of the Nelson patent is as describing an operating quality of the circuit, the invention which plaintiff now claims was made — a rule of thumb for coil selection — was not disclosed with sufficient clarity to permit a reasonably skilled engineer to construct an amplifier in accord with its principles. The patent therefore is invalid.
This conclusion — that the patent describes only a desired consequence and does not clearly state the means by which it is to be obtained — is further strengthened by the course of events during the trial. The trial judge repeatedly pressed the plaintiff for evidence that its own amplifiers were actually produced by employing Nelson's claimed disclosure or were in accordance with it. None was forthcoming. It seems apparent that not even plaintiff was able to make actual use of the claimed invention.
Since the patent must be read as one whose subject matter consists of an interstage coil in a cascode amplifier, whose actual resonance with tube 2's capacitance gives high-level broad band amplification, the patent must also be held invalid because its subject matter was anticipated by the prior art. Plaintiff concedes that the principle of using a cascode circuit to reduce noise was widely recognized. The M. I. T. group working on radar during World War II had recognized this principle, and it was set out at length in the book which made public the fruits of this research, Valley & Wallman, Vacuum Tube Amplifiers (1948), cited in Nelson's 1951 patent application. Thus plaintiff's claim to invention must rest on the patent's recognition that for television frequencies use of (1) a coil or (2) a coil which resonates with tube 2's capacitance in the higher television frequencies will give broad band amplification and offset the effect of higher frequencies on system amplification. But the principle that an interstage coil in a two-stage amplifier will offset the amplification-dampening effect of higher frequencies was clearly recognized by the prior art. This is the teaching of the Landon patent, No. 1,987,687. Landon was concerned with uniform amplification at higher frequencies, including television. He recognized that in two-stage amplifiers with direct couplings signal amplitude declines at high frequencies, and indicated that this effect could be offset by introducing an inductance between the two tubes to match the load impedance with the characteristic impedance of the system through establishment of a series resonant circuit. The subject matter of a patent is anticipated when "the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains." 35 U.S. C. § 103; Zoomar, Inc. v. Paillard Products, Inc., 2 Cir., 258 F.2d 527, certiorari denied 358 U.S. 908, 79 S.Ct. 237, 3 L.Ed.2d 230; American Stainless Steel Co. v. Ludlum Steel Co., 2 Cir., 290 F. 103. While Landon did not discuss the cascode amplifier, there is no indication that his teaching was limited to any one form of linkage in a two-stage amplifier. There is, therefore, nothing in Nelson's patent, as we interpret it, that would not have been "obvious" to a television engineer familiar with the Landon patent, save perhaps the principle of actual resonance between coil 27 and tube 2. This latter principle the trial court labeled "a step backward" from the prior art, as Landon had recognized, as did plaintiff's expert here, that any operational resonance effect would include the total capacitance of the circuit. Since plaintiff admits that the described resonance effect does not occur in even its own amplifier, there is no evidence that this feature constitutes an advance of any kind. Neither plaintiff's own nor defendant's amplifiers incorporate it. Thus we must hold the patent void for want of invention. See, e. g., Zoomar, Inc. v. Paillard Products, Inc., supra, 2 Cir., 258 F.2d 527, certiorari denied 358 U.S. 908, 79 S.Ct. 237, 3 L.Ed.2d 230; Deering, Milliken & Co. v. Temp-Resisto Corp., 2 Cir., 274 F.2d 626; Audio Devices, Inc. v. Armour Research Foundation of Illinois Institute of Technology, 2 Cir., 293 F.2d 102; Lorenz v. F. W. Woolworth Co., 2 Cir., 305 F.2d 102.
A "cascode" amplifier is one in which the grounded cathode of the first tube is connected with the grounded grid of the second tube. In practice this is made as a dual tube in a single glass envelope with a single base
"Noise" was defined by plaintiff's expert as "a name given to unwanted signals." What would appear as distorted sound in sound reproduction would appear as dots — and thus "snow" — on picture reproduction on the television screen
Like the defendant's amplifiers, plaintiff's product works by virtue of a resonance established between the capacitance of both tubes 1 and 2 with the coil. The essence of plaintiff's claim is that defendant uses coils of a value whichtheoretically would resonate in the manner described in the patent, and that the patent covers all amplifiers containing coils of a size corresponding to the calculations which the patent indicates should be made.
These include the output capacitance of the first tube, tube 1, and "stray" capacitances present in all circuits